Kinesins and Cancer

Frank
Kozielski Editor
Kinesins
and Cancer
Kinesins and Cancer
Frank Kozielski
Editor
Kinesins and Cancer

Editor
Prof. Frank Kozielski, FSB
Chair, Department of Pharmaceutical
and Biological Chemistry
School of Pharmacy
University College London
London, WC1N 1AX, UK
ISBN 978-94-017-9731-3 ISBN 978-94-017-9732-0 (eBook)

DOI 10.1007/978-94-017-9732-0
Library of Congress Control Number: 2015932649
Springer Dordrecht Heidelberg New York London

© Springer Science+Business Media Dordrecht 2015
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, express or implied, with respect to the material contained herein or for any errors
or omissions that may have been made.
Printed on acid-free paper
Springer Science+Business Media B.V. Dordrecht is part of Springer Science+Business Media

(www.springer.com)
Preface
The validation of proteins as disease targets is a notoriously difficult and complex

task. In this book, we attempt to validate members of the kinesin superfamily as
potential targets for drug development in cancer chemotherapy.
The first chapter sets out the groundwork for subsequent chapters by summarising
current knowledge and highlighting common principles and features of the kinesin
superfamily, focusing mainly on kinesins involved in mitosis and cytokinesis
(mitotic kinesins). The following three chapters illustrate the present status of the
most advanced kinesin Eg5 in terms of drug development. Chapter 2 describes the
development of highly potent and specific Eg5 inhibitors, a range of which are in
clinical development, whereas Chap. 3 details the mechanism of these Eg5-targeting
drugs. Chapter 4 summarises the outcome of these inhibitors in multiple phase I and
II clinical trials. This chapter also explains some of the reasons why clinical success
has so far been only moderate, and in addition provides some intriguing ideas about
the lessons to be drawn from these trials for the improvement of future clinical trials
on Eg5. In the following eight chapters (Chaps. 5, 6, 7, 8, 9, 10, 11 and 12) experts
in the motor field attempt to systematically validate or exclude the remaining 15
human mitotic kinesins as potential targets for cancer chemotherapy. By summarising
current knowledge on how mitotic kinesins work, the authors provide a balanced
opinion about their usefulness as drug targets, taking into account potential unde-
sired side effects.
Surprisingly, a range of mitotic kinesins seem to represent potential novel targets,
making them worthwhile starting points for the development of hits as tool
compounds pending further validation. However, previously unknown functions
may be discovered or redundant pathways revealed, which may exclude some of the
kinesins as potential targets. Will this initial assessment of kinesins as potential
targets remain valid in the longer term? Only time will tell.
Kinesins are not the only novel prospective targets of the mitotic spindle. Other
non-motor spindle proteins are coming to the fore. Chapter 13, on non-kinesin targets,
is an attempt to broaden the search for potential targets of the mitotic spindle.
v
vi Preface
Finally, the last chapter – on neuronal kinesins – attempts to focus attention on

multifunctional kinesins, in particular those that play important roles in mitosis and
neuronal development. Work on these kinesins will require careful monitoring to
avoid unexpected side effects in the central nervous system, though early results
suggest that these worries are unfounded.
These are very exciting times for scientists working in the motor field. By applying
cutting-edge methodology, the functions of mitotic kinesins are being elaborated
and the mechano-chemistry of these molecular machines revealed in astonishing
detail. At the same time kinesins are emerging as medically relevant proteins. They
may in the future make the leap from potential to validated targets. Very exciting
times lie ahead. I hope that this book will contribute to the initial validation of
mitotic kinesins as targets for drug development in cancer chemotherapy.
London, UK Frank Kozielski

February 2015
Contents
1 The Kinesin Superfamily........................................................................ 1

Kristen J. Verhey, Jared C. Cochran, and Claire E. Walczak
2 The Discovery and Development of Eg5 Inhibitors
for the Clinic ............................................................................................ 27
James A.D. Good, Giacomo Berretta, Nahoum G. Anthony,
and Simon P. Mackay
3 Mechanisms of Action of Eg5 Inhibitors............................................... 53
Robert A. Cross
4 Clinical Trials of Mitotic Kinesin Inhibitors ........................................ 63
Steven S. Rosenfeld
5 Kif15: A Useful Target for Anti-cancer Therapy?................................ 77
Roy G.H.P. van Heesbeen and René H. Medema
6 Down-Regulating CENP-E Activity: For Better
or for Worse ............................................................................................. 87
Julien Espeut and Ariane Abrieu
7 The Human Kinesin-14 Motor KifC1/HSET Is an Attractive
Anti-cancer Drug Target ........................................................................ 101
Vaishali Pannu, Padmashree C.G. Rida, and Ritu Aneja
8 Kinesin-13 Microtubule Depolymerizing Proteins
as Targets for Cancer Therapy .............................................................. 117
Anutosh Ganguly and Fernando Cabral
9 Chromokinesins in Genome Maintenance and Cancer ....................... 135
Manjari Mazumdar and Tom Misteli
vii
viii Contents
10 Kif14: A Clinically Relevant Kinesin and Potential

Target for Cancer Therapy .................................................................... 149
Brigitte L. Thériault and Timothy W. Corson
11 Kinesin-8 Members and Their Potential as Biomarker
or Therapeutic Target ............................................................................. 171
Thomas U. Mayer and Silke Hauf
12 The Kinesin-6 Members MKLP1, MKLP2 and MPP1 ....................... 193
Ryan D. Baron and Francis A. Barr
13 Non-motor Spindle Proteins as Cancer
Chemotherapy Targets ........................................................................... 223
Robert L. Margolis and Mythili Yenjerla
14 Inhibitors of Mitotic Kinesins for Cancer
Treatment: Consequences for Neurons ................................................. 251
Olga I. Kahn and Peter W. Baas
Index ................................................................................................................. 269

Contributors
Ariane Abrieu Université Montpellier, CRBM, Montpellier, France

CNRS UMR 5237, Montpellier, France
Ritu Aneja Department of Biology, Georgia State University, Atlanta, GA, USA
Nahoum G. Anthony Strathclyde Institute of Pharmacy and Biomedical Sciences,
University of Strathclyde, Glasgow, Scotland, UK
Peter W. Baas Department of Neurobiology and Anatomy, Drexel University
College of Medicine, Philadelphia, PA, USA
Ryan D. Baron Department of Molecular and Clinical Cancer Medicine, University
of Liverpool, Liverpool, UK
Francis A. Barr Department of Biochemistry, University of Oxford, Oxford, UK
Giacomo Berretta Strathclyde Institute of Pharmacy and Biomedical Sciences,
Fernando Cabral Department of Integrative Biology and Pharmacology,
University of Texas Medical School, Houston, TX, USA
Jared C. Cochran Department of Molecular and Cellular Biochemistry, Indiana
University, Bloomington, IN, USA
Timothy W. Corson Eugene and Marilyn Glick Eye Institute, Departments of
Ophthalmology, Biochemistry and Molecular Biology, and Pharmacology and
Toxicology, and Simon Cancer Center, Indiana University School of Medicine,
Indianapolis, IN, USA
Robert A. Cross Warwick Medical School, Coventry, UK
Julien Espeut Université Montpellier, CRBM, Montpellier, France
CNRS UMR 5237, Montpellier, France
ix
x Contributors
Anutosh Ganguly Department of Microbiology and Infectious Diseases, Snyder

Institute, University of Calgary, Calgary, AB, Canada
James A.D. Good Department of Chemistry, Umeå University, Umeå, Sweden
Umeå Centre for Microbial Research, Umeå University, Umeå, Sweden
Silke Hauf Department of Biological Sciences and Virginia Bioinformatics
Institute, Virginia Tech, Blacksburg, VA, USA
Olga I. Kahn Department of Neurobiology and Anatomy, Drexel University
College of Medicine, Philadelphia, PA, USA
Simon P. Mackay Strathclyde Institute of Pharmacy and Biomedical Sciences,
Robert L. Margolis Tumor Initiation and Maintenance Program, Sanford-
Burnham Medical Research Institute, La Jolla, CA, USA
Thomas U. Mayer Department of Biology and Konstanz Research School
Chemical Biology, University of Konstanz, Konstanz, Germany
Manjari Mazumdar Kimberley Lane, Houston, TX, USA
René H. Medema Division of Cell Biology, The Netherlands Cancer Institute,
Amsterdam, The Netherlands
Tom Misteli National Cancer Institute, NIH, Bethesda, MD, USA
Vaishali Pannu Department of Biology, Georgia State University, Atlanta, GA,
USA
Padmashree C.G. Rida Department of Biology, Georgia State University, Atlanta,
GA, USA
Steven S. Rosenfeld Department of Cancer Biology and Rose Ella Burkhardt
Brain Tumor Center, Cleveland Clinic Foundation, Cleveland, OH, USA
Brigitte L. Thériault Campbell Family Cancer Research Institute, Ontario Cancer
Institute, Princess Margaret Cancer Centre, Toronto, ON, Canada
Roy G.H.P. van Heesbeen Division of Cell Biology, The Netherlands Cancer
Institute, Amsterdam, The Netherlands
Kristen J. Verhey Department of Cell and Developmental Biology, University of
Michigan Medical School, Ann Arbor, MI, USA
Claire E. Walczak Medical Sciences, Indiana University, Bloomington, IN, USA
Mythili Yenjerla Tumor Initiation and Maintenance Program, Sanford-Burnham
Medical Research Institute, La Jolla, CA, USA
Chapter 1
The Kinesin Superfamily
Kristen J. Verhey, Jared C. Cochran, and Claire E. Walczak
1.1 Introduction
Kinesin motors play important roles in cellular processes such as intracellular

transport, organization of the microtubule cytoskeleton, signal transduction, cell
motility, and cell division. The first kinesin motor to be discovered was initially
called conventional kinesin [1, 2] but is now referred to as kinesin-1 [3]. Further
genetic, bioinformatic, and biochemical approaches over the past 25 years have
uncovered a large number of genes across eukaryotic species that contain the core
kinesin motor domain. Phylogenetic analysis has classified these gene products into
14 families [4] (Fig. 1.1), with more recent analyses suggesting that there may be
17 kinesin families [5]. Many of the kinesin proteins are involved in mitosis in
mammalian cells. In fact, the mitotic kinesins were the first members of the kinesin
superfamily to be discovered after kinesin-1. This is likely due to the fact that
mitotic kinesins play key roles in chromosome segregation, and many of kinesin
superfamily members were discovered through classical genetic screens in model
organisms.
K.J. Verhey (*)

Department of Cell and Developmental Biology, University of Michigan Medical School,
Ann Arbor, MI 48109, USA
e-mail: kjverhey@umich.edu
J.C. Cochran
Department of Molecular and Cellular Biochemistry, Indiana University,
Bloomington, IN 47405, USA
e-mail: jcc6@indiana.edu
C.E. Walczak
Medical Sciences, Indiana University, Bloomington, IN 47405, USA
e-mail: cwalczak@indiana.edu
© Springer Science+Business Media Dordrecht 2015 1

F. Kozielski (ed.), Kinesins and Cancer, DOI 10.1007/978-94-017-9732-0_1
2 K.J. Verhey et al.
kinesin-1 (HsKIF5A, HsKIF5B, HsKIF5C)
kinesin-2 (HsKIF3A, HsKIF3B, HsKIF3C, family also contains HsKIF17)
kinesin-3 (HsKIF14, family also contains HsKIF1, HsKIF13, HsKIF16)
kinesin-4 (HsKIF4, family also contains HsKIF7, HsKIF21, HsKIF27)
kinesin-5 (HsKIF11/Eg5)
kinesin-6 (HsKIF20A/MKLP2, HsKIF20B/MPP1, HsKIF23/MKLP1)
kinesin-7 (HsKIF10/CENP-E)
kinesin-8 (HsKIF18A, HsKIF18B, HsKIF19)
kinesin-10 (HsKIF22/Kid)
kinesin-12 (HsKIF15/HKlp2, family also contains HsKIF12)
kinesin-13 (HsKIF2A, HsKIF2B, HsKIF2C/MCAK, HsKIF24)
kinesin-14 (HsKIFC1/HSET, HsKIFC2, HsKIFC3)
kinesin motor domain forkhead associated (FHA) domain
neck linker coiled coil lipid binding, e.g. PH, PX, StarD9
ATP-indep. microtubule binding DNA binding, e.g. helix-hairpin-helix (HHH)
Fig. 1.1 Domain organization of the kinesin families. The core catalytic motor domain is indi-
cated in blue. Most kinesins have their motor domain at the N-terminus whereas the kinesin-13
family has their motor domain in the middle of the molecule and the kinesin-14 family has it at the
C-terminus. All kinesins contain some region of predicted coiled coil for dimerization
1 The Kinesin Superfamily 3
All kinesins contain a kinesin motor domain of ~350 amino acids that includes
the sequences necessary for converting the chemical energy of ATP hydrolysis into
mechanical work along microtubule filaments. Outside of the motor domain,
kinesins contain unique non‐motor regions that confer specific regulatory and/or
functional properties to the different kinesin families (Fig. 1.1). In this chapter, we
begin with a description of the kinesin motor domain and the general principles of
chemomechanical coupling. We then discuss general principles of kinesin organiza-
tion and regulation. Finally, we provide a brief description of the kinesins that are
known to function in mitosis, which will be the focus of the later chapters of this
book. Much of our discussion will focus on the human genes (denoted by Hs in
front of the gene name indicated by the HUGO gene nomenclature committee),
with information about kinesin family members from other species, for example
Caenorabhditis elegans (Ce) and Drosophila melanogaster (Dm), provided when
relevant. Our initial description of the kinesin families (Sect. 1.2 and Fig. 1.1) and
their mechanochemical properties is organized based on the kinesin phylogeny,
whereas the subsequent discussion of the functions of the mitotic kinesins (Sect. 1.3
and Fig. 1.3) is organized according to the chapters in this book.
1.2 General Principles of Kinesin Motor Proteins
1.2.1 Mechanics of the Kinesin Motor Domain
Since the discovery of kinesin-1 in 1985 [1, 2], many researchers have determined
and rigorously studied the minimal protein fragment that comprises the kinesin
motor domain of many families. The first high-resolution structure of the kinesin-1
motor domain [6] provided a glimpse into how the kinesin motor was built and
showed many structural similarities with the related myosin motor domain [7].
Since then, over 100 kinesin motors from multiple families have been solved by
X-ray crystallography and deposited into the Protein Data Bank (PDB). Although
detailed reviews of kinesin structure have been published [8–11], here we will
briefly introduce the general kinesin motor domain.
The kinesin motor domain possesses the ability to convert the free energy of ATP
hydrolysis during its mechanochemical cycle into directed mechanical motion
during its interaction with the microtubule. Kinesin-1, the most well-studied kinesin
motor, is a processive translocase that can take hundreds of steps along the micro-
tubule while consuming 1 ATP molecule per step [12–15]. Using the law of mass
action and standard thermodynamic equations, the average available free energy
of ATP hydrolysis under typical in vitro steady-state ATP turnover conditions
(5 mM MgATP; 1–10 % hydrolysis products ADP + Pi; 298.15 K; pH 7.0) equals
–75 kJ mol−1 or –124 pN nm. The stall force for kinesin-1 has been measured at
5–7 pN using optical trapping methodologies [13, 16, 17]; therefore, with an aver-
age step size at 8.3 nm [18], kinesin-1 is 42–59 % efficient (e.g. 5 pN × 8.3 nm =
41.5 pN nm = 33 % efficiency) in its use of ATP hydrolysis free energy. For the
processive mitotic motors kinesin-5 and kinesin-7, a similar range of force production
(6–7 pN) was observed, though their motility velocities were much slower and their
run lengths were much shorter than kinesin-1 [19, 20]. Based on these data, it
appears that kinesin motors share an overall similar efficiency for converting the
energy of ATP hydrolysis, and use this energy to carry out different functions in
cellular processes (as discussed in Sect. 1.3). It is also important to note that not all
kinesins are processive, and in fact some kinesins are not transport motors at all.
This highlights the diversity of ways that the energy of ATP hydrolysis can be
converted into mechanical work.
The kinesin motor domain consists of (1) a common core domain that folds into
a layered αβα sandwich with three α-helices on each side of an eight-stranded
β-sheet (Fig. 1.2a, b), (2) a family-specific neck linker sequence immediately
preceding or following the core domain, and (3) a motif called the cover strand at
the opposite terminus relative to the neck linker (e.g. kinesin-5 has its neck linker at
the C-terminus and the cover strand at the N-terminus of the core domain). Note that
although the neck linker and the cover strand are at opposite ends of the core motor
domain in the primary sequence, they are in close spatial apposition in the three-
dimensional structure due to the proximity of the N- and C-termini entering and
exiting the core motor domain (Fig. 1.2a, b). The neck linker is the critical element
in the production of forward motion [21, 22] and is also important for communicating
strain between the two motor domains of dimeric kinesins in order to gate the
mechanochemical cycle [23–30]. The cover strand interacts with the neck linker to
form the cover-neck bundle, which participates in force generation [31, 32].
The core domain contains several key features present in all kinesin motors.
First, the nucleotide binding pocket (viewed in Fig. 1.2a, c) and the microtubule
binding interface (viewed in Fig. 1.2b, d) are located on opposite sides of the eight-
stranded β-sheet. The nucleotide binding pocket contains a Walker A motif (P-loop
consensus: GxxxxGKT/S) and the microtubule binding interface is comprised of
the β5-L8 lobe, L11, and the “Sw2 cluster” (α4, L12, and α5) (Fig. 1.2c, d) [33, 34].
Second, the conserved structural elements switch-1 [Sw1: NxxSSR] and switch-2
[Sw2: Walker B motif, DxxGxE] (Fig. 1.2c, d) undergo conformational changes
driven by the biochemistry of the ATP hydrolysis and microtubule binding cycles.
These structural changes are reciprocally related to and, thus, modulate the confor-
mations of the neck linker/cover strand, and vice versa [21, 31, 32, 35, 36]. Structural
differences among kinesin families and family members result in variations of the
rate and equilibrium constants that govern the ATPase cycle. Therefore, each motor
elicits a different work output that is utilized to perform different tasks inside cells.
Although the detailed structural mechanism of the nucleotide- and microtubule-
driven conformational changes remains largely unknown, a current model exists
for processive kinesins and is largely based on studies of kinesin-1. Briefly, the
nucleotide state at the active site is thought to trigger a switch in conformation that
is transmitted to and influences the adjacent microtubule-binding regions of the
core. Communication from the active site to the microtubule binding region is
accomplished through two pathways: (1) Sw1 (L9) to α3 to β5-L8 lobe [37, 38]
(Fig. 1.2c), and (2) Sw2 (L11) to helix α4 to the remainder of the “Sw2 cluster”
a 0 L11
b L11
0
L2 3a 3a
L9 ADP
ADP L9
1a/b/c 6 6 1a/b/c
L7
1
4 180° L7
6 L2
1 5
2 8 5
4
6 4 3 2
3 L5 3 8
5 7
Neck L8
Linker 1 7 L12 Neck
5a/b 1
2 Linker
5c
Cover
Strand L8 L6
Cover
Strand
L10
L6
L10
P-loop Sw2 P-loop

L11
L11
c Sw1 d Sw1
Sw2
-lobe
-lobe MT-Sensing
Latch
180°
5-L8
5-L8
lobe
lobe
2a-L5- 2b “Sw2 Cluster”
4-L12- 5
Fig. 1.2 Structure of the kinesin motor domain. All panels display ribbon diagrams of the crystal
structure of the human kinesin-1 motor domain (PDB ID 1BG2; [6]) in two orientations rotated
180° relative to the image plane. Panels (a) and (b) highlight the secondary structure elements of
β-strands (gold), α-helices (blue), and loops (crimson) with standard labeling shown. The ADP
nucleotide and Mg2+ (cyan) are represented as space-fill model. The dotted lines represent the
disordered neck-linker and cover-strand, as indicated. Panels (c) and (d) show the presumed struc-
tural communication pathway between the active site and the microtubule-binding region via
switch-1 (Sw1, orange) and switch-2 (Sw2, blue). The regulatory α2a-L5-α2b region (magenta)
and β-lobe (green) are also shown. The ADP nucleotide (gray) is shown as stick representation
[21, 39, 40] (Fig. 1.2d). The Sw2 cluster also controls the orientation of the neck
linker in alternate conformations that result in directed force production and motility
stabilization [41, 42].
Where in the kinesin ATPase cycle is mechanical motion produced? Even for
kinesin-1, this question has received different answers and remains controversial
[35]. Nevertheless, this is a very important and outstanding question for other
kinesin families, particularly the mitotic kinesins since inhibition of force production
would provide a plausible therapeutic strategy. Mechanical output requires
communication between the two motor domains in a dimeric kinesin-1 molecule
(see Sect. 1.2.2 for kinesin oligomerization), and how this communication occurs is
still unclear. Most theories suggest that communication occurs through a “gating”
mechanism where a mechanistic step in one motor domain is blocked until a certain
step is taken in the other motor domain [35]. This gating mechanism could be
chemical in nature, such that ATP binding to the one head is prohibited until the
other head dissociates from or attaches to the microtubule [23, 28, 43, 44], or
mechanical in nature such that a conformational change in one head pulls the
other head off the microtubule [30, 45, 46]. Of course, it is likely that the actual
mechanism of communication utilizes both types of gating, as they are not mutually
exclusive.
Family-specific secondary structure elements and motifs modulate the structural
communication of various kinesins during the mechanochemical cycle. For exam-
ple, specific interactions of the “microtubule-sensing latch” L7 with Sw1 and Sw2
have been implicated in the mechanism of MgADP product release from the active
site of the kinesin-3 motor KIF1A [47]. One loop that has garnered a lot of attention
over the last decade is L5 due to its (i) location in the motor domain (it interrupts the
long α2 helix immediately downstream of the nucleotide binding pocket; Fig. 1.2a, c),
(ii) variability in length across the superfamily (6–21 residues), and (iii) forming
part of the binding pocket for various kinesin-5 inhibitors [48–52]. L5 was first
proposed to undergo conformational changes upon nucleotide binding during the
kinesin-5 ATPase cycle [53], and has since been found to function as a “conforma-
tional latch” to structurally and kinetically regulate its stepping behavior [54–56].
Not all kinesin motors utilize the energy of ATP hydrolysis for processive motion.
Members of the kinesin-8 and kinesin-13 families were found to be microtubule
depolymerases. Kinesin-8 proteins undergo directed motility to reach the end of
the microtubule where they induce microtubule destabilization by an undefined
mechanism [57–60]. In contrast, kinesin-13 proteins are strictly microtubule
depolymerizing enzymes [61]. For kinesin-13 motors, the functional output of
microtubule depolymerization is due to an extended L2 that contains the signature
sequence “KVD” (also called the “KVD finger”), which has been shown to be
crucial for microtubule depolymerization activity of these motors [38, 62, 63].
This extended L2 is located within the conserved β-lobe (green in Fig. 1.2d) near α6,
however the actual function of this β-lobe has not yet been identified.
1.2.2 Oligomerization
The functional unit of a kinesin motor protein is a dimer, usually a homodimer, as

dimerization results in two motor domains that communicate with each other
through a gating mechanism to coordinate their catalytic cycles. Dimerization occurs
through the formation of coiled coil structures, and all kinesin proteins have regions
of predicted coiled coil, although the length of the coiled coil can vary greatly.
The most extensive predicted coiled coil segments can be found in members of
the kinesin-1, kinesin-2, kinesin-4, kinesin-7 and kinesin-12 families where nearly
the entire stalk domain is predicted to form a coiled coil structure (Fig. 1.1). Yet the
minimum sequence for homodimerization appears to be the short coiled coil
segment called the neck coil (NC) that immediately follows the neck linker
[19, 64–67].
Initial analysis of members of the kinesin-3 family indicated that these motors
lack long coiled coil domains and therefore function in a monomeric state. However,
more recent work has demonstrated that kinesin-3 NC peptides can form homodi-
meric coiled coils [68, 69] and that the NC segment is sufficient for dimerization
and processive motility [70–72]. Similar work has demonstrated that members of
the kinesin-8 and kinesin-10 families also contain only a few short coiled coil
regions that are sufficient for homodimerization [73–75].
Two examples of heterodimerization of kinesin polypeptides exist. The first is
the heterotrimeric member of the kinesin-2 family. In mammals, the KIF3A poly-
peptide forms a heterodimer with either a KIF3B or KIF3C polypeptide. Both KIF3
heterodimers are thought to undergo a processive hand-over-hand motion and to
associate with a non-motor subunit called kinesin associated protein (KAP) [76–81].
The KIF3 heterodimers are also unique in terms of the mode of dimer formation.
Unlike other kinesin families where the short coiled coil following the motor domain
is sufficient for dimerization, recent work has indicated that a C-terminal region of
the coiled coil stalk domain determines the dimerization process of these kinesin-2
motors [82–88]. The evolutionary development of a heterodimeric kinesin is not
clear but may enable one motor domain in the dimer to generate torque as it moves
around a microtubule filament [89, 90].
The second example of heterodimerization can be found in a S. cerevisiae mem-
ber of the kinesin-14 family where the Kar3p motor polypeptide heterodimerizes
with one of two partner proteins, Cik1p or Vik1p. Interestingly, the Cik1p and Vik1p
polypeptides contain a domain that exhibits the kinesin fold but lacks a nucleotide
binding site. While Cik1p targets Kar3p to the microtubule plus ends for depoly-
merase activity, Vik1p provides a second microtubule binding site for Kar3p to
undergo its powerstroke [91–94].
Some dimeric kinesin molecules can stably associate with partner proteins to
form higher oligomeric structures. For example, the dimeric kinesin-1 motor forms a
stable association with a dimeric kinesin light chain subunit to form a heterotetrameric
molecule [95, 96]. For kinesin-5 motors, two dimeric Eg5/HsKIF11 molecules
associate to form a homotetramer [97–100] and this unique bipolar architecture
allows kinesin-5 motors to crosslink and slide anti-parallel microtubules [101, 102].
For kinesin-6 motors, the dimeric HsKIF23/MKLP1 complex associates with a
dimer of HsCYK-4 (also known as RACGAP1/HsMgcRacGAP) to form a stable
heterotetramer called centralspindlin [103–105].
1.2.3 Regulation
Kinesin activity must be tightly controlled in cells to prevent futile ATP hydrolysis
and undesirable motility and/or force-generating activities. The regulatory mecha-
nisms that influence kinesins can be roughly divided into two kinds – first, regulatory
mechanisms that are intrinsic to the motor protein itself and second, mechanisms
that exert extrinsic control on motor activity, localization and/or function.
For the majority of kinesin motors, intrinsic regulation consists of an autoinhibi-
tion mechanism that enables the motor to be activated in a temporal and spatial
manner. Autoinhibition generally involves intramolecular interactions that inhibit
the microtubule binding and/or ATPase activity of the motor domain. The best-
understood autoinhibition mechanism is that of the transport motor kinesin-1
where the heterotetrameric molecule adopts a folded conformation that allows the
C-terminal tail domain to interact with and inhibit the ability of the N-terminal
motor domain to bind to microtubules [106–116]. A similar autoinhibitory mechanism
has been described for members of the kinesin-2 and kinesin-3 families [70, 72,
117–120]. For the mitotic motors, autoinhibition has been implicated in regulating
kinesin-6, kinesin-7, and kinesin-8 motors [74, 103, 105, 121, 122].
A wide variety of extrinsic mechanisms can regulate kinesin activity. In broad
terms, motor localization can be regulated in terms of nuclear versus cytoplasmic
localization. For example, members of the kinesin-4, kinesin-6, kinesin-8, kinesin-
10, and kinesin-14 families are sequestered in the nucleus during interphase whereas
members of the kinesin-7 and kinesin-12 families are localized to the cytoplasm
until nuclear envelope breakdown [123–130].
As phosphorylation is an important regulatory mechanism for the initiation,
progression and completion of mitosis, it is no surprise that phosphorylation has
emerged as a major regulatory mechanism of the mitotic kinesins. Phosphorylation
by cyclin B–Cdk1 (cyclin-dependent kinase 1) has been shown to regulate the
kinesin-5 motor Eg5/DmKlp61F, the kinesin-6 motor DmPavarotti, the kinesin-7
motor HsCENP-E, and the kinesin-10 motor HsKIF22/Kid [121, 128, 131–135].
The Aurora B kinase has been shown to phosphorylate and regulate the kinesin-5
motor Eg5/HsKIF11, the kinesin-6 centralspindlin complex, the kinesin-7 motor
CENP-E, and the kinesin-13 motors KIF2A and KIF2C/MCAK [121, 136–143].
A second major regulatory mechanism that drives mitotic progression involves
the controlled synthesis and degradation of specific proteins. Like the cyclins, the
protein levels of a variety of mitotic kinesins have been shown to oscillate over the
cell cycle, including members of the kinesin-6, kinesin-7, kinesin-8 and kinesin-13
and kinesin-14 families. The anaphase promoting complex/cyclosome (APC/C)
plays a key role in driving the degradation of specific proteins as cells progress from
metaphase to anaphase, and the APC/C has been shown to drive the selective degra-
dation of members of the kinesin-7, kinesin-8, kinesin-10 and kinesin-13 families
[123, 127, 144–153].
The activity, localization and/or function of kinesin motors can also be regulated
by binding partners. For example, the kinesin-12 motor Xklp2 requires TPX2 for
microtubule localization [154]. Members of the kinesin-10, kinesin-13 and kinesin-14

families bind to importin proteins, which keep the motors inactive and sequestered
in the nucleus until dissociated by the small G-protein Ran [130, 155–158]. And in
yeast, the targeting of Kar3p to specific spindle locales is driven by its heterodimer-
ization partners Cik1p and Vik1p [91, 94, 159, 160].
Recent work has revealed that kinesin motors can also be regulated by alterations
to the microtubule track. This has most extensively been investigated for the
transport kinesins. For example, the activity and directionality of kinesin-1 can be
regulated by post-translational modifications (PTMs) of tubulin subunits, such as
acetylation, detyrosination, and polyglutamylation (reviewed in [161]). Transport
kinesins can also be positively and negatively regulated by microtubule associated
proteins (MAPs) that bind to the surface of microtubules such as tau, Dcx, and
MAP115/ensconsin (reviewed in [161]). Whether similar mechanisms regulate the
localization and/or activity of mitotic kinesins is less clear, although the detyrosina-
tion of tubulin has been associated with reduced kinesin-13 depolymerization
activity [162]. Recent work has shown that the kinesin-12 motor HsKIF15 partitions
to kinetochore microtubules (K-MTs) and influences kinetochore-fiber (K-fiber)
length although the mechanism by which HsKIF15 recognizes K-MTs is not known
[163, 164].
1.3 Kinesin Families with Functional Roles

in Mitotic Progression
1.3.1 Kinesin-5 Family: HsKIF11 (Also Known as Eg5)
The kinesin-5 family was one of the earliest kinesin-related protein families to be
described through its discovery in genetic screens in A. nidulans and S. cerevisiae,
where the genes are required for nuclear division and chromosome segregation
[165–167]. Structurally, kinesin-5 members are bipolar homotetramers [98] in
which the motor can cross-link and slide anti-parallel microtubules [101]. Kinesin-5s
were originally shown to be plus-end directed motors [133], but an interesting
recent report suggests that kinesin-5 motors can walk to microtubule minus ends on
single microtubules, but walk toward the plus end on cross-linked anti-parallel
microtubules [168–170].
Kinesin-5 proteins are nearly universally required for spindle bipolarity, and
inhibition gives a strong mitotic arrest (reviewed in [171]). Their activity is needed
during spindle assembly to push spindle poles apart via the sliding of anti-parallel
microtubules (Fig. 1.3), but they are not required to maintain spindle bipolarity
[172]. Interestingly in C. elegans and in Dictyostelium [173, 174], the kinesin-5
protein is not essential, and appears to limit pole separation rather than being essen-
tial for this process. Thus we do not yet fully understand the role of this kinesin in
all organisms.
Kinesin Activity Localization Function Inhibitor

(Family)
HsKif11/Eg5 + end motor Spindle pole Numerous

(Kinesin-5) MT sliding separation Phase I, Phase II
HsKif15/Klp2
+ end motor Spindle bipolarity None identified
(Kinesin-12)
HsKif10/Cenp-E Congression GSK923295

+ end motor
(Kinesin-7) Spindle checkpoint Phase I
HsKifC1/HSET - end motor Centrosome

AZ82 and CW069
(Kinesin-14) MT sliding clustering
In vitro studies only
Spindle Assembly
HsKif2A, HsKif2B, Three inhibitors
Congression
HsKif2C/MCAK MT depolymerase In vitro studies
MT Flux
(Kinesin-13) only
Error Correction
Congression
HsKif4 + end motor
Condensation
HsKif22/Kid +end dynamics None identified
Spindle Bipolarity
(Kinesin-4 and 10) regulator
Cytokinesis
HsKif18A,
+ end motor BTB1 and
HsKif18B, Chromosome alignment derivatives
+end dynamics
HsKif19 Spindle positioning In vitro studies
regulator
(Kinesin-8)
only
HsKif20A/MKLP2
Paprotrain
HsKif20B/MPP1 + end motor
Cytokinesis In vitro studies
HsKif23/MKLP1 MT sliding
only
(Kinesin-6)
HsKif14
ATPase Cytokinesis None identified
(Kinesin-3)
Fig. 1.3 Summary of mitotic kinesins. The table includes family members, biochemical activity,
localization (red), function, and whether any chemical inhibitors have been identified. When mul-
tiple family members exist, the localization includes the different localization patterns of the indi-
vidual proteins. The function includes the major mitotic phenotypes that have been reported
Kinesin-5s were the first kinesins to which small molecule inhibitors were
identified [175]. These drugs cause mitotic arrest in which the cells have a monopolar
spindle. A common theme among these inhibitors is that they allosterically affect
kinesin-5 activity [53], accounting for their high specificity in vivo [176]. Multiple
of these drugs have entered clinical trials, but the long-term efficacy of these
compounds remains to be elucidated [177].
1.3.2 The Kinesin-12 Family: HsKIF15

(Also Known as HKLP2)
Similar to the kinesin-5 KIF11/Eg5, the kinesin-12 family member HsKIF15 is also
important in spindle bipolarity (Fig. 1.3). Human KIF15 was originally identified in
a screen for proteins that promote bipolarity in the absence of KIF11/Eg5 [178].
HsKIF15 is the ortholog of Xenopus XKlp2 and C. elegans Klp18, two kinesins
previously shown to be important in spindle bipolarity [164, 179]. Unlike KIF11/
Eg5, KIF15 is required for maintenance of spindle bipolarity [180]. There is signifi-
cant functional redundancy between KIF11/Eg5 and KIF15, and in fact KIF15 can
fully rescue spindle assembly in cells lacking Eg5 [178], although mechanistically
they act on distinct sets of microtubules [163]. One interesting observation is that in
cultured cell models that are resistant to KIF11/Eg5 inhibitors, KIF15 expression is
increased [163]. It is unclear whether this is physiologically relevant, but this could
provide an explanation for why KIF11/Eg5 inhibitors have had limited success in
clinical trials.
1.3.3 The Kinesin-7 Family: HsCENP-E

(Also Known as HsKIF10)
The kinesin-7 family has been most thoroughly characterized in mammalian cells.
CENP-E was the first identified kinetochore-associated kinesin [181, 182] whose
inhibition causes a failure in chromosome congression [183–185]. CENP-E is a
very large molecule (>300 kDa), with an N-terminal plus-end directed motor
domain [184] and a C-terminal tail that contains a second microtubule binding site
and a kinetochore targeting domain [183, 186] (Fig. 1.1).
In addition to its role in chromosome congression, CENP-E has been shown to
be involved in chromosome attachment (Fig. 1.3). CENP-E may also play an impor-
tant role in silencing the spindle assembly checkpoint [185, 187, 188], a role that is
likely important in considering the potential utility of CENP-E as a therapeutic
target. Interestingly, CENP-E heterozygous mice have different responses in
tumorigenesis in different tissue types [189] wherein disruption of CENP-E acts
oncogenically in some tissues and as a tumor suppressor in other tissues. This brings
up the complexities associated with mitotic protein inhibition as potential therapeutics.

The first CENP-E inhibitor has potent anti-tumor activity in mouse Xenograft models
[190] and has entered clinical trials, but the long-term efficacy of this compound is
not yet known.
1.3.4 The Kinesin-14 Family: HsKIFC1

(Also Known as HSET), HsKIFC2, HsKIFC3
The kinesin-14 proteins were first described in S. cerevisiae [191–193] and in

D. melanogaster [194] where they were proposed to antagonize kinesin-5 proteins
(yeast) or to be involved in preventing chromosome non-disjunction in meiosis
(flies). Structurally, the kinesin-14 proteins are unique in that their kinesin-like
catalytic domain is located at the C-terminus of the protein (Fig. 1.1), which
corresponds to their motility towards the minus ends of microtubules [194].
The kinesin-14 proteins can cross-link and slide microtubules due to a second
microtubule binding domain in their N-terminal tails [195–198]. Interestingly, this
binding is inhibited when kinesin-14 proteins are bound by the nuclear transport
receptors importin α/β [157], demonstrating how these motors can be regulated
during spindle assembly.
Initial studies in mammalian cells did not uncover a striking role of the mam-
malian kinesin-14 KIFC1/HSET, which had been implicated in pole formation or
spindle length regulation [130, 199–201]. However, several more recent studies
have uncovered an important role for KIFC1/HSET in clustering centrosomes in
cancer cells containing multiple centrosomes [202, 203] (Fig. 1.3). This may provide
a powerful strategy to identify new drugs that target KIFC1/HSET and could selec-
tively kill tumors with centrosome amplification while leaving normal cells largely
unperturbed. The first kinesin-14 inhibitors were recently reported [204–206], but
there is no evidence yet to support its utility in the clinic. Initial studies on KIFC1/
HSET expression in non-small cell lung carcinomas shows that it correlates with a
high risk for brain metastasis [207], and it will be interesting to see whether this
high expression level correlates with cancers that have centrosome amplification.
These results suggest that targeting KIFC1/HSET should be a high priority for
developing new kinesin therapeutics.
1.3.5 The Kinesin-13 Family: HsKIF2A, HsKIF2B, HsKIF2C

(Also Known as MCAK), HsKIF24
The kinesin-13 family members have been most extensively studied in vertebrate
systems and include KIF2A, KIF2B, KIF2C/MCAK and KIF24 (reviewed in [208]).
Structurally, they are composed of an N-terminal sub-cellular targeting domain,
a centrally located kinesin-like catalytic domain, and a C-terminal tail involved in

dimerization and regulation of activity (Fig. 1.1). Unlike conventional kinesins, the
kinesin-13 proteins are not transport motors; instead they couple the energy of ATP
hydrolysis to catalytically depolymerize microtubules [209, 210].
Kinesin-13 proteins play a variety of roles in cells including regulating micro-
tubule dynamics during spindle assembly [211, 212], error correction [213],
tip-tracking [214], pole formation [215], controlling microtubule flux [216, 217], and
chromosome movement [217–219] (Fig. 1.3). Their inhibition also reduces cell
proliferation, suggesting that they may be viable therapeutic targets [220]. Two
small molecule inhibitors have been identified, and both reduce cell proliferation in
cultured cells [221, 222]. However, neither inhibitor is specific for kinesin-13s, and
they both likely inhibit other ATPases. An allosteric kinesin-13 inhibitor was
recently identified, which is likely more specific than the previously identified drugs
[223]. There are several studies suggesting that kinesin-13 activity, specifically
KIF2C/MCAK, has misregulated expression in a number of cancers, and this mis-
expression correlates with poor prognosis [220, 224, 225]. Therefore there is a lot
of interest in the potential utility of kinesin-13 inhibitors in the clinic.
1.3.6 Chromokinesins (The Kinesin-4

and Kinesin-10 Families)
Chromokinesins are defined based on the observation that they associate with
chromosomes during mitosis (reviewed in [226]). These proteins fall mainly in the
category of the kinesin-10 proteins, including HsKIF22 (also known as KID) and
DmNOD, and the kinesin-4 proteins, including HsKIF4A. Structurally, both fami-
lies have an N-terminal motor domain, a central coiled coil stalk, and a C-terminal
tail that contains the DNA binding domain (Fig. 1.1). Members of both families
have been shown to be plus-end directed motors [227, 228], and can also regulate
microtubule dynamics [37, 229, 230].
Chromokinesins play a variety of functions in cells, including chromosome posi-
tioning [123, 127, 147, 231], chromosome condensation [232], spindle bipolarity
[75], and cytokinesis [233] (Fig. 1.3). KIF4 knockout mice have a dramatic pheno-
type in terms of understanding tumorigenesis, as homozygous null mice form
tumors [234]. Interestingly, KIF4 knockout also has dramatic effects on the tran-
scriptional profile in interphase cells [234], suggesting that KIF4 may play a more
general role in controlling gene expression. It will therefore be interesting to deci-
pher whether the induced tumorigenesis in the knockout mice is due to the mitotic
function of KIF4 or to its more global role in gene expression. In addition, it will be
important to understand whether its kinesin-like motor domain plays a role in the
effects on gene expression, which would be a novel function for a kinesin family
member. There are not yet any inhibitors to target the chromokinesins.
1.3.7 The Kinesin-8 Family: HsKIF18A, HsKIF18B, HsKIF19
The kinesin-8 proteins have been studied most extensively in yeasts (Kip3 and
Klp5/6) and in vertebrate cells (KIF18A, KIF18B and KIF19). Early studies
revealed a role for S. cerevisiae Kip3 in spindle positioning and for S. pombe
Klp5/6 in chromosome alignment [235, 236]. HsKIF18A has been shown to be
important for chromosome alignment [145, 237], and HsKIF18B is important for
regulation of astral microtubule dynamics in mammalian cells [143, 238] (Fig. 1.3).
Disruption of kinesin-8 function rendered cells hypersensitive to microtubule
disruption drugs [236], pointing to a potential role in the regulation of microtubule
dynamics. Structurally, these proteins contain an N-terminal motor domain, which
is necessary for microtubule plus end motility [58, 59] (Fig. 1.1). Kinesin-8s also
control microtubule dynamics by either a direct destabilization mechanism [58, 59,
145] or by microtubule capping [57]. These effects are mediated by the ability of the
kinesin-8 protein to stay associated with microtubules by a second microtubule-
binding domain in their tails [122, 239, 240].
Few studies have been carried out that look at kinesin-8 proteins in tumorigenesis.
One study showed a correlation with increased KIF18A expression and colorectal
cancer [241], and another suggested that it may play a role in breast carcinogenesis
[242]. The first kinesin-8 inhibitor, BTB-1, has been identified [243], but its detailed
effects on mitotic progression and kinesin-8 function in cells have not yet been
elucidated. Such studies may be more useful with the recently generated BTB-1
derivatives, which show greater potency and higher specificity [244].
1.3.8 The Kinesin-6 Family: HsKIF23 (Also Known

as MKLP1), HsKIF20A (Also Known as MKLP2),
HsKIF20B (Also Known as MPP1)
The kinesin-6 family members have been mainly studied in vertebrates, worms, and
flies, and are uniformly involved in cytokinesis. HsKIF23/MKLP1 is part of the
centralspindlin complex and helps bundle microtubules for cytokinesis [245]
(Fig. 1.3). HsKIF20A/MKLP2 is also involved in cytokinesis, but it appears to act
mainly during furrow ingression [246]. Kinesin-6 proteins have an N-terminal kine-
sin like motor domain (Fig. 1.1), and the C elegans kinesin-6 has been shown to
exist in a complex with a Rho GTPase activating protein [247]. HsKIF20A/MKLP1
has been shown to be a plus end-directed motor [248] that can slide apart anti-
parallel microtubules, consistent with a role in late mitosis.
Altered levels of expression of HsKIF20A/MKLP2 have been associated with a
number of cancers [249–251]. Initial inhibitors to kinesin-6 family members include
one compound that is specific to HsKIF20A/MKLP2 [252], but how this affects cell
proliferation and whether it will be suitable for therapeutic development is unknown.
1.3.9 The Kinesin-3 Family: HsKIF14
The kinesin-3 family is largely known for its roles in intracellular trafficking but
contains one member, KIF14, involved in mitosis. The first hints for a mitotic func-
tion for a kinesin-3 motor came when the D. melanogaster ortholog DmKIF38B
was shown to be important for cytokinesis and for accurate chromosome segrega-
tion [253–256]. The mammalian ortholog, HsKIF14, associates with citron kinase
and is required for proper cytokinesis [246, 257] (Fig. 1.3). Structurally, KIF14 has
a kinesin motor domain towards the N-terminus of the protein, but this is preceded
by an N-terminal extension (Fig. 1.1) that binds to the central spindle protein, PRC1
[177, 246]. The C-terminus of the protein contains the neck, coiled coil domains,
and the domain that interacts with citron kinase [246]. Although HsKIF14 has
microtubule-stimulated ATPase activity [257], there are no published biochemical
data to suggest whether KIF14 is a motor given its unique domain structure. KIF14
expression is highly correlated with a number of cancers [177], with a strong link
to retinoblastoma [258–262]. Clearly this is a kinesin family member generating
high interest that is in need of biochemical analysis and potential therapeutic
development.
1.4 Summary and Future Perspectives
Members of the kinesin superfamily play critical roles in nearly every aspect of
mitosis, making them powerful potential candidates for therapeutic development.
Their misregulation in numerous cancers supports the idea that they may be impor-
tant players in disease progression. While this field is still emerging, the remaining
chapters of this book relay the current status of this burgeoning and important area
of research.
References
1. Brady ST (1985) A novel brain ATPase with properties expected for the fast axonal transport
motor. Nature 317:73–75
2. Vale RD, Reese TS, Sheetz MP (1985) Identification of a novel force-generating protein,
kinesin, involved in microtubule-based motility. Cell 42:39–50
3. Lawrence CJ et al (2004) A standardized kinesin nomenclature. J Cell Biol 167:19–22
4. Miki H, Okada Y, Hirokawa N (2005) Analysis of the kinesin superfamily: insights into struc-
ture and function. Trends Cell Biol 15:467–476
5. Wickstead B, Gull K (2006) A “holistic” kinesin phylogeny reveals new kinesin families and
predicts protein functions. Mol Biol Cell 17:1734–1743
6. Kull FJ, Sablin EP, Lau R, Fletterick RJ, Vale RD (1996) Crystal structure of the kinesin
motor domain reveals a structural similarity to myosin. Nature 380:550–555
7. Rayment I et al (1993) Three-dimensional structure of myosin subfragment-1: a molecular

motor. Science 261:50–58
8. Kull FJ (2000) Motor proteins of the kinesin superfamily: structure and mechanism. Essays
Biochem 35:61–73
9. Kull FJ, Endow SA (2002) Kinesin: switch I & II and the motor mechanism. J Cell Sci
115:15–23
10. Marx A, Hoenger A, Mandelkow E (2009) Structures of kinesin motor proteins. Cell Motil
Cytoskeleton 66:958–966
11. Woehlke G (2001) A look into kinesin’s powerhouse. FEBS Lett 508:291–294
12. Hua W, Young EC, Fleming ML, Gelles J (1997) Coupling of kinesin steps to ATP hydrolysis.
Nature 388:390–393
13. Schnitzer MJ, Block SM (1997) Kinesin hydrolyses one ATP per 8-nm step. Nature
388:386–390
14. Schnitzer MJ, Visscher K, Block SM (2000) Force production by single kinesin motors. Nat
Cell Biol 2:718–723
15. Vale RD et al (1996) Direct observation of single kinesin molecules moving along microtu-
bules. Nature 380:451–453
16. Svoboda K, Block SM (1994) Force and velocity measured for single kinesin molecules. Cell
77:773–784
17. Visscher K, Schnitzer MJ, Block SM (1999) Single kinesin molecules studied with a molecu-
lar force clamp. Nature 400:184–189
18. Yildiz A, Tomishige M, Vale RD, Selvin PR (2004) Kinesin walks hand-over-hand. Science
303:676–678
19. Valentine MT, Fordyce PM, Krzysiak TC, Gilbert SP, Block SM (2006) Individual dimers of
the mitotic kinesin motor Eg5 step processively and support substantial loads in vitro. Nat
20. Yardimci H, van Duffelen M, Mao Y, Rosenfeld SS, Selvin PR (2008) The mitotic kinesin
CENP-E is a processive transport motor. Proc Natl Acad Sci U S A 105:6016–6021
21. Rice S et al (1999) A structural change in the kinesin motor protein that drives motility.
Nature 402:778–784
22. Schief WR, Howard J (2001) Conformational changes during kinesin motility. Curr Opin
23. Guydosh NR, Block SM (2006) Backsteps induced by nucleotide analogs suggest the front
head of kinesin is gated by strain. Proc Natl Acad Sci U S A 103:8054–8059
24. Hwang W, Lang MJ (2009) Mechanical design of translocating motor proteins. Cell Biochem
Biophys 54:11–22
25. Hyeon C, Onuchic JN (2007) Internal strain regulates the nucleotide binding site of the kine-
sin leading head. Proc Natl Acad Sci U S A 104:2175–2180
26. Jaud J, Bathe F, Schliwa M, Rief M, Woehlke G (2006) Flexibility of the neck domain
enhances kinesin-1 motility under load. Biophys J 91:1407–1412
27. Miyazono Y, Hayashi M, Karagiannis P, Harada Y, Tadakuma H (2010) Strain through the
neck linker ensures processive runs: a DNA-kinesin hybrid nanomachine study. EMBO J
29:93–106
28. Rosenfeld SS, Fordyce PM, Jefferson GM, King PH, Block SM (2003) Stepping and stretch-
ing. How kinesin uses internal strain to walk processively. J Biol Chem 278:18550–18556
29. Shastry S, Hancock WO (2010) Neck linker length determines the degree of processivity in
kinesin-1 and kinesin-2 motors. Curr Biol 20:939–943
30. Yildiz A, Tomishige M, Gennerich A, Vale RD (2008) Intramolecular strain coordinates kine-
sin stepping behavior along microtubules. Cell 134:1030–1041
31. Khalil AS et al (2008) Kinesin’s cover-neck bundle folds forward to generate force. Proc Natl
Acad Sci U S A 105:19247–19252
32. Hwang W, Lang MJ, Karplus M (2008) Force generation in kinesin hinges on cover-neck
bundle formation. Structure 16:62–71
33. Sosa H et al (1997) A model for the microtubule-Ncd motor protein complex obtained by
cryo-electron microscopy and image analysis. Cell 90:217–224
34. Woehlke G et al (1997) Microtubule interaction site of the kinesin motor. Cell 90:207–216
35. Block SM (2007) Kinesin motor mechanics: binding, stepping, tracking, gating, and limping.
Biophys J 92:2986–2995
36. Zhao YC, Kull FJ, Cochran JC (2010) Modulation of the kinesin ATPase cycle by neck linker
docking and microtubule binding. J Biol Chem 285:25213–25220
37. Cochran JC et al (2009) ATPase cycle of the nonmotile kinesin NOD allows microtubule end
tracking and drives chromosome movement. Cell 136:110–122
38. Ogawa T, Nitta R, Okada Y, Hirokawa N (2004) A common mechanism for microtubule
destabilizers-M type kinesins stabilize curling of the protofilament using the class-specific
neck and loops. Cell 116:591–602
39. Sindelar CV, Downing KH (2007) The beginning of kinesin’s force-generating cycle visual-
ized at 9-A resolution. J Cell Biol 177:377–385
40. Vale RD, Milligan RA (2000) The way things move: looking under the hood of molecular
motor proteins. Science 288:88–95
41. Gigant B et al (2013) Structure of a kinesin-tubulin complex and implications for kinesin
motility. Nat Struct Mol Biol 20:1001–1007
42. Kikkawa M et al (2001) Switch-based mechanism of kinesin motors. Nature 411:439–445
43. Block SM, Asbury CL, Shaevitz JW, Lang MJ (2003) Probing the kinesin reaction cycle with
a 2D optical force clamp. Proc Natl Acad Sci U S A 100:2351–2356
44. Klumpp LM, Hoenger A, Gilbert SP (2004) Kinesin’s second step. Proc Natl Acad Sci
U S A 101:3444–3449
45. Hancock WO, Howard J (1999) Kinesin’s processivity results from mechanical and chemical
coordination between the ATP hydrolysis cycles of the two motor domains. Proc Natl Acad
Sci U S A 96:13147–13152
46. Spudich JA (2006) Molecular motors take tension in stride. Cell 126:242–244
47. Nitta R, Okada Y, Hirokawa N (2008) Structural model for strain-dependent microtubule
activation of Mg-ADP release from kinesin. Nat Struct Mol Biol 15:1067–1075
48. Kaan HY et al (2010) Structural basis for inhibition of Eg5 by dihydropyrimidines: stereose-
lectivity of antimitotic inhibitors enastron, dimethylenastron and fluorastrol. J Med Chem
53:5676–5683
49. Maliga Z et al (2006) A pathway of structural changes produced by monastrol binding to Eg5.
J Biol Chem 281:7977–7982
50. Talapatra SK, Schuttelkopf AW, Kozielski F (2012) The structure of the ternary Eg5-ADP-
ispinesib complex. Acta Crystallogr D Biol Crystallogr 68:1311–1319
51. Yan Y et al (2004) Inhibition of a mitotic motor protein: where, how, and conformational
consequences. J Mol Biol 335:547–554
52. Brier S, Lemaire D, Debonis S, Forest E, Kozielski F (2004) Identification of the protein
binding region of S-trityl-L-cysteine, a new potent inhibitor of the mitotic kinesin Eg5.
Biochemistry 43:13072–13082
53. Cochran JC, Gatial JE 3rd, Kapoor TM, Gilbert SP (2005) Monastrol inhibition of the mitotic
kinesin Eg5. J Biol Chem 280:12658–12667
54. Behnke-Parks WM et al (2011) Loop L5 acts as a conformational latch in the mitotic kinesin
Eg5. J Biol Chem 286:5242–5253
55. Waitzman JS et al (2011) The loop 5 element structurally and kinetically coordinates dimers
of the human kinesin-5, Eg5. Biophys J 101:2760–2769
56. Parke CL, Wojcik EJ, Kim S, Worthylake DK (2010) ATP hydrolysis in Eg5 kinesin involves
a catalytic two-water mechanism. J Biol Chem 285:5859–5867
57. Du Y, English CA, Ohi R (2010) The kinesin-8 Kif18A dampens microtubule plus-end
dynamics. Curr Biol 20:374–380
58. Gupta ML Jr, Carvalho P, Roof DM, Pellman D (2006) Plus end-specific depolymerase activ-
ity of Kip3, a kinesin-8 protein, explains its role in positioning the yeast mitotic spindle. Nat
59. Varga V et al (2006) Yeast kinesin-8 depolymerizes microtubules in a length-dependent

manner. Nat Cell Biol 8:957–962
60. Peters C et al (2010) Insight into the molecular mechanism of the multitasking kinesin-8
motor. EMBO J 29:3437–3447
61. Walczak CE, Gayek S, Ohi R (2013) Microtubule-depolymerizing kinesins. Annu Rev Cell
Dev Biol 29:417–441
62. Shipley K et al (2004) Structure of a kinesin microtubule depolymerization machine. EMBO
J 23:1422–1432
63. Tan D, Rice WJ, Sosa H (2008) Structure of the kinesin13-microtubule ring complex.
Structure 16:1732–1739
64. Huang TG, Suhan J, Hackney DD (1994) Drosophila kinesin motor domain extending to
amino acid position 392 is dimeric when expressed in Escherichia coli. J Biol Chem
269:16502–16507
65. Kozielski F, De Bonis S, Burmeister WP, Cohen-Addad C, Wade RH (1999) The crystal
structure of the minus-end-directed microtubule motor protein ncd reveals variable dimer
conformations. Structure 7:1407–1416
66. Kozielski F et al (1997) The crystal structure of dimeric kinesin and implications for
microtubule-dependent motility. Cell 91:985–994
67. Hizlan D et al (2006) Structural analysis of the ZEN-4/CeMKLP1 motor domain and its
interaction with microtubules. J Struct Biol 153:73–84
68. Rashid DJ, Bononi J, Tripet BP, Hodges RS, Pierce DW (2005) Monomeric and dimeric
states exhibited by the kinesin-related motor protein KIF1A. J Pept Res 65:538–549
69. Shimizu Y, Morii H, Arisaka F, Tanokura M (2005) Stalk region of kinesin-related protein
Unc104 has moderate ability to form coiled-coil dimer. Biochem Biophys Res Commun
337:868–874
70. Hammond JW et al (2009) Mammalian kinesin-3 motors are dimeric in vivo and move by
processive motility upon release of autoinhibition. PLoS Biol 7:e72
71. Huckaba TM, Gennerich A, Wilhelm JE, Chishti AH, Vale RD (2011) Kinesin-73 is a proces-
sive motor that localizes to Rab5-containing organelles. J Biol Chem 286:7457–7467
72. Huo L et al (2012) The CC1-FHA tandem as a central hub for controlling the dimerization
and activation of kinesin-3 KIF1A. Structure 20:1550–1561
73. Mayr MI, Storch M, Howard J, Mayer TU (2011) A non-motor microtubule binding site is
essential for the high processivity and mitotic function of kinesin-8 Kif18A. PLoS One
6:e27471
74. Erent M, Drummond DR, Cross RA (2012) S. pombe kinesins-8 promote both nucleation and
catastrophe of microtubules. PLoS One 7:e30738
75. Tokai-Nishizumi N, Ohsugi M, Suzuki E, Yamamoto T (2005) The chromokinesin Kid is
required for maintenance of proper metaphase spindle size. Mol Biol Cell 16:5455–5463
76. Kondo S et al (1994) KIF3A is a new microtubule-based anterograde motor in the nerve axon.
J Cell Biol 125:1095–1107
77. Yamazaki H, Nakata T, Okada Y, Hirokawa N (1995) KIF3A/B: a heterodimeric kinesin
superfamily protein that works as a microtubule plus end-directed motor for membrane
organelle transport. J Cell Biol 130:1387–1399
78. Yamazaki H, Nakata T, Okada Y, Hirokawa N (1996) Cloning and characterization of KAP3:
a novel kinesin superfamily-associated protein of KIF3A/3B. Proc Natl Acad Sci U S A
93:8443–8448
79. Muresan V et al (1998) KIF3C and KIF3A form a novel neuronal heteromeric kinesin that
associates with membrane vesicles. Mol Biol Cell 9:637–652
80. Muthukrishnan G, Zhang Y, Shastry S, Hancock WO (2009) The processivity of kinesin-2
motors suggests diminished front-head gating. Curr Biol 19:442–447
81. Yang Z, Goldstein LS (1998) Characterization of the KIF3C neural kinesin-like motor from
mouse. Mol Biol Cell 9:249–261
82. Chana M, Tripet BP, Mant CT, Hodges RS (2002) The role of unstructured highly charged
regions on the stability and specificity of dimerization of two-stranded alpha-helical coiled-
coils: analysis of the neck-hinge region of the kinesin-like motor protein Kif3A. J Struct Biol
137:206–219
83. Chana MS, Tripet B, Mant CT, Hodges RS (2008) Stability and specificity of heterodimer
formation for the coiled-coil neck regions of the motor proteins Kif3A and Kif3B: the role of
unstructured oppositely charged regions. J Pept Res 65:209–220
84. De Marco V, Burkhard P, Le Bot N, Vernos I, Hoenger A (2001) Analysis of heterodimer
formation by Xklp3A/B, a newly cloned kinesin-II from Xenopus laevis. EMBO J
20:3370–3379
85. De Marco V, De Marco A, Goldie KN, Correia JJ, Hoenger A (2003) Dimerization properties
of a Xenopus laevis kinesin-II carboxy-terminal stalk fragment. EMBO Rep 4:717–722
86. Vukajlovic M, Dietz H, Schliwa M, Okten Z (2011) How kinesin-2 forms a stalk. Mol Biol
Cell 22:4279–4287
87. Doodhi H, Jana SC, Devan P, Mazumdar S, Ray K (2012) Biochemical and molecular
dynamic simulation analysis of a weak coiled coil association between kinesin-II stalks.
PLoS One 7:e45981
88. Rashid DJ, Wedaman KP, Scholey JM (1995) Heterodimerization of the two motor subunits
of the heterotrimeric kinesin, KRP85/95. J Mol Biol 252:157–162
89. Brunnbauer M et al (2012) Torque generation of kinesin motors is governed by the stability
of the neck domain. Mol Cell 46:147–158
90. Pan X, Acar S, Scholey JM (2010) Torque generation by one of the motor subunits of hetero-
trimeric kinesin-2. Biochem Biophys Res Commun 401:53–57
91. Allingham JS, Sproul LR, Rayment I, Gilbert SP (2007) Vik1 modulates microtubule-Kar3
interactions through a motor domain that lacks an active site. Cell 128:1161–1172
92. Chen CJ, Rayment I, Gilbert SP (2011) Kinesin Kar3Cik1 ATPase pathway for microtubule
cross-linking. J Biol Chem 286:29261–29272
93. Rank KC et al (2012) Kar3Vik1, a member of the kinesin-14 superfamily, shows a novel
kinesin microtubule binding pattern. J Cell Biol 197:957–970
94. Sproul LR, Anderson DJ, Mackey AT, Saunders WS, Gilbert SP (2005) Cik1 targets the
minus-end kinesin depolymerase kar3 to microtubule plus ends. Curr Biol 15:1420–1427
95. DeBoer SR et al (2008) Conventional kinesin holoenzymes are composed of heavy and light
chain homodimers. Biochemistry 47:4535–4543
96. Hackney DD, Levitt JD, Wagner DD (1991) Characterization of alpha 2 beta 2 and alpha 2
forms of kinesin. Biochem Biophys Res Commun 174:810–815
97. Cole DG, Saxton WM, Sheehan KB, Scholey JM (1994) A “slow” homotetrameric kinesin-
related motor protein purified from Drosophila embryos. J Biol Chem 269:22913–22916
98. Kashina AS et al (1996) A bipolar kinesin. Nature 379:270–272
99. Tao L et al (2006) A homotetrameric kinesin-5, KLP61F, bundles microtubules and antago-
nizes Ncd in motility assays. Curr Biol 16:2293–2302
100. Acar S et al (2013) The bipolar assembly domain of the mitotic motor kinesin-5. Nat Commun
4:1343
101. Kapitein LC et al (2005) The bipolar mitotic kinesin Eg5 moves on both microtubules that it
crosslinks. Nature 435:114–118
102. van den Wildenberg SM et al (2008) The homotetrameric kinesin-5 KLP61F preferentially
crosslinks microtubules into antiparallel orientations. Curr Biol 18:1860–1864
103. Mishima M, Kaitna S, Glotzer M (2002) Central spindle assembly and cytokinesis require a
kinesin-like protein/RhoGAP complex with microtubule bundling activity. Dev Cell
2:41–54
104. Somers WG, Saint R (2003) A RhoGEF and Rho family GTPase-activating protein complex
links the contractile ring to cortical microtubules at the onset of cytokinesis. Dev Cell
4:29–39
105. Pavicic-Kaltenbrunner V, Mishima M, Glotzer M (2007) Cooperative assembly of CYK-4/

MgcRacGAP and ZEN-4/MKLP1 to form the centralspindlin complex. Mol Biol Cell
18:4992–5003
106. Cai D, Hoppe AD, Swanson JA, Verhey KJ (2007) Kinesin-1 structural organization and
conformational changes revealed by FRET stoichiometry in live cells. J Cell Biol
176:51–63
107. Coy DL, Hancock WO, Wagenbach M, Howard J (1999) Kinesin’s tail domain is an inhibi-
tory regulator of the motor domain. Nat Cell Biol 1:288–292
108. Dietrich KA et al (2008) The kinesin-1 motor protein is regulated by a direct interaction of its
head and tail. Proc Natl Acad Sci U S A 105:8938–8943
109. Friedman DS, Vale RD (1999) Single-molecule analysis of kinesin motility reveals regulation
by the cargo-binding tail domain. Nat Cell Biol 1:293–297
110. Hackney DD, Baek N, Snyder AC (2009) Half-site inhibition of dimeric kinesin head
domains by monomeric tail domains. Biochemistry 48:3448–3456
111. Hackney DD, Levitt JD, Suhan J (1992) Kinesin undergoes a 9 S to 6 S conformational
transition. J Biol Chem 267:8696–8701
112. Hackney DD, Stock MF (2000) Kinesin’s IAK tail domain inhibits initial microtubule-
stimulated ADP release. Nat Cell Biol 2:257–260
113. Hackney DD, Stock MF (2008) Kinesin tail domains and Mg2+ directly inhibit release of
ADP from head domains in the absence of microtubules. Biochemistry 47:7770–7778
114. Kaan HY, Hackney DD, Kozielski F (2011) The structure of the kinesin-1 motor-tail complex
reveals the mechanism of autoinhibition. Science 333:883–885
115. Verhey KJ et al (1998) Light chain-dependent regulation of Kinesin’s interaction with micro-
tubules. J Cell Biol 143:1053–1066
116. Wong YL, Rice SE (2010) Kinesin’s light chains inhibit the head- and microtubule-binding
activity of its tail. Proc Natl Acad Sci U S A 107:11781–11786
117. Hammond JW, Blasius TL, Soppina V, Cai D, Verhey KJ (2010) Autoinhibition of the
kinesin-2 motor KIF17 via dual intramolecular mechanisms. J Cell Biol 189:1013–1025
118. Imanishi M, Endres NF, Gennerich A, Vale RD (2006) Autoinhibition regulates the motility
of the C. elegans intraflagellar transport motor OSM-3. J Cell Biol 174:931–937
119. Yamada KH, Hanada T, Chishti AH (2007) The effector domain of human Dlg tumor sup-
pressor acts as a switch that relieves autoinhibition of kinesin-3 motor GAKIN/
KIF13B. Biochemistry 46:10039–10045
120. Lee JR et al (2004) An intramolecular interaction between the FHA domain and a coiled coil
negatively regulates the kinesin motor KIF1A. EMBO J 23:1506–1515
121. Espeut J et al (2008) Phosphorylation relieves autoinhibition of the kinetochore motor
Cenp-E. Mol Cell 29:637–643
122. Stumpff J et al (2011) A tethering mechanism controls the processivity and kinetochore-
microtubule plus-end enrichment of the kinesin-8 Kif18A. Mol Cell 43:764–775
123. Antonio C et al (2000) Xkid, a chromokinesin required for chromosome alignment on the
metaphase plate. Cell 102:425–435
124. Vernos I et al (1995) Xklp1, a chromosomal Xenopus kinesin-like protein essential for spin-
dle organization and chromosome positioning. Cell 81:117–127
125. Brown KD, Wood KW, Cleveland DW (1996) The kinesin-like protein CENP-E is
kinetochore-associated throughout poleward chromosome segregation during anaphase-A. J
Cell Sci 109:961–969
126. Tokai N et al (1996) Kid, a novel kinesin-like DNA binding protein, is localized to chromo-
somes and the mitotic spindle. EMBO J 15:457–467
127. Levesque AA, Compton DA (2001) The chromokinesin Kid is necessary for chromosome
arm orientation and oscillation, but not congression, on mitotic spindles. J Cell Biol
154:1135–1146
128. Goshima G, Vale RD (2005) Cell cycle-dependent dynamics and regulation of mitotic
kinesins in Drosophila S2 cells. Mol Biol Cell 16:3896–3907
129. Liu X, Erikson RL (2007) The nuclear localization signal of mitotic kinesin-like protein
Mklp-1: effect on Mklp-1 function during cytokinesis. Biochem Biophys Res Commun
353:960–964
130. Cai S, Weaver LN, Ems-McClung SC, Walczak CE (2009) Kinesin-14 family proteins HSET/
XCTK2 control spindle length by cross-linking and sliding microtubules. Mol Biol Cell
20:1348–1359
131. Liao H, Li G, Yen TJ (1994) Mitotic regulation of microtubule cross-linking activity of
CENP-E kinetochore protein. Science 265:394–398
132. Blangy A et al (1995) Phosphorylation by p34cdc2 regulates spindle association of human
Eg5, a kinesin-related motor essential for bipolar spindle formation in vivo. Cell
83:1159–1169
133. Sawin KE, Mitchison TJ (1995) Mutations in the kinesin-like protein Eg5 disrupting localiza-
tion to the mitotic spindle. Proc Natl Acad Sci U S A 92:4289–4293
134. Ohsugi M et al (2003) Cdc2-mediated phosphorylation of Kid controls its distribution to
spindle and chromosomes. EMBO J 22:2091–2103
135. Cahu J et al (2008) Phosphorylation by Cdk1 increases the binding of Eg5 to microtubules
in vitro and in Xenopus egg extract spindles. PLoS One 3:e3936
136. Severson AF, Hamill DR, Carter JC, Schumacher J, Bowerman B (2000) The aurora-related
kinase AIR-2 recruits ZEN-4/CeMKLP1 to the mitotic spindle at metaphase and is required
for cytokinesis. Curr Biol 10:1162–1171
137. Andrews PD et al (2004) Aurora B regulates MCAK at the mitotic centromere. Dev Cell
6:253–268
138. Lan W et al (2004) Aurora B phosphorylates centromeric MCAK and regulates its localiza-
tion and microtubule depolymerization activity. Curr Biol 14:273–286
139. Ohi R, Sapra T, Howard J, Mitchison TJ (2004) Differentiation of cytoplasmic and meiotic
spindle assembly MCAK functions by Aurora B-dependent phosphorylation. Mol Biol Cell
15:2895–2906
140. Knowlton AL, Lan W, Stukenberg PT (2006) Aurora B is enriched at merotelic attachment
sites, where it regulates MCAK. Curr Biol 16:1705–1710
141. Zhang X, Lan W, Ems-McClung SC, Stukenberg PT, Walczak CE (2007) Aurora B phos-
phorylates multiple sites on mitotic centromere-associated kinesin to spatially and temporally
regulate its function. Mol Biol Cell 18:3264–3276
142. Kim Y, Holland AJ, Lan W, Cleveland DW (2010) Aurora kinases and protein phosphatase 1
mediate chromosome congression through regulation of CENP-E. Cell 142:444–455
143. Tanenbaum ME et al (2011) A complex of Kif18b and MCAK promotes microtubule
depolymerization and is negatively regulated by Aurora kinases. Curr Biol 21:1356–1365
144. Lee YM et al (2010) Cell cycle-regulated expression and subcellular localization of a kinesin-
8 member human KIF18B. Gene 466:16–25
145. Mayr MI et al (2007) The human kinesin Kif18A is a motile microtubule depolymerase
essential for chromosome congression. Curr Biol 17:488–498
146. Brown KD, Coulson RM, Yen TJ, Cleveland DW (1994) Cyclin-like accumulation and loss
of the putative kinetochore motor CENP-E results from coupling continuous synthesis with
specific degradation at the end of mitosis. J Cell Biol 125:1303–1312
147. Funabiki H, Murray AW (2000) The Xenopus chromokinesin Xkid is essential for metaphase
chromosome alignment and must be degraded to allow anaphase chromosome movement.
Cell 102:411–424
148. Fontijn RD et al (2001) The human kinesin-like protein RB6K is under tight cell cycle con-
trol and is essential for cytokinesis. Mol Cell Biol 21:2944–2955
149. Carvalho P, Gupta ML Jr, Hoyt MA, Pellman D (2004) Cell cycle control of kinesin-mediated
transport of Bik1 (CLIP-170) regulates microtubule stability and dynein activation. Dev Cell
6:815–829
150. Feine O, Zur A, Mahbubani H, Brandeis M (2007) Human Kid is degraded by the
APC/C(Cdh1) but not by the APC/C(Cdc20). Cell Cycle 6:2516–2523
151. Ganguly A, Bhattacharya R, Cabral F (2008) Cell cycle dependent degradation of MCAK:
evidence against a role in anaphase chromosome movement. Cell Cycle 7:3187–3193
152. Seguin L et al (2009) CUX1 and E2F1 regulate coordinated expression of the mitotic
complex genes Ect2, MgcRacGAP, and MKLP1 in S phase. Mol Cell Biol 29:570–581
153. Sedgwick GG et al (2013) Mechanisms controlling the temporal degradation of Nek2A and
Kif18A by the APC/C-Cdc20 complex. EMBO J 32:303–314
154. Wittmann T, Boleti H, Antony C, Karsenti E, Vernos I (1998) Localization of the kinesin-like
protein Xklp2 to spindle poles requires a leucine zipper, a microtubule-associated protein,
and dynein. J Cell Biol 143:673–685
155. Tahara K et al (2008) Importin-beta and the small guanosine triphosphatase Ran mediate
chromosome loading of the human chromokinesin Kid. J Cell Biol 180:493–506
156. Trieselmann N, Armstrong S, Rauw J, Wilde A (2003) Ran modulates spindle assembly by
regulating a subset of TPX2 and Kid activities including Aurora A activation. J Cell Sci
116:4791–4798
157. Ems-McClung SC, Zheng Y, Walczak CE (2004) Importin alpha/beta and Ran-GTP regulate
XCTK2 microtubule binding through a bipartite nuclear localization signal. Mol Biol Cell
15:46–57
158. Wilbur JD, Heald R (2013) Mitotic spindle scaling during Xenopus development by kif2a
and importin alpha. Elife 2:e00290
159. Manning BD, Barrett JG, Wallace JA, Granok H, Snyder M (1999) Differential regulation of
the Kar3p kinesin-related protein by two associated proteins, Cik1p and Vik1p. J Cell Biol
144:1219–1233
160. Benanti JA, Matyskiela ME, Morgan DO, Toczyski DP (2009) Functionally distinct isoforms
of Cik1 are differentially regulated by APC/C-mediated proteolysis. Mol Cell 33:581–590
161. Atherton J, Houdusse A, Moores C (2013) Mapping out distribution routes for kinesin
couriers. Biol Cell 105:465–487
162. Peris L et al (2009) Motor-dependent microtubule disassembly driven by tubulin tyrosina-
tion. J Cell Biol 185:1159–1166
163. Sturgill EG, Ohi R (2013) Kinesin-12 differentially affects spindle assembly depending on its
microtubule substrate. Curr Biol 23:1280–1290
164. Segbert C et al (2003) KLP-18, a Klp2 kinesin, is required for assembly of acentrosomal
meiotic spindles in Caenorhabditis elegans. Mol Biol Cell 14:4458–4469
165. Enos AP, Morris NR (1990) Mutation of a gene that encodes a kinesin-like protein blocks
nuclear division in A. nidulans. Cell 60:1019–1027
166. Hoyt MA, He L, Loo KK, Saunders WS (1992) Two Saccharomyces cerevisiae kinesin-
related gene products required for mitotic spindle assembly. J Cell Biol 118:109–120
167. Roof DM, Meluh PB, Rose MD (1992) Kinesin-related proteins required for assembly of the
mitotic spindle. J Cell Biol 118:95–108
168. Roostalu J et al (2011) Directional switching of the kinesin Cin8 through motor coupling.
Science 332:94–99
169. Gerson-Gurwitz A et al (2011) Directionality of individual kinesin-5 Cin8 motors is modu-
lated by loop 8, ionic strength and microtubule geometry. EMBO J 30:4942–4954
170. Fridman V et al (2013) Kinesin-5 Kip1 is a bi-directional motor that stabilizes microtubules
and tracks their plus-ends in vivo. J Cell Sci 126:4147–4159
171. Ferenz NP, Gable A, Wadsworth P (2010) Mitotic functions of kinesin-5. Semin Cell Dev
Biol 21:255–259
172. Kapoor TM, Mayer TU, Coughlin ML, Mitchison TJ (2000) Probing spindle assembly mech-
anisms with monastrol, a small molecule inhibitor of the mitotic kinesin, Eg5. J Cell Biol
150:975–988
173. Saunders AM, Powers J, Strome S, Saxton WM (2007) Kinesin-5 acts as a brake in anaphase
spindle elongation. Curr Biol 17:R453–R454
174. Tikhonenko I, Nag DK, Martin N, Koonce MP (2008) Kinesin-5 is not essential for mitotic
spindle elongation in Dictyostelium. Cell Motil Cytoskeleton 65:853–862
175. Sakowicz R et al (1998) A marine natural product inhibitor of kinesin motors. Science
280:292–295
176. Liu L, Parameswaran S, Liu J, Kim S, Wojcik EJ (2011) Loop 5-directed compounds inhibit
chimeric kinesin-5 motors: implications for conserved allosteric mechanisms. J Biol Chem
286:6201–6210
177. Rath O, Kozielski F (2012) Kinesins and cancer. Nat Rev Cancer 12:527–539
178. Tanenbaum ME et al (2009) Kif15 cooperates with eg5 to promote bipolar spindle assembly.
Curr Biol 19:1703–1711
179. Boleti H, Karsenti E, Vernos I (1996) Xklp2, a novel Xenopus centrosomal kinesin-like
protein required for centrosome separation during mitosis. Cell 84:49–59
180. Vanneste D, Takagi M, Imamoto N, Vernos I (2009) The role of Hklp2 in the stabilization and
maintenance of spindle bipolarity. Curr Biol 19:1712–1717
181. Yen TJ et al (1991) CENP-E, a novel human centromere-associated protein required for
progression from metaphase to anaphase. EMBO J 10:1245–1254
182. Yen TJ, Li G, Schaar BT, Szilak I, Cleveland DW (1992) CENP-E is a putative kinetochore
motor that accumulates just before mitosis. Nature 359:536–539
183. Schaar BT, Chan GK, Maddox P, Salmon ED, Yen TJ (1997) CENP-E function at kineto-
chores is essential for chromosome alignment. J Cell Biol 139:1373–1382
184. Wood KW, Sakowicz R, Goldstein LS, Cleveland DW (1997) CENP-E is a plus end-directed
kinetochore motor required for metaphase chromosome alignment. Cell 91:357–366
185. Yao X, Anderson KL, Cleveland DW (1997) The microtubule-dependent motor centromere-
associated protein E (CENP-E) is an integral component of kinetochore corona fibers that
link centromeres to spindle microtubules. J Cell Biol 139:435–447
186. Chan GK, Schaar BT, Yen TJ (1998) Characterization of the kinetochore binding domain of
CENP-E reveals interactions with the kinetochore proteins CENP-F and hBUBR1. J Cell
Biol 143:49–63
187. Abrieu A, Kahana JA, Wood KW, Cleveland DW (2000) CENP-E as an essential component
of the mitotic checkpoint in vitro. Cell 102:817–826
188. Yao X, Abrieu A, Zheng Y, Sullivan KF, Cleveland DW (2000) CENP-E forms a link between
attachment of spindle microtubules to kinetochores and the mitotic checkpoint. Nat Cell Biol
2:484–491
189. Weaver BA, Silk AD, Montagna C, Verdier-Pinard P, Cleveland DW (2007) Aneuploidy acts
both oncogenically and as a tumor suppressor. Cancer Cell 11:25–36
190. Wood KW et al (2010) Antitumor activity of an allosteric inhibitor of centromere-associated
protein-E. Proc Natl Acad Sci U S A 107:5839–5844
191. Meluh PB, Rose MD (1990) KAR3, a kinesin-related gene required for yeast nuclear fusion.
Cell 60:1029–1041
192. Hoyt MA, He L, Totis L, Saunders WS (1993) Loss of function of Saccharomyces cerevisiae
kinesin-related CIN8 and KIP1 is suppressed by KAR3 motor domain mutations. Genetics
135:35–44
193. O’Connell MJ, Meluh PB, Rose MD, Morris NR (1993) Suppression of the bimC4 mitotic
spindle defect by deletion of klpA, a gene encoding a KAR3-related kinesin-like protein in
Aspergillus nidulans. J Cell Biol 120:153–162
194. McDonald HB, Stewart RJ, Goldstein LS (1990) The kinesin-like ncd protein of Drosophila
is a minus end-directed microtubule motor. Cell 63:1159–1165
195. Chandra R, Salmon ED, Erickson HP, Lockhart A, Endow SA (1993) Structural and
functional domains of the Drosophila ncd microtubule motor protein. J Biol Chem
268:9005–9013
196. Janson ME et al (2007) Crosslinkers and motors organize dynamic microtubules to form
stable bipolar arrays in fission yeast. Cell 128:357–368
197. Oladipo A, Cowan A, Rodionov V (2007) Microtubule motor Ncd induces sliding of micro-
tubules in vivo. Mol Biol Cell 18:3601–3606
198. Fink G et al (2009) The mitotic kinesin-14 Ncd drives directional microtubule-microtubule
sliding. Nat Cell Biol 11:717–723
199. Gaglio T et al (1996) Opposing motor activities are required for the organization of the
mammalian mitotic spindle pole. J Cell Biol 135:399–414
200. Mountain V et al (1999) The kinesin-related protein, HSET, opposes the activity of Eg5 and
cross-links microtubules in the mammalian mitotic spindle. J Cell Biol 147:351–366
201. Zhu C et al (2005) Functional analysis of human microtubule-based motor proteins, the
kinesins and dyneins, in mitosis/cytokinesis using RNA interference. Mol Biol Cell
16:3187–3199
202. Kwon M et al (2008) Mechanisms to suppress multipolar divisions in cancer cells with extra
centrosomes. Genes Dev 22:2189–2203
203. Kleylein-Sohn J et al (2012) Acentrosomal spindle organization renders cancer cells
dependent on the kinesin HSET. J Cell Sci 125:5391–5402
204. Wu J et al (2013) Discovery and mechanistic study of a small molecule inhibitor for motor
protein KIFC1. ACS Chem Biol 8:2201–2208
205. Watts CA, Richards FM, Bender A, Bond PJ, Korb O, Kern O, Riddick M, Owen P, Myers
RM, Raff J, Gergely F, Jodrell DI, Ley SV (2013) Design, synthesis, and biological evalua-
tion of an allosteric inhibitor of HSET that targets cancer cells with supernumerary centro-
somes. Chem Biol 20(11):1399–1410
206. Yang B, Lamb ML, Zhang T, Hennessy EJ, Grewal G, Sha L, Zambrowski M, Block MH,
Dowling JE, Su N, Wu J, Deegan T, Mikule K, Wang W, Kaspera R, Chuaqui C, Chen H
(2014) Discovery of potent KIFC1 inhibitors using a method of integrated high-throughput
synthesis and screening. J Med Chem 57(23):9958–9970
207. Grinberg-Rashi H et al (2009) The expression of three genes in primary non-small cell lung
cancer is associated with metastatic spread to the brain. Clin Cancer Res 15:1755–1761
208. Ems-McClung SC, Walczak CE (2010) Kinesin-13s in mitosis: key players in the spatial and
temporal organization of spindle microtubules. Semin Cell Dev Biol 21:276–282
209. Desai A, Verma S, Mitchison TJ, Walczak CE (1999) Kin I kinesins are microtubule-
destabilizing enzymes. Cell 96:69–78
210. Hunter AW et al (2003) The kinesin-related protein MCAK is a microtubule depolymerase
that forms an ATP-hydrolyzing complex at microtubule ends. Mol Cell 11:445–457
211. Kline-Smith SL, Walczak CE (2002) The microtubule-destabilizing kinesin XKCM1 regu-
lates microtubule dynamic instability in cells. Mol Biol Cell 13:2718–2731
212. Rizk RS et al (2009) MCAK and paclitaxel have differential effects on spindle microtubule
organization and dynamics. Mol Biol Cell 20:1639–1651
213. Kline-Smith SL, Khodjakov A, Hergert P, Walczak CE (2004) Depletion of centromeric
MCAK leads to chromosome congression and segregation defects due to improper kineto-
chore attachments. Mol Biol Cell 15:1146–1159
214. Moore AT et al (2005) MCAK associates with the tips of polymerizing microtubules. J Cell
Biol 169:391–397
215. Zhang X, Ems-McClung SC, Walczak CE (2008) Aurora A phosphorylates MCAK to control
ran-dependent spindle bipolarity. Mol Biol Cell 19:2752–2765
216. Ganem NJ, Compton DA (2004) The KinI kinesin Kif2a is required for bipolar spindle
assembly through a functional relationship with MCAK. J Cell Biol 166:473–478
217. Rogers GC et al (2004) Two mitotic kinesins cooperate to drive sister chromatid separation
during anaphase. Nature 427:364–370
218. Maney T, Hunter AW, Wagenbach M, Wordeman L (1998) Mitotic centromere-associated
kinesin is important for anaphase chromosome segregation. J Cell Biol 142:787–801
219. Wordeman L, Wagenbach M, von Dassow G (2007) MCAK facilitates chromosome
movement by promoting kinetochore microtubule turnover. J Cell Biol 179:869–879
220. Shimo A et al (2008) Involvement of kinesin family member 2C/mitotic centromere-
associated kinesin overexpression in mammary carcinogenesis. Cancer Sci 99:62–70
221. Huang R et al (2013) Intracellular targets for a phosphotyrosine peptidomimetic include the
mitotic kinesin, MCAK. Biochem Pharmacol 86:597–611
222. Maeda N et al (2005) Effects of DNA polymerase inhibitory and antitumor activities of
lipase-hydrolyzed glycolipid fractions from spinach. J Nutr Biochem 16:121–128
223. Talje L, Ben E, Kadhi K, Atchia K, Tremblay-Boudreault T, Carreno S, Kwok BH (2014)
DHTP is an allosteric inhibitor of the kinesin-13 family of microtubule depolymerases.
FEBS Lett 588(14):2315–2320
224. Ishikawa K et al (2008) Mitotic centromere-associated kinesin is a novel marker for
prognosis and lymph node metastasis in colorectal cancer. Br J Cancer 98:1824–1829
225. Nakamura Y et al (2007) Clinicopathological and biological significance of mitotic
centromere-associated kinesin overexpression in human gastric cancer. Br J Cancer
97:543–549
226. Vanneste D, Ferreira V, Vernos I (2011) Chromokinesins: localization-dependent functions
and regulation during cell division. Biochem Soc Trans 39:1154–1160
227. Yajima J et al (2003) The human chromokinesin Kid is a plus end-directed microtubule-based
motor. EMBO J 22:1067–1074
228. Sekine Y et al (1994) A novel microtubule-based motor protein (KIF4) for organelle
transports, whose expression is regulated developmentally. J Cell Biol 127:187–201
229. Bieling P, Telley IA, Surrey T (2010) A minimal midzone protein module controls formation
and length of antiparallel microtubule overlaps. Cell 142:420–432
230. Bringmann H et al (2004) A kinesin-like motor inhibits microtubule dynamic instability.
Science 303:1519–1522
231. Zhang P, Knowles BA, Goldstein LS, Hawley RS (1990) A kinesin-like protein required for
distributive chromosome segregation in Drosophila. Cell 62:1053–1062
232. Mazumdar M, Sundareshan S, Misteli T (2004) Human chromokinesin KIF4A functions in
chromosome condensation and segregation. J Cell Biol 166:613–620
233. Williams BC, Riedy MF, Williams EV, Gatti M, Goldberg ML (1995) The Drosophila
kinesin-like protein KLP3A is a midbody component required for central spindle assembly
and initiation of cytokinesis. J Cell Biol 129:709–723
234. Mazumdar M et al (2006) Tumor formation via loss of a molecular motor protein. Curr Biol
16:1559–1564
235. Cottingham FR, Hoyt MA (1997) Mitotic spindle positioning in Saccharomyces cerevisiae
is accomplished by antagonistically acting microtubule motor proteins. J Cell Biol
138:1041–1053
236. Garcia MA, Koonrugsa N, Toda T (2002) Two kinesin-like Kin I family proteins in fission
yeast regulate the establishment of metaphase and the onset of anaphase A. Curr Biol
12:610–621
237. Stumpff J, von Dassow G, Wagenbach M, Asbury C, Wordeman L (2008) The kinesin-8
motor Kif18A suppresses kinetochore movements to control mitotic chromosome alignment.
Dev Cell 14:252–262
238. Stout JR et al (2011) Kif18B interacts with EB1 and controls astral microtubule length during
mitosis. Mol Biol Cell 22:3070–3080
239. Su X et al (2011) Mechanisms underlying the dual-mode regulation of microtubule dynamics
by Kip3/kinesin-8. Mol Cell 43:751–763
240. Weaver LN et al (2011) Kif18A uses a microtubule binding site in the tail for plus-end
localization and spindle length regulation. Curr Biol 21:1500–1506
241. Nagahara M et al (2011) Kinesin 18A expression: clinical relevance to colorectal cancer
progression. Int J Cancer 129:2543–2552
242. Zhang C et al (2010) Kif18A is involved in human breast carcinogenesis. Carcinogenesis
31:1676–1684
243. Catarinella M, Gruner T, Strittmatter T, Marx A, Mayer TU (2009) BTB-1: a small molecule
inhibitor of the mitotic motor protein Kif18A. Angew Chem Int Ed Engl 48:9072–9076
244. Braun J, Möckel MM, Strittmatter T, Marx A, Groth U, Mayer TU (2014) Synthesis and
biological evaluation of optimized inhibitors of the mitotic kinesin Kif18A. ACS Chem Biol
[Epub ahead of print]
245. Glotzer M (2009) The 3Ms of central spindle assembly: microtubules, motors and MAPs.
Nat Rev Mol Cell Biol 10:9–20
246. Gruneberg U et al (2006) KIF14 and citron kinase act together to promote efficient cytokine-
sis. J Cell Biol 172:363–372
247. Jantsch-Plunger V et al (2000) CYK-4: a Rho family gtpase activating protein (GAP) required
for central spindle formation and cytokinesis. J Cell Biol 149:1391–1404
248. Nislow C, Lombillo VA, Kuriyama R, McIntosh JR (1992) A plus-end-directed motor
enzyme that moves antiparallel microtubules in vitro localizes to the interzone of mitotic
spindles. Nature 359:543–547
249. Gasnereau I et al (2012) KIF20A mRNA and its product MKlp2 are increased during hepa-
tocyte proliferation and hepatocarcinogenesis. Am J Pathol 180:131–140
250. Imai K et al (2011) Identification of HLA-A2-restricted CTL epitopes of a novel tumour-
associated antigen, KIF20A, overexpressed in pancreatic cancer. Br J Cancer 104:300–307
251. Taniuchi K et al (2005) Down-regulation of RAB6KIFL/KIF20A, a kinesin involved with
membrane trafficking of discs large homologue 5, can attenuate growth of pancreatic cancer
cell. Cancer Res 65:105–112
252. Tcherniuk S et al (2010) Relocation of Aurora B and survivin from centromeres to the central
spindle impaired by a kinesin-specific MKLP-2 inhibitor. Angew Chem Int Ed Engl
49:8228–8231
253. Alphey L et al (1997) KLP38B: a mitotic kinesin-related protein that binds PP1. J Cell Biol
138:395–409
254. Molina I et al (1997) A chromatin-associated kinesin-related protein required for normal
mitotic chromosome segregation in Drosophila. J Cell Biol 139:1361–1371
255. Ohkura H et al (1997) Mutation of a gene for a Drosophila kinesin-like protein, Klp38B,
leads to failure of cytokinesis. J Cell Sci 110:945–954
256. Ruden DM, Cui W, Sollars V, Alterman M (1997) A Drosophila kinesin-like protein, Klp38B,
functions during meiosis, mitosis, and segmentation. Dev Biol 191:284–296
257. Carleton M et al (2006) RNA interference-mediated silencing of mitotic kinesin KIF14 dis-
rupts cell cycle progression and induces cytokinesis failure. Mol Cell Biol 26:3853–3863
258. Basavarajappa HD, Corson TW (2012) KIF14 as an oncogene in retinoblastoma: a target for
novel therapeutics? Future Med Chem 4:2149–2152
259. Corson TW et al (2007) KIF14 messenger RNA expression is independently prognostic for
outcome in lung cancer. Clin Cancer Res 13:3229–3234
260. Corson TW, Gallie BL (2006) KIF14 mRNA expression is a predictor of grade and outcome
in breast cancer. Int J Cancer 119:1088–1094
261. Corson TW, Huang A, Tsao MS, Gallie BL (2005) KIF14 is a candidate oncogene in the 1q
minimal region of genomic gain in multiple cancers. Oncogene 24:4741–4753
262. Madhavan J et al (2007) High expression of KIF14 in retinoblastoma: association with older
age at diagnosis. Invest Ophthalmol Vis Sci 48:4901–4906
Chapter 2
The Discovery and Development
of Eg5 Inhibitors for the Clinic
James A.D. Good, Giacomo Berretta, Nahoum G. Anthony,

and Simon P. Mackay
Abbreviations and Definitions
AAG α-1-acid glycoprotein

AML Acute myeloid leukemia
Basal Eg5 inhibition Inhibition of the basal ATPase activity of Eg5
CBR Clinical benefit rate
CYP Cytochrome P450
DMPK Drug metabolism and pharmacokinetics
F Bioavailability
fu Fraction unbound
hERG Human ether-a-go-go-related gene
HHPQ Hexahydropyranoquinoline
HTS High-throughput screening
i.p. Intraperitoneal
Kiapp Estimated apparent Ki value
MCL-1 Antiapoptotic protein myeloid cell leukemia 1
MDR Multidrug resistance
MM Multiple myeloma
MT Microtubules
MT Eg5 inhibition Inhibition of the microtubule stimulated ATPase activity of
Eg5
J.A.D. Good (*)

Department of Chemistry, Umeå University, 901 87 Umeå, Sweden
Umeå Centre for Microbial Research, Umeå University, 901 87 Umeå, Sweden
e-mail: james.good@chem.umu.se
G. Berretta • N.G. Anthony • S.P. Mackay (*)
Strathclyde Institute of Pharmacy and Biomedical Sciences, University of Strathclyde,
161 Cathedral Street, Glasgow G4 0RE, Scotland, UK
e-mail: simon.mackay@strath.ac.uk

28 J.A.D. Good et al.
MTD Maximum tolerated dose

NCI National Cancer Institute
n.i. No inhibition
ORR Overall response rate
PK Pharmacokinetic
PgP P-glycoprotein
RRMM Relapsed/refractory multiple myeloma
SAR Structure activity relationship
STLC S-trityl L-cysteine
2.1 Introduction
Interest in the mitotic kinesins as targets in cancer treatment began following the
discovery of monastrol (Fig. 2.1), the first selective inhibitor of a mitotic kinesin [1].
Monastrol was identified in a high throughput phenotypic screen designed to detect
novel anti-mitotic agents, with exposure to HeLa cells producing a distinctive
monoastral spindle phenotype, comparable to that observed by RNAi mediated Eg5
depletion [1, 2]. Biochemical characterisation of the kinetics of monastrol with
respect to ATP and microtubule (MT) activity [3], along with determination of the
Fig. 2.1 Crystal structure of

Eg5 with monastrol (PDB
entry 1Q0B) [4] highlighting
the allosteric binding site
(circled) delineated by helix
α2 interrupted by loop L5 and
helix α3 and the structure of
monastrol (inset). The
orthosteric nucleotide binding
site occupied by ADP can be
seen above helix α2. The
nucleotide and ligand are
coloured by atom type: green
(C), yellow (S), blue (N),
orange (P) and red (O)
2 The Discovery and Development of Eg5 Inhibitors for the Clinic 29
crystal structure of the ternary monastrol · Eg5 · ADP-Mg2+ complex, revealed that
monastrol was an allosteric inhibitor of Eg5. The allosteric binding site was formed
by helix α2/loop L5 and helix α3, and situated approximately 12 Å from the nucleo-
tide site (Fig. 2.1) [4]. This was the first example of a compound which selectively
disrupted the mitotic spindle apparatus without directly affecting tubulin.
Since the discovery of monastrol, many studies have focused on Eg5, and new
inhibitors have been identified by screening and structure-based design in both aca-
demic and industry-based research groups. Development of these inhibitors has had
two primary objectives: to produce chemical tools to investigate the role of Eg5 in
mitosis and connection to cancer, and to generate chemotherapy drug candidates.
There are a number of comprehensive reviews in print that cover the current Eg5
inhibitor landscape [5–7]. This chapter focuses on the preclinical development and
lead optimisation of Eg5 inhibitors designed for therapeutic use, and in particular
those which reached clinical evaluation, with emphasis on their optimisation, and the
strategies employed in addressing developmental challenges. For four of the series,
the observations are from an external viewpoint limited to the published data, with
the commentary based on general medicinal chemistry principles. In the final case,
the narrative is based on our own experience of carrying a hit through preclinical
development. All of the compounds described in this chapter bind to the allosteric
binding site described for monastrol: for details on how this affects Eg5 processivity
and disrupts assembly of the mitotic spindle, refer to Chap. 3. Similarly, a full analy-
sis of the available clinical data for kinesin inhibitors is presented in Chap. 4.
2.2 Ispinesib and Related Analogues
2.2.1 Ispinesib
Ispinesib (also known as SB-715992; Fig. 2.2) was discovered by researchers at

Cytokinetics using library screening and subsequent optimisation [8]. Ispinesib was
the first Eg5 inhibitor to demonstrate in vivo anti-tumour activity and to enter clini-
cal trials [9, 10]. Biochemical [11] and X-ray crystallography studies [12, 13]
revealed that ispinesib, like monastrol, was an allosteric inhibitor that bound to the
helix α2/loop L5 and helix α3 region to prevent ADP release from the catalytic site
(Fig. 2.3). Ispinesib inhibits Eg5 ATPase activity with an estimated Kiapp of 3.1 and
1.7 nM for the basal and MT stimulated activities respectively (referred to hereafter
as “basal Eg5 inhibition” and “MT Eg5 inhibition”) [11]. It is selective for Eg5
within the kinesin family, with no inhibitory effects observed at 20 µM on the kine-
sins CENP-E, MKLP-1, MKLP-2, MCAK, Kif1A or Kif5B. Initial studies involv-
ing the close analogue CK0106023 (Fig. 2.2) showed robust mitotic arrest in
multiple cancer cell lines, including several multidrug resistant (MDR) variants [8].
Anti-tumour activity comparable to paclitaxel was observed when mice bearing
human ovarian carcinoma SKOV3 xenografts were treated with CK0106023. While
paclitaxel inhibited tumour growth by an average of 73 % at its maximum tolerated
Fig. 2.2 A common-feature pharmacophore model of ispinesib and related compounds
dose (MTD) of 20 mg/kg, CK0106023 delivered an average tumour growth inhibi-

tion of 71 % at 25 mg/kg. The excised tumours displayed the phenotypical monopo-
lar mitotic spindles, implying the response was through an on-target effect. Ispinesib
possesses broad spectrum anti-tumour efficacy in vivo, with complete regression
evident across multiple histologically distinct murine xenograft models [9, 14, 15].
Clinical trials commenced in 2003 in collaboration with GlaxoSmithKline [10], and
ispinesib has since been the subject of multiple phase 1 and 2 evaluations, primarily
in advanced solid tumours [16].
Initial disclosure of the in vivo anti-tumour efficacy of ispinesib generated wider
interest in Eg5 as an oncology target. Numerous related scaffolds were patented by
rival pharmaceutical companies such as Chiron, AstraZeneca, Bristol-Myers Squibb
and Merck & Co, as reviewed by Matsuno and co-workers and more recently Jiang
et al [7, 17]. A common-feature pharmacophore model is presented in Fig. 2.2,
depicting the benzyl substituted bicyclic core with a propylamine side chain and
benzamide moieties linked to the central fragment via a small aliphatic chain sub-
stituted with an isopropyl.
The details of the optimisation of the ispinesib scaffold from the Cytokinetics
programme are scarce. Removal of the para-methylbenzamide reduced activity by
approximately three orders of magnitude, with small lipophilic substituents pre-
ferred on the benzamide (e.g. Me, Br and F) [18, 19]. In the amino side chain, both
ethyl and propyl primary and tertiary amines were active in the low nanomolar
Fig. 2.3 (a) Ispinesib binding to Eg5 (PDB entry 4AP0) [13]; (b) MK0731 binding to Eg5 (PDB
entry 3CJO) [25]; (c) EMD534085 binding to Eg5 (PDB entry 3L9H) [26]; (d) STLC binding to
Eg5 (PDB entry 3KEN) [27]. The ligands are coloured by atom type: green (C), cyan (F), dark
green (Cl), yellow (S), blue (N) and red (O). The amino acid residues are coloured by atom type:
white (C), blue (N) and red (O)
range in the MT-Eg5 enzymatic assay [20]. However, uncharged hydroxyl or

methoxy groups were much less effective. Significant stereoselectivity was also evi-
dent, with (R)-CK0106023 more than 1,000-fold more active against Eg5 activity
than (S)-CK0106023 [8].
Analysis of the crystal structure reveals that the binding of ispinesib with Eg5 is
predominantly hydrophobic (Fig. 2.3a) [13]. The benzyl group occupies a hydropho-
bic pocket bounded by Trp 127, Pro 137, Val 210 and Tyr 211, and is further enclosed
by the benzamide moiety of ispinesib, with which it forms an intramolecular edge-
to-face stacking interaction. The isopropyl group lies in a small hydrophobic patch
between the backbones of Leu 214/Glu215 and Tyr 211, while the quinazolinone
core is situated in deeper hydrophobic pocket formed by Ile 136, Leu160, Leu171,
Gly217, Arg221 and Phe239. The primary hydrophilic contribution to binding is

from the propylamine, which is oriented towards the bulk solvent and forms a salt
bridge with Glu116, thus explaining the preference for a charged moiety in this posi-
tion. Ispinesib displays reasonable overall drug-like properties, with log D7.4 = 2.66,
moderate aqueous solubility and low clearance in both human microsomes and
hepatocytes in vitro [21]. However, it moderately inhibits hERG (4.81 µM), cyto-
chrome P450 (CYP) 3A4 (4.0 µM), and the P-glycoprotein (Pgp) efflux pump [21,
22]. Toxicity is also an issue, with significantly higher mortality rates observed in
large scale mouse studies in treated versus untreated animals: (32 % versus 1 % in a
study involving 1,021 mice) [14].
2.2.2 SB-743921
Cytokinetics developed a more potent second generation inhibitor with

GlaxoSmithKline (SB-743921, Fig. 2.2), in which the quinazolinone core was
replaced with a chromen-4-one [23]. This analogue exhibited improved activity in
enzymatic and cellular assays, increased solubility (turbidimetric solubility pH7.4:
ispinesib = 65 µM; SB-743921 = 100 µM) and decreased plasma protein binding
(ispinesib fu = 8 %; SB-743921 fu = 22 %); however interactions with the hERG
channel (1.6 µM), and inhibition of CYP3A4 (4.0 µM) remained undiminished [15,
21, 23]. SB-743921 progressed to phase 2 trials in patients with advanced solid
tumours and lymphomas, where it showed reduced neutropenia compared to ispine-
sib, but no strong therapeutic responses [24].
2.2.3 AZD-4877
Despite the number of anti-tumour studies performed with both ispinesib and
SB-743921, there is little information in the scientific literature describing their
development. However, the recent publication describing the development of AZD-
4877 (Fig. 2.2) helps provide insight into the strategies employed to optimise this
class of Eg5 inhibitor [28].
AZD-4877 was developed by scaffold hopping through replacing the chlorophe-
nyl moiety of the quinazolinone core of ispinesib with [3,2-d]thiophene (Table 2.1).
This first compound (2.1) had a basal Eg5 inhibitory activity (IC50) of 12 nM and a
cellular EC50 in Colo205 cells of 104 nM. Initial attempts to improve activity by
removing, replacing or derivatising the “northern” benzyl ring were unsuccessful:
heterocyclic alternatives reduced potency by 50–1,000-fold and substituents in the
phenyl ring produced an order of magnitude drop in activity. Modifications to the
propylamino group that included increasing the chain length or removing confor-
mational flexibility led to a 30–140-fold drop in activity, whilst replacing the basic
centre with an acetamide or methylsulfonamide decreased inhibitory activity by 290
Table 2.1 Optimisation of AZD-4877
Compound Ring A R1 R2 Basal Eg5 IC50 (nM) Colo205 cell EC50 (nM)
2.1 H H 12 104
2.2 F H 20 60
2.3 H H 8 30
2.4 H H 16 23
2.5 H H 3 4
2.6 H H 5 45
(+)-(AZD- H Me 2 3
4877)
Compounds 2.1–2.6 were prepared racemically. Compiled from reference with permission [28]
and 600-fold respectively. Attempts to alter the “eastern” benzamide moiety with
alkyl or cycloalkyl replacements reduced activity to micromolar levels, whilst
replacing the phenyl ring with different heterocycles or finding alternatives to the
4-methyl substituent consistently failed to improve potency, with only the 3-fluoro-
4-methylphenyl analogue (2.2) equipotent to the parent compound. Improvements
in potency were only achieved through changes to the heterocyclic core: replace-
ment of the [3,2-d]thiophene with the [2,3-d]thiophene isomer (2.3), an isoxazole
(2.4) or two different isothiazole isomers (2.5 and 2.6) led to compounds with com-
parable basal Eg5 inhibitory activity and potency against Colo205 cells. Evaluation
of drug-like properties revealed 2.1 had poor aqueous solubility (~70 µM) in com-
parison with 2.3 and 2.5, which were both >1 mM. Extending the southern alkyl
group from an ethyl to an isopropyl group provided a further increase in potency,

and afforded a compound (AZD-4877) with essentially the same substituents as
ispinesib, but a different bicyclic core.
AZD-4877 had a lower in vivo clearance than 2.5 (17 mL/min/kg vs. 50 mL/min/
kg) and was therefore selected as the candidate for progression. Pharmacokinetic
(PK) analysis revealed AZD-4877 possessed a modest half-life in rat (t½ = 3.5 h),
and was highly bound to plasma protein in human plasma (fraction bound = 95.4 %).
Unlike ispinesib, AZD-4877 did not inhibit common CYP450 isoforms at 10 µM,
including 3A4, although calcium and sodium channel blocking effects were
observed in vitro in guinea-pig ion channel models, but at levels not likely to be
pharmacologically significant. A pharmacodynamic response was observed in vivo
in a rat hollow fibre model using implanted Colo205 cells and treated with a single
intravenous dose of AZD-4877, and this compound subsequently progressed into
clinical trials in solid tumours, lymphoma and acute myeloid leukemia (AML) [16].
Generally, it was well-tolerated, but despite pharmacodynamic evidence of on-
target inhibition, the trials did not show clear clinical benefit, and development was
discontinued [29].
2.3 MK-0731
2.3.1 Discovery and Early Lead Optimisation
MK-0731 was the second Eg5 inhibitor to enter clinical trials [30]. It is a 2,4-diaryl-
2,5-dihydropyrrole discovered at Merck Research Laboratories by a systematic lead
optimisation programme, which involved the development of four different core
scaffolds (reported as Core A, B, C and D in Table 2.2). They have in common the
presence of an “eastern” phenyl group linked to a chiral carbon of the core scaffold
and another phenyl group linked to the western portion of the heterocyclic core. In
addition, a carbonyl group extends from the southern position of the molecule.
The discovery of MK-0731 arose from HTS of the Merck in-house collection
with an MT-stimulated Eg5 ATPase assay to determine enzyme inhibition. Racemic
3,5-diaryl-4,5-dihydropyrazoles 3.1 and 3.2 were shown to have low micromolar
activity [31]. While the analogue 3.3 devoid of substituents in the phenyl rings was
inactive, the combination of the two substituents present in the HTS hits provided
the hybrid 3.4 with sub-micromolar activity. Extensive substitution patterns were
then investigated for both aryl rings. Most substituents introduced in the 2-position
of the western aryl ring substantially lowered inhibitory activity, with only the
exchange of chlorine to fluorine tolerated (3.5). Introducing more than one substitu-
ent in the western ring led to the 2,5-difluoro analogue (3.6) which had an impres-
sive 40-fold boost in potency, with the improvement at least partially attributed to
enhanced multi-polar contacts [32]. Analogues with variations in the southern
groups showed that large substituents could be accommodated with no detrimental
Table 2.2 Hit-to-lead process and common-feature pharmacophore model of MK-0731 and related
Compound Core R1 R2 R3 R4 R5 Chiral carbon MT Eg5 IC50 (nM)

3.1 A Cl H H CH3 H R/S 3,600
3.2 A H H OH CH3 H R/S 6,900
3.3 A H H H CH3 H R/S >50,000
3.4 A Cl H OH CH3 H R/S 450
3.5 A F H H CH3 H R/S 3,600
3.6 A F F H CH3 H R/S 94
3.7 A Cl H OH t-Bu H R/S 460
3.8 A F F H NMe2 H R/S 84
3.9 A F F OH CH3 H R/S 51
3.10 A F F OH CH3 H S 26
3.11 A F F OH CH3 H R 4,100
3.12 B F F H NMe2 H R/S 84
3.13 B F F H NMe2 H S 38
3.14 B F F H NMe2 H R >50,000
3.15 B F F OH NMe2 H S 7
Compiled from references with permission [31, 33, 34]
effect on potency: for example, the pivaloyl analogue 3.7 had the same activity as
the corresponding acetyl 3.4 [31]. Interestingly, the acetyl group of 3.6 could be
replaced by a dimethylamide group (3.8), whereas methyl ester, methyl amide and
methyl sulfonyl moieties resulted in substantial reductions in potency. In the eastern
aromatic ring, a 3-hydroxyl group gave the derivative 3.9 with an IC50 = 51 nM, but
attempts to optimise further with other substituents were ineffective. Resolution of
the enantiomers of 3.9 established that the (S)-enantiomer 3.10 was solely respon-
sible for activity with an IC50 of 26 nM, whereas the (R)-enantiomer 3.11 was essen-
tially inactive.
Co-crystal structures of 3,5-diaryl-4,5-dihydropyrazoles with Eg5 showed they
occupied the allosteric binding site and that the 3-hydroxyl group of the eastern
aromatic ring interacted with the backbone amide of Glu118, similar to the phenol
group of monastrol, and that the S configuration of the chiral carbon was essential
[31]. The western substituted phenyl occupied a large hydrophobic pocket not uti-
lized by monastrol, while the southern carbonyl moiety extended into solvent,
which accounted for the greater structural diversity tolerated in this region of the
molecule.
In order to expand the scope of the 4,5-dihydropyrazole series (core A), Merck
researchers rationalised that the nucleus of this series could be replaced by a
2,5-dihydropyrrole ring (core B) [33]. The advantage of this heterocycle was its
synthetic potential to facilitate greater diversification whilst retaining potency via
maintenance of the relative aromatic ring orientations. The racemic dihydropyrrole
3.12 was equipotent to the corresponding dihydropyrazole 3.8, and as before, only
the (S)-enantiomer (3.13) was responsible for inhibitory activity, with the (R)-
enantiomer (3.14) was completely inactive. Introduction of the 3-hydroxyl group in
the eastern ring of 3.13 produced 3.15 with 7 nM potency (Table 2.2).
Although the dihydropyrrole 3.15 was a potent and low molecular weight com-
pound, it possessed negligible aqueous solubility in water (<1 mg/mL at pH 7),
which prohibited simple aqueous administration for in vivo studies without the use
of formulation vehicles. Target-therapeutic profiles generally require oral delivery
in solid dose forms, which typically necessitates optimising the intrinsic solubility
of the lead compounds under development. Two approaches were adopted by the
researchers at Merck to address this: preparation of prodrugs or insertion of a salt-
forming amine group in the eastern or southern motifs. Initial efforts led to the
development of the water-soluble prodrug 3.16, designed to undergo hydrolysis to
3.15 by non-specific phosphatases (Fig. 2.4) [34].
Rat and dog PK profiles confirmed the rapid and complete conversion of 3.16 to
3.15. The inhibitory activity was evaluated by an in vivo xenograft tumour model of
mitotic arrest using nude mice bearing human ovarian carcinoma cell line A2780.
However, the modest in vivo efficacy and concerns about the rapid clearance and
possible toxicity of phenol 3.15 by reactive metabolite formation, led to abandoning
the phenolic moiety [30].
2.3.2 Introduction of the Propylamine Moiety
Molecular modelling of the Eg5 L5 allosteric site suggested a water solubilising

group could occupy solvent-accessible space situated above the plane of the central
core [35]. Initial efforts showed a propylamine group branching from the chiral
F F
F F
OH
O OH
P OH
N O Complete conversion N
N in rat (t1/2 = 0.25h)
O N O
in dog(t1/2 = 0.15h)
3.16 3.15
Aq. sol. > 20 mg/mL at pH 7 Aq. sol. < 1 mg/mL at pH 7
Eg5 IC50 >1 µM Eg5 IC50 = 7 µM (Cell EC50 = 22 nM)
in vivo EC90 = 700 nM
Fig. 2.4 In vivo conversion of the water-soluble prodrug 3.16 to the active form
carbon of the core resulted in increased potency, which was attributed to the forma-
tion of an additional hydrogen-bond with the backbone Gly117. Specifically, com-
pound 3.17 had significantly improved potency (IC50 = 2.2 nM, Table 2.3) compared
to 3.13 (IC50 = 38 nM) and favourable aqueous solubility (>12 mg/mL, pH 4) [36].
However, the basic amine side chain increased both the binding affinity for hERG
and susceptibility to Pgp-efflux, as measured by a decrease in cellular potency of
3.17 in the Pgp overexpressing cell line KB-V1, with an MDR ratio of 1,200 over
the parental KB-3-1 cell line. One approach devised to reduce Pgp-efflux and
improve MDR ratio was to attenuate basicity of the amino group by β-fluorination.
Modulation of the pKa of the ethylaminopropyl group to 10.7–7.0 by insertion of
fluorine atoms had the required effect on the MDR ratio whilst maintaining potency
(3.18–3.20; Table 2.3). However, a more marked decrease in pKa (3.21) produced a
drop in potency, which was probably due to a reduced population of the ionised spe-
cies required for good Eg5 affinity. Optimal modulation was achieved with 3.22,
which maintained potency and solubility whilst simultaneously reducing the MDR
ratio and hERG binding.
Dihydropyrazole analogues with different amide groups were also investigated
(3.23–3.28), with 3.23 having a better MDR ratio in comparison to the correspond-
ing dihydropyrrole 3.17 (491 vs. 1,200 respectively) [37]. Furthermore, replace-
ment of the dimethylamide in 3.23 with an acetyl group improved the MDR ratio
(3.24). Replacement of the primary amine with tertiary amines improved both MDR
and passive permeability (3.25–3.27). Interestingly, although the low basicity of
compound 3.27 abolished hERG binding in radioligand competition experiments, a
weak blockade of the IKr current was seen in a classical patch clamp assay. Indeed,
low micromolar plasma levels of 3.27 were found to induce QTc prolongation in a
canine cardiovascular model, indicating a severe interaction with the hERG chan-
nel. Fortunately, primary amine 3.24 had a minor propensity to block the IKr current,
and efforts were refocused on the attenuation of the pKa of this compound by
β-fluorination. Amine 3.28 (Table 2.3) showed excellent enzyme and cellular
potency, low MDR ratio and no hERG interaction in both the patch clamp assay and
dog model.
Table 2.3 Lead optimisation of the propylamine side chain across cores A-D
R1
F F O F
F NHR2 F
1 1
R2 NR3R4 R NR3R4 R NR3R4
N N
N N N
N N
N O R1 O O O
Core B 3.17-3.22 Core A 3.23-3.28 Core C 3.29-3.34 Core D 3.35-3.40
hERG
Eg5 IC50 A2780 MDR IC50/IKr
Cmpd Core R1 R2 NR3R4 pKa (nM) EC50 (nM) ratio IC50(µM)
3.17 B H H – 10.3 2.2 6.0 1,200 7.1/–
3.18 B H Et – 10.7 10.2 – >135 –
3.19 B H CH2CH2F – 8.8 10.2 – 32 –
3.20 B H CH2CHF2 – 7.0 12.1 – 3 –
3.21 B H CH2CF3 – 5.2 110 – 1 –
3.22 B CHF2 H – 7.0 5.2 22.5 5 15.9/–
3.23 A NMe2 H NH2 – 2.0 5.2 491 19.3/–
3.24 A Me H NH2 – 0.9 2.6 88 –
3.25 A Me H NMe2 – 1.4 2.3 4.4 –
3.26 A Me H Morpholine 6.7 1.8 11 1.7 –
3.27 A Me H N-Ac- 7.1 2.8 6.0 2.5 >30/7
piperazine
3.28 A Me F NH2 – 0.8 3.4 5.2 –/>30
3.29 C F – NH2 – 0.5 2.5 1.7 0.9/–
3.30* C F – NMe2 – 1.6 7.4 2.3 5.9/–
3.31* C Cl – NMe2 – 1.0 4.8 2.5 2.0/–
3.32* C Me – NH2 – 1.8 6.4 2.3 2.4/–
3.33 C Cl – N-Ac- – 1.0 9.0 1.6 5.6/–
piperazine
3.34 C F – N-Ac- – 1.6 5.0 1.0 10/>35
piperazine
3.35* D F – N-Ac- – 3.8 78 – –
piperazine
3.36* D Cl – N-Ac- – 3.9 46 – –
piperazine
3.37* D Br – N-Ac- – 4.7 38 – –
piperazine
3.38* D Me – N-Ac- – 2.0 27 – –
piperazine
3.39* D H – N-Ac- – 22 368 – –
piperazine
3.40 D Me – NMe2 – 0.2 3.2 1.6 3.0/–
(continued)
Table 2.3 (continued)

Compounds for in vivo In vivo mitotic arrest EC90
PK/PD Cl (min/min/kg) t½ (h) Vdss (L/kg) (nM)
3.22 16.5 3.2 2.8 1,425
3.28 11.1 9.0 6.3 38
3.34 14 5.6 – 100
3.40 – – – 67
MDR ratio determined using KB-3-1/KB-V1 cell lines [38]. Compiled from references with
permission [35–39]
(–) not available, * racemic
This SAR knowledge was applied to develop an alternative scaffold based upon
ring-fused chromenpyrazole HTS hits structurally related to the 4,5-dihydropyrazole
lead series [38]. Substituting the dihydropyrrole core present in 3.24 and 3.25 with
the dihydropyrazolobenzoxazine scaffold (Core C) produced 3.29 and 3.30. This
new series showed similar enzyme and cellular potency, but a negligible susceptibil-
ity to Pgp-efflux. Replacement of the fluorine of the western phenyl with small
groups had limited influence on potency and Pgp-efflux, but had a detrimental effect
on the hERG profile (3.30–3.34 in Table 2.3). Attenuation of the pKa of the
aminopropyl motif with an N-acetylpiperazine reduced hERG binding in this series
(3.34).
Excision of the central methyleneoxy constraint from Core C afforded a new
series of 1,4-diaryl-4,5-dihydropyrazole inhibitors (Core D) [39]. This fourth class
of compounds had a kinesin inhibitory potency similar to that of the dihydropyr-
azolobenzoxazines, but reduced cellular potency. Specifically, 3.35 maintained the
intrinsic potency of 3.34, but showed a 16-fold loss in cellular potency. In this series,
exchanging the 5-fluoro atom in the western ring with other small substituents pro-
duced an increase in cellular potency, whereas its removal was detrimental to activ-
ity in both assays (3.36–3.39, Table 2.3). Replacement of the N-acetylpiperazine
with other tertiary amines allowed cellular activity to be regained, with a dimethyl-
aminopropyl side chain optimal in this class. In particular, 3.40 showed excellent
in vitro potency and a cellular potency comparable to that of optimised compounds
from the other series. Importantly, although 3.40 showed a high binding affinity for
the hERG channel, its potent in vitro activity produced a better therapeutic window
(hERG IC50/Eg5 IC50 = 15,000) compared to 3.34.
In vivo PK profiling of the optimised compounds from each category was per-
formed in dogs (Table 2.3) and in vivo efficacy was evaluated in A2780 xenograft
tumour model, as for 3.16. Dihydropyrrole 3.22 showed unexpectedly low in vivo
efficacy, which was attributed to non-specific plasma protein binding. The dihydro-
pyrazole 3.28 showed superior PK properties and excellent in vivo efficacy.
2.3.3 Development of the C-2 Hydroxymethyl Dihydropyrrole

Series
While working to overcome limitations in in vivo efficacy and affinity for the hERG
channel, the researchers at Merck revisited the simpler dihydropyrrole scaffold of
3.13 (Table 2.1) [25]. During lead optimisation, they had noted that a hydroxy-
methyl substituent was a viable alternative to the solubilising propylamine side
chain introduced later in the program which did not display the increased propensity
for hERG channel interactions. In order to increase the potency and aqueous solu-
bility of 3.13, efforts were directed at introducing a water solubilising amine group
in the southern region of the molecule. Earlier work had already led to compound
3.41, in which one methyl group of the urea in 3.13 was replaced with a
N-methylpiperidine moiety, which had significantly improved enzymatic potency
(15-fold increase compared to 3.13) and aqueous solubility (>10 mg/mL, pH 5), but
increased potential for hERG channel interactions [25, 33]. Introduction of the polar
hydroxymethyl substituent at the 2-position of the dihydropyrrole ring afforded
3.42 (log P = 1.7), which demonstrated had an 11-fold decrease in hERG binding
compared to 3.41 (log P = 2.5) (Table 2.4) [25]. Structural studies also confirmed
that this motif interacted with the backbone carbonyl oxygen of Glu117 (Fig. 2.3b).
Unfortunately, while 3.41 was not a substrate for Pgp efflux, 3.42 had a higher
MDR ratio in the matched KB-V1/KB-3-1 isogenic cell lines [30]. Efforts were
therefore directed to decrease the pKa of the amine (3.43) by either β-fluorination or
cyclopropylation (3.44–3.47): compounds 3.45 and 3.46 possessed the best profiles
in terms of Eg5 potency and MDR ratio and were selected for further characteriza-
tion. However, studies were suspended when 3.46 demonstrated time-dependent
inhibition of CYP 3A4, attributed to metabolically-unstable cyclopropylamino
intermediates [40] and when acute toxicity in rats were observed with 3.45 due to
the formation of fluoroacetate [41] via metabolic N-dealkylation and subsequent
oxidation. Modification of amine 3.42 was then attempted by β-fluorination within
the piperidine ring. The superior potency of the cis isomer, designated MK-0731,
meant it was chosen for further in vivo studies.
Drug metabolism and pharmacokinetic properties (DMPK) of MK-0731 were
studied in rat, dog and monkey, where a moderate-to-high clearance was observed
and extrapolation to human metabolic turnover was predicted to be slower with a
low-to-moderate clearance [25]. MK-0731 showed no acute toxicity in rat, no QTc
prolongation in dog, selectivity amongst a selection of kinesins at 50 µM, and very
good selectivity against 160 enzyme, receptors and transporters. Evaluation of anti-
tumour activity in mice with tumours developed from injected A2780 cells demon-
strated MK-0731 inhibited tumour growth more effectively than paclitaxel
administered on an optimised schedule. Significantly, MK-0731 also inhibited
tumour growth in more rigorous tumour models developed from Pgp-overexpressing
KB-V1 and paclitaxel resistant PTX10 cell lines. Based on its favourable efficacy,
pharmacokinetic and safety profile, MK-0731 entered into clinical trials against
solid tumours, where in agreement with preclinical models the dose limiting toxic-
ity was neutropenia [25, 42]. It was shown to be well tolerated, but no objective
responses were recorded, and development has not been continued.
Table 2.4 Lead optimisation of the C-2 hydroxymethyl dihydropyrrole series
F F F
F F F
HO HO
4 3
H
5
N 2 N N
N O N O N O
Ra
Re
N N N
R
3.41 3.42 R= Me 3.48 Ra= H; Re = F
3.43 R= H MK-0731 Ra= F; Re= H
3.44 R= CH2cPr
3.45 R= CH2CH2F
3.46 R= cPr
3.47 R= CH2CHF2
MT Eg5 IC50 A2780 EC50 MDR

Compound pKa (nM) (nM) ratio hERG IC50 (µM)
3.41 – 4 – – 1.2
3.42 8.8 6.2 – 21.2 14.6
3.43 9.8 7.4 – 345 19.2
3.44 9.1 5.1 – 29.9 –
3.45 7.6 5.0 – 4.5 13.8
3.46 7.5 5.9 – 1.2 15.2
3.47 4.9 34.1 – 1.2 –
3.48 6.6 11.5 16.5 2.4 36.9
MK-0731 7.6 2.2 5.3 4.5 20.5
MDR ratio determined in matched KB-3-1/KB-V1 cell lines. Compiled from reference with
permission [25]
(–) not available
2.4 ARRY-520
A promising clinical candidate structurally related to MK-0731 is ARRY-520 (also

known as filanesib), discovered by researchers at Array BioPharma. ARRY-520 can
be considered an analogue of 3.23 (Table 2.3), with the dihydropyrazole core
switched to a dihydrothiadiazole, and an N,O-dimethyl hydroxamate group in the
southern portion of the molecule. Although the optimisation process of ARRY-520
has not been published in detail, it began with the identification of a moderately
potent series of pyrrolinone analogues (e.g. 4.1, enzymatic IC50 = 2.7 µM, Fig. 2.5)
[43]. Switching to an oxadiazoline core produced a marked increase in potency (e.g.
4.2), however the aqueous stability of these leads at physiologically relevant pHs
Fig. 2.5 Evolution of ARRY-520 from the initial pyrrolinone hit 4.1 [43]
prompted a switch to the dihydrothiadiazole core. Wide variations in the amide

substituent were found to be permissible during lead optimisation, with small alkyl
and ether groups giving acceptable potency (e.g. 4.3). However, the ether moiety in
4.3 was found to be metabolized to the more rapidly cleared hydroxyl derivative 4.4,
compromising in vivo activity in an HT-29 (colon cancer) xenograft model, and was
thus replaced with a more resilient hydroxamate, leading to ARRY-520.
ARRY-520 inhibits the MT-stimulated Eg5 ATPase activity with a reported IC50
of 6 nM [44] and exhibits favourable “drug-like” properties. No CYP inhibition is
reported, good aqueous solubility is observed at pH 7.4 (4,400 µg/ml) and the log
D7.4 = 1.3 for ARRY-520 lies within the perceived “optimal” range of 1–3 [45].
Selectivity is seen against panels of 8 kinesins (>100 µM), 225 kinases (>10 µM)
and 27 assorted receptors (>10 µM) [43, 46], and this compound demonstrates
excellent efficacy in pre-clinical in vitro cell-based assays, with cell proliferation
EC50 values <10 nM across multiple tumour cell lines. ARRY-520 retains anti-
proliferative activity <15 nM in the MDR Pgp overexpressing tumour cell lines
HCT-15, ovarian NCI/ADR-RES and K562/ADR [44]. When the in vivo activity of
ARRY-520 was evaluated in a histologically diverse panel of 16 tumour xenograft
models, partial or complete responses were recorded across 13 of the models.
Haematological cancers were particularly sensitive, with a 100 % complete response
rate in multiple myeloma (MM), AML and acute promyelocytic leukemia, with
dependence on the anti-apoptotic protein myeloid cell leukemia 1 (MCL-1) identi-
fied as a prospective biomarker for susceptibility to Eg5 inhibition [47]. ARRY-520
has advanced into clinical trials against haematological malignancies, and indica-
tions from phase 1 and phase 2 studies suggest it can provide clinical benefits in
heavily pre-treated patients with relapsed and refractory multiple myeloma
(RRMM), both as a monotherapy, and in combination with existing therapies [48–
50]. In a phase 2 RRMM study, ARRY-520 achieved an overall response rate (ORR)
of 16 % (5/32 patients) and a clinical benefit rate (CBR; minimal response or
greater) of 22 % (7/32 patients) when administered as a monotherapy; this increased
to 24 % and 33 % respectively in the patient subpopulation with low levels of the
acute-phase serum protein α-1-acid glycoprotein (AAG) [48]. Results from a phase
1 study with the proteasome inhibitor carfilzomib suggest an even higher clinical
benefit rate may be achieved when used in combination chemotherapies
(ORR = 37 %; CBR = 63 % in a study with 19 patients) [49]. Further expansion into
phase 3 trials is now planned [51].
2.5 EMD 534085
In 2010, researchers at Merck-Serono reported the development of their clinical

candidate EMD 534085, a hexahydropyranoquinoline (HHPQ) analogue with
potent selective Eg5 inhibitory activity that arrested mitosis, induced apoptosis and
had significant anti-tumour activity in diverse, solid tumour cell lines, including
colon, breast and cervical cancers, as well as lymphomas [26]. A biochemical HTS
of their in-house library based on an Eg5-ATPase assay in the presence of MTs
identified 40 HHPQ hits. The screening results indicated the substituent at position
9 was important for activity, and initial SAR studies found that potency was reduced
upon moving to smaller alkyl groups from the initial t-butyl hits (Table 2.5; com-
pounds 5.1–5.9). Potency was retained with a CF3 group, and subsequent X-ray
crystallographic studies showed this group fitted perfectly into a hydrophobic
pocket in the allosteric binding site flanked by the side chains of Leu160 and
Phe239, with the CF3 probably forming pseudo hydrogen-bond interactions with the
guanidine head group of Arg221 at the base of the pocket (Fig. 2.3c).
In position 5 of the HHPQ scaffold, an aromatic moiety was crucial for activity,
as demonstrated by inactive compounds 5.10 and 5.11, with the 5-phenyl ring occu-
pying a similar position to the phenolic ring in monastrol. Indeed, the phenolic
derivatives 5.12 and 5.13 improved potency, whilst larger substituents reduced
activity, indicating limited steric freedom in this region of the binding site. The
hydroxyl substituent was not included in the optimisation process because of its
poor PK properties, presumably due to high phase II conjugation, while the potent
thiophene analogue 5.14 was also abandoned due to PK concerns.
As with many lead optimisation programmes, increasing potency through
promoting the lipophilic interactions at positions 5 and 9 led to poor aqueous solu-
bility, poor PK properties and metabolic liabilities. The orientation of the tetrahy-
dropyran ring towards the bulk solvent (Fig. 2.3c) offered the potential to introduce
water solubilising moieties to balance the lipophilic character, and was achieved at
position 2 by appending a methylamine group and functionalising with various sul-
fonamides (e.g. 5.17), ureas (e.g. EMD 534085) and carbamates (e.g. 5.18).
Fortuitously, expansion into the aqueous region not only improved solubility, but
also increased the potency and metabolic stability of the lead series, which allowed
in vivo PK studies to commence. EMD 534085 was selected for PK profiling in
mice, presumably because the in vitro metabolic data and physicochemical proper-
ties were more favourable than some of the more potent anti-proliferative com-
pounds. In mice PK studies, EMD 534085 exhibited a relatively short half-life
Table 2.5 Lead optimisation of the HHPQ scaffold to EMD 534085
MT Eg5 IC50 HCT116 EC50

Compound R1 R2 R3 (nM) (nM)
5.1 H Ph – >10,000 n.d.
5.2 Me Ph – 420 840
5.3 Et Ph – 130 370
5.4 i-Pr Ph – 130 170
5.5 t-Bu Ph – 120 250
5.6 CF3 Ph – 110 280
5.7 NMe2 Ph – 5,300 8,600
5.8 COOH Ph – >10,000 n.d.
5.9 CH2OH Ph – >10,000 >10,000
5.10 t-Bu c-Hexyl – >10,000 n.d.
5.11 t-Bu (CH2)4OH – >10,000 n.d.
5.12 t-Bu 3-OH Ph – 40 40
5.13 t-Bu 4-OH Ph – 34 32
5.14 t-Bu 2-Thiophenyl – 25 20
5.15 CF3 Ph H 20 4
5.16 t-Bu Ph Me 4 5
5.17 CF3 Ph SO2(CH2)2NHMe 4 12
5.18 CF3 Ph COO(CH2)2NMe2 14 110
EMD CF3 Ph CONH(CH2)2NMe2 8 30
534085
Adapted from reference with permission [26]
n.d. not determined
(t½ = 2.5 h), moderate clearance (30 mL/min/kg), a relatively high volume of distri-
bution of 7.4 L/kg and a bioavailability of >50 %, and was subsequently taken for-
ward into in vivo efficacy studies in a subcutaneously grown human xenograft
Colo205 colon mouse model. EMD 534085 completely inhibited tumour growth in
the higher dose group (30 mg/kg i.p.) and reduced the growth rate in the lower dose
cohort (15 mg/kg). Consequently, EMD 534085 advanced into clinical trials in
patients with refractory solid tumours, Hodgkin’s lymphoma or non-Hodgkin’s
lymphoma. However following limited activity, coupled with the poor results from
clinical trials of other Eg5 inhibitors, development was ceased [52].
2.6 S-Trityl L-Cysteine and Related Inhibitors
The origins of S-trityl L-cysteine (STLC, Table 2.6) as an anticancer agent began
over 50 years ago when following the observation that leukemic white blood cells
incorporated radiolabelled cysteine [53], various S-alkylated derivatives of cysteine
were prepared as prospective chemotherapeutics [54, 55]. STLC was later found to
exhibit anti-tumour activity in vivo in murine leukemia models during testing at the
National Cancer Institute (NCI) in the 1960s [56]. It was not until 2004, during the
Table 2.6 Selected analogues from the STLC development programmea
Basal Eg5 MT stimulated

inibition Kiapp Eg5 inhibition K562
Compound X R1 R2 R3 (nM)b Kiapp (nM) GI50 (nM)
STLC S NH2 (R)-CO2H H 135.9 129.0 1,452
6.1 S NMe2 (R)-CO2H H n.i. n.d. n.d.
6.2 S NH2 H H 245.0 152.3 1,791
6.3 S OH H H n.i. n.d. n.d.
6.4 S NH2 (R)-CO2H 4-Me <50 13.6 242
6.5 S NH2 (R)-CO2H 4-OMe <50 9.1 240
6.6 S NH2 H 3-Me 80.0 n.d. 698
6.7 S NH2 H 3-Et 5.9 n.d. 680
6.8 S NH2 (R)-CO2H 3,4-Me 1.2 18.6 72
6.9 S NH2 (R)-CO2H 3-Me, 4-Et 4.6 4.9 34
6.10 S NH2 H 3-OH 200.3 n.d. 555
6.11 S NH2 H 3-CONH2 419.7 n.d. 802
6.12 C NH2 (R)-CO2H H 173.5 211.3 776
6.13 C NH2 H 3-Me 8.8 10.8 200
6.14 C NH2 H 4-OMe 15.5 10.1 83
6.15 C NH2 (R)-CO2H 3-Me 12.2 13.7 98
6.16 C NH2 CO2H 3,4-Me 12.4 9.9 23
Compiled from references with permission [15, 19, 54, 67]
n.d. not determined, n.i. no inhibition
a
The minimum pharmacophore for inhibition is indicated in black, where: X = S or C; R1 = NH2;
R2 = H and ≥Pr is present for the third phenyl ring
b
Estimates of the Kiapp values vary depending on the buffer and test conditions used, due to the tight
binding nature of STLC and related analogues [66]
biochemical screening of compounds from the NCI against Eg5, that the Kozielski
group identified STLC as an Eg5 inhibitor [57].
STLC is a tight binding allosteric loop L5 inhibitor that inhibits the in vitro basal
ATPase activity of Eg5 with Kiapp ≈ 150 nM, and in HeLa cells induces the pheno-
typic response of a monopolar mitotic spindle indicative of Eg5 inhibition [58].
Against a panel of human kinesins (Kif5B, MKLP-1, MKLP-2, CENP-E, Kif22,
Kif2A, MCAK and KifC1), STLC proved selective for Eg5 [56, 58] and this
compound has broad spectrum activity in the NCI60 panel of cancer cell lines [57].
As with other Eg5 inhibitors, docetaxel-resistant prostate cancer cell lines remain
sensitive to STLC treatment [59], and in vivo anti-tumour activity has been demon-
strated in various xenograft models [56, 60]. How STLC induces cell death has been
elucidated in detail, with apoptosis occurring primarily via activation of the spindle
checkpoint and the intrinsic apoptotic pathway [61].
Early SAR profiling (Table 2.6) identified the primary amine as essential for
inhibition, with activity abrogated in secondary and tertiary amine derivatives (e.g.
6.1), or by replacement with a hydroxyl (6.3), whereas the carboxylic acid was not
an absolute requirement (6.2) [58]. Incorporation of a lipophilic para- substituent
on one phenyl ring afforded analogues with improved activity in HeLa growth inhi-
bition assays (e.g. 6.4 and 6.5) [58, 62], and inhibitory activity dropped markedly on
removal of a phenyl ring unless an alkyl substituent at least the size of a propyl
group was introduced [58].
The crystal structure of the Eg5 · STLC · ADP complex [63, 64] showed two of
the trityl group rings were in hydrophobic pockets, while the third π-stacked with
Tyr211 (Fig. 2.3d). The cysteine moiety extends towards solvent forming a network
of hydrogen-bond interactions with the protein, with the SAR indicating that the
hydrogen bonds formed by the primary amine are more important than in ispinesib.
Molecular dynamics simulations have verified the importance of the interactions of
the amine to the overall free energy of binding, and found the carboxylic acid is
partially repelled by the anionic diphosphate moiety of ADP [65].
Optimisation by structure-based drug design produced a series of meta- and
para-alkyl substituted inhibitors (e.g. 6.6–6.9), with enhanced affinity attributed to
strengthening the CH-π interaction between Leu214 and the phenyl ring (e.g. 6.7,
Kiapp = 5.9 nM) [15, 21]. Replacing one phenyl ring with benzyl to extend further
into this region delivered analogues which retained potency against Eg5, but were
less effective in cellular assays, probably due to reduced solubility [65]. Phenolic
derivatives mimicking the hydrogen-bond interaction with Glu218 observed in
monastrol also gave improved affinity (e.g. 6.10), but were not pursued due to rapid
clearance in hepatocyte assays [21]. More polar phenyl substituents or a heteroaro-
matic in place of the solvent exposed phenyl proximal to Tyr211 were introduced to
lower log P, but were less successful, and delivered only moderately active inhibi-
tors (e.g. 6.11) [15, 21]. Replacing the sulfur atom with either O or N yielded weak
inhibitors which were ineffective in cell-based assays due to reduced stability, but
replacement with CH2 gave inhibitors with greater potency in cellular assays (6.12)
[21] and improved GI50 values across multiple cancer cell lines versus their cysteine
counterparts (6.13–6.16) [15].
In addition to their improved potency, these triphenylbutanamine inhibitors pos-

sessed favourable drug-like properties, including good aqueous solubility, no inhi-
bition of major CYP isoforms and no interaction with the hERG channel. The
original sulfur based series, in comparison, proved to be moderate inhibitors of
certain CYP isoforms (e.g. CYP 2C19 inhibition: 6.5 = 6.3 µM; 6.8 = 7.9 µM) and in
general were less metabolically resilient in vitro [15]. Furthermore, triphenylbu-
tanamine analogues 6.13 and 6.14 lacking the carboxylic acid possessed several
toxicological and off-target liabilities, including inhibition of the hERG channel
several CYP isoforms and greatly reduced bioavailability (e.g. rac-6.15, F = 51 %;
6.13: F = 4 %) [15, 21]. In vivo evaluation in mice with explanted LXFS 538 lung
cancer tumours found the cysteine based 6.8 had no effect, and treatment with 6.9
only delayed tumour progression, due to their rapid clearance. Significantly, the
triphenylbutanamines induced more robust and sustained responses, with both 6.15
and 6.16 affording complete tumour regression. This positive in vivo result com-
bined with their favourable drug-like properties highlights the excellent potential of
this series for continued development.
2.7 Future Prospects
So what does the future hold for Eg5 inhibitors within the context of drug discovery
and development? Their development represents a classical reductionist approach to
drug-discovery, which has inevitably produced mixed results when evaluated in
terms of clinical success. The availability of suitable biochemical assays enabled
identification of diverse scaffolds through HTS, and combined with the amenability
of Eg5 to crystallisation, further accelerated lead optimisation to give highly potent,
selective inhibitors within the kinesin family. Our case studies clearly show that
given a druggable target, a fit-for-purpose assay cascade that includes assessment of
drug-like properties in parallel with potency, cellular pharmacodynamic and pheno-
type readouts, a lead optimisation team will deliver a clinical candidate given suffi-
cient resource. However, successful drug development is not just about resource; a
knowledge-based approach is essential to address the considerable hurdles encoun-
tered during a hit-to-candidate optimisation process. Iterations towards potency must
be achieved within an integrated platform that includes optimisation of DMPK and
physical chemistry parameters (so called ‘drug-like properties’) to address solubility
issues early on in the programme, and to navigate off-target issues such as hERG
interactions, Pgp-efflux and CYP450 inhibition. This latter point is particularly
important in oncology as drug treatment becomes more sophisticated and complex
through combination-based regimes, which raises the significance of issues such as
drug-interactions through CYP450. To increase the chances of developing effective
candidates, the selection of rigorous in vivo models should also be considered. The
candidates described here that were evaluated in more advanced in vivo systems,
delivered more successful outcomes when translated into clinical trials (vide infra).
Through this review, we have described the key drivers that guide lead optimisa-
tion from hit to clinical candidate specifically related to Eg5. All drug discovery and
development programmes have a certain generic process associated with them to
balance pharmacodynamic and pharmacokinetic requirements. However, it is essen-
tial to appreciate the paramount importance of demonstrating that target engage-
ment and the desired clinical output are intrinsically related. This is often unique to
each programme and can be particularly challenging. The greatest source of attri-
tion in modern drug development is not during the preclinical phase [68, 69], but
once clinical trials in patients have started: this is when the validity of the target can
be critically evaluated in its true clinical context. The majority of Eg5 inhibitors
assessed so far have had very limited success in achieving objective responses in
patients [16, 70]. Pharmacodynamic responses have been noted during treatment,
which suggests target engagement, but it would appear that for the majority of can-
cers, Eg5 inhibition is not an appropriate single therapy. Considering we do not
fully understand how MT-interfering drugs kill cancer cells, which highlights the
complexity of the underlying biology, it is perhaps not surprising that Eg5 inhibition
alone has proved insufficient therapeutically [71]. The reasons for this are open to
interpretation, but may include target redundancy (e.g. through Kif15 supplanting
Eg5 during spindle assembly) [72], the lower replication rate of solid tumours ver-
sus haematological malignancies, and artificially high rates of mitosis in pre-clinical
models [70]. The success of ARRY-520 relative to other Eg5 inhibitors in achieving
objective responses in RRMM correlates with these hypotheses. Specifically, we
attribute the relative success of this particular candidate to a number of reasons: (i)
its better physicochemical properties compared with the other compound classes;
(ii) the selection and use of appropriate and robust preclinical in vivo models; and
(iii) the identification of appropriate biomarkers enabling effective patient stratifica-
tion. Firstly, the lower log D of ARRY-520 is a factor that is known to impact hugely
on a large number of ADME properties [45]. Secondly, during pre-clinical develop-
ment, ARRY-520 (in addition to ispinesib, SB-743921 and STLC derivatives) was
evaluated in explanted tumour xenograft models, which retain the histological com-
plexity from the patient and reflect prior treatment, and successfully induced com-
plete regression. In comparison, according to the published data, EMD-534085 was
evaluated in models derived from implanted cancer cell lines, which lack this com-
plexity, and are known to be less effective predictors of clinical outcomes [73, 74].
Finally, during the optimisation of ARRY-520, appropriate biomarkers were identi-
fied in pre-clinical models, which allowed more effective targeting of patient popu-
lations for clinical trials design [47, 48]. These last two points emphasise the need
for these potential therapeutic agents to be evaluated much more rigorously in
advanced, complex pre-clinical models of cancer that demonstrate target engage-
ment and functionally related outputs to improve survival in phase II clinical trials,
as succinctly demonstrated in Morgan and Van Der Graafs’ recent analysis of the
reasons for attrition in drug development programmes [75].
The reasonable responses to ARRY-520 in patients with intractable multiple
myeloma [48], which are to be followed up by further clinical trials with existing
therapies, suggest that Eg5 inhibitors may be most efficacious when used in combi-
nation [49, 50]. We propose that this combinatorial approach has the best chance of
achieving therapeutic success.
Acknowledgments We apologise to authors whose work we were unable to include due to limita-
tions of space. We thank Prof. Frank Kozielski for helpful comments on the manuscript. We are
grateful to Cancer Research UK for supporting the STLC programme and funding the postdoctoral
positions of NGA and GB on the Small Molecule Drug Discovery Programme, in association with
Prostate Cancer UK. JADG thanks the Umeå Centre for Microbial Research for funding his post-
doctoral research at Umeå University.
Copyright Acknowledgements: Excerpts from this chapter appeared previously in the doctoral
thesis of James A. D. Good [76]. The data appearing in Tables 2.1, 2.4 and 2.6 was adapted with
permission from the cited references and is copyright American Chemical Society [15, 21, 25, 28,
58, 67]. The date appearing in Tables 2.2, 2.3 and 2.5 was adapted with permission from the cited
references and is copyright Elsevier [26, 31, 33–39].
References
1. Mayer TU et al (1999) Small molecule inhibitor of mitotic spindle bipolarity identified in a

phenotype-based screen. Science 286:971–974
2. Zhu C et al (2005) Functional analysis of human microtubule-based motor proteins, the kine-
sins and dyneins, in mitosis/cytokinesis using RNA interference. Mol Biol Cell
16:3187–3199
3. Maliga Z, Kapoor TM, Mitchison TJ (2002) Evidence that monastrol is an allosteric inhibitor
of the mitotic kinesin Eg5. Chem Biol 9:989–996
4. Yan Y et al (2004) Inhibition of a mitotic motor protein: where, how, and conformational con-
sequences. J Mol Biol 335:547–554
5. Bergnes G, Brejc K, Belmont L (2005) Mitotic kinesins: prospects for antimitotic drug discov-
ery. Curr Top Med Chem 5:127–145
6. Knight SD, Parrish CA (2008) Recent progress in the identification and clinical evaluation of
inhibitors of the mitotic kinesin KSP. Curr Top Med Chem 8:888–904
7. Jiang C, You Q (2013) Kinesin spindle protein inhibitors in cancer: a patent review (2008 –
present). Expert Opin Ther Pat 23:1547–1560
8. Sakowicz R et al (2004) Antitumor activity of a kinesin inhibitor. Cancer Res 64:3276–3280
9. Johnson RK et al (2002) SB-715992, a potent and selective inhibitor of the mitotic kinesin
KSP, demonstrates broad-spectrum activity in advanced murine tumors and human tumor
xenografts. Proc Annu Meet Am Assoc Cancer Res 43:269
10. Chu Q et al (2003) A phase I study to determine the safety and pharmacokinetics of IV admin-
istered SB-715992, a novel kinesin spindle protein (KSP) inhibitor, in patients with solid
tumors. Proc Am Soc Clin Oncol 22:525
11. Lad L et al (2008) Mechanism of inhibition of human KSP by Ispinesib. Biochemistry
47:3576–3585
12. Zhang B, Liu J-F, Xu Y, Ng S-C (2008) Crystal structure of HsEg5 in complex with clinical
candidate CK0238273 provides insight into inhibitory mechanism, potency, and specificity.
Biochem Biophys Res Commun 372:565–570
13. Talapatra SK, Schuttelkopf AW, Kozielski F (2012) The structure of the ternary Eg5-ADP-
ispinesib complex. Acta Crystallogr D 68:1311–1319
14. Carol H et al (2009) Initial testing (stage 1) of the kinesin spindle protein inhibitor ispinesib by
the pediatric preclinical testing program. Pediatr Blood Cancer 53:1255–1263
15. Good JAD et al (2013) Optimized S-trityl-L-cysteine-based inhibitors of kinesin spindle pro-
tein with potent in vivo antitumor activity in lung cancer xenograft models. J Med Chem
56:1878–1893
17. Matsuno K, Sawada J, Asai A (2008) Therapeutic potential of mitotic kinesin inhibitors in
cancer. Expert Opin Ther Pat 18:253–274
18. Bergnes G et al (2002) Mitotic kinesin-targeted antitumor agents: discovery, lead optimization
and anti-tumor activity of a series of novel quinazolinones as inhibitors of kinesin spindle
protein (KSP). Abstr Pap Am Chem Soc 223:B140
19. Holland JP, Jones MW, Cohrs S, Schibli R, Fischer E (2013) Fluorinated quinazolinones as
potential radiotracers for imaging kinesin spindle protein expression. Bioorg Med Chem
21:496–507
20. Bergnes G et al (2002) Mitotic kinesin-targeted antitumor agents: discovery, lead optimization
and anti-tumor activity of a series of novel quinazolinones as inhibitors of kinesin spindle
protein (KSP). Proc Annu Meet Am Assoc Cancer Res 43:736
21. Wang F et al (2012) Triphenylbutanamines: kinesin spindle protein inhibitors with in vivo
antitumor activity. J Med Chem 55:1511–1525
22. Ansbro MR, Shukla S, Ambudkar SV, Yuspa SH, Li L (2013) Screening compounds with a
novel high-throughput ABCB1-mediated efflux assay identifies drugs with known therapeutic
targets at risk for multidrug resistance interference. PLoS One 8:e60334
23. Jackson JR et al (2006) A second generation KSP inhibitor, SB-743921, is a highly potent and
active therapeutic in preclinical models of cancer. AACR Meet Abstr 2006:B11
24. Holen K et al (2011) A first in human study of SB-743921, a kinesin spindle protein inhibitor,
to determine pharmacokinetics, biologic effects and establish a recommended phase II dose.
Cancer Chemother Pharmacol 67:447–454
25. Cox CD et al (2008) Kinesin spindle protein (KSP) inhibitors. 9. Discovery of (2S)-4-
(2,5-difluorophenyl)-N-[(3R,4S)-3-fluoro-1-methylpiperidin-4-yl]-2-(hydroxymethyl)-N-met
hyl-2-phenyl-2,5-dihydro-1H-pyrrole-1-carboxamide (MK-0731) for the treatment of taxane-
refractory cancer. J Med Chem 51:4239–4252
26. Schiemann K et al (2010) The discovery and optimization of hexahydro-2H-pyrano[3,2-c]
quinolines (HHPQs) as potent and selective inhibitors of the mitotic kinesin-5. Bioorg Med
Chem Lett 20:1491–1495
27. Kim ED et al (2010) Allosteric drug discrimination is coupled to mechanochemical changes in
the kinesin-5 motor core. J Biol Chem 285:18650–18661
28. Theoclitou M-E et al (2011) Discovery of (+)-N-(3-Aminopropyl)-N-[1-(5-benzyl-3-methyl-
4-oxo-[1,2]thiazolo[5,4-d]pyrimidin-6-yl)-2-methylpropyl]-4-methylbenzamide (AZD4877),
a kinesin spindle protein inhibitor and potential anticancer agent. J Med Chem 54:6734–6750
29. Kantarjian H et al (2012) Phase I/II multicenter study to assess the safety, tolerability, pharma-
cokinetics and pharmacodynamics of AZD4877 in patients with refractory acute myeloid leu-
kemia. Invest New Drugs 30:1107–1115
30. Cox CD, Garbaccio RM (2010) Discovery of allosteric inhibitors of kinesin spindle protein
(KSP) for the treatment of taxane-refractory cancer: MK-0731 and analogs. Anticancer Agents
Med Chem 10:697–712
31. Cox CD et al (2005) Kinesin spindle protein (KSP) inhibitors. Part 1: the discovery of
3,5-diaryl-4,5-dihydropyrazoles as potent and selective inhibitors of the mitotic kinesin
KSP. Bioorg Med Chem Lett 15:2041–2045
32. Bissantz C, Kuhn B, Stahl M (2010) A medicinal chemist’s guide to molecular interactions.
J Med Chem 53:5061–5084
33. Fraley ME et al (2006) Kinesin spindle protein (KSP) inhibitors. Part 2: The design, synthesis,
and characterization of 2,4-diaryl-2,5-dihydropyrrole inhibitors of the mitotic kinesin
34. Garbaccio RM et al (2006) Kinesin spindle protein (KSP) inhibitors. Part 3: Synthesis and
evaluation of phenolic 2,4-diaryl-2,5-dihydropyrroles with reduced hERG binding and
employment of a phosphate prodrug strategy for aqueous solubility. Bioorg Med Chem Lett
16:1780–1783
35. Cox CD et al (2006) Kinesin spindle protein (KSP) inhibitors. Part 4: Structure-based design
of 5-alkylamino-3,5-diaryl-4,5-dihydropyrazoles as potent, water-soluble inhibitors of the
mitotic kinesin KSP. Bioorg Med Chem Lett 16:3175–3179
36. Cox CD et al (2007) Kinesin spindle protein (KSP) inhibitors. Part V: Discovery of
2-propylamino-2,4-diaryl-2,5-dihydropyrroles as potent, water-soluble KSP inhibitors, and
modulation of their basicity by β-fluorination to overcome cellular efflux by P-glycoprotein.
Bioorg Med Chem Lett 17:2697–2702
37. Coleman PJ et al (2007) Kinesin spindle protein (KSP) inhibitors. Part 6: Design and synthesis
of 3,5-diaryl-4,5-dihydropyrazole amides as potent inhibitors of the mitotic kinesin
38. Garbaccio RM et al (2007) Kinesin spindle protein (KSP) inhibitors. Part 7: Design and syn-
thesis of 3,3-disubstituted dihydropyrazolobenzoxazines as potent inhibitors of the mitotic
kinesin KSP. Bioorg Med Chem Lett 17:5671–5676
39. Roecker AJ et al (2007) Kinesin spindle protein (KSP) inhibitors. Part 8: Design and synthesis
of 1,4-diaryl-4,5-dihydropyrazoles as potent inhibitors of the mitotic kinesin KSP. Bioorg Med
Chem Lett 17:5677–5682
40. Cerny MA, Hanzlik RP (2005) Cyclopropylamine inactivation of cytochromes P450: Role of
metabolic intermediate complexes. Arch Biochem Biophys 436:265–275
41. Goncharov NV, Jenkins RO, Radilov AS (2006) Toxicology of fluoroacetate: a review, with
possible directions for therapy research. J Appl Toxicol 26:148–161
42. Holen K et al (2012) A phase I trial of MK-0731, a kinesin spindle protein (KSP) inhibitor, in
patients with solid tumors. Invest New Drugs 30:1088–1095
43. Allen S et al (2012) The discovery and optimization of kinesin spindle protein (KSP) inhibi-
tors: path to ARRY-520. Cambridge Healthtech Institute conference, 4 June 2012. http://array-
biopharma.com/files/1713/9810/7999/PubAttachment526.pdf
44. Woessner R et al (2009) ARRY-520, a novel KSP inhibitor with potent activity in hematologi-
cal and taxane-resistant tumor models. Anticancer Res 29:4373–4380
45. Waring MJ (2010) Lipophilicity in drug discovery. Expert Opin Drug Discov 5:235–248
46. Lemieux C et al (2007) ARRY-520, a novel, highly selective KSP inhibitor with potent anti-
proliferative activity. Proc Am Assoc Cancer Res 48:5590
47. Tunquist BJ, Woessner RD, Walker DH (2010) Mcl-1 stability determines mitotic cell fate of
human multiple myeloma tumor cells treated with the kinesin spindle protein inhibitor ARRY-
520. Mol Cancer Ther 9:2046–2056
48. Lonial S et al (2013) Prolonged survival and improved response rates with ARRY-520 in
relapsed/refractory multiple myeloma (RRMM) patients with low α-1 acid glycoprotein
(AAG) levels: results from a phase 2 study. ASH Annu Meet Abstr 122:285
49. Shah JJ et al (2013) Phase 1 study of the novel kinesin spindle protein inhibitor ARRY-
520 + carfilzomib (car) in patients with relapsed and/or refractory multiple myeloma (RRMM).
ASH Annu Meet Abstr 122:1982
50. Chari A et al (2013) A phase 1 study of ARRY-520 with bortezomib (BTZ) and dexamethasone
(dex) in relapsed or refractory multiple myeloma (RRMM). ASH Annu Meet Abstr 122:1938
51. Owens B (2013) Kinesin inhibitor marches toward first-in-class pivotal trial. Nat Med 19:1550
52. Hollebecque A et al (2013) A phase I, dose-escalation study of the Eg5-inhibitor EMD
534085 in patients with advanced solid tumors or lymphoma. Invest New Drugs
31:1530–1538
53. Weisberger AS, Levine B (1954) Incorporation of radioactive L-cystine by normal and leuke-
mic leukocytes in vivo. Blood 9:1082–1094
54. Goodman L, Ross LO, Baker BR (1958) Potential anticancer agents. V. Some sulfur-substituted
derivatives of cysteine. J Organ Chem 23:1251–1257
55. Theodoropoulos D (1959) Synthesis of certain S-substituted L-cysteines. Acta Chem Scand
13:383–384
56. Zee-Cheng K-Y, Cheng C-C (1970) Experimental antileukemic agents. Preparation and
structure-activity study of S-tritylcysteine and related compounds. J Med Chem 13:414–418
57. DeBonis S et al (2004) In vitro screening for inhibitors of the human mitotic kinesin Eg5 with
antimitotic and antitumor activities. Mol Cancer Ther 3:1079–1090
58. DeBonis S et al (2008) Structure–activity relationship of S-trityl-L-cysteine analogues as
inhibitors of the human mitotic kinesin Eg5. J Med Chem 51:1115–1125
59. Wiltshire C et al (2010) Docetaxel-resistant prostate cancer cells remain sensitive to S-trityl-l-
cysteine–mediated Eg5 inhibition. Mol Cancer Ther 9:1730–1739
60. Zee-Cheng KY, Cheng CC (1972) Structural modification of S-trityl-L-cysteine. Preparation
of some S-(substituted trityl)-L-cysteines and dipeptides of S-trityl-L-cysteine. J Med Chem
15:13–16
61. Kozielski F et al (2008) Proteome analysis of apoptosis signaling by S-trityl-L-cysteine, a
potent reversible inhibitor of human mitotic kinesin Eg5. Proteomics 8:289–300
62. Ogo N et al (2007) Synthesis and biological evaluation of l-cysteine derivatives as mitotic
kinesin Eg5 inhibitors. Bioorg Med Chem Lett 17:3921–3924
63. Kaan HYK, Ulaganathan V, Hackney DD, Kozielski F (2010) An allosteric transition trapped
in an intermediate state of a new kinesin–inhibitor complex. Biochem J 425:55–60
64. Basso AD et al (2010) SCH 2047069, a novel oral kinesin spindle protein inhibitor, shows
single-agent antitumor activity and enhances the efficacy of chemotherapeutics. Mol Cancer
Ther 9:2993–3002
65. Abualhasan MN et al (2012) Doing the methylene shuffle – further insights into the inhibition
of mitotic kinesin Eg5 with S-trityl L-cysteine. Eur J Med Chem 54:483–498
66. Skoufias DA et al (2006) S-Trityl-L-cysteine is a reversible, tight binding inhibitor of the
human kinesin Eg5 that specifically blocks mitotic progression. J Biol Chem 281:
17559–17569
67. Kaan HYK et al (2011) Structure – activity relationship and multidrug resistance study of new
S-trityl-L-cysteine derivatives as inhibitors of Eg5. J Med Chem 54:1576–1586
68. Arrowsmith J, Miller P (2013) Trial watch: phase II and phase III attrition rates 2011–2012.
Nat Rev Drug Discov 12:569
69. Orloff J et al (2009) The future of drug development: advancing clinical trial design. Nat Rev
Drug Discov 8:949–957
70. Komlodi-Pasztor E, Sackett DL, Fojo AT (2012) Inhibitors targeting mitosis: tales of how great
drugs against a promising target were brought down by a flawed rationale. Clin Cancer Res
18:51–63
71. Mitchison TJ (2012) The proliferation rate paradox in antimitotic chemotherapy. Mol Biol Cell
23:1–6
72. Tanenbaum ME et al (2009) Kif15 cooperates with Eg5 to promote bipolar spindle assembly.
Curr Biol 19:1703–1711
73. Voskoglou-Nomikos T, Pater JL, Seymour L (2003) Clinical predictive value of the in vitro cell
line, human xenograft, and mouse allograft preclinical cancer models. Clin Cancer Res
9:4227–4239
74. Humphries M et al (2012) Abstract 1782: human tumor explants are better predictors of clini-
cal trial outcome than cell line xenografts for the KSP inhibitor ARRY-520. Cancer Res
72:1782
75. Morgan P et al (2012) Can the flow of medicines be improved? Fundamental pharmacokinetic
and pharmacological principles toward improving phase II survival. Drug Discov Today
17:419–424
76. Good JAD (2012) The development of S-trityl L-cysteine based inhibitors of Eg5 as anticancer
chemotherapeutics. PhD thesis, The Beatson Institute for Cancer Research, University of
Glasgow, Glasgow
Chapter 3
Mechanisms of Action of Eg5 Inhibitors
Robert A. Cross
3.1 ATPase Mechanism of Eg5
The kinetic mechanism of Eg5 is broadly similar to the widely studied kinesin-1
mechanism, but with some distinct features (Waitzman and Rice 2013). As with all
molecular motors, the most stable state (the ground state) is the apo state, in which
the active site is empty and the motor binds strongly (stably) and stereospecifically
to its microtubule (MT) track. The subsequent steps of MgATP turnover progres-
sively destabilise MT binding. The initial binding of MgATP or MgAMPPNP (a
non-hydrolysable analog) into the empty active site drives the motor into a new
conformational state, but this new motor·MgADP state binds to MTs only slightly
less stably than the apo (empty) state, so that the motor head can still hold substan-
tial force. The subsequent catalytic steps of hydrolysis and Pi release generate a
motor·MgADP state that tends to detach rapidly from the MT. Rebinding of this
motor·MgADP state to MTs triggers MgADP release, regenerating the apo state. In
the motor·MgADP state, spontaneous release of MgADP from the active site is
extremely slow, ~0.005 s−1, and rate-limiting for ATP turnover in the absence of
MTs. The weak binding motor·MgADP state is a “slip-state” that slides along the
MT almost without friction if force is applied [1]. Nonetheless, this motor·MgADP
slip-state is crucial because it provides access to a transient strong binding
motor·MgADP state from which MgADP release is very fast: around three orders of
magnitude faster than in the absence of MTs. The MT binding surface of the motor
head is on the opposite side of the central beta sheet backbone to the active site, so
that much of the structural mechanism of MT activation is necessarily allosteric. As
I discuss below, the binding of allosteric inhibitors to the Eg5 motor head adds
further stability to the motor·MgADP state and thereby inhibits MT-activated
MgADP release. The crucial point therefore is that inhibitors and MTs have
R.A. Cross (*)

Warwick Medical School, Coventry CV4 7AL, UK
e-mail: r.a.cross@warwick.ac.uk

54 R.A. Cross
opposite allosteric effects, with inhibitors tending to inhibit MgADP release, and
MTs tending to accelerate MgADP release.
3.2 Crystal Structures of Eg5 Motor Domains
To understand how inhibitors influence the Eg5 active site, we need to understand
the conformational cycle of the motor. A recent principal component (PC) analysis
of all available Eg5 crystal structures identifies three main classes of conformers,
nominally ATP-like, ADP-like and Inhibitor-bound [2]. Accelerated molecular
dynamics (aMD) simulations suggest that the motor head is driven to interconvert
between these conformations by exchange of the active site nucleotide. Kinesins
appear to differ in this way from RAS-like small G-proteins, which interconvert
readily between GDP-like and GTP-like conformations in their apo state, with one
or the other of these conformations then being selected by the binding of a specific
active site nucleotide.
Figure 3.1 shows representative crystal structures of the ATP-like and ADP-like
states of Eg5. In the AMPPNP (ATP-like) state (Fig. 3.1a), the switch 1 (Sw1)
region of the active site (orange) is formed into an antiparallel beta sheet and the key
salt bridge linking Sw1 and Sw2 (see below) is closed. On the opposite (MT bind-
ing) face of the molecule, the neck linker is docked and helix H4 (the Sw2 helix) lies
clear of the neck linker dock site. By contrast in the ADP-like state (Fig. 3.1b), the
inter-switch salt bridge is broken, the C-terminal portion of Sw1 is helical, and the
N-terminal portion (orange) is recruited to extend helix H3. On the MT binding face
of the molecule, helix H4 has moved down to intrude into the dock site for the neck
linker and the neck linker is undocked.
Figure 3.2 shows inhibitor-bound structures. In all of these structures (details in
figure legend), the active site is in exactly the same conformation, corresponding to
the ADP-like state described above. These inhibitor-bound structures nonetheless
fall into two structural subclasses, one with the neck linker docked and one with the
neck linker undocked.
3.3 Conformational Cycle and Catalytic Mechanism of Eg5
How do the conformations visualised by X-ray crystallography relate to the full

mechanochemical cycle of the Eg5 motor, and to the mechanism(s) of inhibitors?
Figure 3.3 shows the ATP-like and ADP-like conformations of the active site in
more detail. In the Eg5 active site, ATP and ADP both dock against the P-loop,
which is structurally invariant, and the catalytically-active conformation of the site
then forms via the closing-together of the Sw1 and Sw2 elements. This catalytically
active conformation is stabilised by the formation of a critical salt bridge between a
3 Mechanisms of Action of Eg5 Inhibitors 55
Fig. 3.1 Global conformations of the human Eg5 motor domain in (a) “ATP-like” (MgAMPPNP)
and (b) “ADP-like” (MgADP) states. The AMPPNP structure is 3HQD. The ADP structure is a
composite overlay of 4A28, 4A1Z, 4B7B and 1II6. Structures were globally aligned and rendered
using Pymol. Helices H1, H2, H3, H4, loop L5 and the neck linker (NL) are labelled. Switch 1 is
orange, the active site Mg2+ ion is green
conserved Arg234 in Sw1 (SSRSH) and a conserved Glu270 in Sw2 (LAGSE). The
key role of this salt bridge on the detailed catalytic mechanism of ATP turnover by
Eg5 was recently much-clarified by a high resolution crystal structure of Eg5 in a
pre-hydrolysis conformation (Figs. 3.1 and 3.3 [3]). This work showed that the inter-
switch salt bridge, once formed, not only tethers the two switch regions together, but
also positions a pair of water molecules, one of which acts as the catalytic base in
the hydrolysis of the beta-gamma phosphate bond of MgATP. A two-waters
56 R.A. Cross
Fig. 3.2 Global conformations of inhibitor-bound human Eg5 motor domain. Same orientations
as Fig. 3.1 (a) Inhibitor bound with docked NL. This composite overlay is a global alignment of
4A50, 4AP0, 2X7D, 1X88, 2X7C, 3K5E, 2IEH, 3KEN, 2GM1, 4BBG, 2PG2, 1Q0B, 1YRS,
2FKY, 2FL2, 2FL6, 2G1Q, 2Q2Y, 2Q2Z, 3CJO, 2UYI, 2UYM & 3L9H. (b) Inhibitor bound with
undocked NL. This is an alignment of 4AS7, 4A5Y, 4A51, 4BXN, 3K3B, 2XAE, 2X2R, 2FME &
2WOG. Structures were globally aligned and rendered using Pymol
mechanism for ATP hydrolysis by a molecular motor was first proposed by Onishi
[4] for myosin. A recent very high resolution structure of Kif4 also visualises this
pair of water molecules (Chang et al. 2013b). The formation of this salt bridge is
thus a signature for the catalytically active conformation of the active site. Mutation
of this salt bridge in kinesins profoundly inhibits the hydrolysis step of ATP turn-
over. Importantly, in inhibitor-bound structures of Eg5, this salt bridge is always
open (Figs. 3.2 and 3.3).
Fig. 3.3 Detailed view of the active sites for representatives of the ATP-like (a) and “ADP-like”
(b) states. In the AMPNP-bound structure 3HQD (a) the key salt bridge (SB) linking R234 to E270
is closed. The Mg2+ ion is tightly coordinated by the beta and gamma phosphate oxygens of
AMPPNP, Thr112, Ser233, and a network of hydrogen-bonded water molecules. Switch 1 (orange)
is folded into two short, antiparallel beta strands. In the ADP-like inhibitor-bound structure
(2WOG), the salt bridge (SB) is open and the Mg2+ has retracted back towards the beta phosphate
oxygens of ADP. Part of switch 1 (orange) is sequestered into helix H3, whilst the remainder is
folded into a helix. Switch 1 is further from the Mg2+ and the network of hydrogen-bonded water
molecules coordinating the Mg2+ is correspondingly more extensive
The closing-together of the active site Sw1 and Sw2, stabilised by the formation
of the salt bridge, encloses the gamma phosphate of MgATP in the so-called phos-
phate tube. For phosphate to escape from this tube following the hydrolysis step, it
needs either to exit through the end of the tube via a so-called backdoor mechanism,
through the gate created by the opening of the inter-switch salt bridge, or via a trap-
door mechanism, whereby in addition to salt bridge opening, the Sw2 changes con-
formation so as to open up the tube sufficiently for phosphate to escape, coincident
with retraction of the Mg to coordinate to the beta phosphate of ADP [5]. Release of
MgADP follows release of phosphate.
58 R.A. Cross
3.4 The Key Role of Mg
The substrate for kinesins is MgATP. The high charge density of the Mg2+ ion serves
as an organiser of the active site and is the key to stabilising the nucleotide in the
active site. Within the active site, the Mg2+ ion is coordinated by oxygen atoms, two
from the beta and gamma phosphates of ATP, and the remainder from either the side
chains of surrounding residues, or from water molecules that are partially immobil-
ised by a network of hydrogen bonds, as they are in ice. At different stages of
nucleotide processing in the active site, different side chains are involved in Mg2+
coordination (Fig. 3.3).
Sequestering free Mg2+ by adding EDTA to kinesin solutions rapidly strips out
the active site Mg2+, causing immediate ADP release [6]. It was suggested shortly
after the first kinesin crystal structures appeared that MTs might accelerate ADP
release from kinesin by accelerating Mg release [7]. Certainly it is clear that this
must be a major part of the mechanism of MT activation, as discussed by Nitta et al.
[8] for KIF1a and Chang et al. [9] for KIF4.
3.5 Conformational Effects of Inhibitor Binding
Inhibitor-bound structures of Eg5 show a consistent conformation of the active

site, with helix H3 wedged apart from helix H2 by the inhibitor, L5 forming an
induced-fit pocket that encloses the inhibitor, H3 extended by recruitment of part
of Sw1, and the remainder of Sw1 formed into a helix that closely apposes the
active site Mg2+ ion. The “back door” salt bridge is broken (Figs. 3.2 and 3.3). In
the absence of MTs the coupling between the contents of the active site and the
global conformation is so loose that it is essentially absent: across the set of all
kinesin crystal structures there is no detectable correlation between the global con-
formation and the conformation of the active site [2]. For example, inhibitor-bound
structures of Eg5 typically have MgADP in their active sites, but the neck linker
can be either in the “ATP-like” docked position, or the “ADP-like” undocked posi-
tion. Exactly the same conformation of the active site can support up or down H4
positions and docked/ undocked NL positions (Figs. 3.1 and 3.2), showing that the
global conformation in the inhibited state is not tightly coupled to the active site
conformation. Electron microscopy, though necessarily at lower resolution, sug-
gests that MT binding tightens the coupling between the active site and the global
conformation, and a complete scheme for the major nucleotide-dependent confor-
mational shifts in the Eg5-MT complex has recently been proposed [10]. Crucially
however, Eg5-inhibitor complexes are in a slip-state [1] that interacts only weakly
with MTs. The X-ray structures show that inhibitors oppose MT-activation by sta-
bilising the same motor·MgADP slip-state that occurs in the absence of inhibitors
(Figs. 3.1, 3.2, and 3.3). By stabilising this state, inhibitors inhibit both ATP turn-
over in the absence of MTs and the activation of the Eg5 ATPase by MTs.
Consistent with this, Kozielski and colleagues have suggested that allosteric
resistance mechanisms exist, in which mutations do not alter the inhibited state,
but rather alter the rates of entry to and exit from the inhibited state [11].
3.6 Role of Loop 5 in Allosteric Inhibition
Almost all Eg5-specific inhibitors so far described bind into an induced-fit pocket
formed by H2, H3 and Loop 5. All kinesins contain an L5, but the L5 of Eg5 is unusu-
ally long. Mutagenesis of L5 of Eg5 reveals that it has a substantial influence on rates
of ATP binding and nucleotide exchange by the wild type Eg5 motor [12, 13]. L5
changes conformation during the kinetic cycle of ATP turnover, as shown for example
by a shift in the intrinsic fluorescence of a Trp residue at the tip of the loop [14, 15].
It was recently shown [16] that the L5 of human Eg5 could be transplanted into
Klp61f, a kinesin-5 that is normally not inhibited by monastrol, and both drug bind-
ing and allosteric inhibition reconstituted.
Several structural roles for L5 have been proposed. Based on mutagenesis of L5
and on MD simulations, Behnke-Parks et al. [12] propose that L5 is a conforma-
tional latch that stabilises the AMPPNP conformation of Sw1. Latching is proposed
to result from binding of L5 to H3. In Eg5, EM reconstructions provide evidence
that the L5 can interact directly with H3 [17]. In the “ATP-like” structure 3HQD
(Fig. 3.1) L5 interacts with H3 via a hydrophobic (ring stacking) interaction and an
ionic interaction. In “ADP-like” structures H3 moves away from H2 and these con-
tacts are reduced, but in the inhibited structures there is extensive linkage due to the
formation of the induced-fit drug binding pocket. Inhibitor binding thus couples H2
and L5 much more tightly to H3. In a quite different proposal, Harrington and col-
leagues [13] propose that L5 can change conformation to interact directly with the
ribose of the bound nucleotide, and thereby influence MgATP binding and turnover.
In a recent, extensive fluorescence study, Muretta and colleagues [18] found spectro-
scopic evidence for three different, rapidly-exchanging L5 conformations. These
workers suggest that an L5-H3 interface is part of a pathway of communication
between the active site and the MT, and that inhibitors disrupt this pathway. Earlier
work [15] reached related conclusions. Waitzman and colleagues emphasise a role
for the L5 in mechanical processivity: they find that L5 is critical in the coordination
between the two heads of a walking Eg5 dimer [19].
When thinking about the function of L5 it is important to keep in mind that
whilst the L5 of Eg5 is unusually long, all kinesins have an L5. I think it is possible
that the generic role of L5 in kinesins is to stabilise the apo state of the active site.
Insertion of L5 between H2 and H3 in the wild type (uninhibited) enzyme would
serve to lock the structural relationship of H2 and H3, thereby quenching the dynam-
ics of the empty active site and stabilising it against a tendency to collapse.
In many kinesins, the apo state is unstable, so that the motor tends to lose activity
rapidly when not attached to a MT [6]. Eg5 is unusual amongst kinesins in that its
apo state is stable [14], suggesting that this state may play a more significant
biological role than has yet been realised. Perhaps the unusually long L5 of Eg5
60 R.A. Cross
Fig. 3.4 1RY6, a structure

showing the conformation
of the switch 1 and the loop
L5 in the apo state of a
kinesin-13
represents an adaptation to fix the structural relationship of H2 to H3 so as to prevent

an off-pathway collapse of the apo conformation of the active site following MgADP
release. Stabilising the apo state might be expected to slow down ATP binding, as is
observed [19], and potentially might drive MgADP release. Waitzman and Rice [20]
and Muretta et al. [18] both discuss a role of L5 in accelerating MgADP release in
the absence of inhibitors.
There are no apo structures of Eg5, but in 1RY6, the only available apo crystal
structure of a kinesin [21], Sw1 is predominantly random coil and the L5 folds over
to insert a Phe residue (equivalent to the Trp in L5 of human Eg5) into a hydropho-
bic pocket between H2 and H3 (Fig. 3.4). If this is indeed the role of L5 in the wild
type mechanism, then inhibitors work by recruiting and amplifying the native abil-
ity of the L5 to stabilize the empty active site and using it instead to hyperstabilize
the motor·MgADP state of the site.
3.7 Concluding Comments
Interestingly, the structural effects of inhibitor binding to an non-canonical site in Eg5

have recently been visualised [22]. The biochemical mechanism of inhibition none-
theless is similar to L5-binding inhibitors, and in the X-ray structure (3ZCW.pdb), the
active site is in the same motor·MgADP weak binding conformation, even though L5
is not involved in binding the inhibitor. Other inhibitors thought to bind at non-canon-
ical sites may inhibit by a non-canonical mechanism [23, 24], but for the moment the
structural evidence fits with the main speculative hypothesis I have described, that the
current generation of Eg5 inhibitors work by quenching the dynamics of an on-path-
way structural state and thereby inhibiting Mg2+ release. I have also suggested that the
role of L5 in the wild type mechanism of Eg5 may be to provide structural stability in
the as-yet little studied apo state of the Eg5 motor head, and that inhibitors hijack this
role to provide added structural stability to the motor·MgADP state. The apo state of
the Eg5 active site has not yet been visualised by X-ray crystallography, but recent
aMD simulation work has found that the dynamics of L5 in the wild type (un-inhib-
ited) Eg5 motor domain are quenched in the apo state relative to the nucleotide-bound
states (B. Grant, personal communication), which would be consistent.
References
1. Crevel IM-TC, Alonso MC, Cross RA (2004) Monastrol stabilises an attached low-friction
mode of Eg5. Curr Biol 14:R411–R412
2. Scarabelli G, Grant BJ (2013) Mapping the structural and dynamical features of kinesin motor
domains. PLoS Comput Biol 9:e1003329
3. Parke CL, Wojcik EJ, Kim S, Worthylake DK (2010) ATP hydrolysis in Eg5 kinesin involves
a catalytic two-water mechanism. J Biol Chem 285:5859–5867
4. Onishi H, Mochizuki N, Morales MF (2004) On the myosin catalysis of ATP hydrolysis.
Biochemistry 43:3757–3763
5. Kull FJ, Endow SA (2013) Force generation by kinesin and myosin cytoskeletal motor pro-
teins. J Cell Sci 126:9–19
6. Ma YZ, Taylor EW (1995) Kinetic mechanism of kinesin motor domain. Biochemistry
34:13233–13241
7. Vale RD (1996) Switches, latches, and amplifiers: common themes of G proteins and molecu-
lar motors. J Cell Biol 135:291–302
8. Nitta R, Okada Y, Hirokawa N (2008) Structural model for strain-dependent microtubule acti-
vation of Mg-ADP release from kinesin. Nat Struct Mol Biol 15:1067–1075
9. Chang Q, Nitta R, Inoue S, Hirokawa N (2013) Structural basis for the ATP-induced
Isomerization of Kinesin. J Mol Biol 425:1869–1880. doi:10.1016/j.jmb.2013.03.004
10. Goulet A, Major J, Jun Y, GROSS SP, Rosenfeld SS, Moores CA (2014) Comprehensive struc-
tural model of the mechanochemical cycle of a mitotic motor highlights molecular adaptations
in the kinesin family. Proc Natl Acad Sci U S A 111:1837–1842
11. Talapatra SK, Anthony NG, Mackay SP, Kozielski F (2013) Mitotic kinesin Eg5 overcomes
inhibition to the phase I/II clinical candidate SB743921 by an allosteric resistance mechanism.
J Med Chem 56:6317–6329
12. Behnke-Parks WM, Vendome J, Honig B, Maliga Z, Moores C, Rosenfeld SS (2011) Loop L5
acts as a conformational latch in the mitotic kinesin Eg5. J Biol Chem 286:5242–5253
13. Harrington TD, Naber N, Larson AG, Cooke R, Rice SE, Pate E (2011) Analysis of the interac-
tion of the Eg5 Loop5 with the nucleotide site. J Theor Biol 289:107–115
14. Cochran JC, Gilbert SP (2005) ATPase mechanism of Eg5 in the absence of microtubules:
insight into microtubule activation and allosteric inhibition by monastrol. Biochemistry
44:16633–16648
15. Maliga Z, Xing J, Cheung H, Juszczak LJ, Friedman JM, Rosenfeld SS (2006) A pathway of
structural changes produced by monastrol binding to Eg5. J Biol Chem 281:7977–7982
16. Liu L, Parameswaran S, Liu J, Kim S, Wojcik EJ (2011) Loop 5-directed compounds inhibit
chimeric kinesin-5 motors: implications for conserved allosteric mechanisms. J Biol Chem
286:6201–6210
62 R.A. Cross
17. Bodey AJ, Kikkawa M, Moores CA (2009) 9-Angström structure of a microtubule-bound

mitotic motor. J Mol Biol 388:218–224
18. Muretta JM, Behnke-Parks WM, Major J, Petersen KJ, Goulet A, Moores CA, Thomas DD,
Rosenfeld SS (2013) Loop L5 assumes three distinct orientations during the ATPase cycle of
the mitotic kinesin Eg5: a transient and time-resolved fluorescence study. J Biol Chem
288:34839–34849
19. Waitzman JS, Larson AG, Cochran JC, Naber N, Cooke R, Jon Kull F, Pate E, Rice SE (2011)
The loop 5 element structurally and kinetically coordinates dimers of the human kinesin-5,
Eg5. Biophys J 101:2760–2769
20. Waitzman JS, Rice SE (2014) Mechanism and regulation of kinesin-5, an essential motor for
the mitotic spindle. Biol Cell 106:1–12. doi:10.1111/boc.201300054
21. Shipley K, Hekmat-Nejad M, Turner J, Moores C, Anderson R, Milligan R, Sakowicz R,
Fletterick R (2004) Structure of a kinesin microtubule depolymerization machine. EMBO J
23:1422–1432
22. Ulaganathan VV, Talapatra SKS, Rath OO, Pannifer AA, Hackney DDD, Kozielski FF (2013)
Structural insights into a unique inhibitor binding pocket in kinesin spindle protein. J Am
Chem Soc 135:2263–2272
23. Groen AC, Needleman D, Brangwynne C, Gradinaru C, Fowler B, Mazitschek R, Mitchison
TJ (2008) A novel small-molecule inhibitor reveals a possible role of kinesin-5 in anastral
spindle-pole assembly. J Cell Sci 121:2293–2300
24. Tarui Y, Chinen T, Nagumo Y, Motoyama T, Hayashi T, Hirota H, Muroi M, Ishii Y, Kondo H,
Osada H, Usui T (2014) Terpendole E and its derivative inhibit STLC- and GSK-1-resistant
Eg5. Chembiochem 15:934–938
Chapter 4
Clinical Trials of Mitotic Kinesin Inhibitors
Steven S. Rosenfeld
4.1 Introduction
Kinesins subserve a wide variety of important roles in orchestrating the orderly

array of events that underlie mitosis (reviewed in Chap. 1 of this monograph). The
essential, non-redundant roles that many of these mitotic kinesins play, along with
their increased expression in a wide variety of malignancies, has motivated an
extensive amount of work directed toward targeting these molecular motors in order
to interfere with the unregulated proliferation that characterizes cancer [1–10]. This
has in turn led to the synthesis of a number of small molecule inhibitors of mitotic
kinesins and the investigation of a number of these in clinical trials of patients suf-
fering from a variety of malignancies. Indeed, part of the motivation for pursuing
mitotic kinesins has come from the relatively successful experience of another class
of anti-cancer therapeutics, ones that target the physiology of the microtubules that
mitotic kinesins use. Vinca alkaloids as well as taxanes interfere with microtubule
structure and dynamics in cells both during mitosis as well as in interphase. These
microtubule poisons have been shown to be highly active in a wide variety of both
solid and hematologic malignancies [11]. It has been generally assumed that while
the therapeutic efficacy of these drugs is though their disruption of the mitotic spin-
dle, their toxicities, which include significant neurotoxicity, are due to their effects
on both mitotic and non-mitotic microtubules, including those in post-mitotic cells,
such as neurons [12–15]. This has led to a desire to target more mitosis-specific
components of the cellular machinery, in the hope of preserving anti-cancer, anti-
proliferative efficacy while reducing toxicity.
The initial discovery that the mitotic kinesin Eg5 (Kif11) could be targeted with
a small molecule inhibitor—monastrol—served as the proof-of-principle that at
S.S. Rosenfeld (*)

Department of Cancer Biology and Rose Ella Burkhardt Brain Tumor Center, Cleveland
Clinic Foundation, 9500 Euclid Avenue NB40, Cleveland, OH 44195, USA
e-mail: rosenfs@ccf.org

64 S.S. Rosenfeld
least some mitotic kinesins were “druggable” with small molecule therapeutics
[16]. This in turn led to an extensive amount of effort by individual investigators and
biotech/pharma to identify and optimize small molecule inhibitors of mitotic kines-
ins—initially Eg5 and more recently several others [1, 4, 8, 10]. As discussed in
other chapters in this monograph and elsewhere [1–10, 17], studies in vitro and in a
variety of animal models of cancer demonstrated that these drugs were quite active
against a variety of both solid and hematologic malignancies. As a consequence, a
number of phase I and II clinical trials in patients with a variety of cancers have
been carried out. In this chapter, I will discuss: (1) how promising small molecule
inhibitors are evaluated, both pre-clinically and in clinical trials in humans; (2)
review the clinical experience of small molecule inhibitors of two mitotic kinesins—
Eg5 and CENP-E; and (3) discuss that while trials of kinesin inhibitors as single
agents in the treatment of cancer have shown very modest activity, the lessons
learned from this clinical experience point to potentially more effective ways to use
these agents in future human interventions. Thus, in the right context, these drugs
may hold an important place in the therapeutic armamentarium for the treatment of
cancer and other diseases of abnormal proliferation.
4.2 How Do Promising Drugs Get Studied in Humans?
General Concepts Related to Cancer Clinical Trials While there are many features
of clinical trials in cancer that are shared with those in other diseases, there are three
features of malignancies that create particular challenges and opportunities. First,
most cancers are characterized by a significant amount of genetic instability that in
aggregate enables these tumors to eventually adapt to the effect of treatment [18–
21]. Many tumors have amplification or constitutive activation of oncogenes, inac-
tivation or deletion of tumor suppressor genes, as well as changes in the epigenetic
modification of the genome [22, 23]. Furthermore, these genetic and epigenetic
changes can often develop in response to anti-cancer therapy, and can mediate resis-
tance to these treatments [24–26].
Second, tumors with such amplified or dysregulated oncogenes often become
dependent on these gene products for growth, such that inhibition of the oncogene
effectively blocks tumor proliferation—a phenomenon referred to as “oncogene
addiction” [27–29]. The concept of oncogene addiction has been validated in a wide
variety of both hematologic and solid tumors [29], and in turn is increasingly lead-
ing to the design of clinical trials that stratify patient enrollment based on the dys-
regulated expression of targeted oncogenes. While this concept has been generally
applied to well-appreciated oncogenes, including c-myc, H-ras, K-ras, and the epi-
dermal growth factor receptor (EGFR), it could in theory be relevant to any growth-
promoting gene product that tumors depend upon to maintain their malignant
phenotype, including mitotic kinesins.
Finally, tumors frequently harbor amplifications in a variety of growth-promoting
pathways that are either complementary or partially redundant. Thus, blocking any
4 Clinical Trials of Mitotic Kinesin Inhibitors 65
one oncogene may not necessarily produce a therapeutic effect if other mutant
oncogenes can compensate for the loss. Effectively inhibiting tumor growth may
therefore require simultaneously blocking several growth-promoting components,
or using therapies that target a specific function but work only in tumors with muta-
tions relevant to this function. This represents an extension of the concept of “syn-
thetic lethality” [30, 31], and underlies the rationale for combining several targeted
therapies for achieving more effective tumor control. However, combining different
treatments together also increases the risk of unacceptable side effects, and thus,
cancer clinical trials have to continually balance the risk of tumors developing resis-
tance with the risk of producing serious toxicities in patients. As will be discussed
below, the concept of synthetic lethality may also justify combining mitotic kinesin
inhibitors with other agents that affect cell cycle or tumor cell repair mechanisms.
From Pre-clinical Studies to Clinical Trials In the United States, less than 8 % of
promising drug candidates get approved by the Food and Drug Administration
(FDA), and the overall cost of taking a drug through the steps that lead to approval
can be in excess of $800 million [32]. The process of getting approval for use of a
drug in human patients requires that it be validated as a therapeutic through a series
of increasingly rigorous steps. With oncology therapeutics, the first of these are a
series of studies in vitro, using a variety of tumor cell lines or primary tumor speci-
mens derived from patients to examine if the drug reaches its target and has a
growth-inhibitory or cytotoxic effect. Given the genetic heterogeneity noted in the
preceding section, this often means that multiple tumor cell lines need to be used,
representing a variety of genetic lesions and backgrounds that reflect the clinical
reality of patients’ tumors. Drugs that make it through this first step are then evalu-
ated in a series of in vivo animal models of the cancer being studied. These studies
are designed to determine if the drug reaches its target in the organ(s) that are
affected by the malignancy, if it effectively kills the tumor or suppresses its growth,
and if its use provides a survival benefit. Additional studies in both normal and
tumor-bearing animals are also required to determine the types of toxicities that the
drug produces. In the United States, moving a promising therapeutic into the next
phase of evaluation—a clinical trial—requires that the data from these studies be
forwarded to the FDA as part of an Investigational New Drug License (IND)
application.
From Clinical Trials to Approval for Human Use Once an IND is obtained, the next
step is a set of human clinical trials, which occur through four discrete steps, referred
to as phases, that each attempt to address a specific question. In a phase 0 trial, the
primary goal is to determine if the drug reaches and affects its target (referred to as
pharmacodynamics), and how long the drug remains in the body to affect its target
(referred to as pharmacokinetics). This typically involves a small number of patients
who have advanced malignant disease that has progressed in spite of conventional
therapy and who are administered a low dose of the drug to be investigated. This is
then followed by a phase I trial, whose primary purpose is to determine how safe the
drug is and what type of side effects it produces. This phase also collects data on
pharmacokinetics and pharmacodynamics. Patients who have progressed through
66 S.S. Rosenfeld
conventional therapy are enrolled in small cohorts (typically 3) that receive a fixed
dose of drug, and successive cohorts receive progressively higher doses of drug
until a Maximum Tolerated Dose (MTD) is reached—typically defined as a dose
above which unacceptable toxicity is observed. The relatively small number of
patients enrolled (typically 15–30) and the dose escalation aspect of phase I trials
means that these studies are generally not powered statistically to determine how
effective a drug is. However, these trials do collect data on responses, which can be
useful in guiding the next phase. The next step is a phase II trial, which enrolls a
larger number of patients (25–100), who receive the drug at the MTD. The primary
question being addressed in this phase is how often responses are seen when the
drug is administered at the MTD. However, these trials also collect additional data
on toxicities that are observed in this larger collection of patients. Drugs that do
produce measureable responses can then be advanced into a phase III trial. These
trials typically include large numbers of patients (100s–1000s) and are designed
primarily to determine how the drug in question compares to standard therapy.
Patients are typically randomized to receive either standard therapy or the investiga-
tional drug, and the patient cohort is sufficiently large as to allow a statistically
meaningful comparison between the two. Additional information about safety at the
MTD is also obtained during this trial. If a phase III trial shows that the drug in
question is at least as efficacious and safe as standard therapy, an application can
then be made to the FDA for approval. The increasingly larger numbers of patients
required with increasing phase means that taking a drug all the way from pre-clinical
in vitro and animal studies through a phase III trial can take over a decade to com-
plete, and given the costs involved, underscores the “high risk/high reward” nature
of the drug development and approval process. Finally, once a drug is FDA approved,
a phase IV trial will be carried out; typically by the pharmaceutical company that
has obtained the approval. This is essentially a post-marketing study in which data
on side effects from a much larger cohort of patients is acquired in order to mini-
mize the chance of any untoward side effect being missed.
Clinical Trial Oversight and Accessibility It goes without saying that clinical trials
must comply with national and international norms to insure the ethical treatment of
human subjects, including standards to minimize the risk of harm. In the United
States, this function largely resides within Human Investigations Review Boards
(IRBs). These are committees, formed in response to federal law and regulations
from the Department of Health and Human Services, that are composed of experts
in the field along with community representatives. The IRB is responsible for
reviewing all ethical issues as they relate to the trial, including efforts to minimize
risk, clarity of the informed consent procedure, and conflicts of interest. Through
the course of the trial phases, a Data and Safety Monitoring Board reviews all toxic-
ity and side effect results and reports these to the IRB periodically. Since early
1990s, the International Conference on Harmonization of Technical Requirements
for Registration of Pharmaceuticals for Human Use (ICH) has attempted to coordi-
nate clinical trial initiatives and review processes in the EU, Japan, and the US in
order to develop more standardized oversight mechanisms, given the increasing
number of multi-national clinical trials that are being undertaken.
Clinical trials depend on the interest and willingness of patients to participate in

these investigations. In an effort to make clinical investigations more transparent
and information about these trials available to the general public, both the US (www.
clinicaltrials.gov) and the European Union (www.clinicaltrialsregister.eu) have
developed websites with updated information relating to oncology clinical trials.
This has more recently been supplemented by a World Health Organization clinical
trials database (http://apps.who.int/trialsearch/) that includes as well countries not
represented in these websites.
4.3 Phase I Trials of Eg5 Inhibitors
Eg5 has served as the most striking example of a “druggable” mitotic kinesin, with
a large number of small molecule inhibitors that have been discovered and opti-
mized [4]. Furthermore, Eg5, like a number of other mitotic kinesins, is up-regulated
in a variety of cancers, consistent with one of the central requirements of the “onco-
gene addiction” hypothesis [27–29]. Overexpression of Eg5 correlates with a more
aggressive tumor type and poorer prognosis in renal cell, urothelial, and pancreatic
carcinomas [33–35]. Thus, there would a priori be reason for optimism that clinical
trials of high affinity, specific Eg5 inhibitors should be promising. As will be dis-
cussed below, this promise has not been fully realized, but the results may in fact
point to better ways to use these types of drugs.
Ispinesib Six phase I trials with ispinesib have been completed (www.clinicaltrials.
gov). The results of four of these have been published [36–39], all in patients with
advanced, recurrent solid tumors. As noted previously, phase I trials are not designed
or powered to provide statistically meaningful information on tumor responses, an
issue that is even more problematic in oncology clinical trials, since enrolled patients
are typically in the advanced stages of their disease. It is thus not altogether surpris-
ing that none of these trials showed evidence of a complete response, and partial
responses (defined as >30 % reduction in size of the target lesion) were noted in
<10 % of cases. However, “stable disease”, generally defined as changes in target
lesion size <30 %, was noted in 12–60 % of patients. This rate of disease stabiliza-
tion might seem encouraging. However, rates of “stable disease” as high as 55 %
have been reported in patients with renal cell carcinoma treated with placebo in the
context of a clinical trial [40], which raises questions about the value of this response
measure. In each case, an MTD was determined, and the dose-limiting toxicities
included suppression of neutrophil production (neutropenia), anemia, and liver tox-
icity. Unlike the microtubule inhibitors, such as taxanes and vinca alkaloids, neuro-
toxicity was not noted. Pharmacokinetic measurements demonstrated an elimination
half-life of 16–30 h.
SB-743921 One phase I trial has been reported [41] for this agent. Neutropenia was
the dose limiting toxicity, and one of 41 evaluable patients had a partial response.
The elimination half-life for this drug was 29 h.
68 S.S. Rosenfeld
AZD-4877 Three phase I trials have been completed (www.clinicaltrials.gov) and

results of two have been published [42, 43]. As with ispinesib, the major toxicity
was neutropenia, and the elimination half-life was found to range from 7 to 20 h.
No complete or partial responses were noted, with stable disease noted in a few
patients.
MK-0731 One phase I trial has been reported [44]. Neutropenia was the dose-
limiting toxicity, and no complete or partial responses were noted. Pharmacokinetic
studies demonstrated an elimination half-life in the range of 2–22 h.
ARQ-621 A phase I trial has been completed (www.clinicaltrials.gov), and results
were reported in the 2011 American Society of Clinical Oncology (ASCO) meeting.
A total of 48 patients with a variety of solid tumors were enrolled. Anemia and
venous thrombosis were the most serious adverse events, and unlike the other drugs
noted in this section, inhibition of bone marrow derived white blood cells (myelo-
suppression) was not a dominant toxicity. An MTD was defined in this study, and
6/48 patients experienced disease stabilization for >4 months. However, as noted
above, the significance of disease stability as a measureable endpoint remains
unclear.
ARRY-520 Five phase I trials are currently accruing patients (www.clinicaltrials.
gov). The results of one of these have been published [45]. Unlike the other drugs
discussed above, which were examined in patients with a variety of solid malignan-
cies, this trial was limited to patients with advanced myeloid leukemias, and the drug
demonstrated a different spectrum of dose limiting toxicities—inflammation of the
oral mucosa (mucositis), rash, and elevated serum bilirubin. As with the phase I trials
of these other agents, however, there was little evidence of anti-cancer activity.
However, several unpublished phase I trial results of ARRY-520 in multiple
myeloma have been reported at the December, 2013 Annual Meeting of The
American Society of Hematology (ASH) [46–48]. These reports reveal consider-
ably more encouraging results when ARRY-520 is combined with other agents for
the treatment of multiple myeloma. Combining ARRY-520 with the proteasome
inhibitor bortezomib produced response rates as high as 42 %. In another phase I
trial, patients with relapsed and refractory multiple myeloma were treated with a
combination of ARRY-520 with carfilzomib, another proteasome inhibitor. Overall,
37 % of patients responded to treatment, with 5 % achieving a near complete
response and 32 % a partial response, and the combination of these two drugs was
well tolerated.
One feature that may be contributing to the relative success of ARRY-520 is
pharmacokinetic. While the elimination half-lives of ispinesib, SB-743921, AZD-
4877, and MK-0731 are all <30 h, the corresponding half-life of ARRY-520 was
found to be >90 h [45]. This is an important point, since as will be discussed below,
the mitotic fraction of many solid tumors is quite small [49]. This means that for an
anti-mitotic to be effective in vivo, it must remain in contact with the tumor for
considerably longer than the average cell cycle time—a criterion that ARRY-520
appears to satisfy.
4.4 Phase II Trials of Eg5 Inhibitors
Ispinesib Nine phase II trials of ispinesib have been completed (www.clinicaltrials.

gov) and results of five have been published [50–54]. These trials were carried out
in patients with melanoma, head and neck squamous cell carcinoma, hepatocellular
carcinoma, prostate carcinoma, and renal cell carcinoma. While these trials con-
firmed that the toxicities of drug were manageable, none of them demonstrated any
objective responses.
AZD-4877 One phase II trial has been completed (www.clinicaltrials.gov) in
patients with urothelial cancers, and demonstrated an objective response rate of
2.6 % with a progression-free survival of less than 8 weeks. Results of a second
phase II trial, in patients with refractory acute myeloid leukemia have been pub-
lished [55]. No objective responses were observed, and the study was terminated.
ARRY-520 By contrast with the above, results with ARRY-520 in hematologic
malignancies have been very encouraging. At the 2013 ASH meeting, results of a
phase II trial of ARRY-520 in relapsed/refractory multiple myeloma were reported
[46]. This trial contained two arms: 32 patients treated with the phase II dose of
ARRY-520 alone (“single agent”), and 55 patients treated with ARRY-520 in com-
bination with the steroid dexamethasone. Outcomes rated as equal to or better than
a partial response were similar in both arms of the study: 16 % for the single agent
arm and 15 % for the combination of ARRY-520 + dexamethasone. However, these
investigators also found in the course of the study that responsiveness to ARRY-520
correlated with the level of α-1 acidic glycoprotein (AAG). AAG is a serum protein
that can bind to ARRY-520, and presumably reduces the effective concentration of
drug. Patients with low levels of AAG levels had an overall response rate to single
agent ARRY-520 of 24 %, while none of the patients with high AAG levels showed
a response. A similar trend was seen for the ARRY-520 + dexamethasone cohort.
Not surprisingly, overall survival could also be stratified by AAG levels—23.3
months in the low AAG group, and 4.5 months in the high AAG group. Based on
these and other results reported at this meeting, the maker of ARRY-520, Array
BioPharma, is planning on a phase III trial of ARRY-520 in combination with the
proteasome inhibitor carfilzomib in patients with relapsed/refractory multiple
myeloma, which is due to begin enrollment in late 2014 [56].
4.5 Phase I Trial of a CENP-E Inhibitor
A single phase I trial of GSK-923295 has been reported [57]. Like the Eg5 inhibi-
tors, this drug is an allosteric inhibitor of the mitotic kinesin CENP-E. Thirty-nine
patients with recurrent or refractory solid tumors were enrolled in this study. Unlike
the Eg5 inhibitors, the dose limiting toxicities included fatigue, anemia, and eleva-
tion of liver enzymes, while neutropenia was relatively uncommon. The terminal
elimination half-life was 9–11 h, and there was one durable, partial response.
70 S.S. Rosenfeld
4.6 Summary of the Clinical Experience with Mitotic

Kinesin Inhibitors
The recent results of phase I and II trials with ARRY-520 serves as proof-of-principle
that drugs which target mitotic kinesins may have a place in the therapeutic arma-
mentarium of oncology practice. As expected, Eg5 and CENP-E inhibitors do not
produce the neurotoxicity that is commonly seen with another class of drugs that
have been considered anti-mitotics—the vinca alkaloids and taxanes [11–15].
However, with the exception of ARRY-520, the clinical experience with Eg5 and
CENP-E inhibitors in phase I trials has been disappointing, and in the case of ispi-
nesib and AZD-4877, this lack of efficacy has been confirmed by subsequent phase
II trials. There are several explanations for these results:
First, an obvious explanation is the development of tumor resistance, due to point
mutations within L5—a loop in the Eg5 enzymatic domain that constitutes part of
the drug-binding site [58]. The genetic instability that is characteristic of most
malignant tumors might allow for selection of Eg5 mutations that confer treatment-
resistance. In addition, several recent reports have shown that phosphomimetic
mutations of three tyrosine residues within or near L5 may confer resistance to the
Eg5 inhibitor S-trityl cysteine [59]. These tyrosines can be phosphorylated by src
kinase, which is frequently dysregulated in cancers [60]. However, in the phase I
trials that have been reported, tumor progression was typically noted after a rela-
tively brief period of treatment [36–39], making it less likely that treatment failure
reflects mutationally-acquired resistance to the small molecules. Furthermore, while
enhanced src activity might endow tumor cells with resistance to Eg5 inhibitors, it
should also be noted that differentiation of normal myeloid cells also leads to up-
regulation of src kinase activity [61]. This would predict that myeloid cells should
be relatively resistant to the effects of small molecule inhibitors of Eg5, a point that
is not consistent with several of the phase I and II trials (noted above).
A potentially more serious problem relates to two underlying assumptions that
have driven the effort to develop and utilize these mitotic kinesin inhibitors. These
have been extensively discussed in three recent reviews by Komlodi-Pasztor et al.
and Mitchison [49, 62, 63]. The first assumption is that the relatively high activity
that microtubule inhibitors, including vinca alkaloids and taxanes, have against a
variety of solid and hematologic malignancies is because of their effects on mito-
sis—and specifically, on the mitotic spindle. As noted above, this has motivated the
search for inhibitors of other targets that are more specifically expressed during
mitosis, and whose inhibition might produce less toxicity. However, the disparity
between the effectiveness of vinca alkaloids and taxanes, on the one hand, and most
of the Eg5 inhibitors (as well as inhibitors of Aurora Kinase and Polo-Like Kinase)
on the other suggests another explanation—that microtubule integrity outside of
mitosis is just as or even more important than its integrity in mitosis for tumor cells
to function [49, 62, 63]. A large number of oncogenes and signaling molecules
essential for tumor maintenance and for response to injury (induced, for example,
by radiation therapy or chemotherapy) are transported on or associate with
microtubules [62], and it is equally plausible that the effectiveness of microtubule

poisons is due to this broad degree of anti-signaling activity.
The second assumption is that the rapid tumor doubling times and high mitotic
indices that characterize tumor cells in culture and in rodent models applies as well
to the clinical setting of human tumors growing in human patients. It goes without
saying that for a mitosis-specific inhibitor to be effective, it needs to be present in a
sufficient concentration to hit its target during the time a tumor cells is in the mitotic
phase of its cell cycle. However, while the doubling times of solid tumors in vitro
and in animal models are in the range of several days, in human patients, they can
range from several months, in the case of melanoma, to several years, in the case of
prostate cancer [49, 62]. Consequently, one could predict a priori that for a mitotic
kinesin inhibitor to be effective, two features need to be satisfied: (1) the drug needs
to be present for a long enough period of time after administration to affect an
appreciable number of mitotically active tumor cells, and (2) the tumor that is being
treated needs to have a rapid doubling time or high mitotic index, to enhance the
probability that the kinesin inhibitor will hit its target. Both of these requirements
appear to be met in the case of the clinical trials with ARRY-520 in recurrent mul-
tiple myeloma. As noted above, ARRY-520 has a terminal elimination half-life of
>90 h [45], at least fourfold larger than that for the other Eg5 inhibitors. Furthermore,
while the percentage of tumor cells in mitosis at any given time is quite low for solid
tumors from humans (<2 %, Refs. [49] and [62]), it ranges from 14 % to 83 % in
relapsed multiple myeloma [64]. Thus, the mismatch between the cycling kinetics
of malignant cells in vitro or in mouse xenografts and the corresponding value in the
real (human) setting may mean that most of the trials discussed above have used the
wrong drug (e.g., with too short a serum half-life) in the wrong setting (e.g., in
tumors with relatively low mitotic fractions).
4.7 Where Do We Go from Here?
While the clinical experience with most of the mitotic kinesin inhibitors has been
disappointing, the relative success of ARRY-520, coupled to the considerations dis-
cussed in the preceding section, suggest that mitotic kinesin inhibitors may indeed
be very useful and effective anti-cancer drugs, if used in a nuanced way that is con-
sistent with the underlying biology of cancers in humans. I discuss below three
modifications that could be applied to future clinical trials of mitotic kinesin inhibi-
tors and which address the limitations of the prior clinical investigations.
Use Mitotic Kinesin Inhibitors in Cells with a High Mitotic Fraction That Over-
express the Target The oncogene addiction hypothesis would predict that cells
which over-express Eg5 or CENP-E should be particularly sensitive to inhibitors of
these mitotic motors. Oncology clinical trials are increasingly stratifying patients or
limiting eligibility for entry by the level of expression of the drug target, or in some
cases, by the presence of particular target mutations that enhance drug efficacy. The
72 S.S. Rosenfeld
latter, for example, is increasingly being relied upon in guiding the use of EGFR
tyrosine kinase inhibitors in non-small cell lung cancer [65]. Measuring expression
levels of a target protein and/or target protein mutations are becoming increasingly
common in the clinical setting, and future trials of mitotic kinesin inhibitors should
incorporate these types of measurements in determining eligibility and stratifying
patients to investigate the effect on outcome. Likewise, mitotic fraction is routinely
now measured in biopsies of human tumors, by means of Ki67 staining [66], and
future trials could be designed to limit enrollment to patients whose tumors have a
high Ki67 index. The relative success of ARRY-520 in recurrent multiple myeloma
suggests that future studies of mitotic kinesin inhibitors should be used in tumors
with an inherently high mitotic fraction. These could include a number of hemato-
logic malignancies, as well as glioblastoma, a malignant brain tumor that can have
a Ki67 index in excess of 40 % [67, 68].
Combine Mitotic Kinesin Inhibitors with Other Agents That Affect Cell Cycle A
number of FDA-approved drugs for use in cancer can also produce cell cycle arrest,
and their combination might be synergistic with a mitotic kinesin inhibitor. These
include proteasome inhibitors [68], which are already in clinical trials combined
with ARRY-520; topoisomerase inhibitors [69]; and radiation therapy [70].
Utilize Mitotic Kinesin Inhibitors in Tumors That Are Compartmentalized The
dose limiting toxicity of many of the Eg5 inhibitors is on bone marrow-derived
myeloid and erythroid precursor cells, which understandably have a high mitotic
index to continually generate new white and red blood cells, respectively. All of the
clinical trials discussed above have involved systemic administration of kinesin
inhibitors—under conditions where normal tissue such as bone marrow would
experience significant drug concentrations with consequent myelosuppression and/
or anemia. However, some cancers are truly localized diseases, and this may pro-
vide a motivation to administer the mitotic kinesin inhibitors locally, avoiding the
systemic toxicities that may be limiting drug efficacy. In one recent phase I trial
[71], a combination of siRNAs targeting vascular endothelial growth factor (VEGF)
and Eg5 were incorporated into lipid nanoparticles and infused into patients with
liver metastases from a variety of solid tumors, including renal cell, pancreatic neu-
roendocrine, and endometrial carcinomas and angiosarcoma. The authors had previ-
ously shown that these nanoparticles are selectively taken up by the liver and spleen.
Dose limiting toxicities were generally reversible, and the treatment was well toler-
ated, and one of these patients had a substantial partial response. While the relative
contributions of the VEGF and Eg5 targeting siRNAs could not be determined in
this study, the lack of significant bone marrow toxicity does suggest that a more
compartmentalized delivery of an Eg5 inhibitor is less toxic than one involving
systemic delivery.
Glioblastoma, the most aggressive of the primary brain tumors, is a particularly
good example of a highly compartmentalized tumor. Although this tumor can spread
diffusely through the brain, and as noted above, has a high mitotic index, it almost
never metastasizes outside the CNS. The mitotic index of normal brain and spinal
cord is extremely low, reflecting the post mitotic state of neurons and the low mitotic
activity of normal glia. Furthermore, there are several methods of delivering thera-
peutics directly into the brain that provide a way of circumventing systemic delivery
of toxic drugs. The most developed of these is convection-enhanced delivery (CED),
whereby a catheter is inserted into brain tumor and/or surrounding normal brain tis-
sue, and the therapeutic is slowly infused into the brain with an external syringe
pump [72, 73]. Phase I, II, and III trials utilizing CED for the treatment of glioblas-
toma have been performed, and this technology could readily be applied to the use
of kinesin inhibitors if pre-clinical models utilizing CED administration of these
drugs showed safety and efficacy.
The road from identifying a small molecule inhibitor to getting final approval for
human use is a long and arduous one, and the vast majority of promising candidates
fall by the wayside through this process. In hindsight, it should therefore not be
surprising that many of the clinical trials of mitotic kinesin inhibitors have been
disappointing. However, it should be clear from this chapter that this should not
deter investigators from continuing to look for new inhibitors or, perhaps more
importantly, use the inhibitors already available in new ways, guided by the biology
of human tumors in humans. There is every reason to think that better aligning the
use of these drugs with the mitotic biology of specific cancers will enhance the
chances for success and ultimately lead to the introduction of kinesin inhibitors for
human use.
References
2. Exertier P et al (2013) Impaired angiogenesis and tumor development by inhibition of the
mitotic kinesin Eg5. Oncotarget 4(12):1–15
3. Carol H et al (2009) Initial testing (stage 1) of the kinesin spindle protein inhibitor ispinesib by
the pediatric preclinical testing program. Pediatr Blood Cancer 53:1255–1263
4. El-Nassan HB (2014) Advances in the discovery of kinesin spindle protein (Eg5) inhibitors as
antitumor agents. Eur J Med Chem 62:614–631
5. Koller E et al (2006) Use of a chemically modified antisense oligonucleotide library to identify
and validate Eg5 as a target for antineoplastic drug development. Cancer Res 66:2059–2066
6. Muller C et al (2007) Inhibitors of kinesin Eg5: antiproliferative activity of monastrol ana-
logues against human glioblastoma cells. Cancer Chemother Pharmacol 59:157–164
7. Purcell JW et al (2010) Activity of the kinesin spindle protein inhibitor ispinesib (SB- 715992)
in models of breast cancer. Clin Cancer Res 16:566–576
8. Sakowicz R et al (2004) Antitumor activity of a kinesin inhibitor. Cancer Res 64:3276–3280
9. Valensin S et al (2009) KIF11 inhibition for glioblastoma treatment: reason to hope or a strug-
gle with the brain? BMC Cancer 9:196
74 S.S. Rosenfeld
11. Jackson JR et al (2007) Targeted anti-mitotic therapies: can we improve on tubulin agents? Nat
Rev Cancer 7:107–117
12. Canta A et al (2009) Tubulin: a target for antineoplastic drugs into the cancer cells but also in
the peripheral nervous system. Curr Med Chem 16:1315–1324
13. Windebank AJ (1999) Chemotherapeutic neuropathy. Curr Opin Neurol 12:565–571
14. Tuxen MK, Hansen SW (1994) Neurotoxicity secondary to antineoplastic drugs. Cancer Treat
Rev 20:191–214
15. Morris PG, Fornier MN (2008) Microtubule active agents: beyond the taxane frontier. Clin
Cancer Res 14:7167–7172
16. Mayer TU et al (1999) Small molecule inhibitor of mitotic spindle bipolarity identified in a
phenotype-based screen. Science 286:971–974
17. Liu X et al (2013) Oncogenic role of kinesin proteins and targeting kinesin therapy. Cancer Sci
104(6):651–656. doi:10.1111/cas.12138
18. Arnedos M et al (2014) The genetic complexity of common cancers and the promise of person-
alized medicine: is there any hope? J Pathol 232:274–282
19. Blair BG et al (2014) Somatic alterations as the basis for resistance to targeted therapies. J
Pathol 232:244–254
20. Zheng J (2013) Oncogenic chromosomal translocations and human cancer. Oncol Rep
30:2011–2019
21. Bunting SF, Nussenzweig A (2013) End-joining, translocations and cancer. Nat Rev Cancer
13:443–454
22. Weinstein IB et al (1997) Disorders in cell circuitry associated with multistage carcinogenesis:
exploitable targets for cancer prevention and therapy. Clin Cancer Res 3:2696–2702
23. Greenman C et al (2007) Patterns of somatic mutation in human cancer genomes. Nature
446:153–158
24. Gorre ME et al (2001) Clinical resistance to STI-571 cancer therapy caused by BCRABL gene
mutation or amplification. Science 293:876–880
25. Kobayashi S et al (2005) EGFR mutation and resistance of non-small-cell lung cancer to gefi-
tinib. New Engl J Med 352:786–792
26. Pao W et al (2005) Acquired resistance of lung adenocarcinomas to gefitinib or erlotinib is
associated with a second mutation in the EGFR kinase domain. PLoS Med 2:225–235
27. Weinstein IB (2002) Addiction to oncogenes—the Achilles heal of cancer. Science
297:63–64
28. Weinstein IB, Joe AK (2006) Mechanisms of disease: oncogene addiction—a rationale for
molecular targeting in cancer therapy. Nat Clin Pract Oncol 3:448–457
29. Weinstein IB, Joe AK (2008) Oncogene addiction—a rationale for molecular targeting in can-
cer therapy. Cancer Res 68:3077–3080
30. Hartwell LH et al (1997) Integrating genetic approaches into the discovery of anti-cancer
drugs. Science 278:1064–1068
31. Kaelin WG (2005) The concept of synthetic lethality in the context of anti-cancer therapy. Nat
Rev Cancer 5:689–698
32. Morgan S et al (2011) The cost of drug development: a systematic review. Health Policy
100:4–17
33. Liu M, Wang X, Yang Y, Li D, Ren H, Zhu Q, Chen Q, Han S, Hao J, Zhou J (2010) Ectopic
expression of the microtubule-dependent motor protein Eg5 promotes pancreatic tumourigen-
esis. J Pathol 221(2):221–228
34. Ding S, Xing N, Lu J, Zhang H, Nishizawa K, Liu S, Yuan X, Qin Y, Liu Y, Ogawa O,
Nishiyama H (2011) Overexpression of Eg5 predicts unfavorable prognosis in non- muscle
invasive bladder urothelial carcinoma. Int J Urol 18(6):432–438
35. Sun D, Lu J, Ding K, Bi D, Niu Z, Cao Q, Zhang J, Ding S (2013) The expression of Eg5
predicts a poor outcome for patients with renal cell carcinoma. Med Oncol 30(1):476
36. Gomeza HL et al (2012) Phase I dose-escalation and pharmacokinetic study of ispinesib, a
kinesin spindle protein inhibitor, administered on days 1 and 15 of a 28- day schedule in
patients with no prior treatment for advanced breast cancer. Anti Cancer Drugs 23:335–341
37. Blagden SP et al (2008) A phase I trial of ispinesib, a kinesin spindle protein inhibitor, with
docetaxel in patients with advanced solid tumours. Brit J Cancer 98:894–899
38. Burris HA III et al (2011) A phase I study of ispinesib, a kinesin spindle protein inhibitor,
administered weekly for three consecutive weeks of a 28-day cycle in patients with solid
tumors. Invest New Drugs 29:467–472
39. Souid A-K et al (2010) A pediatric phase I trial and pharmacokinetic study of ispinesib: a
children’s oncology group phase I consortium study. Pediatr Blood Cancer 55:1323–1328
40. Escudier B et al (2007) Sorafenib in advanced clear-cell renal-cell carcinoma. New Engl J Med
356:125–134
41. Holen KD et al (2011) A first in human study of SB-743921, a kinesin spindle protein inhibi-
tor, to determine pharmacokinetics, biologic effects and establish a recommended phase II
dose. Cancer Chemother Pharmacol 67:447–454
42. Gerecitano JF et al (2013) A phase I trial of the kinesin spindle protein (Eg5) inhibitor
AZD4877 in patients with solid and lymphoid malignancies. Invest New Drugs 31:355–362
43. Infante JR et al (2012) A phase I study to assess the safety, tolerability, and pharmacokinetics
of AZD4877, an intravenous Eg5 inhibitor in patients with advanced solid tumors. Cancer
Chemother Pharmacol 69:165–172
44. Holen K et al (2012) A phase I trial of MK-0731, a kinesin spindle protein (KSP) inhibitor, in
patients with solid tumors. Invest New Drugs 30:1088–1095
45. Khoury HJ et al (2012) A phase 1 dose-escalation study of ARRY-520, a kinesin spindle pro-
tein inhibitor, in patients with advanced myeloid leukemias. Cancer 118:3556–3564
46. Lonial S et al (2013) Prolonged survival and improved response rates with ARRY-520 in
relapsed/refractory multiple myeloma (RRMM) patients with low α-1 acid glycoprotein
(AAG) levels: results from a phase 2 study. Abstract #285, American Society of Hematology
annual meeting, 2013
47. Chari A et al (2013) A phase 1 study of ARRY-520 with bortezomib (BTZ) and dexamethasone
(dex) in relapsed or refractory multiple myeloma (RRMM). Abstract #1938, American Society
of Hematology annual meeting, 2013
48. Shah J et al (2013) Phase 1 study of the novel kinesin spindle protein inhibitor ARRY-
520 + carfilzomib (Car) in patients with relapsed and/or refractory multiple myeloma (RRMM)
phase 1 study of the novel kinesin spindle protein inhibitor ARRY-520 + carfilzomib (Car) in
patients with relapsed and/or refractory multiple myeloma (RRMM). Abstract #1982,
American Society of Hematology annual meeting, 2013
49. Komlodi-Pasztor E et al (2011) Mitosis is not a key target of microtubule agents in patient
tumors. Nat Rev Clin Oncol 8:244–250
50. Lee CW et al (2008) A phase II study of ispinesib (SB-715992) in patients with metastatic or
recurrent malignant melanoma: a National Cancer Institute of Canada Clinical Trials Group
trial. Invest New Drugs 26:249–255
51. Tang PA et al (2008) Phase II study of ispinesib in recurrent or metastatic squamous cell car-
cinoma of the head and neck. Invest New Drugs 26:257–264
52. Knox JJ et al (2008) A phase II and pharmacokinetic study of SB-715992, in patients with
metastatic hepatocellular carcinoma: a study of the National Cancer Institute of Canada
Clinical Trials. Invest New Drugs 26:265–272
53. Beer TM et al (2008) Southwest oncology group phase II study of ispinesib in androgen inde-
pendent prostate cancer previously treated with taxanes. Clin Genitourin Cancer 6:103–109
54. Lee RT et al (2008) A university of Chicago consortium phase II trial of SB-715992 in
advanced renal cell cancer. Clin Genitourin Cancer 6:21–24
55. Kantarjian HM et al (2012) Phase I/II multicenter study to assess the safety, tolerability, phar-
macokinetics and pharmacodynamics of AZD4877 in patients with refractory acute myeloid
leukemia. Invest New Drugs 30:1107–1115
56. Owens B (2013) Kinesin inhibitor marches toward first-in-class pivotal trial. Nat Med 19:1550
57. Chung V et al (2012) First-time-in-human study of GSK923295, a novel antimitotic inhibitor
of centromere-associated protein E (CENP-E), in patients with refractory cancer. Cancer
76 S.S. Rosenfeld
58. Talapatra SK et al (2013) Mitotic kinesin Eg5 overcomes inhibition to the phase I/II clinical
candidate SB743921 by an allosteric resistance mechanism. J Med Chem 56:6317–6329
59. Gifford KM et al (2014) SRC phosphorylation regulates the human kinesin-5, Eg5, and dis-
rupts the binding of Eg5 inhibitors. Biophys J 106:780a
60. Wheeler DL et al (2009) The role of Src in solid tumors. Oncologist 14:667–678
61. Gee CE et al (1986) Differentiation of myeloid cells is accompanied by increased levels of
pp6Oc-src protein and kinase activity. Proc Natl Acad Sci U S A 83:5131–5135
62. Komlodi-Pasztor E et al (2012) Inhibitors targeting mitosis: tales of how great drugs against a
promising target were brought down by a flawed rationale. Clin Cancer Res 18:51–63
63. Mitchison TJ (2012) The proliferation rate paradox in antimitotic chemotherapy. Mol Biol
Cell 23:1–6
64. Drewinko B (1981) The growth fraction of human myeloma cells. Blood 57:333–338
65. Rosell R et al (2012) Erlotinib versus standard chemotherapy as first-line treatment for
European patients with advanced EGFR mutation-positive non-small-cell lung cancer
(EURTAC): a multicentre, open-label, randomised phase 3 trial. Lancet Oncol 13:239–246
66. Scholzen T, Gerdes J (2000) The Ki-67 protein: from the known and the unknown. J Cell
Physiol 182:311–322
67. Andres G, Chiesa-Vottero AG et al (2003) Comparison of proliferation indices in glioblastoma
multiforme by whole tissue section vs tissue microarray. Am J Clin Pathol 120:902–908
68. Yamaguchi R, Dutta A (2000) Proteasome inhibitors alter the orderly progression of DNA
synthesis during S-phase in HeLa cells and lead to rereplication of DNA. Exp Cell Res
261:271–283
69. Poot M et al (1992) Cell cycle effects of the DNA topoisomerase inhibitors camptothecin and
m-AMSA in lymphoblastoid cell lines from patients with Fanconi anemia. Mutat Res
270:185–189
70. McBride WH et al (2003) The role of the ubiquitin/proteasome system in cellular responses to
radiation. Oncogene 22:5755–5773
71. Tabernero J et al (2013) First in humans trial of an RNA interference therapeutic targeting
VEGF and KSP in cancer patients with liver involvement. Cancer Discov 3:406–417
72. Lidar Z et al (2004) Convection-enhanced delivery of paclitaxel for the treatment of recurrent
malignant glioma: a phase I/II clinical study. J Neurosurg 100:472–479
73. Bruce JN et al (2011) Regression of recurrent malignant gliomas with convection-enhanced
delivery of topotecan. Neurosurgery 69(6):1272–1279
Chapter 5
Kif15: A Useful Target for Anti-cancer
Therapy?
Roy G.H.P. van Heesbeen and René H. Medema
5.1 Introduction
Centrosome separation and bipolar spindle formation is essential for chromosome

segregation during mitosis. The bipolar spindle is a highly dynamic structure
composed of microtubules (MTs) and many microtubule associated proteins
(MAPs). In mammalian cells, the centrosomes play an important role in regulating
spindle bipolarity, by acting as the major microtubule-organizing center (MTOC) in
cells. Before nuclear envelope breakdown (NEB), the centrosomes separate and
move to opposite sites of the nucleus. After NEB, MTs with their minus ends
embedded in the centrosomes, interact with the chromosomes with their plus ends
to form the bipolar spindle [1].
One of the main drivers of bipolar spindle formation is the highly conserved
homotetrameric kinesin Eg5 (also known as kinesin-5/Kif11). Due to its unique
tetrameric configuration, it is thought that Eg5 can crosslink antiparallel MTs and
subsequently slide them apart using its plus-end-directed motor activity. Indeed,
in vitro studies showed that Eg5 preferentially crosslinks MTs in an antiparallel
configuration and moves on both MTs that it crosslinks [2–4]. Depletion or inhibi-
tion of Eg5 blocks centrosome separation both before as after NEB and these cells
arrest in mitosis with a characteristic monopolar spindle [2, 5–8].
While previous data clearly indicate an essential role for Eg5 in bipolar spindle
assembly, recent studies indicate that additional, Eg5-independent pathways exist
that cooperate with Eg5 to drive bipolar spindle formation. It was shown that bipolar
spindle formation could be rescued by simultaneous inhibition of Eg5 and a
minus-end-directed motor (dynein or kinesin-14 [7, 9–12]). In addition, Eg5 activ-
ity is not essential for maintenance of the metaphase spindle, since established
R.G.H.P. van Heesbeen • R.H. Medema (*)

Division of Cell Biology, The Netherlands Cancer Institute,
Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands
e-mail: r.medema@nki.nl

78 R.G.H.P. van Heesbeen and R.H. Medema
bipolar spindles do not collapse into a monopolar state upon Eg5 inhibition [13–16].
Moreover, it was shown that both C. elegans and Dictostelium do not require Eg5
for bipolar spindle formation [17, 18].
The fact that Eg5 is dispensable for bipolar spindle formation under certain con-
ditions led to the suggestion that additional motors might exist that cooperate with
Eg5 to drive bipolar spindle formation. One of these main candidates to cooperate
with Eg5 is the kinesin-12 family member Kif15.
5.2 Identification of Kif15
Kif15 belongs to the kinesin-12 family and was originally identified as a novel
kinesin-like protein (KLP) in Xenopus oocytes [19]. Early studies showed that
Xklp2 (Xenopus Kif15) is a relatively slow plus-end-directed motor [20] and is
localized to the centrosomes and to the spindle MTs. The first evidence for a role in
centrosome separation and bipolar spindle assembly came from in vitro spindle
assembly experiments using injection of dominant negative Xklp2 protein and
inhibitory antibodies to the extracts [20]. Addition of a truncated Xklp2 construct,
lacking the N-terminal motor domain, blocked centrosome separation and led to the
formation of abnormal spindles, including monopolar spindles [20]. Similar to addi-
tion of truncated Xklp2 constructs, injection of Xklp2 inhibitory antibodies also
efficiently blocked centrosome separation and bipolar spindle formation and
resulted in the formation of monopolar spindles [20]. Interestingly, the study of
Boleti et al., showed that the C-terminal domain of Xklp2 was sufficient for its
localization to spindle poles [20]. This led to the suggestion that the C-terminus of
Xklp2, containing a conserved leucine-zipper (Fig. 5.1a), acts a non-motor MT tar-
geting domain. Follow-up studies by the Vernos lab identified a novel MAP, named
TPX2 (Targeting Protein for Xklp2) required for MT targeting of Xklp2 C-terminal
leucine zipper [21].
5.3 Kif15 Domain Structure
Similar to most other kinesins, Kif15 contains an N-terminal motor domain of about
350 amino acids that has about 40 % identity to other kinesins (Fig. 5.1a). The
C-terminal part of the protein is predicted to form a coiled coil structure, hereby
making it a unique member of the kinesin family. In addition to that, Kif15 contains
two conserved leucine zipper domains (Fig. 5.1a). The second, C-terminal leucine
zipper was previously identified to be required for MT targeting of Xenopus Xklp2
[20, 21]. Studies in human cells confirmed that Kif15 is recruited to spindle MTs
due to interaction between the C-terminal Leucine zipper and TPX2 and that the
motor domain is not required for MT targeting [15].
5 Kif15: A Useful Target for Anti-cancer Therapy? 79
#1
#2
a
g
in
er
er
in
nd
pp
pp
nd
bi
zi
zi
bi
u-
u-
TP
Le
Le
M
A
Kif15 / Xklp2 Motor Coiled Coil Coiled Coil
1 1388
g
in
r
in
pe
b
nd
nd
p
bi
zi
bi
u-
TP
Le
M
A
KLP-18 Motor Coiled Coil

1 932
Fig. 5.1 Schematic representation of protein domains present in Kif15. (a) Human Kif15 and
Xenopus Xklp2 are about 61 % identical and contain a N-terminal kinesin motor domain, a long
C-terminal coiled coil structure that is characteristic for kinesin-12 members. Furthermore, the
protein contains two conserved leucine zippers, of which Leucine zipper #2 is required for MT
targeting of the protein by its interaction with the MT binding protein TPX2. (b) The C. elegans
ortholog KLP-18 is less well conserved (about 41 % identical). Similar to other family members,
its C-terminus is predicted to form a coiled coil structure and it contains also a leucine zipper in the
C-terminus. Currently it is unknown if KLP-18 requires a TPX2-like protein for spindle
localization
While human Kif15, mouse Kif15 and Xenopus Xklp2 are highly conserved
(86 % and 61 % identical respectively), the C. elegans kinesin-12 KLP-18 is only
about 41 % identical and lacks a part of the extended C-terminus present in other
kinesin-12 members (Fig. 5.1b). Similar to other kinesin-12 family members, its
entire C-terminus is predicted to form a coiled coil structure. Although the
C-terminus of KLP-18 contains a leucine zipper motif, it is currently unknown if
KLP-18, similar to human and Xenopus Kif15, requires an interaction with a target-
ing protein for its spindle localization. An obvious ortholog of TPX2 has not been
identified in C. elegans, indicating why the C-terminus of KLP-18 might diverge so
much from the Kif15 orthologs in organisms that do have a clear TPX2 ortholog.
5.4 Kif15 Function in Centrosome Separation and Bipolar

Spindle Assembly
While early studies proposed important roles for Kif15 in centrosome separation
and bipolar spindle assembly, subsequent studies in different experimental spindle
assembly systems suggested that Kif15 is not essential for bipolar spindle assembly
[22, 23]. For example, studies using plasmid DNA-coated bead incubated with
Xenopus extracts showed that Xklp2 does not have an important role during spindle
assembly [22]. Only conditions in which Xklp2 was inhibited in combination with
either dynein or the minus-end-directed kinesin-14 XCTK2 led to increased spindle

defects compared to either inhibiting dynein or XCTK2 alone [22]. These results
where somewhat surprising as they contradicted the data from the original study
that led to the identification of Kif15 [20]. However, the experimental setup in the
study by Walczak et al., lacked centrosomes. This might suggest that Xklp2 is
specifically required to push centrosomes apart and is therefore not absolutely
required for acentrosomal spindle formation.
In addition to spindle assembly in Xenopus egg extracts, Kif15 was also found
not to be essential for spindle formation in human cells. This was demonstrated by
Zhu et al., in a systematic RNAi screen to test mitotic functions of all human kine-
sins and the dynein motor complex [23]. This also, suggested that Kif15 function is
either redundant or not required for centrosome separation and bipolar spindle
assembly.
Subsequent evidence for a potential role of Kif15 in bipolar spindle assembly
came from studies conducted in sea urchins and C. elegans [24, 25]. The sea urchin
kinesin-12 KRP180 localizes to interpolar MTs and was shown to be important to
maintain a bipolar spindle in prometaphase embryos [24]. Injection of inhibitory
antibodies or expression of a dominant negative mutant in prometaphase resulted in
bipolar spindle collapse, in which the poles moved towards each other in a linear
fashion [24]. This data suggested that KRP180, similar to Eg5 in other experimental
systems, might slide MTs outward, hereby promoting centrosome separation and
spindle elongation. In the absence of KRP180, antagonistic minus-end-directed
motors, such as dynein and/or kinesin-14 could possibly drive this spindle collapse
and promote monopolar spindle formation.
While Eg5/kinesin-5 was reported to be essential for bipolar spindle assembly in
nearly all organisms tested, the sole C. elegans kinesin-5, BMK-1 was shown not to
be essential [17]. Bmk-1 homozygous deletion mutants do display reduced fertility,
but these animals are otherwise fully viable and do not display any defects in bipolar
spindle formation or cytokinesis, suggesting that in C. elegans, other motors must
be present that drive centrosome separation and bipolar spindle assembly [17].
Interestingly, while Eg5/kinesin-5 is not essential in C. elegans, the C. elegans kine-
sin-12 KLP-18 is essential [25]. Similar to kinesin-12 homologs in other organisms,
KLP-18 localizes to the spindle poles during prometaphase, metaphase and early
anaphase and subsequently transfers to the spindle midzone during late anaphase
and telophase [25]. RNAi depletion of KLP-18 showed that its function is specifi-
cally required for acentrosomal (female meiotic) spindle formation, causing the
mutant to produce embryos with abnormal maternal DNA content [25]. This is in
striking contrast to its proposed function in Xenopus, where Xklp2 seems to be
required for spindle assembly in the presence of centrosomes [20]. The exact molec-
ular mechanism how KLP-18 contributes to meiotic spindle assembly in C. elegans
is not fully understood, but its depletion prevents parallel bundling of MTs that are
nucleated near meiotic chromosomes and as a result, the formation of disorganized
spindles [25]. This might also indicate why KLP-18 is not absolutely required for
mitotic divisions that occur in the presence of centrosomes, as centrosomes can play
a dominant role in the ordering of the MTs, by sequestering the minus ends towards
the centrosomes. MTs in acentrosomal spindles are nucleated in random directions

at or near the chromatin and under these circumstances KLP-18 is likely essential to
organize and bundles these MTs into (anti)-parallel arrays. A more recent study
indicated that KLP-18 is required for the establishment and/or maintenance of spin-
dle bipolarity in oocytes; high-resolution analysis of KLP-18 RNAi-treated worms
indicated the spindles in oocytes are monopolar [26]. Similar to previous studies,
this study shows that KLP-18 is only required for bipolar spindle formation in meio-
sis, suggesting that in mitotic cell divisions of C. elegans, the possible contribution
of KLP-18, but also of Eg5/kinesin-5 in bipolar spindle formation, might be masked
by other motors.
In addition to C. elegans, two recent studies in human cells also provided clear
evidence for a role of Kif15 in bipolar spindle formation [15, 16]. Depletion of
Kif15 in human cells strongly sensitized cells to partial Eg5 inhibition [15, 16],
indicating that Kif15 cooperates with Eg5 during bipolar spindle assembly. In addi-
tion to that, the maintenance of a bipolar spindle in Eg5-inhibited cells was shown
to be fully dependent of Kif15 [15, 16]. This indicates that Kif15 acts redundant
with Eg5 during bipolar spindle formation; cells form normal bipolar spindles in the
absence of Kif15 and do not show major spindle defects [15, 16] while metaphase
spindle length in Kif15-depleted cells was on average only 11 % shorter compared
to control cells [15, 16]. Conversely, inhibition of Eg5 in these human cells prevents
bipolar spindle formation, indicating that the activity of Kif15 is insufficient to pro-
mote centrosome separation and bipolar spindle formation in the absence of Eg5
activity. However, over-expression of Kif15 in Eg5-inhibited cells fully rescued
bipolar spindle formation [15]. These results strongly suggest that Kif15 is, similar
to Eg5, able to generate an outward force during bipolar spindle formation. In line
with data from Xenopus, human Kif15 requires TPX2 for its function [15, 16].
Depletion of TPX2 or expression of Kif15 truncation mutants lacking the C-terminal
leucine zipper, prevents Kif15 from binding to the spindle MTs [15, 16] and these
mutants fail to promote bipolar spindle formation in cells lacking Eg5 activity. It is
however, not completely clear why the endogenous Kif15 is incapable of promoting
bipolar spindle formation under circumstances where Eg5 is inhibited. Possibly, this
is due to low levels of endogenous expression, but it should also be noted that TPX2
is strictly nuclear during interphase and prophase. Due to the physical separation of
Kif15 and TPX2 during prophase, Kif15 is unable to replace Eg5's function during
prophase centrosome separation [15]. In the absence of Eg5 activity cells will there-
fore enter mitosis with the centrosomes in close proximity, regardless of the expres-
sion level of Kif15. Indeed, the capacity of Kif15 to drive bipolar spindle formation
strictly depends on the initial separation of the centrosomes; cells with partial loss
of Eg5 activity only assembled bipolar spindles when their centrosomes were at
least partially separated during prophase [15]. In addition to this, Eg5-independent
human cells (EICs) described in a recent study, also critically depend on prophase
centrosome separation in order to assemble a bipolar spindle [11]. These cells, that
can proliferate in the complete absence of Eg5 activity, require both nuclear enve-
lope (NE)-associated dynein and Kif15 for prophase centrosome separation and
bipolar spindle formation respectively [11, 27]. Thus, the data available at present
+ +
+
+ - - - Kif15
- +
- -
+ - - +
TPX2
- +
+ +
Fig. 5.2 Model of Kif15 function during bipolar spindle formation. TPX2 targets Kif15 to the
spindle through the interaction with the C-terminal leucine zipper of Kif15. The complex of Kif15
and TPX2 can crosslink anti-parallel MTs and the plus-end-directed motility of Kif15 can slide
anti-parallel MTs apart, hereby generating a force similar to Eg5. The arrows indicate the move-
ment of motor (red) and the MTs (green)
are most compatible with a model in which Kif15, in complex with TPX2 can cross-
link and slide antiparallel MTs (Fig. 5.2).
5.5 Kif15 Function in the Regulation of Kinetochore-

Microtubule Dynamics
A recent study suggested that Kif15 and Eg5 act on distinct sets of MTs [28]. While
Eg5 does not seem to discriminate between kinetochore-microtubules (KT-MTs)
and interpolar MTs, Kif15 recruitment to the spindle seems to depend primarily on
KT-MTs and the spindle-associated pool of Kif15 increases in metaphase cells com-
pared to prometaphase cells [28]. Indeed, bipolar spindles collapsed upon Eg5 inhi-
bition in cases where the KT-MT where removed, while unperturbed bipolar
spindles did not collapse upon Eg5 inhibition. These data indicate that Kif15’s con-
tribution to bipolar spindle maintenance is likely mediated via KT-MTs, instead of
interpolar MTs [28]. Strikingly, Kif15-depleted cells initially build a longer meta-
phase spindle, followed by a contraction phase until the spindle reaches the charac-
teristically shorter spindle [28]. This contraction phase is KT-MT-dependent since
cells co-depleted of Kif15 and Nuf2 produced longer spindles that do not show the
subsequent contraction phase [28]. These data suggest that at physiological levels,
Kif15 depends on KT-MTs for force generation. Related to Kif15’s function in reg-
ulating KT-MT length, a recent paper identified Eg5 and Kif15 as key regulators of
chromosome movements, by being part of a possible KT-MT crosslinking structure
[29]. This potential crosslinking structure mediates coupled movement of neighbor-
ing chromosomes during mitosis. Although the molecular mechanism underlying
this coupled movement of neighboring chromosomes is still unknown, Kif15 is sug-
gested to act as a mechanical coupling factor of KT-MTs in metaphase and is lim-
ited in its action by Eg5 [29]. Inhibition of Eg5 leads to a loss of periodic chromosome
oscillations and increased coupling of neighboring chromosomes that can be par-

tially rescued by Kif15 depletion [29]. Similar to the study of Sturgill and Ohi, Eg5
and Kif15 seem to antagonize each other in this process.
5.6 Kif15-Dependent Bipolar Spindle Assembly

in Eg5-Independent Cells
While the study of Sturgill et al. suggests that Kif15 can antagonize the activity of
Eg5, there is a considerable amount of evidence that indicates that Kif15 can
produce a force within the spindle that can separate the centrosomes. The strongest
support in favor of a role for Kif15 during centrosome separation comes from the
fact that Kif15 overexpression can fully rescue bipolar spindle assembly when the
activity of Eg5 is blocked [15]. Additionally, several labs have recently generated
human cell lines that can proliferate in the absence of Eg5, but in contrast to the
parental cells the Eg5-independent derivatives have become absolutely dependent
on Kif15 activity for bipolar spindle assembly [11, 28]. The OL-EIC1 cells,
generated by the Ohi lab depend on the over-expression of Kif15 for survival. The
data generated with these cells indicate that while Kif15, localized to KT-MTs
opposes Eg5 activity during bipolar spindle assembly in normal cells, Kif15
over-expression in OL-EIC1 cells results in mislocalization of Kif15 to interpolar
MTs, hereby generating a force that can separate centrosomes [28]. EICs generated
independently in our own lab were also shown to depend on Kif15 activity, but in
addition, we found that NE-associated dynein is also essential in these cells in order
to promote prophase centrosome separation when Eg5 activity is absent [11]. Due
to the activity of this dynein-dependent prophase pathway, these cells enter pro-
metaphase with separated centrosomes, hereby creating a situation that likely results
in a strong bias towards bipolar spindle assembly as has been shown previously [9,
15]. Importantly, two of the three EICs clones that were generated in our lab did not
show any signs of Kif15 over-expression, which shows that physiological levels of
Kif15 can drive centrosome separation. If this outward force produced by
endogenous Kif15 depends on its binding to KT-MTs, it seems most logical that this
KT-MT-associated Kif15 can crosslink with an anti-parallel interpolar MT and
slide the MTs apart. Further studies need to be conducted to clarify this specific
function of Kif15.
5.7 Kif15 as a Potential Target for Drug Development
Currently, it is unclear if Kif15 plays an important role in cancer. Data from cancer
genome expression show no clear up- or downregulation of Kif15. Also, Kif15
mutations are found in not more than 1 % of cancer genomes present in the COSMIC
database, from which 75 % are missense substitutions, 13 % are synonymous
substitutions and 4 % are nonsense substitution. None of the identified missense

mutations have thus far been shown to interfere with Kif15 function.
In contrast to cancer genome data, recent studies indicate that alteration in Kif15
expression levels might have major implications for anticancer therapies that
employ inhibitors of Eg5 to inhibit cell proliferation [11, 15, 28]. Tanenbaum et al.,
showed that five to tenfold ectopic overexpression of Kif15 is sufficient to drive
bipolar spindle assembly in the absence of Eg5 activity [15]. Follow-up studies in
which EICs were generated using an in vitro ‘directed evolution’ approach indicated
that enhanced expression of the endogenous Kif15 can be induced by prolonged
exposure to Eg5 inhibitors and this might contribute to resistance against these
inhibitors [11, 28]. While it is clear that other mechanisms besides Kif15 over-
expression can result in the growth of EICs [11], all EICs lines that have currently
been tested invariably depend on Kif15 activity, also in cases where the protein is
not overexpressed [11, 28].
In addition to the studies describing EICs, a recent study found that Eg5 activity
becomes dispensable for bipolar spindle assembly in cells with altered MT dynam-
ics [30]. Treatment of cancer cells with low doses of nocodazole (suppresses MT
dynamics) or treatment with siRNAs against TOGp (a MAP required for MT nucle-
ation) rescued bipolar spindle assembly in cells treated with an Eg5 inhibitor [30].
Live cell analysis of spindle formation in the cells with reduced MT dynamics
revealed that these cells enter mitosis with non-separated centrosomes, followed by
the formation of multiple MTOCs that eventually organize into a bipolar spindle
with separated centrosomes [30]. Similar to EICs, these cells now fully depend on
the activity of Kif15 for bipolar spindle assembly by [30].
The identification of these Eg5-independent mechanisms for centrosome separa-
tion and bipolar spindle assembly have important implications for anticancer thera-
pies that employ Eg5 inhibitors to perturb cell division. The data from these studies
indicate that Kif15 is the central player driving Eg5-independent bipolar spindle
assembly. First, Kif15 is required for maintenance of bipolar spindle in metaphase
when Eg5 is inhibited [15, 16]. Second, Kif15 is required for bipolar spindle assem-
bly in Eg5-inhibited cells with reduced MT dynamics [30]. Third, Kif15 over-
expression can bypass the requirement of Eg5 in human cells [15]. In fact these cells
can form a normal bipolar spindle when Eg5 activity is completely blocked, and
their proliferative capacity is not compromised by the inhibition of Eg5 [15]. As
such, mere overexpression of Kif15 appears to be sufficient to promote resistance to
Eg5 inhibitors. Finally, recent results from our lab also indicate that bipolar spindle
assembly in cells co-depleted of both Eg5 and the minus-end-directed motor dynein
depends on Kif15 activity [31]. Thus, while Eg5 inhibitors show potent anti-tumor
activity both in vitro and in vivo (reviewed in [32]), these recent studies indicate that
changes in MT dynamics or upregulation of Kif15 expression can easily bypass Eg5
inhibition and promote drug resistance. Currently, it is unclear if upregulation of
Kif15 occurs in tumors that are treated with Eg5 inhibitors and show signs of resis-
tance. Nonetheless, since multiple Kif15-dependent mechanisms can easily pro-
mote resistance to Eg5 inhibitors, the eventual success of an anti-tumor strategy
involving inhibition of Eg5 will likely require co-administration of Kif15 inhibitors
to prevent rapid acquisition of therapy resistance and the combination might also
prove to have a more potent anti-tumor activity. Obviously, such a combinatorial
regimen might trigger more pleiotropic side effects, especially if Kif15 also has
functions that are not limited to driving bipolar spindle assembly in mitosis. In this
respect it is important to note that the essential co-factor for Kif15's role in spindle
assembly, TPX2, is localized in the nucleus in interphase. Thus, if one were to be
able to specifically target the interaction between Kif15 and TPX2, this could pos-
sible limit interference to the mitotic function of Kif15 only.
Members of the kinesin-12 family play an evolutionary conserved role during bipo-
lar spindle assembly. Although the exact contribution of kinesin-12 members to
bipolar spindle assembly in different organisms is still not fully understood, it is
clear that in human cells, Kif15 is a central player in Eg5-indepdent mechanisms of
bipolar spindle assembly. Inhibitors of Kif15 might therefore increase the efficacy
of Eg5 inhibitors and are likely to reduce the occurrence of resistance to inhibitors
of Eg5. Development of specific Kif15 inhibitors could therefor be of great value in
therapy and will certainly be useful to study the functions of Kif15 in more detail.
References
1. Tanenbaum ME, Medema RH (2010) Mechanisms of centrosome separation and bipolar spin-
dle assembly. Dev Cell 19(6):797–806
2. Kashina AS et al (1996) A bipolar kinesin. Nature 379(6562):270–272
3. Kapitein LC et al (2005) The bipolar mitotic kinesin Eg5 moves on both microtubules that it
crosslinks. Nature 435(7038):114–118
4. van den Wildenberg SMJL et al (2008) The homotetrameric kinesin-5 KLP61F preferentially
crosslinks microtubules into antiparallel orientations. Curr Biol 18(23):1860–1864
5. Blangy A et al (1995) Phosphorylation by p34cdc2 regulates spindle association of human
Eg5, a kinesin-related motor essential for bipolar spindle formation in vivo. Cell
83(7):1159–1169
6. Sawin KE et al (1992) Mitotic spindle organization by a plus-end-directed microtubule motor.
Nature 359(6395):540–543
7. Tanenbaum ME et al (2008) Dynein, Lis1 and CLIP-170 counteract Eg5-dependent centro-
some separation during bipolar spindle assembly. EMBO J 27(24):3235–3245
8. Whitehead CM, Rattner JB (1998) Expanding the role of HsEg5 within the mitotic and post-
mitotic phases of the cell cycle. J Cell Sci 111(Pt 17):2551–2561
9. Ferenz NP et al (2009) Dynein antagonizes Eg5 by crosslinking and sliding antiparallel micro-
tubules. Curr Biol 19(21):1833–1838
cross-links microtubules in the mammalian mitotic spindle. J Cell Biol 147(2):351–366
11. Raaijmakers JA et al (2012) Nuclear envelope-associated dynein drives prophase centrosome
separation and enables Eg5-independent bipolar spindle formation. EMBO J 31(21):4179–4190
12. Sharp DJ et al (1999) Antagonistic microtubule-sliding motors position mitotic centrosomes in

Drosophila early embryos. Nat Cell Biol 1(1):51–54
13. Kapoor TM et al (2000) Probing spindle assembly mechanisms with monastrol, a small mol-
ecule inhibitor of the mitotic kinesin, Eg5. J Cell Biol 150(5):975–988
14. Kollu S, Bakhoum SF, Compton DA (2009) Interplay of microtubule dynamics and sliding
during bipolar spindle formation in mammalian cells. Curr Biol 19(24):2108–2113
15. Tanenbaum ME et al (2009) Kif15 cooperates with Eg5 to promote bipolar spindle assembly.
Curr Biol 19(20):1703–1711
16. Vanneste D et al (2009) The role of Hklp2 in the stabilization and maintenance of spindle
bipolarity. Curr Biol 19(20):1712–1717
17. Bishop JD, Han Z, Schumacher JM (2005) The Caenorhabditis elegans Aurora B kinase AIR-2
phosphorylates and is required for the localization of a BimC kinesin to meiotic and mitotic
spindles. Mol Biol Cell 16(2):742–756
18. Tikhonenko I et al (2008) Kinesin-5 is not essential for mitotic spindle elongation in
Dictyostelium. Cell Motil Cytoskeleton 65(11):853–862
19. Vernos I, Heasman J, Wylie C (1993) Multiple kinesin-like transcripts in Xenopus oocytes.
Dev Biol 157(1):232–239
20. Boleti H, Karsenti E, Vernos I (1996) Xklp2, a novel Xenopus centrosomal kinesin-like protein
required for centrosome separation during mitosis. Cell 84(1):49–59
21. Wittmann T et al (1998) Localization of the kinesin-like protein Xklp2 to spindle poles requires
a leucine zipper, a microtubule-associated protein, and dynein. J Cell Biol 143(3):673–685
22. Walczak CE et al (1998) A model for the proposed roles of different microtubule-based motor
proteins in establishing spindle bipolarity. Curr Biol 8(16):903–913
16(7):3187–3199
24. Rogers GC et al (2000) A kinesin-related protein, KRP(180), positions prometaphase spindle
poles during early sea urchin embryonic cell division. J Cell Biol 150(3):499–512
25. Segbert C et al (2003) KLP-18, a Klp2 kinesin, is required for assembly of acentrosomal mei-
otic spindles in Caenorhabditis elegans. Mol Biol Cell 14(11):4458–4469
26. Wignall SM, Villeneuve AM (2009) Lateral microtubule bundles promote chromosome align-
ment during acentrosomal oocyte meiosis. Nat Cell Biol 11(7):909–913
27. van Heesbeen RGHP et al (2013) Nuclear envelope-associated dynein cooperates with Eg5 to
drive prophase centrosome separation. Commun Integr Biol 6(3):e23841
microtubule substrate. Curr Biol 23(14):1280–1290
29. Vladimirou E et al (2013) Nonautonomous movement of chromosomes in mitosis. Dev Cell
27(1):60–71
30. Florian S, Mayer TU (2011) Modulated microtubule dynamics enable Hklp2/Kif15 to assemble
bipolar spindles. Cell Cycle 10(20):3533–3544
31. van Heesbeen RGHP et al (2014) Balanced activity of three mitotic motors is required for
bipolar spindle assembly and chromosome segregation. Cell Reports 8(4):948–956
32. Rath O, Kozielski F (2012) Kinesins and cancer. Nat Rev Cancer 12(8):527–539
Chapter 6
Down-Regulating CENP-E Activity:
For Better or for Worse
Julien Espeut and Ariane Abrieu
6.1 Introduction
CENP-E (KIF10) is the largest kinesin and unique member of the kinesin-7 family
in human. CENP-E is an essential gene, since knocking it down is lethal very early
during mice embryonic development [1]. Down-regulating CENP-E invariably per-
turbs chromosome congression to the metaphase plate. This was shown by CENP-E
depletion in Xenopus egg extracts [2], micro-injection of CENP-E blocking anti-
bodies in HeLa cells [3], gene inactivation in Drosophila [4], CENP-E antisense
oligonucleotides [5], siRNA in HeLa cells [6], CENP-E gene knock-out in mice [1],
and CENP-E inhibition by an allosteric inhibitor in human cancer cell lines [7]. The
most striking consequence of CENP-E depletion is the occurrence of mono-oriented
kinetochores in the vicinity of the spindle poles. There, the chromosomes are out of
reach for chromokinesin-dependent forces to push CENP-E-/- kinetochores toward
the metaphase plate, independently of kinetochore-to-microtubule attachment [1].
CENP-E can glide along a microtubule track using its N-terminus motor domain,
although being hooked at a kinetochore that is not yet properly attached [8]
(Fig. 6.1a). This model is further supported by electron microscopy analysis of full-
length CENP-E, which reveals a flexible “spaghetti-like” structure, over 200 nm
long that can potentially interact with distant microtubules [9] (Fig. 6.1b). Finally,
it was recently demonstrated that CENP-E can glide bidirectionally [10] (Fig. 6.1c).
It is the combination of these properties that allows CENP-E to drive kinetochores
toward the equatorial plate.
J. Espeut • A. Abrieu (*)

Université Montpellier, CRBM, 34293 Montpellier, France
CNRS UMR 5237, 1919 Route de mende, 34293 Montpellier, France
e-mail: ariane.abrieu@crbm.cnrs.fr

88 J. Espeut and A. Abrieu
Fig. 6.1 CENP-E drives chromosomes to the metaphase plate. (a) CENP-E can glide along
adjacent spindle microtubules in order to displace mono-oriented chromosomes to the equatorial
plate. The plus-end directed motility of CENP-E is necessary for this movement. (b) CENP-E is a
very long and flexible protein that can be over 200 nm long. Its length should allow it to bind to
distant microtubules. (c) CENP-E movement can be bidirectional due to the combination of two
microtubule-binding domains. The plus-end directed motion of its motor domain pulls the chromo-
somes toward the metaphase plate and the fast bidirectional diffusion of its tail domain allows
kinetochore-microtubules tether. Representation not to scale
6.2 CENP-E Mechanics
6.2.1 CENP-E Is a Molecular Machine That Tethers

Kinetochores to Dynamic Microtubules
Despite its gigantic size, CENP-E is a genuine motor whose motility has been mea-
sured using human or Xenopus motor domains purified from prokaryote expression
systems or Xenopus full-length molecules purified from baculovirus-infected insect
cells. Crystal structure of its motor domain revealed how similar it is to conven-
tional kinesin [11], and more recent experiments demonstrated that CENP-E is a
processive motor [9, 10, 12, 13]. Determining its exact speed might be challenging
since in vitro studies reported CENP-E velocities ranging from 8 to 340 nm per
second [2, 9, 10, 12, 14, 15].
Classical techniques using taxol-stabilized microtubules concluded that CENP-E
moves unidirectionally toward the microtubule plus-end [2, 9, 14]. This finding was
convenient because (i) it fitted the dogma that kinesins bearing their motor domain
at their N-terminus are usually plus-end directed (with the notable exception of the
kinesin-14 family), and (ii) it was in keeping with its undisputed function: bringing
kinetochores toward the plus-tips of microtubules at the metaphase plate. The odd
one out persisted for almost 20 years: how then could CENP-E blocking antibodies
inhibit chromosome motion driven by microtubule depolymerization in vitro [16]?
Thanks to a recent technological breakthrough allowing visualization of CENP-E
6 Down-Regulating CENP-E Activity: For Better or for Worse 89
onto more physiologically relevant, dynamic microtubules, it is now known that

CENP-E can move bidirectionally in vitro. This unique property requires (i) plus-
end directed motion of its motor domain, (ii) bidirectional diffusion of its tail
domain, and (iii) fast diffusion rates for CENP-E tail (100 times faster than Kif18A
for instance), compared to microtubule dynamic kinetics [10]. These recent findings
thus support a second proposed function for CENP-E as a tether for kinetochores to
remain hooked onto dynamic and unstable microtubules [2]. In support of this
hypothesis, it was recently established that CENP-E tail binds to microtubules with
an affinity comparable to that of the NDC80 complex, a conserved core component
of the microtubule-kinetochore tether [17].
A mathematical model of CENP-E motility was built, highlighting how slowly
microtubules depolymerize compared to CENP-E kinetic parameters. However, one
should keep in mind that this model assumes that both CENP-E motor and tail
domains follow the same microtubule track [10]. But there is no evidence that this
is the case in vivo. No matter how attractive this model is, the in vivo situation is
likely to be much more complicated. Indeed, the number of microtubules to which
CENP-E motor and tail domains could simultaneously bind to, is very large.
Reported to the external microtubule diameter of 25 nm, the motor domain of
kinetochore-bound CENPE is thus potentially able to interact with dozens of micro-
tubule tracks within the three-dimension environment of the mitotic spindle
(Fig. 6.2). There must thus exist some regulatory mechanisms to prevent CENP-E
from “making knots with its leash”. What could they be?
First of all, the biophysical parameters of CENP-E motility could bias its interac-
tions with stable over dynamic microtubules, therefore favoring its interaction with
kinetochore fibers [18]. Additionally, CENP-E is not alone at the kinetochore: sev-
eral binding partners, such as BUBR1, CLASP1/2 and PP1 could help CENP-E to
choose which microtubule(s) to bind to [5, 19, 20]. Also, proteomic studies revealed
that CENP-E is heavily phosphorylated [21]. The phosphorylation of one site in
CENP-E motor domain by Aurora kinases was shown to be required for CENP-E
dependent motility of the kinetochore to the metaphase plate, but its PP1-dependent
dephosphorylation and subsequent PP1-binding are also essential for kinetochores
to maintain stable attachments [20]. Another possible level of regulation is the one
that we proposed, showing that CENP-E can be auto-inhibited via direct motor-tail
interaction, which we could relieve via CDK1- or MPS1-dependent phosphorylation
in vitro [14]. Besides, CENP-E kinetochore localization depends upon SUMOylation
of its tail [22]. Moreover, the tip of CENP-E C-terminal tail is farnesylated in human
cells, which affects its binding to microtubule but does not alter its recruitment to
kinetochores [23]. This is logical since we now know that CENP-E kinetochore-
binding domain lies more internally than its C-terminal microtubule-binding domain
[14, 24] (Fig. 6.3). Finally, its ubiquitylation (independently of its action on CENP-E
proteasome-dependent degradation) might regulate CENP-E binding to kineto-
chores [25]. Owing to (i) the gigantic size of the CENP-E molecule, (ii) the variety
of its possible post-translational modifications that includes (but might not be lim-
ited to) phosphorylation, SUMOylation, farnesylation and ubiquitinylation, and (iii)
the fast kinetics at which these could be turned on and off, a thorough understanding
of CENP-E dynamics at the kinetochore might never be achieved.
<-nm->
<-nm-> <-nm->
<-nm->
<-nm->
<-nm->
<-nm->
<-nm->
<-nm->
<-nm->
<-nm->
<-nm->
<-nm->
<-nm->
<-nm->
<-nm->
<-nm->
<-nm->
<-nm->
<-nm->
<-nm->
<-nm-> <-nm->
<-nm->
0 20 40 60 80 100 120 140 160 180 200 nm
Fig. 6.2 CENP-E has the potential to interact with several dozens of microtubule tracks
from the kinetochore. Microtubules are represented as empty tubes (circles) observed from the
kinetochore. Each microtubule possess 13 possible tracks between protofilaments. CENP-E could
interact with microtubule bundles constituting the kinetochore fiber (k-fiber in dark green), and to
non-bundled microtubules (lighter green) from the mitotic spindle. The blue scale represents the
length of the CENP-E molecule relative to the 25 nm microtubule diameter
6.2.2 CENP-E and Microtubule Dynamics at the Kinetochore
Owing to its two microtubule binding domains, CENP-E could directly regulate
microtubule dynamics at the kinetochore, but this has not yet been shown. In vitro,
CENP-E motor domain slightly stimulates microtubule elongation [12], but micro-
tubule growth and shortening rates or catastrophe frequency are not affected by
full-length CENP-E [10]. In cellulo, the situation is different, because CENP-E
recruits and/or interacts with several proteins that regulate microtubule dynamics.
For instance, by recruiting CLASPs at the kinetochore, CENP-E promotes microtu-
bule poleward flux and increased turnover rates [19]. In the same vein, an interac-
tion was demonstrated between CENP-E and SKAP, a protein regulating microtubule
flux rate and kinetochore attachment to the mitotic spindle [26, 27]. Also, CENP-E
interacts with Nuf2 [28], a component of the Ndc80 complex, the major complex
binding microtubules at the kinetochore [29]. Finally, CENP-E could participate in
regulation of microtubule dynamics via its interaction with BUBR1, which
NDC80
Complex
Mis 12
Complex KNL-1
CCAN
BubR
Tail Microtubule Motor
CENP-A Kinetochore domain domain
CENP-E
nucleosome binding
domain PP
Nuf2 SKAP CLASP PRC

CENP-A
0 20 40 60 80 100 120 140 160 180 200 220 240 nm
Fig. 6.3 CENP-E is a very large protein at the interface between chromosomes and the
mitotic spindle. Schematic view of CENP-E and its binding partners at the kinetochore. Scale and
location along the kinetochore is respected for CENP-A nucleosome, Mis12 complex, NDC80
complex, KNL-1, CENP-E and the microtubule [84, 85]. CENP-E was represented according to
the location of the human CENP-E antibodies [84] targeting the regions encompassing amino acids
663–973 and 1571–1859. CENP-E interacts with several partners at the kinetochore-microtubule
interface (purple circles not represented to scale). The known interactors are microtubule binding
proteins (Nuf2, SKAP, CLASP, PRC1) or checkpoint regulating elements (BubR1, PP1). The
kinetochore-binding domain of CENP-E is located at the C-terminus of the protein but is distinct
from the second microtubule-binding site that lies in the tail domain. On the other side of the mol-
ecule, the N-terminal motor domain performs 8 nm steps toward the microtubule plus-end [10, 14]
stabilizes microtubules [30]. Whether interaction between CENP-E and SKAP,

Nuf2, or BUBR1 require CENP-E ATPase activity is unknown. On the contrary,
CENP-E dependent CLASPs kinetochore recruitment was clearly shown not to
require CENP-E motor activity [19, 31]. Since there is now an inhibitor of CENP-E
that targets its ATPase activity, it will be possible to address whether these interac-
tions require CENP-E enzymatic activity.
6.2.3 CENP-E and the Checkpoint
Besides its role in regulating chromosome positioning at the metaphase plate,

CENP-E could regulate the spindle assembly checkpoint that monitors the
metaphase-to-anaphase transition. This was clearly shown in Xenopus egg
extracts where CENP-E is required to recruit MAD2 at the kinetochore and to
activate BUBR1 [32, 33]. It was also shown that microtubule capture by CENP-E
could silence checkpoint activation through direct inhibition of BubR1 kinase
activity [34]. Furthermore, CENP-E directly interacts with BUBR1 in human
cells and Xenopus egg extracts. Deletion mutants revealed that it is a region
encompassing CENP-E kinetochore-binding domain that interacts with BUBR1
kinase domain [5, 24, 33]. In human cells, CENP-E is not absolutely essential for
checkpoint activation, but reduction of CENP-E diminishes the strength of the
mitotic checkpoint, as well as BUBR1 activation and recruitment to the kineto-
chore [6, 35, 36]. Although a recent study challenges the view that BUBR1 is a
genuine kinase [37], the role of BUBR1 as an essential player of the checkpoint
remains undisputed [38, 39].
6.2.4 Does CENP-E Fulfill Other Functions Once

Chromosomes Are Separated Post-Metaphase?
Upon anaphase, most CENP-E leaves the kinetochore and switches to spindle mid-
zone microtubules, where it remains localized until the formation of the midbody
[40, 41]. At the end of mitosis, CENP-E is degraded by the ubiquitin-dependent
proteasome pathway [42, 43]. CENP-E degradation occurs roughly 1 h after that of
cyclin B [44], which allows inquiring whether CENP-E could fulfill other functions
post-metaphase, outside of the kinetochore. Evidence that CENP-E could play a
role post-metaphase is rather scarce. One hint could be that CENP-E interacts with
septins, which are mainly known for their role during cytokinesis [45, 46]. However,
these studies showed that septins are required to recruit CENP-E at the kinetochore
in prometaphase, but whether and how the interaction between CENP-E and septins
is relevant to later mitotic stages is not yet understood. Another partner that could
link CENP-E to cytokinesis is PRC1 [47]. Alone, PRC1 bundles antiparallel micro-
tubules in vitro, and is likely to be an essential player in organizing the central
anaphase spindle [48, 49]. Whether or not CENP-E participates at this task remains
to be determined. Finally, CENP-E was shown to interact with Skp1 at the midbody,
leading to its degradation at the end of mitosis. Affecting this interaction perturbs
chromosome segregation and cytokinesis [50]. Investigating the role of CENP-E
post-metaphase has been precluded by the fact that CENP-E inhibition/disruption
halts the cells earlier in mitosis. Hopefully, the recent development of CENP-E
specific inhibitors will soon allow circumventing this impediment.
6.3 CENP-E and Cancer
6.3.1 CENP-E and Aneuploidy
As described above, CENP-E depletion/inhibition halts the cell cycle in metaphase

in mammalian cells. There are two possible outcomes to this long-term arrest. In one
scenario, long-term arrest during mitosis activates a process called “mitotic catastro-
phe”, i.e. a form of cell death that bears most hallmarks of apoptosis. Alternatively,
the cells not being able to sustain mitotic arrest forever, cyclin B will eventually be
degraded and the cells will exit mitosis. This process that is called “mitotic slip-
page” often produces aneuploid cells [51, 52]. If cytokinesis occurs thereafter, the
daughter cells will be pseudo-diploid, while they will be pseudo-tetraploid in case
of subsequent cytokinesis failure. This latter situation occurs for instance after treat-
ment with microtubule-depolymerizing drugs that perturb both chromosome segre-
gation and cytokinesis furrow ingression. Later on, the chromosome segregation
errors that are often associated with DNA double strand break, should be detected
by the DNA damage checkpoint [53]. This is however dependent upon the presence
of p53, whose pathway is actually altered in almost all cancer cells [54].
In short, if mitotic catastrophe is not activated, the most likely outcome of
long-term mitotic arrest is aneuploidy. Evidence is mounting that promoting
aneuploidy could be deleterious to tumor cells: (i) this was first shown in mice when
high levels of aneuploidy were induced [55], (ii) in cellulo experiments revealed
that elevating aneuploidy levels by perturbing simultaneously microtubule dynamics
and the spindle assembly checkpoint activity, could be used as a strategy to kill
cancer cells [56] and (iii) it was elegantly demonstrated in the yeast S. cerevisiae
and in mammalian cells, that aneuploidy confers a proliferative disadvantage, due to
protein imbalance [57, 58]. Indeed, in yeast a common feature of aneuploidy-toler-
ating cells is to up-regulate the ubiquitin-proteasome dependent pathway that is
necessary to degrade extra/misfolded proteins occurring from genetic imbalance
[59]. This important new finding could provide the Achilles’ heel whereby tumor
cells could be targeted without affecting the healthy cells of the organism. Following
this lead, the same team headed by Angelica Amon, showed that inhibiting protein
folding or autophagy were more efficient in preventing growth in xenograft mouse
models of aneuploid versus near-diploid cells [60].
Promoting aneuploidy by inhibiting CENP-E could be a double-edged sword
strategy since it can either inhibit or promote tumorigenesis [55]. It was also shown
in yeast that, as was proposed more than a century ago by Theodor Boveri, aneu-
ploidy actually drives genetic instability [61]. How to push the balance toward cell
death rather than growth of aneuploid cells is an important challenge whose resolu-
tion is likely to affect the efficiency of therapies using anti-mitotic drugs. Addressing
this issue will require better understanding of the pathways leading to cell death
upon mitotic arrest, which is beyond the scope of the current review.
6.3.2 CENP-E Levels Are Perturbed in Various Cancer Cells
CENP-E expression levels are misregulated in a wide range of cancers. While

CENP-E levels are downregulated in hepatocellular carcinoma [62], CENP-E
expression is upregulated in most cancer types [63]. CENP-E overexpression is
associated with bad prognosis in breast cancer [64], sarcoma [65], prolactin pitu-
itary tumors [66], and glioma [67]. CENP-E is also found upregulated in chemore-
sistant epithelial ovarian cancer [68] compared to chemosensitive counterparts. In
addition, whole-exome sequencing revealed a CENP-E missense mutation in one

high-risk family for colorectal cancer [69]. Follow-up studies will be required to
determine to which extent this mutation participates in carcinogenesis.
6.3.3 CENP-E Inhibitors
In 1998, the company Cytokinetics was launched to explore the potential of kinesins
as novel drug targets, based on the recent characterization of the first kinesin inhibi-
tor isolated from a marine sponge [70]. Together with GlaxoSmithKline, this com-
pany developed the first CENP-E inhibitor, currently known as GSK923295. This
compound is an allosteric molecule that inhibits release of inorganic phosphate and
stabilizes the interaction of CENP-E motor domain with microtubules. This pre-
vents CENP-E motility and leads to uncongressed chromosomes in human cells. To
date, GSK923295 is highly specific for CENP-E, as it does not alter the ATPase
activity of six of the main mitotic kinesins [7]. GSK923295 is also very potent as it
inhibits CENP-E in the low nanomolar range in vitro.
Analysis of a panel of 299 cell lines revealed that the vast majority (287/299) of
cell lines tested in each tumor type were sensitive to doses of GSK923295 ranging
between 6 and 219 nM. The median dose was 30 nM for solid tumor cell lines and
52 nM for hematological ones. This survey examined a wide variety of solid tumors
including breast, colon, kidney, liver, lung, ovary, pancreas, prostate, sarcoma, skin,
and several hematological tumors. However, no particular cancer type was found
more sensitive to GSK923295 [71]. Similarly, a study of 23 cell lines from the pedi-
atric preclinical testing program revealed that GSK923295 has significant antitumor
activity against pediatric sarcomas. IC50 values distributed in the 18–39 nM range in
cellulo. Objective responses were noted in roughly a third of solid tumor xenografts
after high doses (125 mg/kg) administered by i.p. (intraperitoneal) [72]. Similar
results were obtained at the same doses in mice bearing xenografts of colon, lung
and breast tumor cell lines [7]. However, GSK923295 Phase I trial was not as suc-
cessful as hoped since only one patient out of 39 showed a partial response when
treated with a dose of 250 mg/kg [73]. As a comparison, the blockbuster drug bort-
ezomib (a proteasome inhibitor also named Velcade) is administered at doses
between 1 and 2 mg/kg to treat multiple myeloma.
More attempts to screen/develop CENP-E inhibitors are under way. For instance,
UA62784 was screened as an inhibitor of CENP-E [74], but later shown to rather
inhibit microtubule polymerization [75]. CENP-E function could be impaired,
although less specifically, by inhibiting its binding to microtubule with farnesyl
transferase inhibitors [76]. More promising is the novel compound syntelin that was
recently described as the first member of a novel class of CENP-E inhibitor that
binds to different sites from those of GSK923295 [77], and for which clinical trials
are scheduled in China (X. Yao, personal communication). Novel lead-compounds
are also being designed thanks to rational modeling of the interaction between
GSK923295 and CENP-E binding pocket [78].
6.4 Conclusion: Is CENP-E a Good Target

for Chemotherapies?
During the last 15 years, there has been intense research, both from academic and
private laboratories, for the development of novel, rationally designed anti-mitotics.
These were primarily targeted towards mitotic kinases or kinesins because (i) these
are essential to cell proliferation, (ii) ATP-analogues provided some level of speci-
ficity in targeting these enzymes, and (iii) they are overexpressed in a wide range of
tumors [79, 80]. The rationale was to design novel therapies that could be as effi-
cient in targeting mitosis as were microtubule-targeting agents such as taxoteres or
vinca alkaloids. The hope was to avoid the side-effects associated with microtubule
inhibition in non-proliferating cells such as neurotoxicity. However, although this
field has been successful in producing dozens of low molecular weight inhibitors
that are active in the nanomolar range, providing outstanding tools for basic research,
it has not yet delivered efficient treatments for cancer patients.
CENP-E and other mitosis-specific genes were found overexpressed in tumor-
initiating cells from multiple myeloma [81]. This is particularly interesting since the
tumor-initiating cells are the ones most difficult to eradicate. Such tumors could
thus get the most benefit from high doses of CENP-E targeting drugs. However
similar reasoning using an Aurora B kinase inhibitor (VX-680) showed that while
VX-680 efficiently reduces the number of tumor-initiating cells, it does not com-
pletely eliminate the population. This suggests that the use of a mitotic inhibitor
alone could be therapeutically insufficient. The results of the only phase I study
involving a CENP-E inhibitor available to date led to a similar conclusion.
One reason for this lack of success could be that phase I trials were performed on
individuals that were refractory to classical treatments. It thus might be hopeless to
test for novel anti-mitotic drugs in patients that did not respond to anti-mitotics such
as vinca alkaloids or taxoteres based therapies. Another explanation could be that
microtubule-targeting agents act by affecting non-mitotic microtubule-associated
functions [82]. Indeed, tumors proliferate much more slowly in vivo than they do in
culture, and hoping to kill tumor cells that would divide every month(s) with once-
weekly schedules of drugs might prove difficult. This might be particularly true in
situations where once-weekly administrations of GSK923295 do not maintain plasma
concentrations above relevant IC50 values for more than 24 h [73]. It could thus be
worth testing these drugs on more frequent administration schedules, and/or in com-
bination therapies. Indeed, since diminishing CENP-E levels were shown to act as a
tumor suppressor in the presence of associated genetic damage [55], future therapies
combining CENP-E inhibitor with DNA damaging agents could be envisioned.
Another reason to explain the low efficiency of CENP-E inhibitor-based thera-
pies could be that it was not conducted against the proper tumor type. No prior data
is available to apprehend which tumor type could best be sensitive to CENP-E inhib-
itors. In the future, some hints might come from unbiased studies such as the one
creating an algorithm that identified CENP-E as a potential drug target in a transgenic
model for neuroblastoma. The efficiency of GSK923295 was then validated in
cellulo and in vivo using xenograft mouse models [83]. Interestingly, CENP-E is
also part of a 67-genes signature called CINSARC whose overexpression predicts
the likelihood of metastasis in human sarcoma with complex genomics [65].
Moreover, GSK923295 demonstrated anti-proliferation activity against a few xeno-
grafted pediatric sarcomas [72]. Sarcoma tumors with metastatic potential could
thus deserve being tested for CENP-E inhibitors. The ungoing rapid expansion of
genome-wide transcriptomic data production during the recent years will certainly
help predicting which cancers could benefit most from CENP-E targeted drugs.
References
1. Putkey FR et al (2002) Unstable kinetochore-microtubule capture and chromosomal instability

following deletion of CENP-E. Dev Cell 3:351–365
kinetochore motor required for metaphase chromosome alignment. Cell 91:357–366
3. Schaar BT, Chan GK, Maddox P, Salmon ED, Yen TJ (1997) CENP-E function at kinetochores
is essential for chromosome alignment. J Cell Biol 139:1373–1382
4. Yucel JK et al (2000) CENP-meta, an essential kinetochore kinesin required for the mainte-
nance of metaphase chromosome alignment in Drosophila. J Cell Biol 150:1–11
5. Yao X, Abrieu A, Zheng Y, Sullivan KF, Cleveland DW (2000) CENP-E forms a link between
attachment of spindle microtubules to kinetochores and the mitotic checkpoint. Nat Cell Biol
2:484–491
6. Tanudji M et al (2004) Gene silencing of CENP-E by small interfering RNA in HeLa cells
leads to missegregation of chromosomes after a mitotic delay. Mol Biol Cell 15:3771–3781
8. Kapoor TM et al (2006) Chromosomes can congress to the metaphase plate before biorienta-
tion. Science 311:388–391
9. Kim Y, Heuser JE, Waterman CM, Cleveland DW (2008) CENP-E combines a slow, proces-
sive motor and a flexible coiled coil to produce an essential motile kinetochore tether. J Cell
Biol 181:411–419
10. Gudimchuk N et al (2013) Kinetochore kinesin CENP-E is a processive bi-directional tracker
of dynamic microtubule tips. Nat Cell Biol 15:1079–1088
11. Garcia-Saez I, Yen T, Wade RH, Kozielski F (2004) Crystal structure of the motor domain of
the human kinetochore protein CENP-E. J Mol Biol 340:1107–1116
12. Sardar HS, Luczak VG, Lopez MM, Lister BC, Gilbert SP (2010) Mitotic kinesin CENP-E
promotes microtubule plus-end elongation. Curr Biol 20:1648–1653
13. Rosenfeld SS et al (2009) The ATPase cycle of the mitotic motor CENP-E. J Biol Chem
284:32858–32868
14. Espeut J et al (2008) Phosphorylation relieves autoinhibition of the kinetochore motor
Cenp-E. Mol Cell 29:637–643
15. Yardimci H, van Duffelen M, Mao Y, Rosenfeld SS, Selvin PR (2008) The mitotic kinesin
CENP-E is a processive transport motor. Proc Natl Acad Sci U S A 105:6016–6021
16. Lombillo VA, Nislow C, Yen TJ, Gelfand VI, McIntosh JR (1995) Antibodies to the kinesin
motor domain and CENP-E inhibit microtubule depolymerization-dependent motion of chro-
mosomes in vitro [see comments]. J Cell Biol 128:107–115
17. Musinipally V, Howes S, Alushin GM, Nogales E (2013) The microtubule binding properties
of CENP-E’s C-terminus and CENP-F. J Mol Biol 425:4427–4441
18. Sardar HS, Gilbert SP (2012) Microtubule capture by mitotic kinesin centromere protein E
(CENP-E). J Biol Chem 287:24894–24904
19. Maffini S et al (2009) Motor-independent targeting of CLASPs to kinetochores by CENP-E

promotes microtubule turnover and poleward flux. Curr Biol 19:1566–1572
20. Kim Y, Holland AJ, Lan W, Cleveland DW (2010) Aurora kinases and protein phosphatase 1
mediate chromosome congression through regulation of CENP-E. Cell 142:444–455
21. Nousiainen M, Sillje HH, Sauer G, Nigg EA, Korner R (2006) Phosphoproteome analysis of
the human mitotic spindle. Proc Natl Acad Sci U S A 103:5391–5396
22. Zhang XD et al (2008) SUMO-2/3 modification and binding regulate the association of
CENP-E with kinetochores and progression through mitosis. Mol Cell 29:729–741
23. Ashar HR et al (2000) Farnesyl transferase inhibitors block the farnesylation of CENP-E and
CENP-F and alter the association of CENP-E with the microtubules. J Biol Chem
275:30451–30457
24. Chan GK, Schaar BT, Yen TJ (1998) Characterization of the kinetochore binding domain of
CENP-E reveals interactions with the kinetochore proteins CENP-F and hBUBR1. J Cell Biol
143:49–63
25. Topper LM, Bastians H, Ruderman JV, Gorbsky GJ (2001) Elevating the level of Cdc34/Ubc3
ubiquitin-conjugating enzyme in mitosis inhibits association of CENP-E with kinetochores
and blocks the metaphase alignment of chromosomes. J Cell Biol 154:707–717
26. Huang Y et al (2012) CENP-E kinesin interacts with SKAP protein to orchestrate accurate
chromosome segregation in mitosis. J Biol Chem 287:1500–1509
27. Wang X et al (2012) Mitotic regulator SKAP forms a link between kinetochore core complex
KMN and dynamic spindle microtubules. J Biol Chem 287:39380–39390
28. Liu D et al (2007) Human NUF2 interacts with centromere-associated protein E and is essen-
tial for a stable spindle microtubule-kinetochore attachment. J Biol Chem 282:21415–21424
29. Cheeseman IM, Chappie JS, Wilson-Kubalek EM, Desai A (2006) The conserved KMN net-
work constitutes the core microtubule-binding site of the kinetochore. Cell 127:983–997
30. Lampson MA, Kapoor TM (2005) The human mitotic checkpoint protein BubR1 regulates
chromosome-spindle attachments. Nat Cell Biol 7:93–98
31. Maiato H, Logarinho E (2011) Motor-dependent and -independent roles of CENP-E at kineto-
chores: the cautionary tale of UA62784. Chem Biol 18:679–680
32. Abrieu A, Kahana JA, Wood KW, Cleveland DW (2000) CENP-E as an essential component
of the mitotic checkpoint in vitro. Cell 102:817–826
33. Mao Y, Abrieu A, Cleveland DW (2003) Activating and silencing the mitotic checkpoint
through CENP-E-dependent activation/inactivation of BubR1. Cell 114:87–98
34. Mao Y, Desai A, Cleveland DW (2005) Microtubule capture by CENP-E silences BubR1-
dependent mitotic checkpoint signaling. J Cell Biol 170:873–880
35. Weaver BA et al (2003) Centromere-associated protein-E is essential for the mammalian mitotic
checkpoint to prevent aneuploidy due to single chromosome loss. J Cell Biol 162:551–563
36. Guo Y, Kim C, Ahmad S, Zhang J, Mao Y (2012) CENP-E–dependent BubR1 autophosphory-
lation enhances chromosome alignment and the mitotic checkpoint. J Cell Biol 198:205–217
37. Suijkerbuijk SJ et al (2012) The vertebrate mitotic checkpoint protein BUBR1 is an unusual
pseudokinase. Dev Cell 22:1321–1329
38. Han JS et al (2013) Catalytic assembly of the mitotic checkpoint inhibitor BubR1-Cdc20 by a
Mad2-induced functional switch in Cdc20. Mol Cell 51:92–104
39. Park I et al (2013) Loss of BubR1 acetylation causes defects in spindle assembly checkpoint
signaling and promotes tumor formation. J Cell Biol 202:295–309
40. Brown KD, Wood KW, Cleveland DW (1996) The kinesin-like protein CENP-E is kinetochore-
associated throughout poleward chromosome segregation during anaphase-A. J Cell Sci
109(Pt 5):961–969
41. Cooke CA, Schaar B, Yen TJ, Earnshaw WC (1997) Localization of CENP-E in the fibrous
corona and outer plate of mammalian kinetochores from prometaphase through anaphase.
Chromosoma 106:446–455
42. Brown KD, Coulson RM, Yen TJ, Cleveland DW (1994) Cyclin-like accumulation and loss of
the putative kinetochore motor CENP-E results from coupling continuous synthesis with spe-
cific degradation at the end of mitosis. J Cell Biol 125:1303–1312
43. Savoian MS, Earnshaw WC, Khodjakov A, Rieder CL (1999) Cleavage furrows formed
between centrosomes lacking an intervening spindle and chromosomes contain microtubule
bundles, INCENP, and CHO1 but not CENP-E. Mol Biol Cell 10:297–311
44. Lindon C, Pines J (2004) Ordered proteolysis in anaphase inactivates Plk1 to contribute to
proper mitotic exit in human cells. J Cell Biol 164:233–241
45. Spiliotis ET, Kinoshita M, Nelson WJ (2005) A mitotic septin scaffold required for mamma-
lian chromosome congression and segregation. Science 307:1781–1785
46. Zhu M et al (2008) Septin 7 interacts with centromere-associated protein E and is required for
its kinetochore localization. J Biol Chem 283:18916–18925
47. Kurasawa Y, Earnshaw WC, Mochizuki Y, Dohmae N, Todokoro K (2004) Essential roles of
KIF4 and its binding partner PRC1 in organized central spindle midzone formation. EMBO J
23:3237–3248
and length of antiparallel microtubule overlaps. Cell 142:420–432
49. Subramanian R et al (2010) Insights into antiparallel microtubule crosslinking by PRC1, a
conserved nonmotor microtubule binding protein. Cell 142:433–443
50. Liu D et al (2006) Interaction of Skp1 with CENP-E at the midbody is essential for cytokine-
sis. Biochem Biophys Res Commun 345:394–402
51. Gascoigne KE, Taylor SS (2008) Cancer cells display profound intra- and interline variation
following prolonged exposure to antimitotic drugs. Cancer Cell 14:111–122
52. Topham CH, Taylor SS (2013) Mitosis and apoptosis: how is the balance set? Curr Opin Cell
Biol 25:780–785
53. Janssen A, van der Burg M, Szuhai K, Kops GJ, Medema RH (2011) Chromosome segregation
errors as a cause of DNA damage and structural chromosome aberrations. Science
333:1895–1898
54. Vogelstein B, Lane D, Levine AJ (2000) Surfing the p53 network. Nature 408:307–310
55. Weaver BA, Silk AD, Montagna C, Verdier-Pinard P, Cleveland DW (2007) Aneuploidy acts
both oncogenically and as a tumor suppressor. Cancer Cell 11:25–36
56. Janssen A, Kops GJ, Medema RH (2009) Elevating the frequency of chromosome mis-
segregation as a strategy to kill tumor cells. Proc Natl Acad Sci U S A 106:19108–19113
57. Torres EM et al (2007) Effects of aneuploidy on cellular physiology and cell division in hap-
loid yeast. Science 317:916–924
58. Williams BR et al (2008) Aneuploidy affects proliferation and spontaneous immortalization in
mammalian cells. Science 322:703–709
59. Torres EM et al (2010) Identification of aneuploidy-tolerating mutations. Cell 143:71–83
60. Tang YC, Williams BR, Siegel JJ, Amon A (2011) Identification of aneuploidy-selective anti-
proliferation compounds. Cell 144:499–512
61. Sheltzer JM et al (2011) Aneuploidy drives genomic instability in yeast. Science
333:1026–1030
62. Liu Z et al (2009) Reduced expression of cenp-e in human hepatocellular carcinoma. J Exp
Clin Cancer Res 28:156
63. Torres JZ et al (2011) The STARD9/Kif16a kinesin associates with mitotic microtubules and
regulates spindle pole assembly. Cell 147:1309–1323
64. Agarwal R et al (2009) Integrative analysis of cyclin protein levels identifies cyclin b1 as a
classifier and predictor of outcomes in breast cancer. Clin Cancer Res 15:3654–3662
65. Chibon F et al (2010) Validated prediction of clinical outcome in sarcomas and multiple types
of cancer on the basis of a gene expression signature related to genome complexity. Nat Med
16:781–787
66. Raverot G et al (2010) Prognostic factors in prolactin pituitary tumors: clinical, histological,
and molecular data from a series of 94 patients with a long postoperative follow-up. J Clin
Endocrinol Metab 95:1708–1716
67. Bie L et al (2011) The accuracy of survival time prediction for patients with glioma is improved
by measuring mitotic spindle checkpoint gene expression. PLoS One 6:e25631
68. Ju W et al (2009) Identification of genes with differential expression in chemoresistant epithe-

lial ovarian cancer using high-density oligonucleotide microarrays. Oncol Res 18:47–56
69. DeRycke MS et al (2013) Identification of novel variants in colorectal cancer families by high-
throughput exome sequencing. Cancer Epidemiol Biomarkers Prev 22:1239–1251
70. Sakowicz R et al (1998) A marine natural product inhibitor of kinesin motors. Science
280:292–295
71. Sutton D et al (2007) GSK923295A, a potent and selective CENP-E inhibitor, has broad spec-
trum activity against human tumor xenografts in nude mice. In Proceedings of the American
Association for Cancer Research, Los Angeles, Abstract #1522
72. Lock RB et al (2012) Initial testing of the CENP-E inhibitor GSK923295A by the pediatric
preclinical testing program. Pediatr Blood Cancer 58:916–923
73. Chung V et al (2012) First-time-in-human study of GSK923295, a novel antimitotic inhibitor
of centromere-associated protein E (CENP-E), in patients with refractory cancer. Cancer
74. Henderson MC et al (2009) UA62784, a novel inhibitor of centromere protein E kinesin-like
protein. Mol Cancer Ther 8:36–44
75. Tcherniuk S, Deshayes S, Sarli V, Divita G, Abrieu A (2011) UA62784 is a cytotoxic inhibitor
of microtubules, not CENP-E. Chem Biol 18:631–641
76. Schafer-Hales K et al (2007) Farnesyl transferase inhibitors impair chromosomal maintenance
in cell lines and human tumors by compromising CENP-E and CENP-F function. Mol Cancer
Ther 6:1317–1328
77. Ding X et al (2010) Probing CENP-E function in chromosome dynamics using small molecule
inhibitor syntelin. Cell Res 20:1386–1389
78. Hirayama T et al (2013) Synthetic studies of centromere-associated protein-E (CENP-E)
inhibitors: 1. Exploration of fused bicyclic core scaffolds using electrostatic potential map.
Bioorg Med Chem 21:5488–5502
79. Wood KW, Chua P, Sutton D, Jackson JR (2008) Centromere-associated protein E: a motor
that puts the brakes on the mitotic checkpoint. Clin Cancer Res 14:7588–7592
80. Malumbres M (2011) Physiological relevance of cell cycle kinases. Physiol Rev
91:973–1007
81. Nara M et al (2013) Bortezomib reduces the tumorigenicity of multiple myeloma via down-
regulation of upregulated targets in clonogenic side population cells. PLoS One 8:e56954
82. Komlodi-Pasztor E, Sackett D, Wilkerson J, Fojo T (2011) Mitosis is not a key target of micro-
tubule agents in patient tumors. Nat Rev Clin Oncol 8:244–250
83. Balamuth NJ et al (2010) Serial transcriptome analysis and cross-species integration identifies
centromere-associated protein E as a novel neuroblastoma target. Cancer Res 70:2749–2758
84. Wan X et al (2009) Protein architecture of the human kinetochore microtubule attachment site.
Cell 137:672–684
85. Welburn JP, Cheeseman IM (2008) Toward a molecular structure of the eukaryotic kineto-
chore. Dev Cell 15:645–655
Chapter 7
The Human Kinesin-14 Motor KifC1/HSET Is
an Attractive Anti-cancer Drug Target
Vaishali Pannu, Padmashree C.G. Rida, and Ritu Aneja
7.1 Background
Human spleen, embryo, and testes protein (HSET, also known as KifC1) belongs to
the kinesin-14 family which is unique both structurally and functionally. A key
feature is that the members of this family have their motor domains in the C-terminus
as opposed to conventional kinesins (whose motor domain resides in the N-terminus),
thus imparting them their distinct minus-end directed motility. Additionally,
kinesin-14s bear a central coiled-coil stalk, and an N-terminal globular domain. The
motor domain binds microtubules in an ATP-sensitive manner, and the tail pos-
sesses a second ATP-insensitive microtubule binding site that enables kinesin-14s to
cross-link and slide microtubules [1–5]. KifC1/HSET, KifC2 and KifC3 are the
only known human members of this family. The native human HSET is an active
microtubule motor with a microtubule gliding rate of ~5 µm/min [6]. The recombi-
nant forms of the family members Kar3 (budding yeast ortholog), CHO2 (hamster
ortholog), and Ncd (Drosophila ortholog) all exhibit minus-end directed microtu-
bule gliding activity at the rate of 1–8 µm/min [7–9]. These motors are proposed to
generate the “inward” directed forces to oppose the “outward” directed forces pro-
vided by kinesin-5 family members such as Eg5 [10, 11].
The human HSET gene is located on the short arm of human Chromosome 6
(6p21.3), on the centromeric side of the Major Histocompatibility Complex class II
gene region. The 2.4 kilobase (kb) transcript of the HSET gene has been shown to
be abundantly expressed in testis, spleen and ovary [12]. HSET expression is known
to be regulated by the p110 Cux1 transcription factor [13], which interacts with E2F
family of transcription factors resulting in transcriptional regulation of a slew of cell
cycle-regulated genes [14]. Thus, HSET expression is presumably cell cycle-regulated
V. Pannu • P.C.G. Rida • R. Aneja (*)

Department of Biology, Georgia State University, Atlanta, GA, USA
e-mail: raneja@gsu.edu

102 V. Pannu et al.
by p110 Cux1 whose transcriptional targets are known to include several genes that
play crucial roles in S-phase and mitosis.
7.2 Localization of HSET Changes Dynamically During Cell

Cycle Progression
In addition to the cell cycle-regulated transcription of HSET, the localization of

HSET is also tightly regulated during cell cycle progression. During interphase,
HSET is sequestered in the nucleus, presumably to avoid untimely microtubule
cross-linking; upon nuclear envelope breakdown at the onset of mitosis, HSET is
released into the cytoplasm to resume its microtubule cross-linking and sliding
activities [15]. During this time, it is localized both on the spindle poles and along
the spindle length. Its distinct pole localization is prominent on shorter 1–2 µm
spindles. With mitotic spindle breakdown in telophase, HSET is localized on the
minus-end of microtubules near the spindle poles before being shuttled back into
the nucleus [15, 16]. HSET transport in and out of the nucleus is regulated by Ran
GTPase via association of the bipartite Nuclear Localization Signal of HSET with
nuclear import receptors importin α/β [15].
7.3 HSET Plays a Key Role in Morphogenesis of a Bipolar

Mitotic Spindle
The organization of a bipolar spindle requires precise orchestration of three distinct

types of dynamic microtubules: astral microtubules originating from centrosomal
poles and emanating towards the cell cortex, interpolar spindle microtubules inter-
spersing the spindle midzone in an antiparallel fashion and K-fibers (kinetochore
fibers) extending from kinetochores to spindle poles [17, 18]. Two distinct models
have been proposed to explain mitotic spindle assembly. In centrosome-mediated
spindle assembly (also referred to as “search and capture” model), dynamic plus
ends of microtubules nucleated by centrosomes extend towards the spindle midzone
until they are captured and stabilized by a kinetochore [19, 20]. In the chromatin-
mediated spindle assembly, microtubules are nucleated around the chromosomes;
these microtubules then get integrated with the centrosome-nucleated spindle
microtubules by the action of microtubule-associated motors [18, 21].
The motor activity of HSET is required for spindle assembly and pole organiza-
tion in both mitotic and meiotic systems. HSET function has been demonstrated to
regulate (a) establishment of the meiotic spindle in mouse oocytes, and (b) mainte-
nance of mitotic spindle in cultured mammalian cells; however, it is not essential for
the establishment of mitotic spindle per se in mammalian cells where the centro-
somes perform this vital function [22]. HeLa cells depleted of HSET exhibit multipolar
7 The Human Kinesin-14 Motor KifC1/HSET Is an Attractive Anti-cancer Drug Target 103
spindles and chromosome misalignment in mitosis. Owing to these aberrations,

these cells undergo anaphase with considerable chromosome missegregation and
form micro-nuclei upon cytokinesis. In the absence of NuMA and other motors
such as dynein and Eg5, HSET alone is not sufficient to organize astral microtu-
bules. However, HSET motor activity is sufficient to assemble microtubule asters in
the presence of NuMA [23]. Studies have shown that HSET functionally fulfills
Pkl1 requirement (yeast ortholog of HSET) to regulate the establishment of a bipo-
lar spindle in S. pombe. This finding strongly demonstrates the inherent ability of
HSET to organize a bipolar spindle, a crucial function which is rendered redundant
by centrosomes in human mitotic cells. The elements most integral to the spindle
organization activity of HSET have been shown to lie in the stalk region [16].
7.4 Motor Activity of HSET Is Crucial

for Averting Polyploidy
A key step for ensuring genetic stability is the proper execution of cytokinesis.
Successful cytokinesis relies on the assembly of a robust central spindle to specify
the cleavage furrow position along a plane that bisects the axis of chromosome seg-
regation [24, 25]. The onus of assembling and maintaining a proper central spindle
rests on microtubule-associated factors, including motor proteins that sort microtu-
bules into bundles with their plus-ends overlapping at the center. A study that exam-
ined the phenotypes produced by RNAi-mediated knockdown of HSET revealed
that HSET motor function is required for proper cytokinesis progression to maintain
a normal karyotype, and perturbation of HSET increases the frequency of polyploid
cells arising from cytokinetic failure [26]. A closer inspection showed that HSET
knockdown resulted in severe loss of midzone microtubule organization and central
spindle stability, as well as mislocalization of several midbody-associated proteins.
HSET thus contributes to the sorting and organization of central spindle microtu-
bules to ensure the normal distribution of many central spindle components.
7.5 Role of HSET in Vesicular Trafficking

and DNA Transport
Kinesins are known to have a role in anterograde trafficking of vesicular complexes

carrying macromolecules [27] and organelles such as mitochondria [28] and lyso-
somes [28, 29] in murine systems. It has been shown that in mouse liver, the
microtubule-based motility and processing of early endocytic vesicles involves the
opposing and coordinated motor activities of the plus-end directed kinesin Kif5B
and the minus-end directed KifC1/HSET [30]. Data from this study also suggest
that these motors may be part of a yet to be elucidated protein complex that
104 V. Pannu et al.
mediates their coordinate regulation. It is presently unclear if HSET performs a

similar function in human cells and whether a 19-amino acid sequence previously
shown to be required for binding to membrane-bound organelles [12] is required for
HSET to mediate its function in vesicular trafficking, since functions of these
motors can differ substantially between species. Recent studies have also suggested
a role of HSET in intracellular active transport of bare double-stranded DNA [31].
Further experiments are needed to clarify the in vivo relevance, if any, of HSET’s
ability to bind bare DNA. Nevertheless, these data underscore the potential key role
that HSET could play in gene therapy by facilitating delivery of exogenous DNA
fragments to the cell nucleus.
7.6 Role of HSET in Rat Spermatogenesis
HSET function is required for acrosome biogenesis and nuclear reshaping during
spermatogenesis in rats. Essentially, spermatogenesis is accompanied by a remark-
able degree of cytoskeletal remodeling, particularly the transformation of
microtubule-based structures such as mitotic and meiotic spindles, manchette and
sperm axoneme. Motility of the manchette, the extension of the acrosome, and con-
densation and elongation of the spermatid nucleus are tightly concerted events
requiring intricate interplay of microtubule motors such as kinesin and dynein.
HSET has been shown to perform several crucial functions in this process spanning
facilitation of Golgi-originated vesicular transport and the recycling of vesicles
back to the Golgi, to spermatid nucleus remodeling via interaction of HSET with
nucleoporins and β-importin [32, 33]. In sexually mature rats, HSET is initially
observed to be associated with punctate vesicles located between the Golgi and the
spermatid nucleus, before it eventually gets associated with the growing acrosome
[32]. Several studies in rats have demonstrated the role of HSET, in concert with
nuclear factor β-importin, in the growth of the spermatid acrosome. The association
of HSET with the complex containing nucleoporin NUP62 in the testis is partially
controlled by Ran-GTP. This association of HSET with a component of the nuclear
membrane supports the notion that HSET plays a role in acrosome transport along
the nuclear membrane rather than in transport of vesicles along the outer surface of
the manchette. However, it is presently unclear if HSET plays an analogous essen-
tial role in human spermatogenesis.
7.7 HSET Function Is Essential for Acentrosomal Pole

Focusing
Although its role in spindle bipolarity is dispensable in human somatic cells, HSET
and its evolutionarily conserved orthologs play vital roles in organizing spindle
poles in acentrosomal cells [34–36]. In cells lacking centrosomes, such as those in higher
plants (e.g. Arabidopsis), human germ cells or somatic cells with artificial centrosome
ablation, spindle pole formation is accomplished via centrosome-independent

mechanisms [37–41]. In these cells, accumulation of elevated levels of Ran-GTP
around the chromosomes promotes microtubule bundling and cross-linking via
minus end-directed motor activity of kinesin-14 motors; thus actualizing the forma-
tion of an acentrosomal bipolar spindle. HSET has been proposed as a target of
Ran-GTP [3, 5, 15], a key regulator of chromatin-driven microtubule-polymerization
in mitotic cells [38, 42]. Following nuclear envelope breakdown in mitosis, elevated
Ran-GTP levels regulate the dissociation of various microtubule-associated proteins
including HSET from β-importin, thereby causing their release in the cytoplasm.
This stimulates acentrosomal microtubule-polymerization and -organization in the
proximity of chromosomes. Ran-dependent microtubule-polymerization is highly
active until Ran-target proteins are degraded by the Anaphase Promoting Complex
(APC/C) during anaphase [43–50]. The key involvement of minus-end directed
motors in centrosome-independent pole-focusing is most conspicuously apparent
in centrosome-lacking Xenopus egg extracts and Drosophila S2 cells, where
experimentally-induced defects in the function of kinesin-14 orthologs result in
multipolar spindles and misaligned chromosomes [3, 8, 22, 51–53].
By combining microtubule cross-linking activity with minus-end motor activity,
HSET presents a unique competence to slide one microtubule along its adjacent
one, thereby moving it towards the minus-end of the microtubule. This idea is sub-
stantially supported by the observation that defective HSET leads to disorganized
spindle poles during spindle assembly in mitotic/meiotic cells, presumably due to
the disruption of its spindle-pole focusing activity [3, 8, 22, 52, 54–57]. Nonetheless,
this pole-focusing role of HSET is imperative exclusively to spindle assembly in the
absence of functional centrosomes. Conclusive experimental evidence indicates that
HSET’s role is essentially compensated for by the presence of functional centro-
somes, rendering it dispensable for mitotic spindle assembly in human somatic cells
[22]. Additional evidence suggests that acentrosomal spindle organization regulated
by HSET is required for bipolar spindle formation in cancer cells, irrespective of
normal or supernumerary centrosome number [58].
Another facet of HSET’s multiple activities in mammalian cells relates to the
tracking of microtubule plus-ends. Microtubule tip-tracking of kinesin-14 family
members is believed to be important for their various intracellular functions, particu-
larly for assembly and maintenance of dynamic microtubule arrays. Various kine-
sin-14 motors such as Klp2 (S. pombe), Ncd (D. melanogaster), and ATK5 (A.
thaliana), have been observed to track the tips of dynamic microtubules [51, 59, 60].
Tip-tracking of Klp2 and Ncd is proposed to be involved in the stabilization of micro-
tubule arrays in S. pombe interphase cells and stabilization of kinetochore-fibers
focusing towards spindle poles in D. melanogaster mitotic cells, respectively [51, 60].
In D. melanogaster, Ncd tip-tracking is shown to be mediated by EB1 in vivo [51]. A
recent study has substantially demonstrated microtubule tip-tracking ability of human
kinesin-14 HSET in an in vitro system and established the requirement of EB1 to
regulate HSET’s dynamic microtubule end localization [61]. Microtubule end-track-
ing role of kinesin-14 motors including HSET and their dependence on EB family
proteins is ostensibly conserved across species, thus, disclosing important functional
roles of minus-end kinesins yet to be elucidated in various species including human.
106 V. Pannu et al.
7.8 HSET’s Essential Role in Clustering of Supernumerary

Centrosomes and Acentrosomal Poles in Cancer Cells
Structural and numerical centrosome aberrations are well recognized as characteristics

of various solid and hematologic malignancies [62–64]. Strikingly however, mitotic
division in cancer cells with supernumerary centrosomes has been observed to be
typically bipolar [65, 66]. Furthermore, a comparison of the frequency of amplified
centrosomes with the frequency of bipolar mitoses in tumors reveal the preponder-
ance of centrosome clustering mechanisms in maintaining bipolarity in mitotic can-
cer cells [67, 68]. Recent studies have demonstrated that clustering of supernumerary
centrosomes into two polar groups occurs during a prolonged prometaphase and
metaphase and requires pole-focusing activity of HSET along with the spindle-
assembling activity of Augmin-HAUS complex [69, 70]. HSET-dependent centro-
some clustering thus paves the way for cancer cells carrying excess centrosomal
load to erect a pseudobipolar spindle and evade the catastrophic consequences of a
multipolar mitosis. RNAi-mediated HSET knockdown in MDA-MB-231 breast
cancer cells (20–30 % of which possess amplified centrosomes) and PC-3 prostate
cancer cells (8–10 % of which harbor supernumerary centrosomes) results in ram-
pant spindle multipolarity (Fig. 7.1), demonstrating HSET’s essential function in
mitotic centrosome clustering in cancer cells.
Recent research has also uncovered rampant production of acentrosomal micro-
tubule organizing centers (MTOCs) in cancer cells. These acentrosomal MTOCs,
which are functional in nucleating microtubules, are rapidly incorporated into the
assembling bipolar spindle via HSET-dependent mechanisms, sometimes even
resulting in bipolar spindles with acentrosomal poles and free unincorporated cen-
trosomes. It has been proposed that oncogenic transformation instigated by or
accompanied with underlying DNA damage is instrumental in triggering hyperac-
tive acentrosomal microtubule nucleation in cancer cells with inherently compro-
mised cell-cycle checkpoints [58]. These observations position HSET, a key driver
Fig. 7.1 HSET function is essential for the assembly of a pseudobipolar mitotic spindle in
cancer cells with supernumerary centrosomes. Confocal immunomicrographs show mitotic
spindle morphology in control siRNA- and HSET siRNA-transfected MDA-MB-231 breast cancer
cells and PC-3 prostate cancer cells. Anti-gamma tubulin antibody identified centrosomes (green),
and anti-alpha tubulin antibody stained microtubules (red). DNA is stained using the intercalating
dye, DAPI (blue)
of acentrosomal spindle organization, as an attractive target for anti-cancer therapy.

The reliance of cancer cells on centrosomal and acentrosomal pole clustering mech-
anisms for survival makes HSET and other clustering molecules an Achilles' heel in
cancer cells, and this vulnerability can be exploited for chemotherapeutic purposes
[71]. Given that extra centrosomes and acentrosomal poles are frequently produced
in cancer cells, and that these anomalies are cancer cell-specific, agents that induce
spindle multipolarity by functionally inhibiting HSET hold tremendous promise as
the next generation of cancer cell-selective, non-toxic chemotherapeutics. An inter-
esting avenue for investigation would be to examine whether the enhanced depen-
dence of cancer cells on centrosome- and acentrosomal pole clustering proteins
translates into an upregulation of clustering protein expression and/or activity.
Indeed, the overexpression of clustering arsenal selectively in cancer cells with
excess centrosomes opens exciting avenues for exploring the potential of these clus-
tering proteins as invaluable cancer biomarkers.
7.9 Implication in Drug Resistance
Development of drug resistance due to the overexpression of certain proteins

remains a significant impediment to successful chemotherapy. Significant eleva-
tions in the expression of three kinesins KifC1/HSET, Kif1a, Kif5a has been
observed in docetaxel-resistant tumors compared with docetaxel-sensitive tumors,
suggesting the possible role of these kinesins in docetaxel resistance in breast
cancer cell lines [72]. In this study, overexpression of KifC3, KifC1/HSET, or
Kif5a considerably increased the pools of free tubulin in breast cancer cells treated
with docetaxel, suggesting that these kinesins may antagonize the effect of docetaxel
at least in part by promoting the dissociation of tubulin from microtubules. In light
of these observations, it will be interesting to evaluate how kinesin overexpression
in breast cancer patients correlates with clinical outcomes of patients treated with a
docetaxel-containing regimen. Efforts of this nature will aid in the development of
kinesins such as HSET, as new prognostic and diagnostic markers. Hereafter, it may
be worthwhile to explore and develop specific inhibitors of the kinesins that are
particularly responsible for taxane resistance in tumors, with the hope that such
inhibitors may prove to be synergistic with taxanes by interfering with the drug
resistance mechanism [73].
7.10 Implication in Metastasis and Tumor Progression
Gene expression studies suggest a striking correlation of KifC1/HSET expression

with brain metastasis. Although its precise mechanistic role in metastatic dissemina-
tion to brain remains elusive, HSET seemed to be the strongest predictor of brain
metastasis in the study by Grinberg-Rashi et al [74]. Primary breast tumors have also
108 V. Pannu et al.
been observed to overexpress HSET as compared to their normal adjacent tissue

[58]. Moreover, findings from bioinformatic analysis of existing databases reveal
that triple negative breast cancer, associated with the poorest prognosis among all
breast cancers, exhibits higher HSET expression levels compared to non-triple nega-
tive breast cancers [75]. Given the distinct overexpression of HSET in aggressive
and invasive tumors, a more direct and causal involvement of HSET in driving tumor
progression does not seem to be a far cry and merits further investigation.
Presumably HSET has distinct functions corresponding to different phases of the
cell cycle as well as its specific subcellular localization during the various phases.
Upon nuclear membrane breakdown in mitosis, HSET is released into the cyto-
plasm, where it localizes to the minus-end of microtubules and acts as a key player
in enforcing the spindle pole focusing of microtubule minus-ends, thus inciting
clustering of supernumerary centrosomes in cancer cells [69]. In doing so, HSET
fulfils a twofold role: it ensures the survival of cancer cells that can now evade the
catastrophic consequences of dispersed supernumerary centrosomes, and in paral-
lel, it fosters low-grade aneuploidy by promoting chromosome missegregation at a
“moderate” level, and thus potentially facilitates tumor evolution and consequent
generation of karyotypic heterogeneity [76, 77]. Based on these lines of evidence,
the notion that HSET overexpression is responsible for driving evolution of more
aggressive phenotypes rather than merely enabling centrosome clustering in mito-
sis, is plausible. In interphase, HSET evidently contributes in early endocytic vesi-
cle transport and fission in murine hepatocytes [30]. Work from our lab shows a
differential enrichment in the nuclear expression of HSET in breast tumors when
compared to their normal adjacent tissues (unpublished observations), suggesting
the possibility of an additional nucleus-specific role of HSET in interphase. Work
from our group has also demonstrated that HSET significantly accelerates cell cycle
kinetics when overexpressed in HeLa cells (unpublished observations). Given the
role of HSET in the transport of DNA along microtubules [31], its ability to bind
nuclear transport factors such as β-importin [32] and its affinity for nucleoporins
NUP50 and NUP153 [31] and NUP62 complex [33], we speculate that HSET might
possess distinct roles during its nuclear localization in interphase, perhaps in the
regulation of gene expression. However, the primary sequence of this kinesin does
not reveal the presence of any putative DNA-binding domains. Thus, the likelihood
of its role as a transcription factor is imperceptible. However, we cannot rule out the
possibility of HSET being a transcriptional co-activator or an aid in chromatin
remodeling. Thus, although highly obscure at this point, HSET’s regulatory func-
tion in the interphase nucleus, if any, might contribute to its ability to promote and
drive tumor aggressiveness.
7.11 HSET as a Potential Anti-cancer Therapeutic Target
HSET upregulation provides an edge to cancer cells in (i) managing extra centro-
somes to avert multipolarity, (ii) imparting low grade whole chromosome missegre-
gation (“optimal aneuploidy”) to foster karyotypic heterogeneity and thus spawn
superior and advantageous karyotypes, (iii) conferring drug resistance to cancer

cells (docetaxel resistance), (iv) facilitating clustering of centrosomes at spindle
poles to ensure that each progeny cell inherits supernumerary centrosomes and the
accompanying advantages; for example, inherited extra centrosomes may confer
cytoskeletal advantages that may enhance cell migration and invasion to fuel metas-
tasis [78]. We therefore envision that inhibition of HSET could potentially thwart
tumor progression in manifold ways.
Essentially, a valid anti-cancer target is one whose pharmacological modulation
provides meaningful efficacy and acceptable safety upon prolonged clinical use.
HSET presents several appealing characteristics owing to which it qualifies as a
potential chemotherapeutic target. Firstly, HSET function is necessary for proper
spindle assembly, stable pole-focusing and survival of cancer cells irrespective of
normal or supernumerary centrosome numbers. Wartmann and group has shown
that HSET depletion in DNA damage repair (DDR)-deficient cancer cells (BT549
and IGR39) exhibit synthetic lethal effects even in the absence of amplified centro-
somes. Interestingly, corresponding to the effect of HSET depletion in cells harbor-
ing supernumerary centrosomes, DDR-deficient cells were observed to be
completely dependent on HSET for focusing acentrosomal spindle poles into a
bipolar spindle. These observations accentuate the dependency of cancer cells on
HSET in multiple ways and indicate the potential application of selective HSET
inhibitors to a broad range of human cancers.
Secondly, as far as we know, HSET function is dispensable for the viability of
healthy somatic cells and as a result, HSET inhibition can be expected to cause
almost no undesirable side effects on these cells. A clarification of HSET’s role in
spermatogenesis in humans would be greatly instrumental in furthering our under-
standing of the potential for side-effects when HSET function is inhibited in germ
cells. Thirdly, HSET is overexpressed in a wide range of cancers including lung
cancer where it is associated with increased risk of metastatic dissemination to the
brain. Owing to its significant association with drug resistance and possibly, tumor
metastasis, HSET inhibitors might be able to hit multiple blows thus simultaneously
attacking various aspects of tumor biology. Finally, inhibition of HSET could lead
to a reduction in resistance to some drugs such as docetaxel thus opening up syner-
gistic therapeutic avenues.
In recent years, the concept of what constitutes a good cancer target has been
honed more finely, and this advancement has been driven by both a deeper under-
standing of cancer cell biology and the development of new technologies. In order
to unequivocally validate HSET as a chemotherapeutic target, the following key
areas will need to be thoroughly addressed:
• Establish a causal role of HSET in cancer progression: Although several lines of
evidence suggest a key role for HSET overexpression in tumor progression and/
or metastasis, unequivocal experimental evidence demonstrating a determining
and causal role in driving tumor evolution is still far from complete. Mouse mod-
els wherein HSET can be overexpressed (constitutively or conditionally) would
allow in vivo evaluation of whether an elevation in HSET expression can enhance
the kinetics of tumor growth or disease progression. Similarly, development of
110 V. Pannu et al.
conditional HSET knock-out mice would allow us to establish beyond all doubt
whether HSET has causal relevance to tumor progression in mammals. These
in vivo studies using appropriate mice models in conjunction with in vitro stud-
ies, would position us to delineate the HSET-dependent molecular mechanisms
responsible for HSET’s effects on tumor evolution. It would also be essential to
understand (i) the pattern of HSET overexpression as the tumor progresses, (ii)
when the tumor’s dependence on HSET is maximal and (iii) whether this depen-
dence varies during tumor progression.
• Safety: Experiments in animal models would play a major pre-clinical role in
ensuring that inhibition of HSET is a safe, non-toxic and efficacious approach to
combating cancer.
• Mechanism of action: Elucidation of the molecular mechanisms underlying
HSET’s role in cancer progression would shed light on other potentially drug-
gable targets, upstream or downstream of HSET. It is also vital to understand
better (i) whether HSET’s role in tumor progression is absolutely dependent on
centrosome clustering or whether HSET also plays centrosome clustering-
independent roles in fuelling tumor progression, and (ii) the genetic changes that
underlie HSET overexpression, viz., gene amplification or transcriptional upreg-
ulation, in tumors.
• Molecular pharmacology: Studies aimed at providing insights into how the
HSET-dependent signaling pathway can be efficiently targeted by a potential
lead drug to overcome disease would also vastly facilitate target-based rational
design of more effective analogs of the putative HSET-inhibitory lead compound.
Animal models would also be invaluable in ensuring that any HSET-targeting
agents operate specifically through their unique target. The recent publication of
the first two HSET inhibitors by AstraZeneca [75] and research groups from
Cambridge, UK [79] has been timely, and undoubtedly a major stride to take the
HSET subject matter to the next level. We keenly look forward to the mechanism-
based validation and optimization of these inhibitors by employing diverse
in vitro and in vivo models in the times ahead.
• Bioassay design: Identification of putative HSET inhibitors would hinge criti-
cally on the design of robust and well-controlled bioassays to measure the biologi-
cal activity of HSET, viz., its microtubule-dependent and microtubule-independent
ATPase activities. Such an assay is a prerequisite for the pre-clinical development
of a high-throughput screen to identify HSET-inhibitory “lead compounds”. The
initial hits will then have to be extensively evaluated in vitro and in animal mod-
els to ensure efficacy and safety. Two recently identified HSET inhibitors have
undergone a panel of rigorous pre-clinical tests to examine specificity for HSET,
and selectivity and efficacy in cancer cells. For example, the small molecule
inhibitor of HSET, AZ82, has been shown to specifically bind to the HSET-
microtubule complex and functionally inhibit HSET’s ATPase activity, thus lead-
ing to an effective inhibition of HSET function in BT549, HeLa and MCF-7 cells.
AZ82 has been shown to significantly arrest BT-549 cells in mitosis subsequent
to extensive centrosome declustering and eventually lead to the induction of ram-
pant mitotic catastrophe [75]. More recently, an allosteric inhibitor of HSET,
CW069, has been designed, synthesized and elaborately evaluated for biological
activity using a chemogenomics-based compound selection procedure followed

by exploration of candidate “hit” molecules [79]. For example, the IC50 value of
this inhibitor for HSET was assessed and the molecule was found to exhibit sta-
tistically significant selectivity for HSET over the control protein, kinesin spindle
protein (KSP, with which HSET shares 80 % homology over the motor domain),
through in silico ligand-based binding mode prediction utilizing the structures of
well-characterized allosteric inhibitors of KSP. Moreover, it was determined that
treatment with CW069 evoked catastrophic multipolar mitoses in cells with
supernumerary centrosomes, but cells with normal centrosome numbers were left
unscathed. CW069 exposure also induced cell death in cancer cells but not in
healthy cells. Time-lapse microscopy confirmed that CW069 recapitulates the
multipolar meta/anaphase phenotypes and cell death induced by the siRNA-
mediated knockdown of HSET function in cells bearing excess centrosomes.
Additionally, it was shown that CW069 did not exacerbate multipolarity and cell
death in cells wherein HSET was absent, nor did CW069 cause the mitotic delays
and the mitotic arrest phenotypes that are characteristic of inhibition of KSP or
CENP-E functions. Finally, CW069 was demonstrated, through differential scan-
ning fluorimetry, to bind to HSET in a site-specific manner. Thus, rigorous pre-
clinical testing of putative HSET inhibitors is imperative before they can advance
to the next level of testing in animal models and in clinical trials.
• Clinical efficacy: Following pre-clinical testing, HSET inhibitors would then
have to be rigorously profiled for safety and efficacy in animal models, before the
impact of intervention through HSET can be evaluated in large-scale, carefully-
designed clinical trials. For instance, drugs may undergo pharmacodynamics,
pharmacokinetics, ADME (absorption, distribution, metabolism and elimination),
and toxicity testing in appropriate animal models to allow researchers to
allometrically estimate a safe starting dose of the drug for clinical trials in humans.
• Establish HSET dependence in different patient sub-groups: Along with the
recent progress in the development of HSET inhibitors, it is also important to
identify the right patient population that might most benefit from HSET-targeted
therapy and determine whether there exist specific tumor types/sub-types that are
more susceptible to HSET loss of function than others. For optimal therapeutic
outcomes, clinical development of HSET inhibitors must also consider various
factors such as centrosomal profile, cell cycle profile, tumor stage/grade etc.
while stratifying patient sub-groups that might most favorably respond to HSET
inhibition. Evaluating the potential benefits and efficacy of putative HSET inhib-
itors in this susceptible patient population might yield the more efficacious and
less toxic first-of-its-kind class of centrosome-based therapeutics.
7.12 Summary
Since centrosome amplification and clustering are cancer cell-specific phenomena,

centrosome clustering pathways may serve as attractive theranostic targets. The
non-essential role of HSET in normal somatic cells and its crucial requirement for
112 V. Pannu et al.
the viability of cancer cells, together paint a picture of HSET as a seemingly ideal
cancer-selective drug target that is pending validation in animal studies. This propo-
sition is bolstered by the observation that HSET depletion using RNAi induces
robust spindle multipolarity and subsequent apoptosis in aggressive breast cancer
cells and prostate cancer cells. Unquestionable dependence of cancer cells on HSET
function and the association of HSET overexpression with the aggressiveness of
certain cancers offer a strong rationale for the development of HSET as a risk-
predictive biomarker. HSET depletion can also potentially open up manifold thera-
peutic avenues for single agent or combination therapies, particularly for tumors
harboring extensive centrosome amplification. As of today, there are two commer-
cially available HSET inhibitors (AZ82 and CW069) whose effectiveness in animal
models and in the clinic are yet to be demonstrated. From a therapeutic standpoint,
it is important that studies scrutinize the entire spectrum of repercussions following
the pharmacological manipulation of a crucial cytoskeletal motor such as
HSET. Thus, there is a compelling need to launch a systematic search for putative
HSET inhibitors for selective tumor targeting and exploit HSET’s potential in
enabling risk-based patient stratification, thus ensuring a more focused chemothera-
peutic approach as we step into the era of personalized medicine.
References
1. Chandra R, Salmon ED, Erickson HP, Lockhart A, Endow SA (1993) Structural and functional
domains of the Drosophila ncd microtubule motor protein. J Biol Chem 268:9005–9013
2. Walker RA (1995) Ncd and kinesin motor domains interact with both alpha- and beta-tubulin.
Proc Natl Acad Sci U S A 92:5960–5964
3. Walczak CE, Verma S, Mitchison TJ (1997) XCTK2: a kinesin-related protein that promotes
mitotic spindle assembly in Xenopus laevis egg extracts. J Cell Biol 136:859–870
4. Karabay A, Walker RA (1999) The Ncd tail domain promotes microtubule assembly and sta-
bility. Biochem Biophys Res Commun 258:39–43. doi:S0006-291X(99)90572-7 [pii]10.1006/
bbrc.1999.0572
5. Ems-McClung SC, Zheng Y, Walczak CE (2004) Importin alpha/beta and Ran-GTP regulate
XCTK2 microtubule binding through a bipartite nuclear localization signal. Mol Biol Cell
15:46–57. doi:10.1091/mbc.E03-07-0454E03-07-0454 [pii]
6. DeLuca JG, Newton CN, Himes RH, Jordan MA, Wilson L (2001) Purification and character-
ization of native conventional kinesin, HSET, and CENP-E from mitotic hela cells. J Biol
Chem 276:28014–28021. doi:10.1074/jbc.M102801200M102801200 [pii]
7. Walker RA, Salmon ED, Endow SA (1990) The Drosophila claret segregation protein is a
minus-end directed motor molecule. Nature 347:780–782. doi:10.1038/347780a0
8. Endow SA et al (1994) Yeast Kar3 is a minus-end microtubule motor protein that destabilizes
microtubules preferentially at the minus ends. EMBO J 13:2708–2713
9. Kuriyama R et al (1995) Characterization of a minus end-directed kinesin-like motor protein
from cultured mammalian cells. J Cell Biol 129:1049–1059
10. Sharp DJ et al (2000) Functional coordination of three mitotic motors in Drosophila embryos.
Mol Biol Cell 11:241–253
11. Mogilner A, Wollman R, Civelekoglu-Scholey G, Scholey J (2006) Modeling mitosis. Trends
Cell Biol 16:88–96. doi:S0962-8924(05)00315-6 [pii]10.1016/j.tcb.2005.12.007
12. Zhang Y, Sperry AO (2004) Comparative analysis of two C-terminal kinesin motor proteins:
KIFC1 and KIFC5A. Cell Motil Cytoskeleton 58:213–230. doi:10.1002/cm.20008
13. Sansregret L et al (2011) Cut homeobox 1 causes chromosomal instability by promoting bipolar
division after cytokinesis failure. Proc Natl Acad Sci U S A 108:1949–1954. doi:1008403108
[pii]10.1073/pnas.1008403108
14. Truscott M, Harada R, Vadnais C, Robert F, Nepveu A (2008) p110 CUX1 cooperates with
E2F transcription factors in the transcriptional activation of cell cycle-regulated genes. Mol
Cell Biol 28:3127–3138. doi:MCB.02089-07 [pii]10.1128/MCB.02089-07
15. Cai S, Weaver LN, Ems-McClung SC, Walczak CE (2009) Kinesin-14 family proteins HSET/
XCTK2 control spindle length by cross-linking and sliding microtubules. Mol Biol Cell
20:1348–1359. doi:E08-09-0971 [pii]10.1091/mbc.E08-09-0971
16. Simeonov DR et al (2009) Distinct kinesin-14 mitotic mechanisms in spindle bipolarity. Cell
Cycle 8:3563–3575. doi:9970 [pii]
17. Compton DA (2000) Spindle assembly in animal cells. Annu Rev Biochem 69:95–114.
doi:69/1/95 [pii]10.1146/annurev.biochem.69.1.95
18. Walczak CE, Heald R (2008) Mechanisms of mitotic spindle assembly and function. Int Rev
Cytol 265:111–158. doi:S0074-7696(07)65003-7 [pii]10.1016/S0074-7696(07)65003-7
19. Kirschner M, Mitchison T (1986) Beyond self-assembly: from microtubules to morphogene-
sis. Cell 45:329–342. doi:0092-8674(86)90318-1 [pii]
20. McIntosh JR, Grishchuk EL, West RR (2002) Chromosome-microtubule interactions during
mitosis. Annu Rev Cell Dev Biol 18:193–219. doi:10.1146/annurev.cell-
bio.18.032002.132412032002.132412 [pii]
21. McKim KS, Hawley RS (1995) Chromosomal control of meiotic cell division. Science
270:1595–1601
cross-links microtubules in the mammalian mitotic spindle. J Cell Biol 147:351–366
23. Chakravarty A, Howard L, Compton DA (2004) A mechanistic model for the organization of
microtubule asters by motor and non-motor proteins in a mammalian mitotic extract. Mol Biol
Cell 15:2116–2132. doi:10.1091/mbc.E03-08-0579E03-08-0579 [pii]
24. Inoue YH et al (2004) Mutations in orbit/mast reveal that the central spindle is comprised of
two microtubule populations, those that initiate cleavage and those that propagate furrow
ingression. J Cell Biol 166:49–60. doi:10.1083/jcb.200402052jcb.200402052 [pii]
25. Straight AF et al (2003) Dissecting temporal and spatial control of cytokinesis with a myosin
II Inhibitor. Science 299:1743–1747. doi:10.1126/science.1081412299/5613/1743 [pii]
26. Cai S, Weaver LN, Ems-McClung SC, Walczak CE (2010) Proper organization of microtubule
minus ends is needed for midzone stability and cytokinesis. Curr Biol 20:880–885. doi:S0960-
9822(10)00442-2 [pii]10.1016/j.cub.2010.03.067
27. Sheetz MP (1999) Motor and cargo interactions. Eur J Biochem 262:19–25
28. Tanaka Y et al (1998) Targeted disruption of mouse conventional kinesin heavy chain, kif5B,
results in abnormal perinuclear clustering of mitochondria. Cell 93:1147–1158. doi:S0092-
8674(00)81459-2 [pii]
29. Hollenbeck PJ, Swanson JA (1990) Radial extension of macrophage tubular lysosomes sup-
ported by kinesin. Nature 346:864–866. doi:10.1038/346864a0
30. Nath S et al (2007) Kif5B and Kifc1 interact and are required for motility and fission of early
endocytic vesicles in mouse liver. Mol Biol Cell 18:1839–1849. doi:E06-06-0524 [pii]10.1091/
mbc.E06-06-0524
31. Farina F et al (2013) Kinesin KIFC1 actively transports bare double-stranded DNA. Nucleic
Acids Res 41:4926–4937. doi:gkt204 [pii]10.1093/nar/gkt204
32. Yang WX, Sperry AO (2003) C-terminal kinesin motor KIFC1 participates in acrosome bio-
genesis and vesicle transport. Biol Reprod 69:1719–1729. doi:10.1095/biolreprod.102.014878
biolreprod.102.014878 [pii]
33. Yang WX, Jefferson H, Sperry AO (2006) The molecular motor KIFC1 associates with a com-
plex containing nucleoporin NUP62 that is regulated during development and by the small
114 V. Pannu et al.
GTPase RAN. Biol Reprod 74:684–690. doi:biolreprod.105.049312 [pii]10.1095/

biolreprod.105.049312
34. Loughlin R, Riggs B, Heald R (2008) SnapShot: motor proteins in spindle assembly. Cell
134:548, e541. doi:S0092-8674(08)00958-6 [pii]10.1016/j.cell.2008.07.038
35. Verhey KJ, Hammond JW (2009) Traffic control: regulation of kinesin motors. Nat Rev Mol
Cell Biol 10:765–777. doi:nrm2782 [pii]10.1038/nrm2782
36. Wordeman L (2010) How kinesin motor proteins drive mitotic spindle function: lessons from
molecular assays. Semin Cell Dev Biol 21:260–268. doi:S1084-9521(10)00019-4
[pii]10.1016/j.semcdb.2010.01.018
37. Gadde S, Heald R (2004) Mechanisms and molecules of the mitotic spindle. Curr Biol
14:R797–R805. doi:10.1016/j.cub.2004.09.021S0960982204006979 [pii]
38. Kalab P, Heald R (2008) The RanGTP gradient – a GPS for the mitotic spindle. J Cell Sci
121:1577–1586. doi:121/10/1577 [pii]10.1242/jcs.005959
39. O’Connell CB, Khodjakov AL (2007) Cooperative mechanisms of mitotic spindle formation.
J Cell Sci 120:1717–1722. doi:120/10/1717 [pii]10.1242/jcs.03442
40. Schuh M, Ellenberg J (2007) Self-organization of MTOCs replaces centrosome function dur-
ing acentrosomal spindle assembly in live mouse oocytes. Cell 130:484–498. doi:S0092-
8674(07)00792-1 [pii]10.1016/j.cell.2007.06.025
41. Wadsworth P, Khodjakov A (2004) E pluribus unum: towards a universal mechanism for spin-
dle assembly. Trends Cell Biol 14:413–419. doi:10.1016/j.tcb.2004.07.004S0962-
8924(04)00169-2 [pii]
42. Clarke PR, Zhang C (2008) Spatial and temporal coordination of mitosis by Ran GTPase. Nat
Rev Mol Cell Biol 9:464–477. doi:nrm2410 [pii]10.1038/nrm2410
43. Gruss OJ et al (2001) Ran induces spindle assembly by reversing the inhibitory effect of
importin alpha on TPX2 activity. Cell 104:83–93. doi:S0092-8674(01)00193-3 [pii]
44. Koffa MD et al (2006) HURP is part of a Ran-dependent complex involved in spindle forma-
tion. Curr Biol 16:743–754. doi:S0960-9822(06)01353-4 [pii]10.1016/j.cub.2006.03.056
45. Nachury MV et al (2001) Importin beta is a mitotic target of the small GTPase Ran in spindle
assembly. Cell 104:95–106. doi:S0092-8674(01)00194-5 [pii]
46. Ribbeck K et al (2006) NuSAP, a mitotic RanGTP target that stabilizes and cross-links micro-
tubules. Mol Biol Cell 17:2646–2660. doi:E05-12-1178 [pii]10.1091/mbc.E05-12-1178
47. Sillje HH, Nagel S, Korner R, Nigg EA (2006) HURP is a Ran-importin beta-regulated protein
that stabilizes kinetochore microtubules in the vicinity of chromosomes. Curr Biol 16:731–
742. doi:S0960-9822(06)01277-2 [pii]10.1016/j.cub.2006.02.070
48. Song L, Rape M (2010) Regulated degradation of spindle assembly factors by the anaphase-
promoting complex. Mol Cell 38:369–382. doi:S1097-2765(10)00317-5 [pii]10.1016/j.
molcel.2010.02.038
49. Stewart S, Fang G (2005) Anaphase-promoting complex/cyclosome controls the stability of
TPX2 during mitotic exit. Mol Cell Biol 25:10516–10527. doi:25/23/10516 [pii]10.1128/
MCB.25.23.10516-10527.2005
50. Wiese C et al (2001) Role of importin-beta in coupling Ran to downstream targets in microtu-
bule assembly. Science 291:653–656
51. Goshima G, Nedelec F, Vale RD (2005) Mechanisms for focusing mitotic spindle poles by
minus end-directed motor proteins. J Cell Biol 171:229–240. doi:jcb.200505107 [pii]10.1083/
jcb.200505107
52. Hatsumi M, Endow SA (1992) The Drosophila ncd microtubule motor protein is spindle-
associated in meiotic and mitotic cells. J Cell Sci 103(Pt 4):1013–1020
53. Skold HN, Komma DJ, Endow SA (2005) Assembly pathway of the anastral Drosophila
oocyte meiosis I spindle. J Cell Sci 118:1745–1755. doi:jcs.02304 [pii]10.1242/jcs.02304
54. Kimble M, Church K (1983) Meiosis and early cleavage in Drosophila melanogaster eggs:
effects of the claret-non-disjunctional mutation. J Cell Sci 62:301–318
55. Endow SA, Komma DJ (1996) Centrosome and spindle function of the Drosophila Ncd micro-
tubule motor visualized in live embryos using Ncd-GFP fusion proteins. J Cell Sci 109
(Pt 10):2429–2442
56. Endow SA, Komma DJ (1997) Spindle dynamics during meiosis in Drosophila oocytes. J Cell
Biol 137:1321–1336
57. Matthies HJ, McDonald HB, Goldstein LS, Theurkauf WE (1996) Anastral meiotic spindle
morphogenesis: role of the non-claret disjunctional kinesin-like protein. J Cell Biol
134:455–464
58. Kleylein-Sohn J et al (2012) Acentrosomal spindle organization renders cancer cells depen-
dent on the kinesin HSET. J Cell Sci 125:5391–5402. doi:jcs.107474 [pii]10.1242/jcs.107474
59. Ambrose JC, Li W, Marcus A, Ma H, Cyr R (2005) A minus-end-directed kinesin with plus-
end tracking protein activity is involved in spindle morphogenesis. Mol Biol Cell 16:1584–
1592. doi:E04-10-0935 [pii]10.1091/mbc.E04-10-0935
60. Janson ME et al (2007) Crosslinkers and motors organize dynamic microtubules to form stable
bipolar arrays in fission yeast. Cell 128:357–368. doi:S0092-8674(07)00048-7 [pii]10.1016/j.
cell.2006.12.030
61. Braun M et al (2013) The human kinesin-14 HSET tracks the tips of growing microtubules
in vitro. Cytoskeleton (Hoboken) 70:515–521. doi:10.1002/cm.21133
62. Giehl M et al (2005) Centrosome aberrations in chronic myeloid leukemia correlate with stage
of disease and chromosomal instability. Leukemia 19:1192–1197. doi:2403779 [pii]10.1038/
sj.leu.2403779
63. Pihan GA et al (1998) Centrosome defects and genetic instability in malignant tumors. Cancer
Res 58:3974–3985
64. Zyss D, Gergely F (2009) Centrosome function in cancer: guilty or innocent? Trends Cell Biol
19:334–346. doi:S0962-8924(09)00113-5 [pii]10.1016/j.tcb.2009.04.001
65. Godinho SA, Kwon M, Pellman D (2009) Centrosomes and cancer: how cancer cells divide
with too many centrosomes. Cancer Metastasis Rev 28:85–98. doi:10.1007/
s10555-008-9163-6
66. Nigg EA (2002) Centrosome aberrations: cause or consequence of cancer progression? Nat
Rev Cancer 2:815–825. doi:10.1038/nrc924nrc924 [pii]
67. Levine DS, Sanchez CA, Rabinovitch PS, Reid BJ (1991) Formation of the tetraploid interme-
diate is associated with the development of cells with more than four centrioles in the elastase-
simian virus 40 tumor antigen transgenic mouse model of pancreatic cancer. Proc Natl Acad
Sci U S A 88:6427–6431
68. Lingle WL, Salisbury JL (1999) Altered centrosome structure is associated with abnormal
mitoses in human breast tumors. Am J Pathol 155:1941–1951. doi:S0002-9440(10)65513-7
[pii]10.1016/S0002-9440(10)65513-7
69. Kwon M et al (2008) Mechanisms to suppress multipolar divisions in cancer cells with extra
centrosomes. Genes Dev 22:2189–2203. doi:gad.1700908 [pii]10.1101/gad.1700908
70. Leber B et al (2010) Proteins required for centrosome clustering in cancer cells. Sci Transl
Med 2:33ra38. doi:2/33/33ra38 [pii]10.1126/scitranslmed.3000915
71. Ogden A, Rida PC, Aneja R (2012) Let’s huddle to prevent a muddle: centrosome declustering
as an attractive anticancer strategy. Cell Death Differ 19:1255–1267. doi:cdd201261
[pii]10.1038/cdd.2012.61
72. De S, Cipriano R, Jackson MW, Stark GR (2009) Overexpression of kinesins mediates
docetaxel resistance in breast cancer cells. Cancer Res 69:8035–8042. doi:0008-5472.CAN-
09-1224 [pii]10.1158/0008-5472.CAN-09-1224
73. Muller C et al (2007) Inhibitors of kinesin Eg5: antiproliferative activity of monastrol ana-
logues against human glioblastoma cells. Cancer Chemother Pharmacol 59:157–164.
doi:10.1007/s00280-006-0254-1
74. Grinberg-Rashi H et al (2009) The expression of three genes in primary non-small cell lung
cancer is associated with metastatic spread to the brain. Clin Cancer Res 15:1755–1761.
doi:1078-0432.CCR-08-2124 [pii]10.1158/1078-0432.CCR-08-2124
75. Wu J et al (2013) Discovery and mechanistic study of a small molecule inhibitor for motor
protein KIFC1. ACS Chem Biol 8:2201–2208. doi:10.1021/cb400186w
76. Ganem NJ, Godinho SA, Pellman D (2009) A mechanism linking extra centrosomes to chro-
mosomal instability. Nature 460:278–282. doi:nature08136 [pii]10.1038/nature08136
116 V. Pannu et al.
77. Silkworth WT, Nardi IK, Scholl LM, Cimini D (2009) Multipolar spindle pole coalescence is
a major source of kinetochore mis-attachment and chromosome mis-segregation in cancer
cells. PLoS One 4:e6564. doi:10.1371/journal.pone.0006564
78. Ogden A, Rida PC, Aneja R (2013) Heading off with the herd: how cancer cells might maneu-
ver supernumerary centrosomes for directional migration. Cancer Metastasis Rev 32:269–287.
doi:10.1007/s10555-012-9413-5
79. Watts CA et al (2013) Design, synthesis, and biological evaluation of an allosteric inhibitor of
HSET that targets cancer cells with supernumerary centrosomes. Chem Biol 20:1399–1410.
doi:S1074-5521(13)00348-7 [pii]10.1016/j.chembiol.2013.09.012
Chapter 8
Kinesin-13 Microtubule Depolymerizing
Proteins as Targets for Cancer Therapy
Anutosh Ganguly and Fernando Cabral
8.1 Introduction
The kinesin-13 family of molecular motors contains three closely related members:
Kif2a, Kif2b, and Kif2c (also commonly called mitotic centromere associated
kinesin or MCAK). Unlike most kinesin motors that use ATP to power the transport
of cargo along microtubules, kinesin-13 proteins use ATP to catalyze the loss of
tubulin subunits from microtubule ends [1]. These motors differ structurally from
conventional kinesins by having an internal, rather than N- or C-terminal, motor
domain leading to their alternate designation as Kin I (for internal) or M-type (for
middle) kinesins [2, 3]. All three kinesin-13 motors are present during interphase,
but their main activity is seen during mitosis when they are believed to play overlapping
yet distinct roles in cell division [4].
Elevated levels of kinesin-13 family members have been found in an increasing
number of patients with a variety of cancers, and the levels of these proteins appear
to predict tumor aggressiveness and poor patient survival. Altered levels of Kif2c
have also been shown to affect the sensitivity of cells to microtubule inhibitors, a
class of chemotherapeutic drugs that has been successfully used against various
forms of cancer for the last 50 years. Because a number of excellent reviews on
kinesin-13 motors have recently been published [5–7], this article will focus on the
A. Ganguly
Department of Microbiology and Infectious Diseases, Snyder Institute, University of Calgary,
Calgary, AB T2N4Z6, Canada
F. Cabral (*)
Department of Integrative Biology and Pharmacology, University of Texas Medical School,
Houston, TX 77030, USA
e-mail: fernando.r.cabral@uth.tmc.edu

118 A. Ganguly and F. Cabral
potential for using these proteins as markers for tumorigenesis and metastasis, as
well as on the possibility of targeting them to treat cancer and overcome some forms
of drug resistance.
8.2 Cellular Localization
During interphase Kif2a is primarily found in the centrosome. In mitosis it appears on

spindle microtubules but remains concentrated in the spindle poles (centrosomes).
Kif2a localization was also reported in the spindle midzone and midbody during
anaphase and telophase. Trace amounts of Kif2a were detected in kinetochores, but
its presence was only apparent after microtubules were depolymerized with drugs
like nocodazole [8].
Kif2b also localizes most strongly to centrosomes during interphase and to spindle
poles during mitosis, but can additionally be found associated with microtubules [4].
There is some staining of kinetochores that is most clearly seen during prophase
and diminishes as cells progress to metaphase. In the later stages of mitosis, it is
associated with the midbody region as well as separating spindle poles. Because
Kif2b is a low abundance protein in most cells, studies were carried out using
GFP-Kif2b and thus, the localization should be regarded as tentative because other
studies have shown that GFP fusions with proteins such as EB1 or MCAK can lead
to alterations in localization and/or degradation [9, 10]. Although kinesins generally
bind to microtubules via a direct interaction with their motor domains, it was
recently shown that Kif2b binding to spindle fibers is substantially reduced when
Cep170 is depleted [11]. This finding opens up the possibility that the localization
and activity of the kinesin-13 family of motor proteins may be regulated in part by
binding partners.
During interphase, MCAK (Kif2c) is transported into the nucleus by an import
signal-mediated process [12]; however, a substantial fraction remains in the cyto-
plasm where it is attached to the centrosome and is weakly bound to microtubules
[13]. In prophase, it is found predominantly in kinetochores and spindle poles but is
again weakly associated with microtubules. It remains bound to these structures
during prometaphase, but once chromosomes are aligned at metaphase, kinetochore
staining becomes almost undetectable and staining of spindle poles and fibers is
greatly reduced. Analysis of synchronized CHO cells stably transfected with
flag-tagged human MCAK indicates that the protein is synthesized throughout the
cell cycle, reaches its highest accumulation in mitosis, and is then rapidly degraded
at the metaphase to anaphase transition [13]. Immunofluorescence microscopy
demonstrated that FLAG-tagged MCAK parallels the behavior of the endogenous
protein and is only weakly associated with the mitotic spindle and midbody following
metaphase. In contrast, GFP-tagged MCAK has been shown to be strongly associated
with the spindle and chromosomal structures throughout mitosis [14]. These differences
in localization can potentially be explained by the finding that fusion with GFP
prevents the degradation of MCAK [10].
8 Kinesin-13 Microtubule Depolymerizing Proteins as Targets for Cancer Therapy 119
8.3 Function
Kif2a has been shown to catalyze microtubule depolymerization [1], and its
depletion arrests cells in mitosis [8]. Cells depleted of Kif2a are defective in spindle
pole separation and accumulate monopolar spindles. In addition, very long microtu-
bules are formed that include those that are found in the central spindle [8, 15].
Interestingly, it was discovered that nocodazole treatment of cells lacking Kif2a
partially restores spindle bipolarity, suggesting that Kif2a depleted cells have some
form of hyperstabilized microtubules [4, 8]. Depletion of MCAK partially restores
bipolar spindle assembly in Kif2a-depleted cells [8]. This observation suggests that
Kif2a acts together with MCAK to control bipolar spindle assembly, but each
performs a distinct and potentially antagonistic function that necessitates an optimal
balance between the two proteins.
Kif2b is a very low abundance protein that is primarily located in the outer kinet-
ochore during prometaphase where it has been shown to correct kinetochore micro-
tubule attachment errors during the early stages of mitosis [4, 16]. The recruitment
of Kif2b to kinetochores is mediated by CLASP1 in a process that appears to be
regulated by Aurora B kinase. A high concentration of Aurora B on the kinetochore
during prometaphase promotes the binding of Kif2b to CLASP1; but as kineto-
chores on sister chromatids stretch apart when the chromosomes congress at the
metaphase plate, they are exposed to a lower Aurora B activity that allows astrin to
displace Kif2b from CLASP1. Under these high tension conditions, kinetochore
microtubules become more stable and MCAK may then provide the activity needed
to correct any further improper attachments as cells approach metaphase [17].
In support of this idea, it was recently shown that MCAK is required in the final step
of a mechanism needed to convert a kinetochore that is bound to the lateral wall
of a microtubule to an end-on microtubule attachment [18]. It should be noted
that another study did not find a stable association of Kif2b with CLASP1 but did
find an association with Cep170, a microtubule binding protein that could provide
an additional binding target for Kif2b [11]. It thus appears that Kif2b localization
may involve multiple proteins and the regulation of its activity may be controlled by
phosphorylation and perhaps other posttranslational modifications.
In addition to its role in correcting aberrant chromosome attachments, Kif2b is
needed for bipolar spindle assembly. Depletion of Kif2b leads to a high incidence of
monopolar spindles that arise transiently but then progress to bipolarity. Nonetheless,
Kif2b depletion produces cells with an increased frequency of lagging chromo-
somes, presumably due to persistent errors in kinetochore microtubule attachment
[16]. Similar to the ability of MCAK depletion to restore spindle bipolarity in kif2a
depleted cells [8], depletion of MCAK was again able to partially restore bipolarity
in Kif2b depleted cells, but it was not able to prevent the increased frequency of
lagging chromosomes in those cells. Although the mechanism for this effect is
unknown, the observations indicate that Kif2b and MCAK may have overlapping
functions in correcting kinetochore microtubule attachment errors but could
potentially also have antagonistic functions needed for bipolar spindle assembly.
The possibility of overlapping functions in correcting chromosome misalignment is

supported by the observation that overexpression of Kif2b or MCAK (but not Kif2a)
has been shown to suppress the incidence of lagging chromosomes in tumor cell
lines [16].
Kif2c/MCAK is the most well studied member of the kinesin-13 family and is
named for its prominent location in the centromeres of mitotic chromosomes.
Like other members of this family, it is a homodimeric protein with microtubule
depolymerizing activity. Unlike conventional kinesins that have N- or C-terminal
motor domains, MCAK lacks the ability to transport cargo. Instead, it binds to
microtubules and diffuses bidirectionally to the microtubule ends where it can
catalyze microtubule depolymerization by promoting a curved conformation of the
protofilaments [19–21]. The presence of lagging chromosomes in MCAK-inhibited
cells led to an early suggestion that, during anaphase, the protein might be responsible
for poleward movement of sister chromatids by catalyzing microtubule depolymer-
ization at the kinetochores using a Pacman-like mechanism [14]. However, later
studies indicated that lagging chromosomes result from improper kinetochore
microtubule attachments formed earlier in mitosis [22], and the role of MCAK is to
sever the improper microtubule attachments and thus allow chromosomes to correctly
align at the metaphase plate [23]. This view is consistent with the observation that
the protein is largely degraded at the metaphase-to-anaphase transition [13].
Given its prominent role in ensuring faithful chromosome segregation, it is not
surprising that MCAK is essential for normal mitotic progression. Depletion of
MCAK increases the mitotic index and produces cells with longer and more abundant
astral fibers, resulting in a “hairy” spindle morphology [24, 25]. These abnormal
spindles may potentially be explained by the observations that MCAK stimulates
kinetochore microtubule turnover [26] and stimulates microtubule detachment from
centrosomes and spindle poles [27]. Although depletion of MCAK hinders mitotic
progression, many cells nevertheless escape the mitotic checkpoint with missegre-
gated chromosomes resulting in an increase in aneuploidy [23, 24].
For normal mitosis to occur, MCAK must be tightly regulated and this is
accomplished by posttranslational modifications such as phosphorylation and by
regulating the degradation of the protein. Although moderate levels of overexpression
(e.g. two times endogenous) can be tolerated, higher levels (6–8 times endogenous or
more) inhibit cell proliferation [27]. These high levels of expression delay CHO
cells in prometaphase, deplete microtubules, disrupt spindle assembly, and lead to
the formation of unstable multinucleated cells. MCAK was found to continuously
accumulate starting in G1 phase and reach its maximum level during the early stages
of mitosis. It is then degraded at the metaphase to anaphase transition by the protea-
some [13]. The signal for degradation appears to involve phosphorylation of residue
S628 by an unknown kinase, and this phosphorylation may additionally trigger
release of MCAK from the kinetochore once the chromosome has aligned on the
metaphase plate [10]. It was speculated that degradation of MCAK serves not only
to maintain a proper protein levels, but also to prevent the carryover of multiphos-
phorylated species of the protein created during mitosis into the next cell cycle.
MCAK activity is also regulated by phosphorylation through the action of

multiple kinases including Aurora A, Aurora B, cdk1/cyclin B1, and Plk1 [28].
For example, it has been found that Aurora B phosphorylates human MCAK at S95
and S192, sites that were originally identified as T95 and S196 in Xenopus MCAK [29].
The mutation S95A changes kinetochore binding affinity while the mutation S192A
reduces the preference of MCAK for the kinetochore, increases its localization
along chromosome arms, and inhibits its microtubule depolymerase activity [29–31].
Recently, it was shown that Aurora A kinase also phosphorylates MCAK in its
C-terminal region to control ran-dependent spindle bipolarity [32]. A more detailed
discussion of the effects of MCAK phosphorylation can be found in several excellent
reviews [5, 33, 34].
8.4 Clinical Studies of Kinesin-13 Proteins
There is growing evidence that kinesin-13 family proteins could play a role in
tumorigenesis. Kinesin-13 s are needed to assemble a normal spindle apparatus and
to ensure the fidelity of chromosome separation during mitosis. Abnormalities in
mitotic progression can cause aneuploidy that can further lead to transformation of
normal somatic cells into cancer cells. It is therefore not surprising that overexpres-
sion of kinesin-13 s has been reported in a number of metastatic tumors. For exam-
ple, Kif2a was found to be overexpressed in 70 % of 44 patients with oral carcinoma,
and high expression was associated with lymph node metastasis [35]. A possible
role for the protein in cancer cell invasion was strengthened by an ex-vivo experiment
showing that a tongue squamous cell carcinoma cell line (Tca8113) transfected with
si-RNA to deplete Kif2a exhibited decreased motility [35]. Similarly, transfection
with miR-183, a micro RNA that decreased Kif2a levels and inversely correlated
with the metastatic potential of lung cancer cells, inhibited the migration and
invasion capacity of HeLa cells [36].
Overexpression of MCAK has also been reported in tumors, most notably those
of the breast and gastrointestinal track. MCAK was among the proteins found to be
overexpressed in breast tumor cells [37] and a later microarray analysis of samples
derived from 81 patients showed high levels of MCAK in breast tumors but unde-
tectable levels in normal tissues except the testis [38]. The latter authors further
demonstrated that siRNA-mediated depletion of MCAK inhibited the proliferation
of T47D and HBC5 breast cancer cell lines and that MCAK expression could be
suppressed by p53.
Experiments using RT-PCR revealed elevated MCAK in 66 % of 65 patients with
gastric cancer and the high levels were linked to lymphatic invasion, lymph node
metastasis, and poor prognosis [39]. Moreover, a gastric cancer cell line stably
transfected with MCAK exhibited increased proliferation and greater motility in
trans-well assays. Similarly, MCAK was overexpressed in 91 of 120 patients with
colon cancer where it was linked to node metastasis, venous invasion, peritoneal
dissemination, and poor survival [40]. A second study involving 176 patients also
saw increased MCAK expression in colorectal cancer as well as in pancreatic,

gastric, breast, and head and neck cancer [41]. The authors examined expression of
NY-CO-58, a tumor antigen derived from MCAK [42], and found it to be elevated
in all tumor types, but most strongly in breast, colorectal, and gastric cancer.
Moreover, nearly 50 % of colorectal patient peripheral blood mononuclear cells
(PBMCs) exposed to peptide antigens exhibited CD4+ T cell responses, suggesting
that MCAK could be used for immunological approaches to treatment. This possi-
bility was further strengthened by direct experiments showing that a cytotoxic T-cell
response against cells with high MCAK expression could be elicited by ex-vivo
exposure of PBMCs to peptide antigens derived from MCAK [43].
More recent studies point to the possibility that elevated expression of MCAK
and perhaps other kinesin-13 proteins will be associated with many other types of
cancer. High MCAK levels have been reported in castration-resistant and
chemotherapy-resistant prostate cancer, and the knockdown of MCAK led to the
growth arrest of prostate cancer cells in culture [44]. MCAK was also shown to be
upregulated an average of fourfold in glioma samples taken from 40 patients [45].
High levels were associated with a poor survival rate and it was suggested that
MCAK could be used as a marker for prognosis.
Despite the increasing number of reports showing elevated kinesin-13 proteins in
tumor samples, a causal link between these changes and tumor progression has not
been firmly established. It was shown that MCAK levels rise during the cell cycle
and become maximal at mitosis [13]. Other kinesin-13 family members may exhibit
similar patterns of expression. Thus, the elevated levels of these proteins in tumors may
simply reflect the higher mitotic and proliferative activity of tumor versus normal
tissue. Nonetheless, high expression of MCAK in tumors and the appearance of
peptides on the cell surface could open the possibility of targeting MCAK using
immunotherapy. In addition, the strong correlation between elevated MCAK and
poor survival in cancer patients could make this protein a useful prognostic marker.
8.5 Role of Kinesin-13 Proteins in Cancer Progression
From the preceding discussion it is evident that kinesin-13 proteins play important
roles in cell division by maintaining spindle integrity and normal spindle function.
Overproduction as well as depletion of kinesin-13 proteins can affect spindle assem-
bly demonstrating that the levels and activity of these proteins must be tightly
regulated to ensure normal cell division. Kif2b and MCAK appear to act primarily
at kinetochores to correct aberrant microtubule attachments. Thus, changes in the
activities of these proteins would be expected to delay the congression of chromo-
somes to the metaphase plate, engage the mitotic checkpoint and ultimately lead to
apoptosis. However, depending on the severity of the defect, some cells might slip
through the checkpoint and continue to progress through the cell cycle as aneuploid
cells with unstable genomes that ultimately lead to cancer. The potential link
between kinesin-13 proteins and cancer progression was recently strengthened by

the identification of a nonsynonymous single-nucleotide polymorphism in the
coding region of MCAK that is strongly associated with colorectal cancer [46] and
by the finding that overexpression of Kif2a is associated with tumor progression and
metastasis [35]. Kif2a is primarily located at spindle poles and it appears to mainly
act to promote spindle bipolarity. Thus, alterations in Kif2a activity can also cause
defective mitotic progression and the inaccurate segregation of chromosomes that
ultimately promote tumorigenesis.
8.6 A New Role for Kinesin-13 s in Microtubule

Detachment from Spindle Poles
Kinesin-13 family proteins are thought to act by altering microtubule dynamics that
may be necessary for bipolar spindle assembly and proper microtubule attachment
to the kinetochores of segregating chromosomes. This view appears to be consistent
with experiments showing that kinesin-13 proteins increase the frequency of catas-
trophe and reduce the frequency of rescue at microtubule plus-ends [25] and that
MCAK or Kif2b inhibition reduces kinetochore fiber microtubule turnover [16, 26,
47]. However, some recent observations have begun to question the role of microtu-
bule dynamics in mitosis. While it remains likely that some level of microtubule
dynamicity is needed for successful cell division, a scan of the literature reveals that
cells can have widely different levels of dynamicity yet all the cells divide normally
[48–51]. More recent studies show that cells treated with low concentrations of
microtubule drugs exhibit suppressed microtubule dynamics but no change in their
ability to proliferate [52]. Conversely, paclitaxel-dependent cell lines with near
normal microtubule dynamics cannot proliferate when the drug is absent, but are
able to proliferate with suppressed dynamics when the drug is present [53]. These
observations led the authors to conclude that microtubule dynamics are not as
important for mitosis as is currently thought, and to look for other actions of the
drugs that might explain their ability to block cell division. It was found that low drug
concentrations that suppressed microtubule dynamics did not inhibit cell division
but strongly inhibited cell migration. The higher drug concentrations needed to
inhibit cell division in normal cells and promote cell division in paclitaxel-dependent
mutants, on the other hand, affected the detachment of microtubules from centro-
somes and spindle poles.
Microtubule detachment is a poorly studied but important phenomenon that is
cell cycle regulated. The frequency of detachment is low during interphase but
greatly increases when cells enter mitosis [52, 54]. The frequency of detachment is
also increased by treatment with drugs such as vinblastine, colchicine, nocodazole,
and others that inhibit microtubule assembly; but it is suppressed by drugs such as
paclitaxel and epothilones that promote microtubule assembly [52, 53]. Detached
microtubules are relatively short-lived as they can depolymerize from both
plus- and minus-ends, and they are able to translocate away from the centrosome
by an as yet unknown mechanism [52, 55]. These actions generate microtubule
fragments of variable length that may play a critical role in microtubule turnover
and spindle formation.
Although the mechanism of microtubule detachment is not yet well understood,
recent experiments have implicated kinesin-13 family members that are located at
the centrosome. It is noteworthy that MCAK function at kinetochores is well
studied, but its function at the spindle poles is still largely unexplored. There is
now strong evidence that one of its roles is to catalyze microtubule detachment.
This finding is consistent with the ability of MCAK to catalyze catastrophe from
both plus- and minus-ends of the microtubule, and with its elevated activity during
mitosis due to cell cycle dependent accumulation. Direct evidence that MCAK is
involved in microtubule detachment came from experiments showing that the
frequency of detachment was elevated in cells overexpressing MCAK [27] and
suppressed in cells depleted of MCAK [24]. Moreover, depletion of MCAK was
able to reverse the paclitaxel-dependence phenotype and allow the proliferation of
CHO cell lines with elevated frequencies of detachment resulting from mutations in
tubulin or overexpression of a neuronal and testis specific β-tubulin isotype. It is
noteworthy that growth of the paclitaxel dependent cell line with the stronger
phenotype was only partially restored by MCAK depletion, suggesting that MCAK
may be only one of multiple proteins at spindle poles that are involved in microtubule
detachment. Given its strong localization to spindle poles, Kif2a may be another
kinesin involved in detachment, but this prediction is yet to be tested.
8.7 Kinesin-13 Affects Sensitivity to Microtubule Inhibitors
The involvement of kinesin-13 proteins such as MCAK in microtubule detachment

has important therapeutic implications. While most current models of the mitotic
spindle envision continuous microtubules connecting spindle poles with chromo-
somes, some older work on mitotic spindles along with more recent studies of
meiotic spindles have detected the presence of numerous microtubule fragments and
this has led to the idea that spindle fibers consist of both continuous and overlapping
fragmented microtubules connected by motor proteins [56–59]. Some of these frag-
ments appear to be nucleated from the sides of existing microtubules catalyzed by a
protein called augmin [60], but it is likely that microtubule detachment also plays an
important role in their formation. The premise for therapy directed at the mitotic
spindle is that a proper mixture of fragmented and continuous microtubules is
needed for successful cell division. Too many, very short fragments may produce a
weak unstable spindle; but too few fragments may produce an overly stable spindle
that is unable to remodel during the various stages of mitosis.
Given the ability of proteins like MCAK to catalyze microtubule detachment,
one would predict that inhibition of its protein level or activity would lead to
hyperstable spindles whereas overexpression or activation of MCAK would lead

to unstable spindles. A number of studies support this prediction. For example,
MCAK depletion leads to the formation of “hairy” spindles that have abnormally
long astral fibers [4, 24, 25, 61, 62], Kif2a regulates microtubule length in the central
spindle as well as the overall spindle size [15, 63], and Kif2b depletion leads to
disorganized spindles with a three to fourfold increase in microtubule polymer [4].
These observations suggest that targeting kinesin-13 activity to disrupt spindle
function should be an effective strategy to kill tumor cells. Unfortunately, such a
direct approach currently suffers from a lack of drugs with the needed specificity.
On the other hand, drugs that target microtubules are in widespread use and they
also affect microtubule detachment suggesting that kinesin-13 manipulation may
affect the sensitivity of cells to microtubule inhibitors. Thus, tumors that overex-
press kinesin-13 proteins would be predicted to have a high frequency of microtu-
bule detachments making the cells more sensitive to drugs like vinblastine that
increase detachment, but more resistant to drugs like paclitaxel that suppress detach-
ment. In fact, support for these predictions has already appeared. It was reported
that the effects of Kif2a depletion (expected to produce fewer detachments) can be
counteracted by the presence of nocodazole (an inducer of detachment) [4, 8],
and that depletion of a kinesin-13-like protein in Drosophila S2 cells increased
resistance to colchicine whereas overexpression of the protein increased sensitivity
to the drug [64]. Similarly, overexpression of Kif2b has been reported to increase
the sensitivity of spindle microtubules to nocodazole-induced depolymerization
whereas Kif2b deficient cells experienced delayed kinetochore microtubule depoly-
merization [16]. It was further reported that paclitaxel increases the density of
microtubules at spindle poles at the expense of kinetochore fibers and that MCAK
depletion increases astral but reduces K-fiber fluorescence [47]. Consistent with
these observations, it was directly shown that overexpression of MCAK increased
the frequency of microtubule detachment, decreased microtubule polymer levels,
and made cells resistant to paclitaxel and epothilone A, but more sensitive to
colcemid [27]. Depletion of MCAK lowered the frequency of detachment and made
cells more sensitive to paclitaxel [24].
The results argue that all three known kinesin-13 proteins can affect the sensitivity
of cells to microtubule inhibitors, likely by a mechanism that affects the detachment
of microtubules from spindle poles. Thus, efforts to therapeutically target kinesin-13
proteins may potentially be used in combination with existing microtubule inhibitors
to enhance tumor eradication. Treatments to enhance kinesin-13 levels or activity
would be expected to increase microtubule detachment and thereby make cells
more sensitive to drugs that also increase detachment (vinca alkaloids, colchicine
analogs, etc.); but treatments to inhibit kinesin-13 would decrease detachment and
increase sensitivity to drugs such as the taxanes and epothilones that also decrease
detachment. Levels or activity of kinesin-13 proteins could also potentially be used
as a predictive marker for tumor sensitivity to microtubule inhibitors. Tumors with
elevated kinesin-13 proteins are predicted to be relatively resistant to taxanes and
epothilones, but tumors with low kinesin-13 proteins are expected to be relatively
resistant to vinca alkaloids and colchicine analogs. Many of these predictions are
still in need of experimental and clinical verification, but a clearer picture on the
role of kinesin-13 proteins in cancer therapy is now emerging.
8.8 Drugs That Bind Kinesin-13 Proteins
Unfortunately, drugs that specifically bind to Kif2a or Kif2b and interfere with their
function have not yet been identified, but a group of compounds known as sulfoqui-
novosylacylglycerols (SQAGs, Fig. 8.1) have been shown to interact with MCAK. In
vitro analysis suggested that the binding site is a 14 amino acid sequence present in
the neck linker domain that together with the motor domain is involved in MCAK
activity and binding to microtubules [65]. However, cells treated with SQAGs are
blocked in both mitotic and S phases of the cell cycle indicating that the drugs are
likely to have additional cellular targets [66, 67]. Although none of these drugs are
currently in clinical trials, SQAGs may yet prove to be useful drugs for cancer
chemotherapy and may lead to the development of compounds that more specifically
inhibit MCAK.
Another potential group of inhibitors is represented by thiazole compounds that
were originally introduced as Eg5 inhibitors but have been shown to strongly inhibit
the MT-simulated ATPase activity of MCAK. The structures of some of these agents
are shown in Fig. 8.2. Compound A is about tenfold more selective for MCAK
compared to Eg5, and does not appear to inhibit other kinesins like Kif5A, Kif5B,
Kif1B, CENP-E, Kif3A, and MKLP-1 [68]. Compound B, with an IC50 value of
160–170 nM, is the most potent inhibitor of MCAK [68] (reviewed by Good and
Kozielski [67]). The activities of these compounds against other members of the
kinesin-13 family have not yet been evaluated and thus the specificity for MCAK is
not established.
SO3Na
HO
O
HO O OR2
OH OR1
Fig. 8.1 Structures of the
Sulfoquinovosylacylglycerol
group of kinesin-13 R1=H, R2=CO(CH2)16CH3
inhibitors. Members of this R1=H, R2=CO(CH2)7CH=CH(CH2)7CH3
group of compounds R1=H, R2=CO(CH2)16CH3
differ primarily at positions
R1 and R2 R1,R2=CO(CH2)16CH3
a b
S S
Cl
N N
N
N
N
Fig. 8.2 Structures of thiazole inhibitors of kinesin-13 s
8.9 A Model for Predicting the Effects of Kinesin-13

Intervention to Treat Cancer
The spindle can be viewed as a metastable structure that requires rapid remodeling
in order to carry out its mission of segregating chromosomes prior to cell division.
The model in Fig. 8.3 assumes that spindle function relies on dynamic events that
occur not only at the plus ends (growth and shrinkage), but also, and perhaps more
importantly, on sliding of microtubule fragments present in the kinetochore fibers
connecting spindle poles to kinetochores. A proper mixture of fragmented and
continuous microtubules ensures a level of plasticity that promotes efficient
chromosome capture and segregation as well as successful cell division. However,
tubulin mutations that stabilize microtubules, treatment with drugs like paclitaxel
that promote microtubule assembly, or depletion of kinesin-13 proteins such as
MCAK can inhibit the production of microtubule fragments leading to stabilized
spindles that are functionally less efficient. Conversely, destabilized spindles can
arise from destabilizing tubulin mutations, treatment with drugs like colcemid or
vinca alkaloids that inhibit microtubule assembly, or by overexpression of kine-
sin-13 proteins like MCAK. If the defect is relatively small (weak mutations, lower
drug concentrations, small changes in kinesin-13 production) cell division may still
occur, but there may be a higher incidence of aneuploidy. Stabilized spindles would
make the cells relatively resistant to drugs like colcemid and vinca alkaloids;
whereas the destabilized spindles would confer resistance to drugs like paclitaxel
and epothilones.
More severe defects (stronger mutations, higher drug concentrations, larger
changes in kinesin-13 production) can produce very stable or unstable spindles
unable to efficiently segregate chromosomes. This would lead to prolonged mitotic
checkpoint activation and cell death or to cells that exit mitosis but fail to divide.
The highly aneuploid cells could themselves die in subsequent cell cycles or
contribute to the formation or aggressiveness of malignant cells as has been inferred
by clinical observations linking overexpression of kinesin-13 proteins to tumor cell
metastasis and poor patient prognosis. Cells that form very unstable spindles
frequently have a paclitaxel dependence phenotype in which spindle structure and
successful cell division can be restored by the presence of paclitaxel (or other drugs
Too Many Fragments Normal Cell Too Few Fragments

Increasing Severity of Defect
Cmd Treatment Ptx Treatment

Tubulin Mutation Tubulin Mutation
MCAK MCAK
PtxR CmdR
Abnormal spindle Normal spindle Abnormal spindle
Ptx Treatment Cmd Treatment

MCAK MCAK
PtxD CmdD
No cell division Successful cell division No cell division
Fig. 8.3 Model for the effects of various mitotic spindle modifiers on cell proliferation.
The top row shows the effects of various treatments on the production of microtubule fragments at
the spindle poles. Minor effects on cell division but significant effects on drug sensitivity (e.g.
paclitaxel resistance, PtxR; or colcemid resistance, CmdR) can occur when the perturbations are not
too severe. The bottom row shows the expected effects of more severe perturbations on spindle
structure and cell division that can lead to drug dependent phenotypes (e.g. paclitaxel dependence,
PtxD; or colcemid dependence, CmdD). Treatments potentially able to correct those defects and
restore cell division are indicated
that inhibit the production of microtubule fragments), or by depleting MCAK, and

perhaps other kinesin-13 proteins. Cells that form very stable spindles, on the other
hand, can be rescued by treating with drugs like colcemid that destabilize microtu-
bules, or by overexpressing MCAK (and perhaps other kinesin-13 proteins) that
increase the production of microtubule fragments.
Although the model emphasizes treatments that have already been shown to
affect fragment production through detachment of microtubules from spindle poles,
it provides a framework for predicting the effects of modulating other proteins or
drugs that may also interfere with microtubule detachment. For example, overex-
pression of β3-tubulin has been shown to destabilize microtubules and produce
paclitaxel resistant/dependent cells that can be rescued by depleting MCAK [24, 69].
Similarly, overexpression of β5-tubulin can increase microtubule detachment, produce
microtubule fragments, and confer a paclitaxel dependence phenotype [70, 71].
On the other hand, overexpression of β6-tubulin produces microtubule fragments by
a different mechanism and does not confer paclitaxel resistance or dependence [72].
It will be of interest to test whether other proteins such as augmin, katanin, and
spastin that have been implicated in microtubule fragment formation can affect cell
sensitivity to microtubule inhibitors.
8.10 Summary and Perspectives
The previous discussion focused on the possibility of targeting kinesin-13 proteins

to kill tumor cells, alter their sensitivity to microtubule inhibitors, or reverse
resistance to drugs that target microtubules. Although there are a few drugs that can
inhibit these kinesins, they lack specificity and there is a pressing need to discover
better agents. Indirect methods of targeting kinesin-13 s could potentially include
inhibiting kinases that are known to activate, inhibit, alter the localization, or affect
the degradation of the proteins, but these approaches are likely to produce unwanted
side effects due to the fact that kinases typically have multiple substrates. In fact,
mitotic kinase inhibitors have thus far had very limited success in treating cancer [73].
Molecular genetic approaches to directly alter kinesin-13 levels and interfere with
tumor growth is theoretically possible, but current methods for achieving this in
patients are also limited.
On the other hand, kinesin-13 status has very significant implications for the
sensitivity of cells to currently used and future mitotic inhibitors that target micro-
tubules and spindle assembly. Thus, assaying the levels of kinesin-13 proteins in
tumors may aid in predicting the dosage of microtubule inhibitors that will be effec-
tive in patients. It is also possible that weak kinesin-13 inhibitors that by themselves
are unable to prevent tumor growth, may work synergistically to enhance the effec-
tiveness of microtubule inhibitors. As mentioned earlier, kinesin-13 s may provide
effective targets for developing immunotherapeutic approaches to treat tumors that
overexpress those proteins; and they may potentially be used as markers for patient
prognosis.
In addition to targeting kinesin-13 s to kill tumor cells, inhibiting their activity
may reduce tumor malignancy. Several studies cited earlier demonstrated that tumor
cells that overexpress kinesin-13 proteins tend to be more aggressive and that reduc-
ing the production of Kif2a or MCAK can inhibit cell motility [35, 36, 39, 74].
These actions are likely to be mediated by changes in the microtubule cytoskeleton.
Low concentrations of microtubule drugs have been shown to suppress microtubule
dynamics and inhibit cell migration without affecting cell division [52, 75].
The need for dynamic microtubules appears to be most acute at the trailing edge of
migrating cells where the cytoskeleton needs to be remodeled in order to allow tail
retraction and forward movement [76]. Given that kinesin-13 proteins have been
shown to catalyze catastrophe and inhibit rescue at microtubule ends, they are likely
to have important effects on the ability of microtubules to remodel. Thus, treatments
aimed at inhibiting kinesin-13 proteins would be expected to interfere with tumor
cell metastasis and invasion of neighboring tissues. In addition, it has been shown
that suppressing microtubule dynamics in vascular endothelial cells inhibits their
migration, suggesting that therapies aimed at kinesin-13 proteins may also limit
tumor growth by inhibiting angiogenesis [77].
References
destabilizing enzymes. Cell 96:69–78
2. Lawrence CJ et al (2004) A standardized kinesin nomenclature. J Cell Biol 167:19–22
3. Miki H, Okada Y, Hirokawa N (2005) Analysis of the kinesin superfamily: insights into
structure and function. Trends Cell Biol 15:467–476
4. Manning AL et al (2007) The kinesin-13 proteins Kif2a, Kif2b, and Kif2c/MCAK have
distinct roles during mitosis in human cells. Mol Biol Cell 18:2970–2979
5. Ems-McClung SC, Walczak CE (2010) Kinesin-13 s in mitosis: key players in the spatial and
temporal organization of spindle microtubules. Semin Cell Dev Biol 21:276–282
6. Moores CA, Milligan RA (2006) Lucky 13-microtubule depolymerisation by kinesin-13
motors. J Cell Sci 119:3905–3913
7. Walczak CE, Gayek S, Ohi R (2013) Microtubule-depolymerizing kinesins. Annu Rev Cell
Dev Biol 29:417–441
8. Ganem NJ, Compton DA (2004) The KinI kinesin Kif2a is required for bipolar spindle
assembly through a functional relationship with MCAK. J Cell Biol 166:473–478
9. Skube SB, Chaverri JM, Goodson HV (2010) Effect of GFP tags on the localization of EB1
and EB1 fragments in vivo. Cytoskeleton 67:1–12
10. Ganguly A, Bhattacharya R, Cabral F (2012) Control of MCAK degradation and removal from
centromeres. Cytoskeleton 69:303–311
11. Welburn JPI, Cheeseman IM (2012) The microtubule-binding protein Cep170 promotes the
targeting of the kinesin-13 depolymerase Kif2b to the mitotic spindle. Mol Biol Cell
23:4786–4795
12. Wordeman L, Wagenbach M, Maney T (1999) Mutations in the ATP-binding domain affect the
subcellular distribution of mitotic centromere-associated kinesin (MCAK). Cell Biol Int
23:275–286
13. Ganguly A, Bhattacharya R, Cabral F (2008) Cell cycle dependent degradation of MCAK:
evidence against a role in anaphase chromosome movement. Cell Cycle 7:3187–3193
14. Maney T, Hunter AW, Wagenbach M, Wordeman L (1998) Mitotic centromere-associated
kinesin is important for anaphase chromosome segregation. J Cell Biol 142:787–801
15. Uehara R et al (2013) Aurora B and Kif2A control microtubule length for assembly of a
functional central spindle during anaphase. J Cell Biol 202:623–636
16. Bakhoum SF, Thompson SL, Manning AL, Compton DA (2009) Genome stability is ensured
by temporal control of kinetochore-microtubule dynamics. Nat Cell Biol 11:27–35
17. Manning AL et al (2010) CLASP1, astrin and Kif2b form a molecular switch that regulates
kinetochore-microtubule dynamics to promote mitotic progression and fidelity. EMBO J
29:3531–3543
18. Shrestha RL, Draviam VM (2013) Lateral to end-on conversion of chromosome-microtubule
attachment requires kinesins CENP-E and MCAK. Curr Biol 23:1514–1526
19. Helenius J, Brouhard G, Kalaidzidis Y, Diez S, Howard J (2006) The depolymerizing kinesin
MCAK uses lattice diffusion to rapidly target microtubule ends. Nature 441:115–119
20. Ogawa T, Nitta R, Okada Y, Hirokawa N (2004) A common mechanism for microtubule
destabilizers-M type kinesins stabilize curling of the protofilament using the class-specific
neck and loops. Cell 116:591–602
21. Shipley K et al (2004) Structure of a kinesin microtubule depolymerization machine. EMBO J
23:1422–1432
22. Cimini D et al (2001) Merotelic kinetochore orientation is a major mechanism of aneuploidy
in mitotic mammalian tissue cells. J Cell Biol 153:517–527
23. Kline-Smith SL, Khodjakov A, Hergert P, Walczak CE (2004) Depletion of centromeric
MCAK leads to chromosome congression and segregation defects due to improper kineto-
chore attachments. Mol Biol Cell 15:1146–1159
24. Ganguly A, Yang H, Pedroza M, Bhattacharya R, Cabral F (2011) Mitotic centromere

associated kinesin (MCAK) mediates paclitaxel resistance. J Biol Chem 286:36378–36384
25. Kline-Smith SL, Walczak CE (2002) The microtubule destabilizing kinesin XKCM1 regulates
microtubule dynamic instability in cells. Mol Biol Cell 13:2718–2731
26. Wordeman L, Wagenbach M, von Dassow G (2007) MCAK facilitates chromosome movement
by promoting kinetochore microtubule turnover. J Cell Biol 179:869–879
27. Ganguly A, Yang H, Cabral F (2011) Overexpression of mitotic centromere-associated kinesin
stimulates microtubule detachment and confers resistance to paclitaxel. Mol Cancer Ther
10:929–937
28. Sanhaji M, Friel CT, Wordeman L, Louwen F, Yuan J (2011) Mitotic centromere-associated
kinesin (MCAK): a potential cancer drug target. Oncotarget 2:935–947
29. Zhang X, Lan W, Ems-McClung SC, Stukenberg PT, Walczak CE (2007) Aurora B phosphory-
lates multiple sites on mitotic centromere-associated kinesin to spatially and temporally
regulate its function. Mol Biol Cell 18:3264–3276
6:253–268
31. Lan W et al (2004) Aurora B phosphorylates centromeric MCAK and regulates its localization
and microtubule depolymerization activity. Curr Biol 14:273–286
32. Zhang X, Ems-McClung SC, Walczak C (2008) Aurora A phosphorylates MCAK to control
ran-dependent spindle bipolarity. Mol Biol Cell 19:2752–2765
33. Moore A, Wordeman L (2004) The mechanism, function and regulation of depolymerizing
kinesins during mitosis. Trends Cell Biol 14:537–546
34. Gorbsky GJ (2004) Mitosis: MCAK under the aura of aurora B. Curr Biol 14:R346–R348
35. Wang C-Q et al (2010) Overexpression of Kif2a promotes the progression and metastasis of
squamous cell carcinoma of the oral tongue. Oral Oncol 46:65–69
36. Li G, Luna C, Qiu J, Epstein DL, Gonzalez P (2010) Targeting of integrin beta1 and kinesin
2alpha by microRNA 183. J Biol Chem 285:5461–5471
37. Perou CM et al (1999) Distinctive gene expression patterns in human mammary epithelial cells
and breast cancers. Proc Natl Acad Sci U S A 96:9212–9217
38. Shimo A et al (2008) Involvement of kinesin family member 2C/mitotic centromere-associated
kinesin overexpression in mammary carcinogenesis. Cancer Sci 99:62–70
39. Nakamura Y et al (2007) Clinicopathological and biological significance of mitotic centromere-
associated kinesin overexpression in human gastric cancer. Br J Cancer 97:543–549
40. Ishikawa K et al (2008) Mitotic centromere-associated kinesin is a novel marker for prognosis
and lymph node metastasis in colorectal cancer. Br J Cancer 98:1824–1829
41. Gnjatic S et al (2010) NY-CO-58/KIF2C is overexpressed in a variety of solid tumors and
induces frequent T cell responses in patients with colorectal cancer. Int J Cancer 127:381–393
42. Scanlan MJ et al (2002) Cancer-related serological recognition of human colon cancer: identi-
fication of potential diagnostic and immunotherapeutic targets. Cancer Res 62:4041–4047
43. Kawamoto M et al (2011) Identification of HLA-A *0201/-A*2402-restricted CTL epitope-
peptides derived from a novel cancer/testis antigen, MCAK, and induction of a specific antitu-
mor immune response. Oncol Rep 25:469–476
44. Sircar K et al (2012) Mitosis phase enrichment with identification of mitotic centromere-
associated kinesin as a therapeutic target in castration-resistant prostate cancer. PLoS One
7:e31259
45. Bie L, Zhao G, Wang Y-P, Zhang B (2012) Kinesin family member 2C (KIF2C/MCAK) is a
novel marker for prognosis in human gliomas. Clin Neurol Neurosurg 114:356–360
46. Kumar A, Rajendran V, Sethumadhavan R, Purohit R (2013) Evidence of colorectal cancer-
associated mutation in MCAK: a computational report. Cell Biochem Biophys 67:837–851
47. Rizk RS et al (2009) MCAK and paclitaxel have differential effects on spindle microtubule
organization and dynamics. Mol Biol Cell 20:1639–1651
48. Cassimeris L, Pryer NK, Salmon ED (1988) Real-time observation of microtubule dynamic
instability in living cells. J Cell Biol 107:2223–2231
49. Honore S et al (2004) Synergistic suppression of microtubule dynamics by discodermolide and

paclitaxel in non-small cell lung carcinoma cells. Cancer Res 64:4957–4964
50. Panda D, DeLuca K, Williams D, Jordan MA, Wilson L (1998) Antiproliferative mechanism
of action of cryptophycin-52: kinetic stabilization of microtubule dynamics by high-affinity
binding to microtubule ends. Proc Natl Acad Sci U S A 95:9313–9318
51. Yvon AM, Wadsworth P, Jordan MA (1999) Taxol suppresses dynamics of individual microtu-
bules in living human tumor cells. Mol Biol Cell 10:947–959
52. Yang H, Ganguly A, Cabral F (2010) Inhibition of cell migration and cell division correlates
with distinct effects of microtubule inhibiting drugs. J Biol Chem 285:32242–32250
53. Ganguly A, Yang H, Cabral F (2010) Paclitaxel dependent cell lines reveal a novel drug activity.
Mol Cancer Ther 9:2914–2923
54. Rusan NM, Wadsworth P (2005) Centrosome fragments and microtubules are transported
asymmetrically away from division plane in anaphase. J Cell Biol 168:21–28
55. Keating TJ, Peloquin JG, Rodionov VI, Momcilovic D, Borisy GG (1997) Microtubule release
from the centrosome. Proc Natl Acad Sci U S A 94:5078–5083
56. Burbank KS, Groen AC, Perlman ZE, Fisher DS, Mitchison TJ (2006) A new method reveals
microtubule minus ends throughout the meiotic spindle. J Cell Biol 175:369–375
57. Mastronarde DN, McDonald KL, Ding R, McIntosh JR (1993) Interpolar spindle microtubules
in PTK cells. J Cell Biol 123:1475–1489
58. Tulu US, Rusan NM, Wadsworth P (2003) Peripheral, non-centrosome-associated microtubules
contribute to spindle formation in centrosome-containing cells. Curr Biol 13:1894–1899
59. Yang G et al (2007) Architectural dynamics of the meiotic spindle revealed by single-
fluorophore imaging. Nat Cell Biol 9:1233–1242
60. Goshima G, Mayer M, Zhang N, Stuurman N, Vale RD (2008) Augmin: a protein complex
required for centrosome-independent microtubule generation within the spindle. J Cell Biol
181:421–429
61. Tournebize R et al (2000) Control of microtubule dynamics by the antagonistic activities of
XMAP215 and XKCM1 in Xenopus egg extracts. Nat Cell Biol 2:13–19
62. Walczak CE, Mitchison TJ, Desai A (1996) XKCM1: a Xenopus kinesin-related protein that
regulates microtubule dynamics during mitotic spindle assembly. Cell 84:37–47
63. Wilbur JD, Heald R (2013) Mitotic spindle scaling during Xenopus development by kif2a and
importin alpha. Elife 2:e00290
64. Schimizzi GV, Currie JD, Rogers SL (2010) Expression levels of a kinesin-13 microtubule
depolymerase modulates the effectiveness of anti-microtubule agents. PLoS One 5:e11381
65. Aoki S, Ohta K, Yamazaki T, Sugawara F, Sakaguchi K (2005) Mammalian mitotic centromere-
associated kinesin (MCAK): a new molecular target of sulfoquinovosylacylglycerols novel
antitumor and immunosuppressive agents. FEBS J 272:2132–2140
66. Ohta K et al (2000) Studies on a novel DNA polymerase inhibitor group, synthetic sulfoquino-
vosylacylglycerols: inhibitory action on cell proliferation. Mutat Res 467:139–152
67. Good JA, Skoufias DA, Kozielski F (2011) Elucidating the functionality of kinesins: an
overview of small molecule inhibitors. Semin Cell Dev Biol 22:935–945
68. Rickert KW et al (2008) Discovery and biochemical characterization of selective ATP com-
petitive inhibitors of the human mitotic kinesin KSP. Arch Biochem Biophys 469:220–231
69. Hari M, Yang H, Zeng C, Canizales M, Cabral F (2003) Expression of class III β-tubulin
reduces microtubule assembly and confers resistance to paclitaxel. Cell Motil Cytoskeleton
56:45–56
70. Bhattacharya R, Yang H, Cabral F (2011) Class V β-tubulin alters dynamic instability and
stimulates microtubule detachment from centrosomes. Mol Biol Cell 22:1025–1034
71. Bhattacharya R, Cabral F (2004) A ubiquitous β-tubulin disrupts microtubule assembly and
inhibits cell proliferation. Mol Biol Cell 15:3123–3131
72. Yang H, Ganguly A, Yin S, Cabral F (2011) Megakaryocyte lineage-specific class VI β-tubulin
suppresses microtubule dynamics, fragments microtubules, and blocks cell division.
Cytoskeleton (Hoboken) 68:175–187
73. Pitts TM, Davis SL, Eckhardt SG, Bradshaw-Pierce EL (2014) Targeting nuclear kinases in
cancer: development of cell cycle kinase inhibitors. Pharmacol Ther 142:258–269
74. Zaganjor E et al (2014) Ras regulates kinesin 13 family members to control cell migration
pathways in transformed human bronchial epithelial cells. Oncogene 33:5457–5466
75. Mikhailov A, Gundersen GG (1998) Relationship between microtubule dynamics and lamel-
lipodium formation revealed by direct imaging of microtubules in cells treated with nocodazole
or Taxol. Cell Motil Cytoskeleton 41:325–340
76. Ganguly A, Yang H, Sharma R, Patel KD, Cabral F (2012) The role of microtubules and their
dynamics in cell migration. J Biol Chem 287:43359–43369
77. Ganguly A, Yang H, Zhang H, Cabral F, Patel KD (2013) Microtubule dynamics control tail
retraction in migrating vascular endothelial cells. Mol Cancer Ther 12:2837–2846
Chapter 9
Chromokinesins in Genome Maintenance
and Cancer
Manjari Mazumdar and Tom Misteli
9.1 Introduction
Kinesin proteins are multi-functional molecular motors that are associated with the
microtubule cytoskeleton [1–4]. The vast majority of the well-classified subfamilies
of kinesin motors is cytoplasmic and is involved in two critical cellular functions:
transport of intracellular cargo and mitotic cell division, including cytokinesis [5, 6].
The latter function has made several mitotic kinesins interesting potential targets
for cancer therapy [7]. However, their multi-functionality complicates clinical
approaches because of the potential of undesired cellular responses.
A prominent class of kinesin motors is the chromokinesins. In contrast to most
other kinesins, these proteins possess both microtubule and chromatin-binding
activities and are nuclear during interphase [8–11]. This subclass of kinesins associ-
ates with chromosomes both during interphase and mitosis and with the central
spindle during cell division [12–16]. Chromokinesins are microtubule (MT) plus
end–directed motors and were proposed to contribute to generation of polar ejection
forces to promote chromosome movements away from the spindle pole to help
form the metaphase plate [8, 12, 14, 17–23]. However, while some human chromo-
kinesins contribute to anti-polar movements by generating polar ejection forces [12,
22, 24], others do not exhibit any such motor activity, implying that generation
of polar ejection force is not a universal feature of chromokinesins and that these
M. Mazumdar (*)
Kimberley Lane, Houston, TX 77079, USA
e-mail: m_mazumdar_56006@yahoo.com
T. Misteli
National Cancer Institute, NIH, Bethesda, MD 20892, USA
e-mail: mistelit@mail.nih.gov
© Springer Science+Business Media Dordrecht (outside the USA) 2015 135

136 M. Mazumdar and T. Misteli
proteins exert their function by other means. We summarize here some of the generic
functions of chromokinesins during interphase and mitosis, and, given their promi-
nent role in cell division, we discuss their potential role in cancer and as anti-cancer
agents.
9.2 Chromokinesins KIF4 and KIF10
The most prominent chromokinesins are the kinesin-4 family member KIF4
(HsKIF4A in human) and the kinesin-10 family member KID (Fig. 9.1). These proteins
are paradigmatic for chromokinesins and represent many of the diverse properties
of the family.
a
NLS Cys-rich
Kinesin-4 Motor Domain Coiled Coil Divergent C-terminus
DNA-binding
Kinesin-10 Motor Domain Divergent C-terminus
Fig. 9.1 Chromokinesins KIF4 and KID. (a) Schematic representation of the domain structure
of human KIF4 and KID shows various regions and the short motifs present in the molecules. Both
KIF4 (1232 aa) and KID (675 aa) contain highly conserved kinesin motor domain but the regions
beyond the motor domain are significantly different in the two molecules. While the DNA-binding
region of KID is well defined, it is not for KIF4, although experimental data suggests that the
C-terminal globular domain of KIF4 is sufficient for DNA binding. The long coiled-coil leucine
zipper in KIF4 is a potential protein-protein interaction domain and has been identified to be part
of the PRC1-binding domain during cytokinesis. (b) KIF4 plays major roles in mitosis. (Left)
Various stages of mitotic cell division in mock-transfected normal human lung fibroblast MRC-5
cells (Right) RNAi-mediated hKIF4A-depletion in MRC-5 cells leads to defects in spindle func-
tion and causes chromosome missegregation. MRC-5 cells were fixed after 48 h of transfection and
stained for microtubules (red) and DNA (blue). In hKIF4A-depleted cells, the spindle is less
focused at the poles (white arrow) and the spindle may lose its integrity. Microtubule organization
of both prometaphase and metaphase spindles is abnormal and accompanied by chromosome
alignment defects (white arrows). HKIF4A depletion also causes frequent anaphase bridges,
elongated spindles and defective cytokinesis
9 Chromokinesins in Genome Maintenance and Cancer 137
KIF4 has essential roles in diverse facets of mitosis [13]. KIF4A controls MT
dynamics during anaphase and telophase and is important for proper spindle
midzone formation and cytokinesis [11, 25–28]. RNAi-based analysis in normal
human MRC-5 and HeLa cells showed that hKIF4A is also involved in chromo-
some congression and cytokinesis [11, 25, 27, 29]. In Caenorhabditis elegans, its
homologue KLP-19 is required for metaphase plate formation and is involved in
generating polar ejection force [30, 31]. In Drosophila, the KIF4 homolog Klp3A
controls mitotic spindle length [31] and is essential for cytokinesis [32]. The
Xenopus homologue Xklp1 is required for chromosome congression [33], in vitro-
spindle assembly in egg extract, regulates MT dynamics, and controls MT density
of spindles [34, 35].
The protein domain structure of human KIF4A and its vertebrate homologs have
highly conserved ATPase motifs within a conventional kinesin motor domain at the
N-terminal, a long coiled-coil in the middle, a 6 amino acid nuclear localization
motif, followed by a cys-rich globular domain at the C-terminal tail (Fig. 9.1).
The Drosophila Kinesin-4 homologue KLP3A is substantially divergent from these
isoforms beyond the motor domain. It lacks the coiled coil region and the NLS, but
has a TSP type-3 domain at the C-terminus. In all vertebrate homologues, the
leucine zipper forms the stalk and the cys-rich C-terminal domains of hKIF4A are
important for most of the interphase functions and for targeting the protein to mitotic
chromosomes [36, 37]. The N-terminal motor domain is required to execute the full
complement of mitotic chromosome functions [11, 36–38].
KID is a DNA-binding monomeric plus end-directed motor [9, 15, 39]. The pro-
tein contains a conventional microtubule-binding region in the motor domain and a
DNA-binding motif as well as a second microtubule-binding region reside in its tail
domain. During mitosis, KID binds to chromosome arms and spindle MTs and this
association is regulated by CDK1 [16]. The Xenopus laevis Xkid, which has 54 %
identity to the human homologue, is essential for the proper alignment of chromo-
somes and spindle orientations in egg extracts [8, 14]. RNAi-based studies in human
cells or gene-targeted mice have revealed minor problems in chromosome congres-
sion and alignment at the metaphase plate upon loss of KID [12, 16, 29, 40, 41].
Human KID appears to be required for determining spindle length [41] and to
control chromosome arm orientation and oscillation [12, 42]. Although the function
of hKID as a motor, which pushes chromosome arms and exert forces on the bipolar
spindle [43, 44], is well established, a report of hKID as an organizer of the spindle
pole in the absence of NUMA suggests that hKID has redundant functions in
mitosis in somatic cells [40, 41].
Both KIF4 and KID appear to be highly multi-functional proteins involved in
numerous nuclear and mitotic functions, including spindle organization, chromosome
arm orientation, chromosome condensation, and segregation during the mitotic
cycle [13] (Fig. 9.1). It is possible that the myriad of cellular functions reported for
KIF4 may reflect species-specific divergent functions of KIF4 homologues,
especially since the sequence of the C-terminal domain differs between the various
species. Moreover, KIF4 has two homologues in humans: KIF4A and KIF4B share
98 % sequence homology, but have different chromosomal locations with KIF4A
located on chromosome Xq13.1 and KIF4B on chromosome 5q33.1, likely arisen

by gene duplication during evolution [45]. The extent to which the functions of the
two homologues overlap has not been determined and it is possible that some of the
diverse intracellular functions that have been reported for KIF4, such as neuronal
vesicle transport [46, 47] or the association with immunodeficiency viral protein
Gag [48, 49], may be performed exclusively by KIF4B. KIF4A in humans is mostly
nuclear and, unless there is a pool of protein shuttling to the cytoplasm, it is unlikely
that cytoplasmic transport is carried out by a predominantly nuclear protein. While
both KIF4 and KID are functional motors [12, 46], it cannot be excluded that some
of the proteins’ functions do not involve motor activity.
9.3 Chromokinesins Shape Mitotic Chromosome Structure
Chromokinesins KIF4A and KID are unique amongst kinesins in their localization
during mitosis (Fig. 9.1). Both proteins associate along the axis of condensed chro-
mosome arms during mitosis [11]. This prominent localization points to a possible
role of these motor proteins in chromosome architecture and chromosome movement
during cell division [13]. In support, loss of KID function compromises chromo-
some compaction during anaphase and delays metaphase to anaphase transition
[24, 41]. Furthermore, KID has been demonstrated to produce polar ejection force
which contributes to chromosome motion [22]. In contrast, KIF4 does not appear to
exhibit any motor properties related to chromosome movement [22], allowing for
the possibility that the protein, despite its motor domain, exerts its function by other
means, possibly through affecting chromosome structure.
KIF4 is essential for maintaining mitotic chromosome integrity and appears to act
in parallel to condensins, the major non-histone proteins of the chromosome
condensation machinery [11, 38]. Loss of KIF4 leads to shortened and widened
chromosomes [11, 38]. Interestingly, KIF4 interacts with a number of condensin
complex components in addition to its binding to DNA and various non-histone
chromatin proteins such as PARP-1 and Asf1 [11, 36, 50]. The unique property
of KIF4 to bind to DNA and other chromatin-associated proteins may serve to mod-
ulate higher-order structure of the chromatin and to shape mitotic chromosomes
by lateral compaction [13, 36, 38]. Although condensins and KIF4 are involved in
chromosome compaction in apparently parallel pathways, KIF4 is one of the very
few chromosome-associated proteins that are required for the integrity of mitotic
chromosomes [11, 38].
Although very little is known about how KIF4, condensin and other chromatin
architectural proteins interact in the interphase nucleus, it is possible that global
higher-order organization, including chromatin loops and chromosome territory
formation, is facilitated by these interactions [13, 38]. KIF4 may act as a key factor
in maintaining the integrity of mitotic chromosome structure because faithful
chromosome segregation, MT spindle integrity, and chromosome attachment to the
mitotic spindle are severely disrupted upon loss of KIF4 [11, 27, 29].
In contrast to KIF4, KID, although similarly localized in the interphase nucleus

and on the axis of condensed chromosome arms during mitosis, does not appear to
play any significant role in chromatin structure or function except during the very
early stages of development [16]. It has been observed in mice lacking KID
that anaphase/telophase chromosome compaction is compromised in oocyte
meiosis II and the first two mitoses, leading to multinucleated cells or embryonic
death [16]. This function of KID is not observed in male meiosis or in somatic cells
[16]. Thus, it seems unlikely that KID has any significant role in either interphase or
mitotic chromatin structure or function.
9.4 KIF4 as a Guardian of Genome Integrity
Consistent with the role of KIF4 in chromosome condensation, mitosis and cytoki-
nesis, cells lacking KIF4 manifest increased genomic instability including chromo-
some breaks, translocations and aneuploidy both in vitro and in vivo [11, 51].
Moreover, deletion of KIF4 from normal human cells generates multinucleated
cells probably due to failure of cytokinesis [11]. Supernumerary centrosomes and
lagging chromosomes are early hallmarks of carcinogenesis. These defects are
observed when KIF4A is depleted from normal human fibroblasts emphasizing a
critical role of KIF4 in maintaining genome stability [11, 51] (Fig. 9.2). While some
of these effects are likely due to defective mitosis, KIF4 also appears to have a more
direct influence on genome integrity since it has been observed that shRNA-
mediated depletion of KIF4A impairs formation of ionizing-radiation induced foci
(IRIF) of Rad51 [37], suggesting reduced DNA repair efficiency. Moreover, DNA
damage repair pathways are triggered upon RNAi-mediated depletion or genetic
deletion of KIF4 from cells as indicated by phosphorylation of H2AX and activation
of kinases ATM, Chk1 and Chk2 [51]. KIF4 has also been reported to interact with
the BRCA2-associated protein BRAF-35 in a yeast two-hybrid assay [26]. Xenografts
of KIF4-null cells rapidly form tumors in nude mice [51] and gene trap deletion or
inactivation of KIF4 causes defects in immunoglobulin Class Switching (CS) during
B cell development in mice [52]. Furthermore, homozygous deletion or inactivation of
KIF4 in mice causes lymphoma and other hematologic abnormalities that are often
characterized by translocations and aneuploidy in vivo (Mazumdar, unpublished, 2011).
9.5 Chromokinesins in Human Cancer
Recent studies are emerging to implicate KID chromokinesin in carcinogenesis,

although no reports demonstrating causality of KID dysfunction in cancer are
available. The E3 Ubiquitin Ligase SIAH-1 induces the degradation of KID, by the
ubiquitin proteasome pathway [53]. Differential expression patterns of SIAH-1
were found in paired normal and breast tumor tissues and when mRNA expression
Loss of KIF
Changes in chromatin structure

Replication Chromosome DNA repair
defects condensation defects
defects
Aberrant replication Aberrant Mitosis DNA damage
Checkpoint Activation Aneuploidy DNA Damage
Abnormal cellular proliferation / Cancer
Fig. 9.2 The role of KIF4 in interphase chromatin structure, mitosis and cancer. KIF4 performs
multiple parallel functions in the cell and its loss produces a number of cellular phenotypes affecting
genome stability. The varied effects of lack of KIF4 are interlinked and the very first step is the
local and global disruption of chromatin architecture, making the genome susceptible to DNA
damage. Once the spatial organization of the chromatin is lost, it leads to cascades of deleterious
events throughout the cell cycle as manifested in aberrant DNA replication, loss of mitotic chromo-
some integrity and defective mitosis and DNA damage that persists throughout the end of cell
division. As a consequence, checkpoint proteins are activated and aneuploidy is generated, all of
which in turn leads to abnormal cell proliferation and are hallmarks of various human cancers
of SIAH-1 and Kid was analyzed by real-time PCR in normal and cancer breast
tissues from the same patient, variations in the KID mRNA levels were similar to
those seen for SIAH-1 mRNAs [54]. Thus it is possible that KID functions in carci-
nogenesis via its interaction with SIAH-1.
Both loss and upregulation of KIF4 have been reported in various human cancers
[7, 51, 55–57]. KIF4 was predicted to be a proto-oncogene as differential expression
was observed in premalignant human trophoblasts [58]. Partial or complete loss of
KIF4 occurs in ~35 % of cell lines in the NCI-60 panel of human cancer cells [51]
and KIF4A has been identified by microarray analysis as one of the most dramati-
cally down-regulated genes in human tumors with metastatic potential [59]. More
recently it has been observed that KIF4 is one of the most down-regulated genes in
cultures of fresh bone marrow samples from patients with multiple myeloma or
acute myeloid leukemia [60]. The differential expression of KIF4 in cancers of vari-
ous tissues is in line with a role of KIF4 homeostasis and function in maintaining
genome stability either by affecting cell division or DNA damage repair capacity.
KIF4 is highly expressed in hematopoietic tissues, fetal liver, spleen, thymus
and adult thymus and bone marrow, pointing to a particular relevance of KIF4 in
the immune system. A number of human hematological malignancies are linked to
X chromosome aneuploidy [61, 62] and deletion of XIST, the X-chromosome

inactivation gene in the blood compartments of mice causes specific hematologic
neoplasm and aberrations in the development of hematopoietic stem cells [63].
These observations are of interest since KIF4 is an X-linked gene, highly expressed
in normal hematopoietic tissues and loss of KIF4 leads to B-cell lymphoma in
knockout mice [52]. It remains to be investigated whether KIF4 is a potential driver
of hematologic cancers in humans. As of present, no mutation in KIF4 has been
reported in tumors, although mutations in X chromosome genes have not been
mapped yet and GWAS for lymphomas excluded the X chromosome [64].
There have been reports of misregulation of KIF4 expression in several other
human cancers. Cross-species comparative genomics analysis of progression from
pre-invasive ductal carcinoma in situ (DCIS) to invasive ductal carcinoma (IDC) in
breast revealed 16 genes with concomitant genomic alterations, including KIF4A,
that may be involved in tumorigenesis and in the processes of invasion and progres-
sion of disease [56]. These observations were further supported by the finding that
elevated levels of Kif4A, Kif15, Kif20A, and Kif23 correlate with that of ANCCA
(AAA nuclear co-regulator cancer-associated) in estrogen receptor (ER)-positive
breast cancer cells with poor relapse-free survival of patients with ER-positive and
tamoxifen-resistant breast cancer [65]. Furthermore, treatment of lung cancer cells
with small interfering RNAs against KIF4A suppressed growth of these cells [57].
It has also been observed that induction of exogenous expression of KIF4A by
stable expression of GFP-KIF4 in gastric carcinoma cells inhibits tumor growth
both in vitro and in vivo [55]. In breast and bladder cancer, glioma and melanoma,
KIF4A is involved in carcinogenesis and has prognostic value [66]. Moreover, KIF4A
has been proposed to be a potential predictor of human oral cancer [67]. KIF4A
expression levels in seven OSCC-derived cells were up-regulated significantly
compared to human normal oral keratinocytes [67]. These observations suggest that
KIF4 gene dosage is a critical factor in preventing various human cancers. Thus, much
like the mitotic kinesin CENPE, KIF4 may act both as an oncogene and as a tumor
suppressor [68, 69].
9.6 Chromokinesins in Higher-Order Chromatin

Organization: Significance for Oncogenesis
Given their prominent nuclear localization during interphase, it is intriguing to

speculate that chromokinesins contribute to the maintenance of mammalian nuclear
architecture [36]. In support of such a role, KIF4 interacts with DNA and several
chromatin-associated proteins including chromatin assembly factors Asf1, RbAp 48
and PARP-1 during interphase [36]. Furthermore, morphologically discernible
heterochromatin blocks and peripheral heterochromatin are conspicuously absent
from cells lacking KIF4 and global chromatin structure assumes a largely “open”
configuration [36]. Likely as a functional consequence of these defects, normally
silenced satellite repeat elements are highly expressed and global gene expression
is misregulated upon KIF4 depletion [36]. Furthermore, DNA replication is sub-
stantially slowed down in the absence of KIF4 and several histone chaperones,
chromatin remodeling and assembly proteins with which KIF4 interacts, such as
Asf1 and RbAp48, fail to be recruited to replicated chromosomes [36]. Histone
chaperones have also been shown to regulate the localization of important chroma-
tin modifying enzymes. These observations suggest that chromokinesin KIF4 may
be a critical factor for nuclear and chromosomal functions during interphase.
Chromatin structure and chromatin remodeling activities are critical for regula-
tion of global gene expression and function [70] and aberrations in architectural
proteins responsible for maintaining chromatin organization are critically involved
in tumorigenesis [71]. Defects in global genome organization, such as the ones
brought about by lack of KIF4 function, may predispose genomes for tumorigenesis
by affecting faithful replication and mitosis. In this context it is important to remem-
ber that chromosome assembly during S-phase and mitotic chromosome segrega-
tion are not isolated events, but are linked. It is well established that faithful
chromosome segregation in mitosis and genome stability throughout the cell cycle
depends on specific chromatin structures that are assembled during DNA replica-
tion or S-phase [72]. KIF4 may play a central role in this interplay since, in addition
to its demonstrated role in mitosis, it interacts with a number of replication
assembly and chromatin remodeling complex components such as Asf1, RbAp48,
PARP-1, condensins and DNA Methyl Transferase 3B (DNMT 3B) and appears to
have critical roles in S-phase assembly [11, 36, 73]. It is intriguing that condensin
II, with which KIF4 interacts, is required for initiation of chromatin reorganization
and resolution of duplicated sister chromatids during S-phase [74]. The observation
that loss of KIF4 results in premature chromosome condensation [11] is consistent
with the fact that S-phase assembly is affected and progression through S-phase is
delayed due to loss of KIF4 [36]. It is tantalizing to speculate that KIF4, in collabo-
ration with condensin and other chromatin protein complexes, plays a fundamental
role not only in shaping the mitotic chromosome structure, but also in global genome
organization during interphase, and particularly in S-phase chromosome organiza-
tion [36]. It is possible through this dual role in interphase and mitosis that KIF4
leads to genome instability, and possibly contributes to tumorigenesis.
The observed emergence of spontaneous B cell lymphoma and other hemato-
logic malignancies in KIF4-null mice is possibly due to its role in chromatin struc-
ture [52]. In humans, the majority of lymphoid malignancies have their origins in B
cells and it has been speculated that the plasticity of chromatin, which is needed to
carry out the complex process of VDJ- and CS-recombination, makes these cells
more susceptible to genomic instability [75–77]. It is plausible that altered chroma-
tin plasticity and mis-regulation of gene expression in lymphocytes in the absence
of KIF4 similarly contribute to changes in B-cell fate and affects the immune system
adversely leading to lymphomagenesis.
9.7 Chromokinesins as Targets for Cancer Therapy
One of the hallmarks of most types of cancer is the mis-regulation of mitosis and
subsequent DNA damage or aneuploidy. The microtubule component tubulin and
the mitotic spindle apparatus proteins that are involved in cell division are thus
used as targets for cancer chemotherapy, although the specificity of the drugs are
often called into question because of toxicity and non-specific side effects [7, 78].
While intensive research is ongoing to discover novel targets and drugs with high
specificity, mitotic kinesins, including chromokinesins, have emerged as attractive
targets because of their important roles in both maintaining mitotic spindle dynam-
ics well as for their function as components of mitotic chromosome structures such
as kinetochores [7]. Although the complete mechanistic details of the function of
chromokinesins at the molecular level have not been deciphered yet, the novel roles
of these unique kinesins in maintaining higher-order chromosome structure and
genomic stability [11, 36, 38, 51] may make chromokinesins potential targets for
chemotherapy. Several strategies can be envisioned.
KIF4’s function as a chromatin platform that mediates the assembly of multiple
other chromatin protein complexes at various stages of the cell cycle, including
during DNA replication and mitosis and its potential involvement in local chromatin
looping, suggest that KIF4 acts as a guardian of genome stability by potentially
preventing chromosomes from random breaks, DNA damage and translocations
that are deleterious. It is thus attractive to speculate that synthetic tethers that mimic
KIF4 function could serve as potential drugs in tumors that lack KIF4. Although
abundant expression of KIF4 in normal interphase cells indicate that it has a direct
role in modulating global chromatin organization [36], further studies are needed to
resolve whether KIF4 acts as a chromatin structural scaffold or as an adaptor for
other DNA-protein complexes before such therapeutic strategies are adopted.
Another potential therapeutic strategy is based on the ability of KIF4 to recruit at
least two chromatin modifying enzymes, DNA Methyl transferases (DNMT3B) and
PARP-1 [36, 73]. Importantly, PARP-1 enzymatic activity, but not protein levels,
increases in the absence of KIF4 [36, 50]. Specific inhibitors of these two enzymes
could be interesting agents to test in KIF4-overexpressing cancers because both
methylation and PAR-lyation alters chromatin organization at the nucleosomal level.
While these strategies attempt to exploit KIF4 to ameliorate cellular function, the
opposite strategy of using KIF4 as a cell-death inducer can also be anticipated. It is
tantalizing to speculate that KIF4, because of its potential to bind to DNA and a
number of chromatin proteins to maintain higher-order chromatin structure, especially
in its ability to shape the mitotic chromosome, could well be a chemotherapeutic
target akin to DNA supercoil-unwinding topoisomerase 1 (Top1) [79]. It would be
interesting to identify drugs that disrupt the functional association of KIF4 with
DNA and other chromatin proteins such as condensins or PARP-1 that are essential
for its mitotic function. Induction of mitotic defects in rapidly dividing tumor cells
may lead to their elimination. For these approaches, it will be important to determine
how precisely KIF4 binds to specific regions of DNA or whether it requires a
specific DNA structure for binding that could topologically orient KIF4 to bind to
various chromatin proteins.
To evaluate the suitability of chemotherapeutics, a systematic genetic screening
approach may be necessary to determine whether KIF4 is a tumor suppressor or
driver oncoprotein in various human cancers.
9.8 Conclusions
Chromokinesins are essential mitotic motors that perform important functions in

maintaining chromosome structure, spindle dynamics and in the movement of
chromosomes during mitosis. Moreover, in the final stages of cell division, chromo-
kinesins play a significant role in central spindle formation and cytokinesis. In addi-
tion to their mitotic roles, chromokinesins also contribute to interphase chromatin
and chromosome structure. These diverse roles of chromokinesin proteins during
the cell cycle have raised the possibility of involvement of the proteins in genomic
instability and carcinogenesis when deregulated in various human tissues. Not sur-
prisingly a spate of recent reports implicated chromokinesins in human cancers. The
analysis of gene expression patterns in various tumor tissues is a significant first
step in implicating chromokinesins in cancer. It is now important to carefully
analyze and understand the mechanisms that underlie chromokinesin function
in vivo in each step in the various nuclear and mitotic processes in the context of
various tissues. It is obvious that the protein and mRNA levels of chromokinesins
have differential effects on producing tumors in various tissues, possibly due to
various, tissue-specific functions. The emerging role of chromokinesins in cancer
also raises the tantalizing possibility of designing chromokinesins-based therapeu-
tic strategies.
References
kinesin, involved in microtubule-based motility. Cell 42(1):39–50
2. Brady ST (1985) A novel brain ATPase with properties expected for the fast axonal transport
motor. Nature 317(6032):73–75
3. Hirokawa N, Noda Y, Okada Y (1998) Kinesin and dynein superfamily proteins in organelle
transport and cell division. Curr Opin Cell Biol 10(1):60–73
ture and function. Trends Cell Biol 15(9):467–476
5. Hirokawa N et al (2009) Kinesin superfamily motor proteins and intracellular transport. Nat
Rev Mol Cell Biol 10(10):682–696
molecular assays. Semin Cell Dev Biol 21(3):260–268
7. Rath O, Kozielski F (2012) Kinesins and cancer. Nat Rev Cancer 12(8):527–539
8. Funabiki H, Murray AW (2000) The Xenopus chromokinesin Xkid is essential for metaphase
chromosome alignment and must be degraded to allow anaphase chromosome movement. Cell
102(4):411–424
9. Tokai N et al (1996) Kid, a novel kinesin-like DNA binding protein, is localized to chromo-
somes and the mitotic spindle. EMBO J 15(3):457–467
10. Wang SZ, Adler R (1995) Chromokinesin: a DNA-binding, kinesin-like nuclear protein. J Cell
Biol 128(5):761–768
11. Mazumdar M, Sundareshan S, Misteli T (2004) Human chromokinesin KIF4A functions in
chromosome condensation and segregation. J Cell Biol 166(5):613–620
12. Levesque AA, Compton DA (2001) The chromokinesin Kid is necessary for chromosome
arm orientation and oscillation, but not congression, on mitotic spindles. J Cell Biol
154(6):1135–1146
13. Mazumdar M, Misteli T (2005) Chromokinesins: multitalented players in mitosis. Trends Cell
Biol 15(7):349–355
14. Antonio C et al (2000) Xkid, a chromokinesin required for chromosome alignment on the
metaphase plate. Cell 102(4):425–435
15. Shiroguchi K et al (2003) The second microtubule-binding site of monomeric kid enhances the
microtubule affinity. J Biol Chem 278(25):22460–22465
16. Ohsugi M et al (2003) Cdc2-mediated phosphorylation of Kid controls its distribution to
spindle and chromosomes. EMBO J 22(9):2091–2103
17. Cassimeris L, Rieder CL, Salmon ED (1994) Microtubule assembly and kinetochore
directional instability in vertebrate monopolar spindles: implications for the mechanism of
chromosome congression. J Cell Sci 107(Pt 1):285–297
18. Rieder CL et al (1986) Oscillatory movements of monooriented chromosomes and their posi-
tion relative to the spindle pole result from the ejection properties of the aster and half-spindle.
J Cell Biol 103(2):581–591
19. Ke K, Cheng J, Hunt AJ (2009) The distribution of polar ejection forces determines the ampli-
tude of chromosome directional instability. Curr Biol 19(10):807–815
20. Bieling P, Kronja I, Surrey T (2010) Microtubule motility on reconstituted meiotic chromatin.
Curr Biol 20(8):763–769
21. Rieder CL, Salmon ED (1994) Motile kinetochores and polar ejection forces dictate chromo-
some position on the vertebrate mitotic spindle. J Cell Biol 124(3):223–233
22. Brouhard GJ, Hunt AJ (2005) Microtubule movements on the arms of mitotic chromosomes:
polar ejection forces quantified in vitro. Proc Natl Acad Sci U S A 102(39):13903–13908
23. Stumpff J et al (2012) Kif18A and chromokinesins confine centromere movements via
microtubule growth suppression and spatial control of kinetochore tension. Dev Cell
22(5):1017–1029
24. Santamaria A et al (2008) The spindle protein CHICA mediates localization of the chromoki-
nesin Kid to the mitotic spindle. Curr Biol 18(10):723–729
25. Kurasawa Y et al (2004) Essential roles of KIF4 and its binding partner PRC1 in organized
central spindle midzone formation. EMBO J 23(16):3237–3248
26. Lee YM, Kim W (2004) Kinesin superfamily protein member 4 (KIF4) is localized to midzone
and midbody in dividing cells. Exp Mol Med 36(1):93–97
27. Zhu C, Jiang W (2005) Cell cycle-dependent translocation of PRC1 on the spindle by Kif4 is
essential for midzone formation and cytokinesis. Proc Natl Acad Sci U S A 102(2):343–348
28. Hu B et al (2011) ATP hydrolysis is required for relocating cohesin from sites occupied by its
Scc2/4 loading complex. Curr Biol 21(1):12–24
29. Zhu C et al (2005) Functional analysis of human microtubule-based motor proteins, the
kinesins and dyneins, in mitosis/cytokinesis using RNA interference. Mol Biol Cell
16(7):3187–3199
30. Powers J et al (2004) Loss of KLP-19 polar ejection force causes misorientation and misseg-
regation of holocentric chromosomes. J Cell Biol 166(7):991–1001
31. Kwon M et al (2004) The chromokinesin, KLP3A, dives mitotic spindle pole separation during
prometaphase and anaphase and facilitates chromatid motility. Mol Biol Cell 15(1):219–233
32. Williams BC et al (1995) The Drosophila kinesin-like protein KLP3A is a midbody component
required for central spindle assembly and initiation of cytokinesis. J Cell Biol 129(3):709–723
33. Vernos I et al (1995) Xklp1, a chromosomal Xenopus kinesin-like protein essential for spindle
organization and chromosome positioning. Cell 81(1):117–127
34. Bringmann H et al (2004) A kinesin-like motor inhibits microtubule dynamic instability.
Science 303(5663):1519–1522
35. Castoldi M, Vernos I (2006) Chromokinesin Xklp1 contributes to the regulation of microtubule
density and organization during spindle assembly. Mol Biol Cell 17(3):1451–1460
36. Mazumdar M, Sung MH, Misteli T (2011) Chromatin maintenance by a molecular motor
protein. Nucleus 2(6):591–600
37. Wu G et al (2008) A novel role of the chromokinesin Kif4A in DNA damage response. Cell
Cycle 7(13):2013–2020
38. Samejima K et al (2012) Mitotic chromosomes are compacted laterally by KIF4 and condensin
and axially by topoisomerase IIalpha. J Cell Biol 199(5):755–770
39. Yajima J et al (2003) The human chromokinesin Kid is a plus end-directed microtubule-based
motor. EMBO J 22(5):1067–1074
40. Levesque AA et al (2003) A functional relationship between NuMA and kid is involved in both
spindle organization and chromosome alignment in vertebrate cells. Mol Biol Cell
14(9):3541–3552
41. Tokai-Nishizumi N et al (2005) The chromokinesin Kid is required for maintenance of proper
metaphase spindle size. Mol Biol Cell 16(11):5455–5463
42. Magidson V et al (2011) The spatial arrangement of chromosomes during prometaphase facili-
tates spindle assembly. Cell 146(4):555–567
43. Oshimori N, Ohsugi M, Yamamoto T (2006) The Plk1 target Kizuna stabilizes mitotic centro-
somes to ensure spindle bipolarity. Nat Cell Biol 8(10):1095–1101
44. Logarinho E et al (2012) CLASPs prevent irreversible multipolarity by ensuring spindle-pole
resistance to traction forces during chromosome alignment. Nat Cell Biol 14(3):295–303
45. Ha MJ et al (2000) Assignment of the kinesin family member 4 genes (KIF4A and KIF4B) to
human chromosome bands Xq13.1 and 5q33.1 by in situ hybridization. Cytogenet Cell Genet
88(1–2):41–42
46. Sekine Y et al (1994) A novel microtubule-based motor protein (KIF4) for organelle
transports, whose expression is regulated developmentally. J Cell Biol 127(1):187–201
47. Peretti D et al (2000) Evidence for the involvement of KIF4 in the anterograde transport of
L1-containing vesicles. J Cell Biol 149(1):141–152
48. Kim W et al (1998) Binding of murine leukemia virus Gag polyproteins to KIF4, a microtubule-
based motor protein. J Virol 72(8):6898–6901
49. Martinez NW et al (2008) Kinesin KIF4 regulates intracellular trafficking and stability of the
human immunodeficiency virus type 1 Gag polyprotein. J Virol 82(20):9937–9950
50. Hirokawa N (2006) mRNA transport in dendrites: RNA granules, motors, and tracks. J
Neurosci 26(27):7139–7142
16(15):1559–1564
52. Mazumdar M et al. (2011) Chromokinesin KIF4 facilitates chromatin organization critical for
immunoglobulin class switch recombination and prevents lymphoma formation. Cancer Res
71(8): Supplement 1:934
53. Germani A et al (2000) SIAH-1 interacts with alpha-tubulin and degrades the kinesin Kid by
the proteasome pathway during mitosis. Oncogene 19(52):5997–6006
54. Bruzzoni-Giovanelli H et al (2010) Distinct expression patterns of the E3 ligase SIAH-1 and
its partner Kid/KIF22 in normal tissues and in the breast tumoral processes. J Exp Clin Cancer
Res 29:10
55. Gao J et al (2011) Overexpression of chromokinesin KIF4 inhibits proliferation of human
gastric carcinoma cells both in vitro and in vivo. Tumour Biol 32(1):53–61
56. Colak D et al (2013) Age-specific gene expression signatures for breast tumors and
cross-species conserved potential cancer progression markers in young women. PLoS One
8(5):e63204
57. Taniwaki M et al (2007) Activation of KIF4A as a prognostic biomarker and therapeutic target
for lung cancer. Clin Cancer Res 13(22 Pt 1):6624–6631
58. Lee BP et al (2001) Differential gene expression in premalignant human trophoblast: role of
IGFBP-5. Int J Cancer 94(5):674–684
59. Weil RJ et al (2005) Breast cancer metastasis to the central nervous system. Am J Pathol
167(4):913–920
60. Cohen Y et al (2014) The proliferation arrest of primary tumor cells out-of-niche is associated with
widespread downregulation of mitotic and transcriptional genes. Hematology 19(5):286–292
61. Heinonen K et al (1999) Acquired X-chromosome aneuploidy in children with acute lympho-
blastic leukemia. Med Pediatr Oncol 32(5):360–365
62. Pageau GJ et al (2007) The disappearing Barr body in breast and ovarian cancers. Nat Rev
Cancer 7(8):628–633
63. Yildirim E et al (2013) Xist RNA is a potent suppressor of hematologic cancer in mice. Cell
152(4):727–742
64. Conde L et al (2013) Integrating GWAS and expression data for functional characterization of
disease-associated SNPs: an application to follicular lymphoma. Am J Hum Genet
92(1):126–130
65. Zou JX et al (2014) Kinesin family deregulation coordinated by bromodomain protein ANCCA
and histone methyltransferase MLL for breast cancer cell growth, survival, and tamoxifen
resistance. Mol Cancer Res 12(4):539–549
66. Rouam S, Moreau T, Broet P (2010) Identifying common prognostic factors in genomic cancer
studies: a novel index for censored outcomes. BMC Bioinf 11:150
67. Minakawa Y et al (2013) Kinesin family member 4A: a potential predictor for progression of
human oral cancer. PLoS One 8(12):e85951
68. Weaver BA, Cleveland DW (2007) Aneuploidy: instigator and inhibitor of tumorigenesis.
Cancer Res 67(21):10103–10105
69. Weaver BA et al (2007) Aneuploidy acts both oncogenically and as a tumor suppressor. Cancer
Cell 11(1):25–36
70. Misteli T (2013) The cell biology of genomes: bringing the double helix to life. Cell
152(6):1209–1212
71. Scaffidi P, Misteli T (2010) Cancer epigenetics: from disruption of differentiation programs to
the emergence of cancer stem cells. Cold Spring Harb Symp Quant Biol 75:251–258
72. Groth A et al (2007) Chromatin challenges during DNA replication and repair. Cell
128(4):721–733
73. Robertson AK et al (2004) Effects of chromatin structure on the enzymatic and DNA binding
functions of DNA methyltransferases DNMT1 and Dnmt3a in vitro. Biochem Biophys Res
Commun 322(1):110–118
74. Ono T, Yamashita D, Hirano T (2013) Condensin II initiates sister chromatid resolution during
S phase. J Cell Biol 200(4):429–441
75. Hernandez-Munain C, McMurry MT, Krangel MS (1999) Regulation of chromatin accessibil-
ity for V(D)J recombination. Cold Spring Harb Symp Quant Biol 64:183–189
76. Vanasse GJ, Concannon P, Willerford DM (1999) Regulated genomic instability and neoplasia
in the lymphoid lineage. Blood 94(12):3997–4010
77. Zhang Y et al (2010) The role of mechanistic factors in promoting chromosomal translocations
found in lymphoid and other cancers. Adv Immunol 106:93–133
78. Domenech E, Malumbres M (2013) Mitosis-targeting therapies: a troubleshooting guide. Curr
Opin Pharmacol 13(4):519–528
79. Pommier Y, Marchand C (2012) Interfacial inhibitors: targeting macromolecular complexes.
Nat Rev Drug Discov 11(1):25–36
Chapter 10
Kif14: A Clinically Relevant Kinesin
and Potential Target for Cancer Therapy
Brigitte L. Thériault and Timothy W. Corson
10.1 KIF14 Gene and Protein
KIF14 is a mitotic kinesin and microtubule-associated molecular motor that plays

an essential role in the last stages of cytokinesis. The gene is located on chromo-
some 1q32.1, spanning 68.5 kbp on the minus strand. To date, only one 6,586 bp
transcript is known, comprised of 30 coding exons, and no pseudogenes have yet
been identified [1]. The KIF14 mRNA encodes a 1648 amino acid (aa), 185 kDa
protein containing two major effector domains (Fig. 10.1). The first is the highly
conserved 352 aa kinesin motor domain containing both an ATP-binding site
involved in microtubule-dependent ATPase activity and a microtubule binding site
involved in ATP-dependent protein transport. The second is a 94 aa forkhead-
associated (FHA) domain (aa 803–900) which has similarity to the SMAD Mad
Homology 2 (MH2) domain, involved in mediating protein-phosphoprotein interac-
tions, although no interactions through this domain have yet been documented for
KIF14 [2]. In addition to the highly conserved N-type kinesin neck region adjacent
to the motor domain, KIF14 also contains four C-terminal regions predicted to form
coiled-coil structures [2].
Phosphorylation sites have been identified in high-throughput studies on Ser-12,
Tyr-196, Thr-240, Ser-242, Ser-378, Ser-384, Ser-670, Ser-1200, Ser-1292, Ser-1631,
B.L. Thériault
Campbell Family Cancer Research Institute, Ontario Cancer Institute,
Princess Margaret Cancer Centre, Toronto, ON, Canada
T.W. Corson (*)
Eugene and Marilyn Glick Eye Institute, Departments of Ophthalmology,
Biochemistry and Molecular Biology, and Pharmacology
and Toxicology, and Simon Cancer Center, Indiana University School
of Medicine, Indianapolis, IN, USA
e-mail: tcorson@iupui.edu

150 B.L. Thériault and T.W. Corson
Fig. 10.1 Schematic representation of the KIF14 protein (not to scale). KIF14 contains two major
effector domains. The first is a highly conserved 352 amino acid (aa) kinesin motor domain con-
taining both an ATP-binding site and a microtubule binding site. The second is a forkhead-
associated (FHA) domain (aa 803–900). In addition to the highly conserved neck region (N)
adjacent to the motor domain, KIF14 also contains four other C-terminal regions predicted to form
coiled-coil structures (1–4). High-throughput studies have identified phosphorylation (P) sites on
Ser-12, Tyr-196, Thr-240, Ser-242, Ser-378, Ser-384, Ser-670, Ser-1200, Ser-1292, Ser-1631, Ser-
1636 and Thr-1641, and a ubiquitination site (U) on Lys-275. The N-terminal extension (354 aa) is
involved in the binding of PRC1 (protein-regulating cytokinesis 1), and Citron kinase interacts
with the C-terminal stalk and tail of KIF14. Supervillin associates directly with the distal C-terminal
tail of KIF14 (aa 1522–1648), while Radil interacts with the last 4 amino acids of KIF14 (aa
1645–1648). Inset: Fluorescence in situ hybridization reveals increased copy number of the KIF14
genomic locus (red) over ploidy (chromosome 1 centromere [Cen 1], blue) in retinoblastoma com-
pared with normal retina (Reproduced with permission from Ref. [31])
Ser-1636 and Thr-1641 [3–6], and a ubiquitination site identified on Lys-275 [5, 6].
The kinesin motor and FHA domains are flanked by a 354 aa N-terminal extension,
and a 758 aa C-terminal stalk and tail region.
Despite its unique, large N-terminal extension, KIF14 has been considered a
member of the N-3 family of kinesins [7]. These kinesins all have different tail
domains and expression patterns (although several share FHA domains) [7]. There
are KIF14 orthologs in several model species. The closest ortholog is chimpanzee
(sharing 99 % nucleotide (n) and amino acid (a) sequence identity), followed by dog
(87 % n, 82 % a), mouse and rat (77 % n, 73 % a; and 76 % n, 66 % a; respectively),
zebrafish (59 % n, 57 % a), and fruit fly (Drosophila melanogaster, 45 % n, 40 % a)
(Homologene, NCBI) [8]. The best-studied non-mammalian KIF14 ortholog
(see below) is the Drosophila gene nebbish/tiovivo, encoding Klp38B (kinesin-like
protein 38B) [9].
The biochemistry and crystal structure of the murine Kif14 motor domain reveal
some unique features [10]. These include sequence divergence in loops L8a, L10,
L11, and L12 compared to other N-3 kinesins, and a small difference in orientation
of Kif14 on the microtubule, as observed by cryo-electron microscopy. Kif14 has
a high basal ATPase activity and very high microtubule affinity that is largely
insensitive to the nucleotide state of the active site. However, it has relatively slow,
plus-end directed movement along microtubules. In the solved ADP-bound structure,
10 Kif14: A Clinically Relevant Kinesin and Potential Target for Cancer Therapy 151
Kif14’s ATP-binding pocket is remarkably open relative to other kinesins, due to a

“twist” of the central β sheet that is reminiscent of the “rigor-like” conformation
seen in myosins, but not previously documented in other kinesin crystal structures
[11]. In keeping with these biochemical properties, Kif14 can stabilize microtubules
against cold-induced depolymerization, suggestive of a bundling function.
10.2 KIF14 Function and Regulation
10.2.1 Role of KIF14 in Cytokinesis
KIF14 was first cloned from a human myeloid cell line, KG-1 [4]. As the kinesin
motor domain is highly conserved among species, the sequence for the Kif14 motor
domain was subsequently identified from a screen of the mouse genome [12]. Kif14
is highly expressed in mouse liver and brain, suggesting important roles in these
tissues [12]. As mentioned, human KIF14 shares approximately 40 % homology
with the Drosophila kinesin-like protein Klp38B [7]. Mutation of Klp38B has been
shown to lead to cytokinesis failure [9, 13, 14], and to chromosomal segregation
defects during cell division [15].
The first functional analysis of KIF14 was conducted by Zhu et al., who employed
an RNA interference-based strategy to assess function of various kinesins during
HeLa cell division [16]. In this study, depletion of KIF14 had no effect on the
formation of bipolar mitotic spindles, but resulted in abnormal chromosome
congression and alignment during mitosis, resulting in delayed metaphase to anaphase
transition [16]. Shortly thereafter, two studies more extensively defined the role of
KIF14 in cell division. Gruneberg et al. identified KIF14 through binding of its
N-terminal extension to PRC1 (protein regulating cytokinesis 1), which is a central
regulator of mitotic spindle dynamics within the last stages of cytokinesis [17]
(Fig. 10.2a). In HeLa cells, they found that KIF14 was responsible for the proper
localization of citron kinase (CIT) to the cleavage furrow, playing an essential role
in formation of the furrow. The interaction between CIT and KIF14 occurs via the
C-terminal stalk and tail region of KIF14 [17]. These results indicate that KIF14 is
involved in both the mitotic spindle and the process of cytokinesis. A second study
provided a thorough investigation of KIF14 expression and cellular localization
during cell division [18]: cytoplasmic during interphase, but at the midbody during
later stages of mitosis [18].
Crucially, this latter work first confirmed that KIF14 contained an N-type motor
that exhibited microtubule-dependent ATPase activity [18]. On siRNA treatment,
different mitotic phenotypes were seen depending on the level of KIF14 knockdown.
Strong silencing of KIF14 induced cytokinesis failure due to failure of midbody
cleavage, resulting in endoreduplication and the formation of multinucleated cells,
while less efficacious siRNAs caused binucleation and apoptosis following mitotic
exit [18]. In contrast to Zhu et al., these authors could not confirm the role of
Fig. 10.2 Examples of KIF14 function and interactions. (a) KIF14 (green) localizes to the mid-
body during cytokinesis and partially overlaps with PRC1 (red). Bar, 10 µM. (b) Knockout (KO)
of Kif14 causes a reduction in brain size compared to wild-type (WT), notably of the cerebral and
cerebellar cortices and olfactory bulb. Bar, 5 mm. (c) EGFP-KIF14 recruits mCherry-Radil to
microtubules. (d) EGFP-KIF14 colocalizes with mRFP-supervillin at intracellular punctae
(Figures reprinted with permission from: (a) Ref. [17], (b) Ref. [23], (c) Ref. [56], (d) Ref. [19])
KIF14 in chromosome congression and alignment, as delayed metaphase to anaphase

transition was not observed.
Subsequent studies have elaborated and defined the role of KIF14 in cytokinesis.
The Kif14 crystal structure and biochemistry described above suggest a role for
Kif14 in microtubule bundling [10]; how this may be important in vivo remains
unknown at this time. The membrane protein supervillin associates with KIF14 during
cytokinesis and contributes to the establishment or maintenance of the cytokinetic
cleavage furrow, possibly by transporting other cytokinesis effectors from the
membrane compartment [19] (Fig. 10.2d). However, the modulation of cleavage
dynamics by supervillin or KIF14 remains unclear. Building on previous work [17],
CIT was shown to interact with KIF14 via the second coiled-coil domain of KIF14
(between amino acids 901 and 1233; Fig. 10.1) to mediate localization of KIF14 at
the central spindle and with midbody microtubules [20]. Furthermore, the interaction
between KIF14 and CIT is essential in regulating the transfer of CIT to the midbody,
effectively regulating the function of CIT in a spatiotemporal manner. Thus KIF14
may contribute to the proper accumulation of CIT and other cytokinesis proteins
(PRC1, Rho GTPases) required for abscission [20]. However, whether any of these
proteins are true cargos for KIF14 has not been demonstrated.
Similar interactions were documented in Drosophila, where Sticky (CIT ortho-
log) interacts directly with Nebbish (KIF14 ortholog) and Pavarotti (MKLP1
(mitotic kinesin-like protein 1) ortholog), and all of these proteins interact with
Fascetto (PRC1 ortholog) [21]. Sticky recruits nebbish to the cleavage furrow, and
both proteins are required for proper localization and function of Pavarotti and
Fascetto, indicating the involvement of these genes in a highly conserved process [21].
Study of meiosis in Xenopus oocytes revealed that NabKin, a paralog of KIF14,
is involved in meiotic cytokinesis and formation of the polar body by direct association
with both F-actin bundles and microtubules through a specific actin binding domain,
located in the N-terminal extension [22]. Interestingly, in vitro assays confirmed
that human KIF14 also bound to actin bundles through its N-terminal extension,
revealing a previously unknown interaction for KIF14 [22]. Indeed, KIF14 can
localize to both the mitotic spindle via cytoplasmic microtubules and the cleavage
furrow via the F-actin contractile ring of human cultured somatic cells [22]. In contrast
to KIF14, a nuclear import signal was found on the N-terminal extension of NabKin,
where Importin-β binds to and imports NabKin to actin bundles, followed by
disengagement from Importin-β and association with RanGTP, which strongly
promoted NabKin-actin interaction. These results further solidify the crucial role
of KIF14 during cell division, but also expand the possible cellular processes within
which KIF14 may participate.
10.2.2 Kif14 Knockouts in Mouse and Human
One cellular process for which Kif14 is essential was revealed in a report of a novel,
spontaneous mouse mutant, laggard (lag), which exhibited cerebellar ataxia and
was traced back to a splice mutation in the mouse Kif14 gene [23]. This phenotype
was confirmed through the generation of a Kif14 knockout mouse (Fig. 10.2b)
and rescued with a Kif14 transgenic. Lag mice exhibited a small head, including
significantly reduced brain size, tremors, ataxic gait, and uncoordinated locomotion
during the early postnatal weeks; all mice died by P21 [23]. This spontaneous muta-
tion preferentially affected the brain, specifically the cerebral and cerebellar cortices,
and the olfactory bulb [24]. The mutation was found to be a G → A substitution at
the 3′ acceptor site in exon 5 of the Kif14 gene, causing exon skipping and the
generation of truncated and unstable Kif14 protein products [23].
In lag and Kif14 knockout mice, similar phenotypes were seen (Fig. 10.2b),
including severe hypomyelination defects in the central nervous system and dis-
rupted cytoarchitecture in the olfactory bulb, cerebral and cerebellar cortices, as
well as unmyelinated axons within the optic nerve. This latter effect may be due to
a defect in astroglial maturation. Interestingly, “late-born” neurons were more
sensitive to Kif14 loss than earlier-developing cells, suggesting a window of devel-
opmental necessity. This and other phenotypes were similar to those seen in citron
kinase knockout mice [25]. Consistent with in vitro knockdown studies of Kif14
(see below), decreased proliferation and widespread apoptosis were also seen in
Kif14 knockout or lag mice, contributing to reduced brain size, although some binu-
cleated cells in the olfactory bulb survived [23, 24]. These results indicate that Kif14
participates in fundamental neuronal developmental pathways, likely through its
function in cell division. However, as severe myelination defects were observed,
Kif14 may also play an important role in myelination, adding to the complexity of
possible Kif14 functions within the normal cell.
Some of the features of the Kif14 knockout mouse were also observed in a human
fetal congenital anomaly syndrome [26]. Two non-viable fetuses in a family showed
intrauterine growth restriction, microcephaly, renal cystic dysplasia/agenesis, and
significant genitourinary and brain malformations. These latter malformations,
including cerebral and cerebellar hypoplasia and a flat head, were very similar to
those in lag mice [23]. Whole-exome sequencing showed the familial syndrome to
be caused by autosomal recessive, truncating mutations in KIF14 [26]. In keeping
with this, histology revealed a high proportion of binucleated cells in these fetuses.
The collected clinical features of this syndrome are consistent with a ciliopathy
(a disease of the primary cilium). Since other kinesins and cytokinetic machinery
are involved in ciliogenesis and function, it seems probable that KIF14 may also
have an as-yet-uncharacterized role in this intriguing organelle.
10.2.3 Regulation of KIF14 Expression
KIF14 is expressed at low levels in normal adult tissues and at higher levels in
placenta and fetal tissues. In normal tissues, the highest expression is in fetal
thymus and liver [27]. KIF14 expression varies with the cell cycle, with the highest
expression occurring at the G2-M transition [18]. In many cancers, KIF14 shows
tumor-specific overexpression that in many cases correlates with stage, aggressive-
ness and poor patient outcomes (see later). Beyond the germline mutations described
above, several recurrent somatic mutations in KIF14 have been identified in meta-
static melanomas, breast and ovarian cancers [28–30], however none of them have
yet been associated with function.
Genomic gain is one mechanism of KIF14 upregulation (see below). Low-level
genomic gain of KIF14 is seen in retinoblastoma [27] (Fig. 10.1), hepatocellular
carcinoma (HCC) [31, 32], breast [31] and ovarian [33] cancers, and associates with
high mRNA overexpression. However many tumors analyzed in these studies show
high KIF14 expression without genomic gain, suggesting other mechanisms regu-
lating KIF14 expression. A recent expression analysis in gliomas pointed to a strong
correlation between KIF14 mRNA and protein expression, suggesting that the
upregulation of KIF14 may be primarily caused by transcriptional activation [34].
Transcriptional regulation of the KIF14 gene was first identified using genome-
wide analysis of transcription factor occupancy, where the KIF14 promoter was
bound by p130/E2F4 under growth arrest conditions [35]. But functional studies of
transcription factor regulation of KIF14 have been lacking until recently, when we
showed that KIF14 overexpression in ovarian cancers is regulated by multiple
genomic, transcriptional and epigenetic mechanisms [36]. Deletion analysis of
the human KIF14 promoter revealed a cis-regulatory region containing binding sites
for Sp1, HSF1 and YY1. siRNA-mediated knockdown of these transcription factors
in ovarian cancer cell lines demonstrated endogenous regulation of KIF14 by Sp1
and YY1, but not HSF1. Chromatin immunoprecipitation confirmed enrichment for
both Sp1 and YY1 binding to the endogenous promoter. We saw a strong correlation
in primary serous ovarian cancers between both Sp1 and YY1 expression and KIF14
expression, suggesting that these factors are important in maintaining KIF14 over-
expression. A CpG island in the KIF14 promoter was differentially methylated in
tumors with high KIF14 expression, suggesting that methylation could also be at
play in KIF14’s tumor-specific regulation. Finally, miRNA expression analysis of
these same tumors determined that miR-93, miR-144 and miR-382 had significantly
lower levels of expression than in normal tissues, and shared an inverse correlation
with KIF14 expression. In particular, miR-382 was found to regulate KIF14 levels
in vitro. Although in need of further validation, these results offer multiple mecha-
nisms of KIF14 expression regulation, and offer potential avenues for the development
of therapeutic interventions for the reduction of KIF14 levels in cancer cells, such
as reducing SP1 or YY1 expression, or increasing miR-382 levels.
10.3 Identification of KIF14 as a Potential Oncogene
Beyond its basic biology, there is considerable interest in KIF14 as a possible

oncogene (Fig. 10.3). The KIF14 locus at chromosome 1q32.1 is in a common
region of genomic gain in multiple cancers [27, 31–33, 37, 38]. High resolution
genomic characterization and subsequent gene expression validation identified
KIF14 as a potential oncogene within this region of gain [27].
10.3.1 Genomic Copy Number Studies
Our group first identified KIF14 as a potential oncogene by studying genomic

alterations in the prototypic genetic cancer retinoblastoma [27]. We identified in
over 70 % of retinoblastoma tumors (39/55) a hotspot region of gain (up to 5 copies)
at chromosome 1q32.1 (3 Mbp). This region contained 14 named or predicted genes,
including KIF14. KIF14 was the only gene within this hotspot of gain to exhibit
high mRNA overexpression in retinoblastoma tumor tissues in comparison to normal
retinal tissues [27]. In a followup study, we showed that the KIF14 locus itself was
gained in 15/26 retinoblastoma tumor samples [31] (Fig. 10.1). Overexpression
of KIF14 was also seen in breast and medulloblastoma cell lines, and primary
lung cancers compared to normal tissue counterparts. Moreover, the KIF14 locus
Fig. 10.3 Examples of KIF14’s involvement in cancer. (a) Fluorescence in situ hybridization for
the KIF14 genomic locus shows increasing copy number over ploidy in the progression from nor-
mal retina to retinoma to retinoblastoma (RB). (b) KIF14 protein level increases with increasing
grade in glioma. (c) Knockdown of KIF14 (siKIF14) blocks the migration of HepG2 HCC cells
through Transwell chambers. (d) Knockdown of KIF14 (siKIF14) blocks colony formation by
HeLa cervical carcinoma cells in soft agar. (e) Overexpression of KIF14 with either EGFP
(KIF14-E) or Myc (KIF14-M) tags promotes colony formation of SKOV3 ovarian cancer cells.
(f) Knockdown of KIF14 (shKIF14) decreases intrabursal xenograft formation of OvCa429
ovarian cancer cells in mice; FP fat pad, Ov ovary, OD oviduct, UH uterine horn. (g) Breast cancer
patients with high KIF14 mRNA levels in their tumors (solid line) have poorer overall survival than
those with low KIF14 levels (dashed line), Log-rank p = 0.026 (Figures reprinted with permission
from: (a) Ref. [37], (b) Ref. [34], (c) Ref. [45], (d) Ref. [44], (e) Ref. [33], (f) BLT, unpublished data,
(g) Ref. [43])
showed gain in small cohorts of breast cancer cell lines and HCC [31], providing
the first evidence that KIF14 may be an important gene in the progression of
multiple cancers.
In support of this, gain of KIF14 was the most prevalent copy number change
following the loss of the RB1 gene in the benign precursor lesion retinoma [37], and
KIF14 copy number increased during the progression from retinoma to malignant
retinoblastoma (Fig. 10.3a). In a murine retinoblastoma model in which Simian
Virus 40 large and small T antigens are overexpressed in the retina, KIF14 was also
highly overexpressed in comparison to normal retinal tissues [39]. These results
begin to provide evidence that gain of KIF14 may be a driving force in the develop-
ment of malignancy. Recent work also supports this: KIF14 expression was
increased in premalignant breast tissue of BRCA1 or 2 mutation carriers [40].
Kim et al. studied recurrently altered regions in 76 primary HCC using whole-
genome CGH [32], and delineated a common region of gain (1q21.1–1q32.1)
containing the KIF14 gene. KIF14 mRNA and protein was also overexpressed in
most samples (subset of 20), and correlated highly with genomic copy number
status, pointing to the potential importance of KIF14 in HCC. In papillary renal cell
carcinomas, KIF14 was also identified as one of two oncogenic target genes [38].
In this study, duplication of chromosome 1q was significantly associated with fatal
progression, and correlated with high overexpression of KIF14 mRNA. Quantitative
multiplex PCR (QM-PCR) was used to determine KIF14 copy number in ovarian
cancers [33]. Genomic gain (2.26–5 copies) of KIF14 was seen in all major histo-
logical subtypes, with the highest proportion (30 %) in serous ovarian cancers, and
gain was highly correlated with high mRNA overexpression [33].
Mining of the Progenetix database (www.progenetix.net) [41] for gene-specific
genomic gain of the KIF14 locus (positions 198787930–198856485 on chromo-
some 1; HG18/NCBI Build 36) also revealed that over 22 types of cancers exhibit
gain of KIF14 (Table 10.1). These include breast (30–80 %), HCC (40–60 %),
leukemias and lymphomas (21–88 %), retinoblastoma (46 %), endometrioid (35 %),
ovarian (25–31 %), lung (25 %), pancreatic (21 %) and colorectal (10 %) cancers.
Common gain of KIF14 implies its potential relevance as a mechanism of tumor
development and progression worthy of further study.
10.3.2 Microarray Expression Studies
Mining of the microarray database Oncomine (www.oncomine.org) [42] reveals

studies that demonstrated significant overexpression of KIF14 mRNA in comparison
to normal tissues (over twofold, P < 0.05) such as work on breast (5/11 studies),
lung (3/10), and renal carcinomas (2/5) (Table 10.2). However, for most tumor sites,
studies also show significantly decreased KIF14 expression (over −2-fold, P < 0.05)
compared to normal tissues, such as in breast (6/11), lung (4/10), and renal (1/5).
No significant change in KIF14 expression is also seen in some studies, including
in lung (3/10), ovarian (5/5) and renal (2/5) cancers (Table 10.2). Some of these
microarray results agree with individual qRT-PCR studies, such as for lung, breast,
and liver carcinomas [43–45], however some studies are contradictory, such as for
ovarian cancers [33]. Given the heterogeneity of most solid tumors, and the fact that
most of these studies combined expression results from different histological
Table 10.1 Gene-specific gain of KIF14 according to the Progenetix database [34]
Cases with
Tumor designation KIF14 gain Total cases % gain
Breast (ductal carcinoma in situ) 26 76 34.2
Breast (lobular carcinoma situ) 26 29 89.7
Breast (invasive lobular carcinoma) 65 78 83.3
Breast (intraductal micropapillary carcinoma) 5 16 31.2
Breast (infiltrating ductal carcinoma) 793 1739 45.6
Breast carcinoma 991 2108 47
Cervix (squamous intraepithelial neoplasia) 6 22 27.3
Colorectal carcinomas 105 989 10.6
Endometrioid carcinoma 72 208 35.6
Esophageal carcinoma 118 402 29.4
Gastric adenocarcinoma 39 206 18.9
Glioblastoma 175 1241 14.1
Head and neck carcinomas 111 714 15.5
Hepatocellular carcinoma, fibrolamellar 6 10 60
Hepatocellular carcinoma 419 848 49.4
Hodgkin lymphoma 9 19 47.4
Lung (carcinoma) 20 78 25.6
Lung (small cell carcinoma) 20 79 25.3
Lymphoma (malignant B-cell) 14 61 23
Lymphoma (non-Hodgkins B-cell) 14 61 23
Malignant melanoma 44 230 19.1
Malignant meningioma 19 60 31.7
Medulloblastoma 185 960 19.3
Nasopharyngeal carcinoma 545 202 26.7
Osteosarcoma 20 139 14.4
Ovary (mucinous adenocarcinoma) 15 58 25.9
Ovary (papillary serous cystadenocarcinoma) 43 135 31.9
Ovary (serous adenocarcinoma) 35 134 26.1
Pancreas (acinar cell carcinoma) 2 6 33.3
Pancreas (exocrine carcinomas) 69 326 21.2
Plasma cell leukemia 8 9 88.9
Retinoblastoma 71 153 46.4
Thymus carcinoma 17 39 43.6
Thyroid (follicular carcinoma) 12 45 26.7
Uterine carcinoma 60 192 31.2
subtypes signals a cautious interpretation. The study and validation of any potential
cancer target gene requires rigorous authentication of expression using multiple
methods, taking into account tumor heterogeneity, followed by functional validation
via knockdown/inhibitory strategies to uncover biological significance.
Table 10.2 Microarray studies of tumor vs. normal tissue KIF14 expression according to the
Oncomine database [35]
Studies with Studies with
significant increase significant decrease
% of % of
samples samples Total Total
Number with Number with No number samples
Cancer type of studies increase of studies decrease change of studies tested
Bladder 2 (269) 21 2 (41) 7 1 (89) 5 399
Brain 5 (309) 39 1 (60) 21 6 (253) 12 622
Breast 5 (271) 22 6 (286) 25 0 11 557
Cervical 2 (62) 50 0 N/A 2 (71) 4 133
Colorectal 6 (201) 54 1 (18) 11 0 7 219
Esophageal 2 (107) 26 1 (18) 57 2 (69) 4 194
Gastric 3 (115) 45 0 N/A 1 (92) 4 207
Leukemia 4 (2487) 10 4 (60) 46 2 (160) 10 2707
Liver 2 (260) 28 2 (138) 11 1 (22) 5 420
Lung 3 (282) 23 4 (127) 18 3 (70) 10 479
Lymphoma 3 (384) 70 1 (6) 33 2 (213) 6 603
Melanoma 1 (80) 32 0 N/A 3 (93) 4 173
Nasopharyngeal 1 (31) 16 0 N/A 0 1 31
Ovarian 0 N/A 0 N/A 5 (369) 5 369
Pancreatic 4 (93) 19 2 (35) 27 1 (38) 7 156
Prostate 4 (119) 21 8 (385) 39 2 (100) 14 604
Renal 2 (63) 33 1 (10) 20 2 (68) 5 141
Number of samples tested in brackets
N/A not applicable
10.3.3 KIF14 Expression as a Tumor Marker and Prognostic

Indicator in Multiple Cancers
The first documented study to link expression of KIF14 and clinical outcome
demonstrated a positive correlation between KIF14 mRNA expression levels (as part
of a metastatic gene signature) and the appearance of distant metastases in breast
cancer patients [46]. Subsequent characterization of KIF14 mRNA expression in
primary breast carcinomas solidified the concept that a mitotic kinesin could predict
poor outcome in cancer patients. Our analysis revealed that the majority of breast
cancers demonstrate significant overexpression of KIF14 mRNA in relation to
normal controls (on average 10-fold in 99 mixed breast cancer tissues vs. 10 normal
breast tissues), and expression correlated with tumor grade and invasiveness [43].
Univariate Cox regression showed that KIF14 mRNA levels were prognostic for
short overall and progression-free survival (Fig. 10.3g).
We went on to show that KIF14 is also a prognostic indicator in lung and ovarian
cancers [33, 44]. KIF14 mRNA was significantly overexpressed in lung carcinomas,
and decreased with increasing differentiation [44]. Multivariate Cox regression
revealed that KIF14 overexpression was an independent prognostic factor for
disease-free survival. In ovarian cancers, KIF14 was overexpressed in 91 % of all
samples (109), with the highest expression in serous ovarian cancers, the most
common and lethal histologic subtype [33]. Interestingly, KIF14 genomic gain and
overexpression were seen in early stage ovarian cancers, consistent with KIF14
being an early oncogenic event in ovarian cancer tumorigenesis, as also appears
likely in retinoblastoma and some breast cancers (see above). Patients with high
tumor KIF14 overexpression demonstrated a higher rate of recurrence, and signifi-
cantly shorter overall and progression-free survival than patients with KIF14 levels
close to normal tissue expression. Furthermore, KIF14 mRNA was independently
prognostic for poor overall survival, pointing to KIF14 expression as a potential
marker for progression.
Many other groups subsequently identified KIF14 expression to be significantly
associated with tumor cell specificity and/or clinical progression of cancer.
Madhavan et al. studied KIF14 mRNA expression in 30 retinoblastoma samples,
and found that KIF14 mRNA and protein were overexpressed in all tumors in
comparison to normal fetal retinas, and was significantly associated with older age
at diagnosis [47]. However no association was found between KIF14 mRNA and
clinical variables such as differentiation, invasion or duration of disease. In a larger
cohort of retinoblastoma tumors (57) KIF14 mRNA and protein were overexpressed,
as was another important oncogene in retinoblastoma development and progression,
E2F3 [48]. High KIF14 and E2F3 mRNA were also both associated with older age
at diagnosis, but no other clinical associations were found.
A high-density oligonucleotide microarray analysis of 14 squamous cell laryn-
geal carcinomas and matched adjacent normal mucosal controls revealed KIF14
overexpression as a molecular marker for laryngeal carcinomas [49]. A follow-up
study of 24 samples confirmed this finding [50]. Yang et al. studied a cohort of 30
HCCs and found that KIF14 mRNA was overexpressed in most primary tumors
(23/30), and correlated significantly with increased gene-specific DNA copy
number [45]. When assessed in a larger cohort of HCCs (102) with associated
clinicopathological variables, KIF14 mRNA expression increased with stage and
grade, and was associated with poor outcome (decreased disease-free survival),
indicating that KIF14 is an important gene involved in HCC, and a potential
prognostic and therapeutic target. KIF14 is also upregulated in several HCC cell
lines versus “normal” THLE-2 liver cells [51]. Given the genomic gain of KIF14 in
HCC (see above), it will be interesting to see if DNA copy number is also predictive
for outcome.
In endometrioid adenocarcinomas, KIF14 overexpression was identified as a
marker for late-stage, advanced cancers [52], indicating that with further validation,
KIF14 expression could be employed as a potential indicator of poor patient outcomes
for this common gynecological malignancy. More recently, a genomic profiling
study of both adult and pediatric synovial sarcomas revealed that chromosomal
instability accounts for worse metastatic outcomes experienced by adult as opposed

to pediatric patients. Within this chromosomal instability, KIF14 was identified as
one of two top-ranked oncogenes contributing to metastatic outcome [53].
In gliomas, KIF14 expression was also identified as a candidate prognostic
marker [34] (Fig. 10.3b). A large cohort (128) of chemotherapy-naïve human gliomas
was profiled for both KIF14 mRNA and protein expression. When compared to
the KIF14 expression in non-neoplastic brain tissues (10 patients), KIF14 was
significantly overexpressed at both the mRNA (5.5-fold) and protein (4.2-fold) levels,
and increased with increasing pathological grade (Fig. 10.3b). Furthermore, KIF14
protein expression (via immunohistochemical staining) was strongly associated
with low Karnofsky performance score (a measure of quality of life, where a score
of 100 is “perfect health” and 0 is “death”), high Ki-67 index, and high mitotic
index, indicating that KIF14 may promote advanced disease progression [34].
Importantly, Cox univariate and multivariate regression determined that KIF14 protein
overexpression was prognostic for decreased overall survival in glioma patients,
pointing to the potential of KIF14 histological expression as a prognostic marker.
10.3.4 KIF14 as a Predictive Biomarker for Chemotherapy

Response
KIF14 has been identified as a potential predictive biomarker in the response or

resistance to chemotherapy in breast and ovarian cancers. Singel et al. conducted an
RNA interference screen for genes whose loss-of-function enhanced chemosensitivity
to docetaxel in an estrogen, progesterone and Her2-negative (“triple-negative”),
claudin-low breast cancer cell line, MDA-MB-231. KIF14 was identified as one of
the two highest ranked genes associated with conferring docetaxel chemosensitivity;
this sensitivity was subsequently confirmed in 3 additional triple-negative breast
cancer (TNBC) cell lines [54]. In a larger cohort of 52 breast cancer cell lines with
known clinical characteristics, KIF14 was highly overexpressed in TNBC cell lines
more often than in other clinical subtypes [54].
Subsequent analysis of five publicly available breast cancer expression array
datasets (over 900 patients) revealed that KIF14 mRNA expression significantly
correlated with overall and relapse-free survival. Furthermore, the relapse-free and
overall survival prognostic power of estrogen receptor and node status was signifi-
cantly improved when KIF14 expression was added as a variable [54], strongly
indicating that KIF14 expression is an important marker for chemotherapy resistance
in TNBC. This was supported by evidence that KIF14 expression is increased in
treatment-resistant tumors [40].
In a taxane-resistant cohort (51) of ovarian cancers, KIF14 was identified as
highly overexpressed in comparison to normal ovary tissues (almost 20-fold,
16 samples) or matched normal peritoneal epithelial tissues (sevenfold, 51 samples)
[55]. KIF14 mRNA expression was significantly associated with Ki-67 and citron
kinase (CIT) expression, but showed no associations with clinical characteristics

such as stage. Interestingly, a significant association between CIT, PRC1 (KIF14
interactors; see above), Ki-67 expression and shorter time to progression were seen
in this cohort of chemotherapy-treated ovarian cancers, suggesting that KIF14 is
closely related to proliferative status of the tumor and may indeed be a surrogate
marker of progression, as previously suggested in breast cancer [43]. However,
these results are in opposition to our study of chemotherapy-naïve ovarian cancers,
where a lack of association between high KIF14 expression and CIT, PRC1 and
Ki-67 expression was seen, suggesting that KIF14 may be an initial driver of
progression irrespective of proliferative status [33]. These seemingly contradictory
results could simply reflect context-dependent roles for KIF14 in primary,
chemotherapy-naïve tumors, versus chemotherapy-resistant, recurrent disease
(see later).
Taken together, these results identify KIF14 as an important tumor marker in
multiple cancers and possibly an indicator of tumor progression, poor outcome, and
recurrent disease. This is especially important in the evolving era of precision
medicine, where stratification markers and indicators of therapeutic response are
becoming increasingly sought modalities to help guide treatment decisions.
10.4 The Role of KIF14 in Cancer Cells
The discovery of KIF14 as an important cancer gene in multiple malignancies

strongly suggests that it could become a viable therapeutic target. In support of this,
several studies have characterized the effect of either expression or knockdown in
cancer cell lines using either stable or transient methods (Fig. 10.3). These studies
determined that knockdown in lung, breast, cervical, ovarian and HCC cell lines
significantly decreased cell proliferation and induced apoptosis, while overexpres-
sion in HCC and ovarian cancer cell lines strongly increased proliferative capacity
(Fig. 10.3e), consistent with its known role in cytokinesis [27, 33, 43–45, 51].
Overexpression in immortalized mammary epithelial cells was sufficient to confer
increased proliferation and anchorage-independent growth, but not enhanced migra-
tion or tumorigenesis [40]. Knockdown of KIF14 reduced colony formation in lung
(Fig. 10.3d), HCC, and ovarian cancer cell lines, and significantly reduced migra-
tion (Fig. 10.3c) and xenograft formation of HCC and ovarian cancer (Fig. 10.3f)
cell lines, pointing to a tumorigenic role for KIF14 in cancer cells [27, 33, 43–45, 51].
The hypothesis that KIF14 could have cancer-specific, proliferation-independent
functions initially stemmed from the analysis of clinical data from KIF14-
overexpressing serous ovarian cancers [33]. It was determined that in poor-
prognosis, high KIF14-overexpressing tumors, no correlation could be seen between
KIF14 expression levels and the levels of KIF14 cytokinesis binding partners CIT
and PRC1, nor with the clinical proliferation marker Ki-67 (in contrast to what was
observed in breast cancer [43] and in another ovarian cancer study [55]). Additionally,
the change in the anchorage-independent phenotype in ovarian cancer cell lines in
response to either KIF14 overexpression or knockdown was much more marked

than changes in their proliferative rate [33], as was also observed in HeLa and a lung
cancer cell line [44]. Additional proof was provided in HCC cell lines, where knock-
down of KIF14 significantly reduced xenograft formation in mice through apoptosis
mediated by the PI3K pathway [45].
Other recent work in HCC suggests regulation of the cell cycle by KIF14 [51].
KIF14 knockdown was associated with a decrease in the Skp1-Cul1-F-box (SCF)
complex members Skp2 and Cks1, leading to stabilization of cyclin-dependent
kinase inhibitor p27Kip1. p27Kip1 is usually ubiquitinated and tagged for degradation
by the SCF complex, an E3 ubiquitin ligase. p27Kip1 stabilization led to concomitant
repression of cyclins E1, D1, and B1. Skp2 overexpression could rescue the binu-
cleation phenotype of KIF14 knockdown. However, how KIF14 modulates Skp2
and Cks1 levels remains unknown.
A novel interaction between KIF14 and Radil could mediate a cytokinesis-
independent role for KIF14 in breast cancer [56] (Fig. 10.2c). Radil is a crucial
mediator of Rap1a-mediated integrin inside-out signaling. Radil interacts specifi-
cally with the KIF14 C-terminal tail via KIF14’s last 4 amino acids, and co-localizes
to microtubules with KIF14 in breast cancer cell lines. Although shuttling of Radil
by KIF14 on microtubule tracks was not observed, it was determined that KIF14
strongly associates with Radil and microtubules at the cell membrane to mediate
increased cell adhesion and migration. KIF14 controls the amount of Radil-Rap1a
activity at the cell membrane to promote cell adhesion and migration behaviors,
favoring metastatic progression in breast cancer cells. This was the first concrete
evidence demonstrating KIF14 controlling cancer-specific cell behaviours indepen-
dent of cytokinesis [56], offering opportunities for therapeutic disruption of specific
cancer-causing protein interactions [57].
KIF14 may also have an important role in conferring chemotherapy resistance.
In TNBC cell lines, KIF14 knockdown significantly reduced the tumor mass of
mammary fat pad xenografts compared to controls following docetaxel chemother-
apy, effectively chemosensitizing the cancer cells [54]. Conversely, KIF14 overex-
pression in immortalized mammary epithelial cells conferred docetaxel resistance
[40]. The chemosensitization in response to KIF14 knockdown may be modulated
by RIP2 kinase, a dual specificity kinase that controls innate and adaptive immune
responses and activates NF-κB, and was found as a potential KIF14 interacting
partner through a large-scale mapping of human protein-protein interactions by
mass spectrometry [58]. Knockdown of RIP2 also altered in vitro chemosensitivity
to docetaxel in TNBC cell lines, indicating a possible functional relationship
between KIF14 and RIP2 in the survival and chemoresistant phenotype of TNBC
cell lines [54].
KIF14’s chemosensitizing effects are also at least partially mediated through
phosphorylation of Akt [40], which is increased in TNBC cells by KIF14 overex-
pression and decreased on KIF14 knockdown or treatment with a putative KIF14
small molecule inhibitor (see below); constitutively active Akt can rescue the
chemosensitization of KIF14 knockdown. Interestingly, KIF14 interacts with total
Akt and co-localizes to the plasma membrane when TNBC cells are stimulated
with insulin [40].
As mentioned, in primary ovarian cancers, KIF14 expression was associated
with taxane resistance, and strongly associated with the expression of cytokinesis
markers CIT and PRC1 [55]. These results highlight potentially pleomorphic roles
for KIF14 in carcinogenesis, depending on the stage of tumor progression. In the
early stages of tumor formation and metastatic progression, KIF14 could act as an
important driver of tumor progression through cytokinesis-independent pathways;
in response to chemotherapy, KIF14 could promote cellular survival and therapeutic
resistance through cytokinesis-dependent (CIT, PRC1) and -independent (Radil,
Akt, RIP2, NF-κB, Skp2) pathways. Indeed, the overexpression of kinesins in tumor
cells has been proposed as one mechanism through which tumors evade the anti-
mitotic effect of conventional microtubule-targeting chemotherapies [59].
10.4.1 Is KIF14 an Oncogene?
Increased and uncontrolled proliferative capacity is a hallmark of cancer cells [60],

so it is logical to assume that the upregulation of genes involved in cell proliferation
like mitotic kinesins might simply be a consequence of one or more oncogenic
insults. However, mounting evidence suggests that KIF14 itself may indeed be an
oncogene capable of driving tumor progression. As mentioned previously, the iden-
tification of gene-specific KIF14 genomic gain in benign retinoma lesions [37] and
overexpression in premalignant breast tissue [40] suggest that the gain of KIF14 is
an early event in tumorigenesis. Furthermore, the observation of low-level amplifi-
cation in one retinoblastoma tumor strongly points to KIF14 as being a possible
oncogene [31] (Fig. 10.1).
Numerous studies in lung, breast, cervical, ovarian, and HCC cell lines employing
transient (siRNA) or stable (shRNA) knockdown of KIF14 significantly decreased
cell proliferation and induced apoptosis in vitro [27, 33, 43–45] (Fig. 10.3), strength-
ening the idea that KIF14 overexpression is crucial for malignancy. In support of this,
serous ovarian cancers display genomic gain in both early and late stage tumors in
similar proportions, while stable shRNA knockdown of KIF14 in ovarian cancer cell
lines significantly decreases in vitro tumorigenicity [33]. As mentioned, although
KIF14 overexpression is not sufficient to transform normal mammary epithelial
cells, it does promote proliferation and anchorage-independent growth [40]. Moreover,
the fact that KIF14 expression in multiple cancers is tumor cell-specific, can inde-
pendently predict worse outcome for patients, is associated with chemoresistance, and
is necessary for in vitro and in vivo tumorigenesis strongly points towards an
oncogenic function for KIF14. Furthermore, the recent identification of a cytokinesis-
independent function for KIF14 in promoting metastatic behaviors in breast cancer
cells [56] strongly supports this postulate.
Two studies have however defined a possible tumor suppressive function for
KIF14. In pancreatic ductal adenocarcinomas, perineural invasion is a hallmark of
metastatic progression [61]. By expression microarray analysis, Abiatari et al. found
a significant decrease in KIF14 expression in highly invasive clones of pancreatic

cell lines using an ex vivo rat vagal nerve model [62]. Furthermore, transient
(siRNA) knockdown of KIF14 in pancreatic cell lines significantly increased inva-
siveness in Matrigel, anchorage-independent survival, and perineural invasion in
multiple pancreatic cell lines, suggesting a tumor-suppressive role for KIF14 in
pancreatic cancer. Surprisingly, these authors found increased KIF14 expression in
bulk pancreatitis and pancreatic cancer tissue compared with normal pancreas; this
overexpression was absent in cancer cells invading nerves.
A study in lung adenocarcinomas found a 30 % loss of chromosome 1q in a
cohort of 53 patients [63]. The authors confirmed that approximately 30 % of lung
adenocarcinomas exhibited a decrease in KIF14 protein expression compared to
adjacent normal bronchial epithelial cells. In this same cohort, low KIF14 mRNA
expression was associated with poor overall survival, while increased KIF14
expression was an independent factor associated with longer overall survival, and
negatively correlated with lung metastasis [63]. Extending their analysis to a cohort
of 122 patient samples, they confirmed that low immunocytochemical KIF14
expression was independently prognostic for poor patient outcome. Overexpression
of KIF14 in lung adenocarcinoma cell lines significantly inhibited anchorage-
independent growth in vitro, as well as xenograft formation in vivo, implying a
tumor suppressive role for KIF14 in lung adenocarcinomas [63].
Although these two studies are contradictory to the many other studies demon-
strating an oncogenic role for KIF14, validation of these findings by other groups is
crucial to extend the understanding of the role of KIF14 in cancer. The lung adeno-
carcinoma findings [63] contrast with our earlier work showing KIF14 overexpres-
sion as prognostic for poor outcome in a mixed population of non-small cell lung
carcinomas [44]. Aside from the possibility of inter-cohort variation explaining this
difference, in our study, the squamous cell carcinomas had the highest KIF14
expression; perhaps KIF14 has an oncogenic role primarily in this histological
subtype. Independent replication of the discrepant clinical findings is needed. As
well, further in vitro work, with the two cell lines used in the respective studies and
others, is crucial. Additionally, in the pancreatic cancer context, another perineural
invasion study failed to validate KIF14 as a gene of importance [64]. Further study
is warranted before assigning KIF14 as a bona fide tumor suppressor, whereas there
is a wealth of data accumulated to date in multiple studies assigning an oncogenic
role to this gene.
10.5 KIF14 as a Therapeutic Target
10.5.1 Current Evidence
Mounting pre-clinical evidence in cell and animal models provides a solid basis for
further validation of KIF14 as a therapeutic target (Fig. 10.3). Firstly, its very low
expression in many normal tissues, as opposed to specifically high tumor cell
expression (rarely seen with other kinesins) raises the possibility of selective
inhibition with minimal side effects [65]. The strong clinical association between
KIF14 expression, tumor progression, and chemoresistance, along with the crucial
role of KIF14 in both cytokinesis-dependent and -independent cellular pathways
suggests KIF14 overexpression as an essential tumorigenic pathway in many cancer
cells. Functional in vitro and in vivo evidence in cell lines and xenografts reveals
that inhibition of KIF14 significantly reduces tumor burden through apoptosis, a
highly desirable outcome as opposed to a cytostatic effect [33, 45]. The recent
description of Kif14’s crystal structure, discovery of multiple interacting proteins,
mechanisms of tumor overexpression, and functional pathways for KIF14 in cancer
cells provides many opportunities for rational drug design. The relevance of
KIF14 in multiple solid tumor types would also provide a good incentive for the
development of targeted therapies, as they would be applicable to multiple patient
populations.
Lastly, a screen for inhibitors of the motor domain (V432 – K720) of KIF14 has
been described [66]. Four thiosemicarbazone compounds were found to modulate
KIF14 motor domain ATPase activity in the sub-micromolar range, demonstrating
selectivity in comparison to other kinesin motor domains, namely KIF3A, CENP-E,
MCAK, and MKLP-1 [66]. These compounds exerted growth inhibitory effects in
HeLa and A2780 cells in the low micromolar range [66]; one had similar effects in
retinoblastoma cell lines (TWC, unpublished data) and in TNBC cell lines. It also
sensitized these latter lines to docetaxel, consistent with the effects of KIF14
knockdown [40]. Unfortunately, these thiosemicarbazones are poorly soluble and
potentially “dirty” compounds, so further screening is warranted to find other
classes of KIF14 inhibitors.
10.5.2 Future Evidence Needed for Target Validation
As discussed widely in this book, kinesins have recently emerged as exciting novel
therapeutic targets in cancer [59], and have prompted intense efforts to develop
specific inhibitors. Although the kinesin motor domains of several kinesins have
been selectively targeted, one needs to be mindful of the potential challenges
gleaned from targeting the kinesin motor domain, including problems of specificity
and toxicity. Furthermore, there is complementarity in functions between kinesin
family members; for example, when Eg5 is inhibited, KIF15 can substitute for it
and perform all essential functions pertaining to bipolar spindle assembly [67].
Further characterization of KIF14 function and binding partners in normal and cancer
cells will be crucial to rationally design selective, minimally toxic inhibitors.
Specifically, further study of the KIF14 protein structure in complex with known
interactors CIT, PRC1, supervillin, RIP2, Radil, and others will be important. More
information is required on how KIF14 participates in cytokinesis and tumorigenic
signaling, so that inhibition of cancer cell-specific interactions (rather than ATPase
activity alone) can be targeted with the goal of reducing off-target effects. Further
investigations into the transforming ability of KIF14 when overexpressed in various
types of normal cells will further refine the tumorigenic pathways to be targeted.
Additionally, it will be crucial to validate KIF14 as a tumor biomarker, prognostic
indicator and a chemoresistance marker in larger, prospective patient cohorts, in
order for expression to be useful not only as an indicator of initial patient prognosis,
but also as a bioindicator of patient response to inhibitory therapy, an essential
modality in guiding individualized treatment decisions.
10.5.3 Is KIF14 a Good Therapeutic Target?
KIF14 is a viable and very exciting candidate for drug discovery: it is an important
cancer gene with relevance to multiple cancers, it possesses tumor-specific expression,
its expression is prognostic for patient outcomes, it demonstrates tumor-specific
functions and interacting partners, and its motor domain is selectively targetable
with inhibitors. Although there is much more to learn about the basic biology and
tumorigenic functions of KIF14, the data available already point to tractable protein
interactions and signaling pathways to be investigated for therapeutic intervention.
The doors are wide open for the rational design of selective inhibitors of KIF14’s
tumorigenic functions – it is now only a matter of time.
Acknowledgments We thank Brenda Gallie (Ontario Cancer Institute) for critical comments on
the manuscript. Related work in the laboratory of TWC has been supported by an American Cancer
Society Institutional Research Grant, the Knights Templar Eye Foundation and the Alcon Research
Institute. This publication was also made possible in part by Grant Number KL2TR001106 from
the National Institutes of Health, National Center for Advancing Translational Sciences, Clinical and
Translational Sciences Award and by an unrestricted grant from Research to Prevent Blindness, Inc.
References
1. Thériault BL, Corson TW (2012) Kinesin family member 14 (KIF14). Atlas Genet Cytogenet
Oncol Haematol 16:695–699
2. Durocher D et al (2000) The molecular basis of FHA domain: phosphopeptide binding speci-
ficity and implications for phospho-dependent signaling mechanisms. Mol Cell 6:1169–1182
3. Dephoure N et al (2008) A quantitative atlas of mitotic phosphorylation. Proc Natl Acad Sci U
S A 105:10762–10767
4. Nomura N et al (1994) Prediction of the coding sequences of unidentified human genes. II. The
coding sequences of 40 new genes (KIAA0041-KIAA0080) deduced by analysis of cDNA
clones from human cell line KG-1. DNA Res 1:251–262
5. Olsen JV et al (2006) Global, in vivo, and site-specific phosphorylation dynamics in signaling
networks. Cell 127:635–648
6. Vasilescu J et al (2007) The proteomic reactor facilitates the analysis of affinity-purified pro-
teins by mass spectrometry: application for identifying ubiquitinated proteins in human cells.
J Proteome Res 6:298–305
7. Miki H, Setou M, Kaneshiro K, Hirokawa N (2001) All kinesin superfamily protein, KIF,
genes in mouse and human. Proc Natl Acad Sci U S A 98:7004–7011
8. Geer LY et al (2010) The NCBI BioSystems database. Nucleic Acids Res 38:D492–D496
9. Ohkura H et al (1997) Mutation of a gene for a Drosophila kinesin-like protein, Klp38B, leads
to failure of cytokinesis. J Cell Sci 110(Pt 8):945–954
10. Arora K et al (2014) KIF14 binds tightly to microtubules and adopts a rigor-like conformation.
J Mol Biol 426:2997–3015
11. Rice S (2014) Structure of Kif14: an engaging molecular motor. J Mol Biol 426:2993–2996
12. Nakagawa T et al (1997) Identification and classification of 16 new kinesin superfamily (KIF)
proteins in mouse genome. Proc Natl Acad Sci U S A 94:9654–9659
13. Alphey L et al (1997) KLP38B: a mitotic kinesin-related protein that binds PP1. J Cell Biol
138:395–409
14. Ruden DM, Cui W, Sollars V, Alterman M (1997) A Drosophila kinesin-like protein, Klp38B,
functions during meiosis, mitosis, and segmentation. Dev Biol 191:284–296
15. Molina I et al (1997) A chromatin-associated kinesin-related protein required for normal
mitotic chromosome segregation in Drosophila. J Cell Biol 139:1361–1371
16:3187–3199
17. Gruneberg U et al (2006) KIF14 and citron kinase act together to promote efficient cytokinesis.
J Cell Biol 172:363–372
18. Carleton M et al (2006) RNA interference-mediated silencing of mitotic kinesin KIF14 dis-
rupts cell cycle progression and induces cytokinesis failure. Mol Cell Biol 26:3853–3863
19. Smith TC, Fang Z, Luna EJ (2010) Novel interactors and a role for supervillin in early cytoki-
nesis. Cytoskeleton 67:346–364
20. Watanabe S, De Zan T, Ishizaki T, Narumiya S (2013) Citron kinase mediates transition from
constriction to abscission through its coiled-coil domain. J Cell Sci 126:1773–1784
21. Bassi ZI, Audusseau M, Riparbelli MG, Callaini G, D’Avino PP (2013) Citron kinase controls
a molecular network required for midbody formation in cytokinesis. Proc Natl Acad Sci U S A
110:9782–9787
22. Samwer M et al (2013) The nuclear F-actin interactome of Xenopus oocytes reveals an
actin-bundling kinesin that is essential for meiotic cytokinesis. EMBO J 32:1886–1902
23. Fujikura K et al (2013) Kif14 mutation causes severe brain malformation and hypomyelin-
ation. PLoS One 8:e53490
24. Yunus J, Setsu T, Kikkawa S, Sakisaka T, Terashima T (2014) Cytoarchitecture of the olfactory
bulb in the laggard mutant mouse. Neuroscience 275:259–271
25. Di Cunto F et al (2000) Defective neurogenesis in citron kinase knockout mice by altered
cytokinesis and massive apoptosis. Neuron 28:115–127
26. Filges I et al (2014) Exome sequencing identifies mutations in KIF14 as a novel cause of an
autosomal recessive lethal fetal ciliopathy phenotype. Clin Genet 86:220–228
27. Corson TW, Huang A, Tsao MS, Gallie BL (2005) KIF14 is a candidate oncogene in the 1q
minimal region of genomic gain in multiple cancers. Oncogene 24:4741–4753
28. Bell D et al (2011) Integrated genomic analyses of ovarian carcinoma. Nature 474:609–615
29. Wei X et al (2011) Exome sequencing identifies GRIN2A as frequently mutated in melanoma.
Nat Genet 43:442–446
30. Wood LD et al (2007) The genomic landscapes of human breast and colorectal cancers.
Science 318:1108–1113
31. Bowles E et al (2007) Profiling genomic copy number changes in retinoblastoma beyond loss
of RB1. Genes Chromosomes Cancer 46:118–129
32. Kim TM et al (2008) Clinical implication of recurrent copy number alterations in hepatocellular
carcinoma and putative oncogenes in recurrent gains on 1q. Int J Cancer 123:2808–2815
33. Thériault BL, Pajovic S, Bernardini MQ, Shaw PA, Gallie BL (2012) Kinesin family member
14: An independent prognostic marker and potential therapeutic target for ovarian cancer. Int J
Cancer 130:1844–1854
34. Wang Q et al (2013) Kinesin family member 14 is a candidate prognostic marker for outcome
of glioma patients. Cancer Epidemiol 37:79–84
35. Cam H et al (2004) A common set of gene regulatory networks links metabolism and growth
inhibition. Mol Cell 16:399–411
36. Thériault BL et al (2014) Transcriptional and epigenetic regulation of KIF14 overexpression
in ovarian cancer. PLoS One 9:e91540
37. Dimaras H et al (2008) Loss of RB1 induces non-proliferative retinoma: increasing genomic
instability correlates with progression to retinoblastoma. Hum Mol Genet 17:1363–1372
38. Szponar A, Zubakov D, Pawlak J, Jauch A, Kovacs G (2009) Three genetic developmental
stages of papillary renal cell tumors: duplication of chromosome 1q marks fatal progression.
Int J Cancer 124:2071–2076
39. Pajovic S et al (2011) The TAg-RB murine retinoblastoma cell of origin has immunohisto-
chemical features of differentiated Müller glia with progenitor properties. Invest Ophthalmol
Vis Sci 52:7618–7624
40. Singel SM et al (2014) KIF14 promotes AKT phosphorylation and contributes to chemoresis-
tance in triple-negative breast cancer. Neoplasia 16:247–256, 256.e2
41. Baudis M, Cleary ML (2001) Progenetix.net: an online repository for molecular cytogenetic
aberration data. Bioinformatics 17:1228–1229
42. Rhodes DR et al (2004) ONCOMINE: a cancer microarray database and integrated data-
mining platform. Neoplasia 6:1–6
43. Corson TW, Gallie BL (2006) KIF14 mRNA expression is a predictor of grade and outcome in
breast cancer. Int J Cancer 119:1088–1094
44. Corson TW et al (2007) KIF14 messenger RNA expression is independently prognostic for
outcome in lung cancer. Clin Cancer Res 13:3229–3234
45. Yang T, Zhang XB, Zheng ZM (2013) Suppression of KIF14 expression inhibits hepatocellular
carcinoma progression and predicts favorable outcome. Cancer Sci 104:552–557
46. van’t Veer LJ et al (2002) Gene expression profiling predicts clinical outcome of breast cancer.
Nature 415:530–536
47. Madhavan J et al (2007) High expression of KIF14 in retinoblastoma: association with older
age at diagnosis. Invest Ophthalmol Vis Sci 48:4901–4906
48. Madhavan J et al (2009) KIF14 and E2F3 mRNA expression in human retinoblastoma and its
phenotype association. Mol Vis 15:235–240
49. Markowski J et al (2009) Gene expression profile analysis in laryngeal cancer by high-density
oligonucleotide microarrays. J Physiol Pharmacol 60(Suppl 1):57–63
50. Markowski J et al (2009) Metal-proteinase ADAM12, kinesin 14 and checkpoint suppressor 1
as new molecular markers of laryngeal carcinoma. Eur Arch Otorhinolaryngol
266:1501–1507
51. Xu H et al (2014) Silencing of KIF14 interferes with cell cycle progression and cytokinesis by
blocking the p27Kip1 ubiquitination pathway in hepatocellular carcinoma. Exp Mol Med 46:e97
52. Mhawech-Fauceglia P et al (2010) Microarray analysis reveals distinct gene expression profiles
among different tumor histology, stage and disease outcomes in endometrial adenocarcinoma.
PLoS One 5:e15415
53. Lagarde P et al (2013) Chromosome instability accounts for reverse metastatic outcomes of
pediatric and adult synovial sarcomas. J Clin Oncol 31:608–615
54. Singel SM et al (2013) A targeted RNAi screen of the breast cancer genome identifies KIF14
and TLN1 as genes that modulate docetaxel chemosensitivity in triple-negative breast cancer.
Clin Cancer Res 19:2061–2070
55. Ehrlichova M et al (2013) The association of taxane resistance genes with the clinical course
of ovarian carcinoma. Genomics 102:96–101
56. Ahmed SM et al (2012) KIF14 negatively regulates Rap1a-Radil signaling during breast cancer
progression. J Cell Biol 199:951–967
57. Thériault B, Dimaras H, Gallie B, Corson T (2014) The genomic landscape of retinoblastoma:
a review. Clin Experiment Ophthalmol 42:33–52
58. Ewing RM et al (2007) Large-scale mapping of human protein-protein interactions by mass
spectrometry. Mol Syst Biol 3:89
60. Hanahan D, Weinberg RA (2011) Hallmarks of cancer: the next generation. Cell 144:646–674
61. Dai H et al (2007) Enhanced survival in perineural invasion of pancreatic cancer: an in vitro
approach. Hum Pathol 38:299–307
62. Abiatari I et al (2009) Consensus transcriptome signature of perineural invasion in pancreatic
carcinoma. Mol Cancer Ther 8:1494–1504
63. Hung PF et al (2013) The motor protein KIF14 inhibits tumor growth and cancer metastasis in
lung adenocarcinoma. PLoS One 8:e61664
64. Koide N et al (2006) Establishment of perineural invasion models and analysis of gene expres-
sion revealed an invariant chain (CD74) as a possible molecule involved in perineural invasion
in pancreatic cancer. Clin Cancer Res 12:2419–2426
65. Basavarajappa HD, Corson TW (2012) KIF14 as an oncogene in retinoblastoma: a target for
novel therapeutics? Future Med Chem 4:2149–2152
66. Mao M, Linsley PS, Buser CA, Marshall CG, Kim AS (2004) Methods for identifying modula-
tors of kinesin activity. WO 2004/109290 A2
67. Tanenbaum ME et al (2009) Kif15 cooperates with eg5 to promote bipolar spindle assembly.
Curr Biol 19:1703–1711
Chapter 11
Kinesin-8 Members and Their Potential
as Biomarker or Therapeutic Target
Thomas U. Mayer and Silke Hauf
11.1 The Cellular Role of Kinesin-8 Family Members
11.1.1 The Kinesin-8 Family
Among the 14 kinesin families (Chap. 1, [1, 2]), Kinesin-8 stands out due to its
unusual properties [3]. Whereas most kinesins walk directionally along microtu-
bules and some kinesins influence microtubule dynamics, Kinesin-8 combines both
these properties. Kinesin-8 members move towards the plus-end of microtubules in
an exceptionally processive fashion [4–7] and influence microtubule dynamics
specifically at the plus-end [3, 8, 9]. Recently, the budding yeast Kinesin-8 family
member Kip3p has been additionally shown to slide microtubules against each other
[10], making Kinesin-8 the ‘jack-of-all-trades’ of kinesins (Fig. 11.1). Kip3p also
has the ability to switch microtubule protofilaments while walking [11]. In which
way, if any, this latter feature contributes to its function in vivo is unknown.
By amino acid sequence the Kinesin-8 family is most closely related to the
Kinesin-13 family [2, 12]. Kinesin-13 proteins are efficient microtubule depolymer-
ases, but they only diffuse along microtubules and do not ‘walk’ in a directed
fashion, as Kinesin-8 proteins do [13]. Whereas it is clear that Kinesin-8 proteins
influence microtubule plus-end dynamics, it is controversial whether the capacity to
depolymerize microtubule plus-ends, which has been observed for some Kinesin-8
T.U. Mayer (*)

Department of Biology and Konstanz Research School Chemical Biology,
University of Konstanz, Konstanz, Germany
e-mail: thomas.u.mayer@uni-konstanz.de
S. Hauf
Department of Biological Sciences and Virginia Bioinformatics Institute,
Virginia Tech, Blacksburg, VA, USA
e-mail: silke.hauf@vt.edu

172 T.U. Mayer and S. Hauf
tail links to microtubule
modulating
motor activity plus-end dynamics
Kif18A
microtubule Kif18B
sliding EB1
Kip3
tail links via EB1 to microtubule
Fig. 11.1 Kinesin-8 members show directional movement and modulate microtubule dynamics.
The Kinesin-8 Kif18A forms a homodimer that shows both plus-end directed motility and influences
microtubule dynamics at the plus-end. Contacts made between the Kif18A tail and microtubules
promote the processivity of the movement and are crucial for the activity at the plus-end. The
Kif18A homolog Kif18B also connects to microtubules through its tail, but in an indirect fashion
by binding to the microtubule plus-end tracking protein EB1. The budding yeast Kinesin-8 Kip3p
can cross-link and slide microtubules against each other. Other members of the Kinesin-8 family
may have this activity as well
family members, is a conserved feature of all Kinesin-8s [7, 14–19]. Those studies
that reported a depolymerization activity found it to be in detail different from the
Kinesin-13 mechanism of action [14, 17, 20, 21]. Hence, while Kinesin-8 and
Kinesin-13 members share some features, Kinesin-13 proteins are specialized
microtubule depolymerases that have entirely lost walking capability, whereas
Kinesin-8 proteins retain the ability to walk along microtubules but additionally
alter the dynamics of the microtubule plus-end.
11.1.1.1 Members of the Human Kinesin-8 Family
The human genome codes for three Kinesin-8 family members, Kif18A (chromo-
some 2), Kif18B (chromosome 11) and Kif19 (chromosome 17), with possibly
additional variants created by alternative splicing [22–25]. Kif18A and Kif18B are
about 60 % identical in the kinesin domain, whereas Kif19 differs more from the
two (with 45 % and 47 % identity in the kinesin domain compared to either Kif18A
or Kif18B). All three members have the kinesin domain in the N-terminus of
the protein, followed by the neck-linker region, the coiled-coil neck and a tail
(Chap. 1). Although this has not been formally proven for all members of the family,
11 Kinesin-8 Members and Their Potential as Biomarker or Therapeutic Target 173
Kinesin-8 molecules are strongly expected to dimerize through the coiled-coil neck
and act as homodimers [4]. The kinesin domain is required for the plus-end directed
mobility along microtubules and is crucial for the function of the protein in vivo
[26, 27].
In many kinesins, the tail connects to cargo that is being transported. Kinesin-8
proteins are poor carriers that do not support much load [28]. Instead, their walking
seems to mostly serve the purpose of delivering themselves to the microtubule
plus-end, where they influence microtubule dynamics. Accordingly, the tail has
been repurposed: Kinesin-8 proteins use it to connect to microtubules. The Kif18A
tail binds directly to microtubules (Fig. 11.1), which is necessary for the high walking
processivity of this kinesin and is essential for Kif18A function in vivo [4–6, 29].
In Kif18B, the tail binds to the microtubule plus-end tracking protein EB1 and
therefore connects indirectly to microtubules [30, 31] (Fig. 11.1). Sensitive bioin-
formatics methods find evidence for potentially important motifs in the tail of
Kif19 [6], but their function has not been examined. In addition to connecting to
microtubules, the Kinesin-8 tail also connects to other binding partners in at least
some members of the family (see below).
The structure of the kinesin motor domain of Kif18A has been solved by X-ray
crystallography and has been analyzed in association with the microtubule by
cryo-electron microscopy [17]. The structure shows canonical features of kinesin
domains, but also unusual properties – in particular a bent, discontinuous α4 helix.
This helix is a central point of contact to the microtubule. Several loops that are
additionally involved in microtubule binding were disordered in the high-resolution
crystal structure, leaving room for speculation on how Kinesin-8 members interact
with microtubules at the molecular level. Like the depolymerizing Kinesin-13 class,
Kinesin-8 proteins possess an elongated loop L2. Yet, the sequence of the loop is
distinct and the two kinesin classes seem to influence microtubule dynamics in a
different manner [20, 21].
Localization and Function of Kif18A and Kif18B
Kif18A has been more intensely studied than Kif18B and Kif19 – probably because
initial RNA-interference (RNAi) experiments in cell culture indicated the strongest
phenotype for knock down of Kif18A [32]. Both Kif18A and Kif18B are expressed
in a cell cycle-specific manner with a high abundance in mitosis [15, 22, 31, 33, 34].
In early mitosis, Kif18A localizes to spindle microtubules and enriches at the
plus-end of kinetochore microtubules [15, 26]. Efficient localization requires
both its own motor activity and microtubule-binding through the tail [4, 5, 26, 29].
The preference of Kif18A to decorate kinetochore-bound microtubules rather than
astral microtubules may be a result of its stronger accumulation on more stable
microtubules [5, 35]. In anaphase, Kif18A seems to accumulate in the spindle
midzone before disappearing [15, 26]. Kif18B localizes to many, but not all astral
microtubule plus-ends [22, 30, 31]. Again, localization requires the motor domain,
but also the link through the tail to EB1 [30, 31].
a b
low Kif18A abundance high Kif18A abundance
long spindle pronounced short spindle suppressed

chromosome chromosome
oscillations oscillations
c d
Kif18A RNAi wild type klp5/klp6 deletion wild type
Fig. 11.2 The level of Kif18A affects functionality of the mitotic spindle. (a) If Kif18A abun-
dance is low, the spindle is elongated and chromosomes show pronounced oscillations with high
amplitude. (b) If Kif18A abundance is high, the spindle becomes shorter and chromosome oscilla-
tions are dampened. (c) RNAi of Kif18A in HeLa cells leads to longer spindles and impaired
chromosome congression. Microtubules are shown in green, chromosomes in blue and CREST
signal (marking centromeres) in red. (d) Deletion of fission yeast klp5 or klp6 (the orthologs of
Kif18A) also results in longer spindles and impaired chromosome congression. Microtubules are
shown in green, chromosomes in blue
The most prominent phenotype after knocking down Kif18A in cultured human
cells is a longer mitotic spindle with chromosomes that make large oscillations from
the metaphase plate (Fig. 11.2) [15, 26, 29, 32, 36]. This is accompanied by a con-
siderable delay in mitosis [15, 26, 32]. Given this prominent cellular phenotype, it
was surprising that Kif18A knock-out mice were viable [23]. Yet, male mice were
sterile, and both mitosis and meiosis in spermatogenesis were strongly perturbed.
This correlates with findings that Kif18A expression is high in testis [23]. In this
regard it is interesting to note that Kif18A abundance often seems high in cancer
([37, 38], Oncomine database), giving rise to the idea that these cancer cells may
similarly depend on Kif18A functionality for proper cell division, a circumstance
which could be exploited in therapy. An amino-acid altering SNP within the Kif18A
gene has been associated with differences in human height [39], but the cellular
phenotype of changing this amino acid has not been examined to our knowledge.
Less is known about Kif18B. Knock-down of Kif18B in cultured cells yields more
and longer astral microtubules, which fits with the localization of Kif18B to astral
microtubules [30, 31]. If and how Kif18B function is important for mammalian
organisms is unclear as far as we know.
Localization and Function of Kif19
Kif19 seems to have a specialized function in some tissues. Like the Kif18A
knock-out mice, Kif19 knock-out mice were viable, but they showed retarded
growth and a shorter life span [9]. The animals suffered from hydrocephalus and
female mice were infertile, which was attributed to an obstruction of the fallopian
tubes. Interestingly, elongated cilia in the Kif19-/- mice were identified as underlying
basis for these defects. The elongated cilia cannot promote proper fluid flow, which can
explain the hydrocephalus and the obstruction of fallopian tubes [9]. The amount of
Kif19 correlated negatively with cilia length – the less Kif19, the longer the cilia
[9]. In agreement with a role for Kif19 in cilia, the protein was found to localize to
the tips of cilia. In vitro experiments suggested that Kif19 regulates cilia length by
influencing the dynamics of ciliary microtubules [9], probably analogous to the way
in which Kif18A and Kif18B influence spindle microtubules. A recent large scale
study revealed that Kif19 has an expression pattern that is clearly distinct from
Kif18A and Kif18B (Fig. 11.3) and that Kif19 mRNA is particularly abundant in
brain tissue [40]. Whether this only reflects its function in motile cilia, or whether
Kif19 has additional functions in brain cells has not been addressed. In cultured
cancer cells, knock-down of Kif19 seemed to slightly increase the frequency of
monopolar spindles [41]. The relevance of this observation for organismal physiology
is unclear.
In contrast to the poor correlation between Kif19 and Kif18A/B mRNA levels
across diverse cell types, the levels of Kif18A and Kif18B mRNA correlate well
(Fig. 11.3) [40]. In addition their expression correlates with expression of the
well-known cell cycle regulator cyclin B1 (Fig. 11.3), which suggests that higher
levels mostly reflect a higher fraction of dividing cells. Whether there is a tissue-
preference beyond this effect is unclear. Additional data on tissue-specific
expression of Kinesin-8 members from low-throughput studies is sparse and the
results are sometimes conflicting [22, 42, 43]. More information would be important
to judge the influence of Kinesin-8 members on organismal physiology and the
potential (side) effects of Kinesin-8 inhibitors.
11.1.1.2 Frequently Studied Kinesin-8 Family Members in Other

Organisms
Kinesin-8 proteins are present across eukaryotes, and their function has often been
studied in organisms other than humans, with the implicit assumption that the
general features are conserved and knowledge gained from other organisms can
be applied to human cell biology. Particularly well characterized are the single
Kinesin-8 member in budding yeast, Kip3p, the two Kinesin-8 members in fission
yeast, Klp5 and Klp6, as well as the Drosophila melanogaster protein Klp67A. All
of these are more related to Kif18A and Kif18B than to Kif19 [44]. Fission yeast
Klp5 and Klp6 form heterodimers rather than homodimers [45, 46], which seems to
be an exception within the family.
a Kif18A vs. Kif19 expression

cerebellum, corpus callosum,
globus pallidus, substantia nigra,
90 thalamus 100
80
70 MS-1 MKL-1
10
60
Kif19
Kif19
50 Y79 1
40
30
20 0.1
10
0 0.01
0 10 20 30 40 50 60 70 80 90 0.01 0.1 1 10 100
Kif18A Kif18A
b Kif18A vs. Kif18B expression
60 100
50
10
40
Kif18B
Kif18B
30 1
20
0.1
10
0 0.01
0 10 20 30 40 50 60 70 80 90 0.01 0.1 1 10 100
Kif18A Kif18A
c Kif18A vs. cyclin B1 (CCNB1) expression

500 1000
400
cyclin B1
cyclin B1
100
300
10
200
1
100
0 0.1
0 10 20 30 40 50 60 70 80 90 0.01 0.1 1 10 100
Kif18A Kif18A
Fig. 11.3 Correlated expression of Kif18A and Kif18B, but different expression pattern of Kif19.
Expression of Kif18A, Kif18B, Kif19, and cyclin B1 (CCNB1) in different tissues and cell lines,
analyzed by cap analysis of gene expression (CAGE) in a recent large-scale study [40]. Panels on
the left and right show the same data on a linear and log10 axis, respectively. (a) Kif19 expression
does not correlate with Kif18A expression and is high in brain tissue. Very high expression of both
Kif18A and Kif19 is only seen in tumor cell lines: MS-1 and MKL-1 are Merkel cell carcinoma
cell lines, Y79 is a retinoblastoma cell line. Kif18B is high in these cell lines as well (not shown).
(b) Expression of Kif18A and Kif18B tends to correlate. (c) Expression of Kif18A also correlates
with expression of cyclin B1 (CCNB1) (All data from http://fantom.gsc.riken.jp/zenbu/)
Despite occasional differences, the findings in these organisms largely parallel

the findings in human cells: these Kinesin-8 proteins localize to the mitotic spindle,
enrich close to the kinetochores in early mitosis and at the spindle midzone in late
mitosis [14, 45, 47–54]. Their depletion causes chromosome congression defects
[45, 48, 51, 53, 55, 56], elongated spindles [45, 48, 51, 53, 55, 56–59] (Fig. 11.2),
and overly stable, long microtubules that are resistant to drug- or cold-induced
depolymerization [52, 54, 55, 60, 61]. Furthermore, in both yeast and Drosophila, a
delay in mitosis was observed when Kinesin-8 activity was reduced [45, 48, 53, 55,
61]. There is also evidence from Drosophila that male meiosis is particularly sensi-
tive to Kinesin-8 levels [55], and work in fission yeast has indicated that meiosis is
more sensitive to the loss of Kinesin-8 function than mitosis [54]. Hence, the
assumption that many of the findings for these Kinesin-8 molecules can be extrapo-
lated to human cell biology seems justified.
11.1.2 Molecular and Cellular Functions of Kinesin-8
11.1.2.1 Influence of Kinesin-8 Members on Microtubules
Kinesin-8 proteins have a unique combination of features within the kinesin class
(Fig. 11.1), which makes it intriguing to understand their mechanism of action on a
molecular level. Two major lines of work have contributed to elucidating the func-
tion of Kinesin-8: on the one hand, elaborate in vitro assays have been used to study
the relationship between Kinesin-8 molecules and microtubules. On the other hand,
researchers have analyzed the consequences of knocking down or overexpressing
Kinesin-8 in cells.
Although there is ample evidence that Kinesin-8 members influence microtubule
plus-end dynamics, it is debated whether all members of the family use the same
underlying mechanism(s) [3]. Early work with budding yeast Kip3p found that the
protein depolymerizes GMP-CPP stabilized microtubules in vitro [7, 14] and that
this occurs in a length-dependent fashion with depolymerization of long microtu-
bules being faster than depolymerization of short ones [7]. Because of Kip3p’s
exceptional processivity (with run lengths of up to 12 µm [6, 7]), motors that land
on the microtubule typically reach the plus-end. According to the so-called antenna
model [7], long microtubules therefore may collect more motor molecules than
short ones, possibly explaining the stronger depolymerization activity on long
microtubules. How, on a molecular level, Kip3p promotes depolymerization is not
entirely clear. One study [20] suggested an interesting cooperative mechanism,
where depolymerization activity depends on the flux of Kinesin-8 molecules along
the microtubule. Single Kinesin-8 molecules pause at the microtubule plus-end for
prolonged periods of time [7, 18, 20], and the model proposes that additionally
incoming molecules may ‘kick-off’ the molecule sitting at the end, together with
one or two tubulin dimers, thereby causing shortening in a manner that depends on
Kip3p flux [20]. This is entirely different from the processive depolymerization
activity of Kinesin-13 proteins [62].
GMP-CPP stabilized microtubules are thought to mimic the GTP-cap of growing
microtubules [3]. Inside cells, the depolymerization activity that was observed on
stabilized microtubules in vitro should therefore translate into a removal of the
GTP-cap, which will trigger rapid microtubule depolymerization (catastrophe).

Indeed, Kip3p seems to increase catastrophe rate in vivo [14] and there is evidence
that it does so in a length-dependent manner [63]. It has been pointed out, though,
that the length-dependence may only become manifest on overly long microtubules
[3]. In addition, the cellular effects on microtubule dynamics are clearly more
complex than that. Kip3p also slows the microtubule shortening rate, increases the
frequency of rescue and increases pause time in vivo [14] – although not all studies
see the same effects [52]. In vitro, Kip3p slows microtubule growth in a length-
dependent manner [21]. To reconcile all these observations, it has been proposed
that high concentrations of Kip3p at the plus-end may slow growth and promote
catastrophe, whereas lower concentrations along the microtubule may slow the
microtubule shortening rate and promote rescue [3].
Extending the observations to other organisms, yields a yet more complex
picture. For fission yeast Klp5/Klp6, depolymerization activity on stabilized micro-
tubules was not detected in vitro [16, 19]. Yet, Klp5/Klp6 increase catastrophe rate
in vivo [64], apparently even in a length-dependent manner [65]. Like for Kip3p, the
effects of depleting Klp5/Klp6 in vivo are complex [64]. Overall, Klp5/Klp6 seems
to increase both the catastrophe and the rescue frequency [64].
In human cells, there is controversy whether Kif18A has the capacity to
depolymerize stabilized microtubules in vitro. Some studies observed (even
length-dependent) depolymerization activity [15, 17], whereas others did not [18].
Taxol-stabilized microtubules are shortened by Kif18A in vivo [35], but no effect
was seen in vitro [18]. The overall in vivo effect of Kif18A seems to be a dampening
of microtubule dynamics [15, 18, 26, 66]. Although several studies reported larger
chromosome oscillations on the metaphase plate in Kif18A RNAi, even the precise
dynamic of these oscillations is debated [15, 26, 36, 66]. Kif19 depolymerized
GMP-CPP stabilized microtubules from the plus-end in vitro [9], analogous to what
has been observed for Kip3p and sometimes for Kif18A. Since the microtubule
“skeleton” of cilia, the axoneme, is very stable, it is unclear when Kif19 exerts its
effect to regulate the length of cilia – this may happen during a nascent state when
cilia are formed [9].
In summary, the effects of perturbing Kinesin-8 function in vivo are complex and
difficult to interpret – probably partly because the microtubule state influences the
localization and action of Kinesin-8 members like Kif18A, which in turn affect the
microtubule state. As an example, the intensity of Kif18A was found to differ
between the two sister kinetochores of one chromosome, which are in different
states (growing and shrinking) [26], and hence the effect of Kif18A on these
kinetochore-microtubule bundles may differ [26, 36]. Furthermore, Kinesin-8
members interact with and influence additional microtubule regulators (see below),
which will cause a “mixed” phenotype after Kinesin-8 depletion that partially
reflects Kinesin-8’s own effect on microtubules and partially reflects the effect of its
interactors. In such a complex situation, it is not surprising that phenotypes are
difficult to decipher. The discrepancies that have been found between – or even
within – organisms could be true biological differences, could result from the
precise circumstances (e.g. cell cycle state or post-translational modifications) or

could be due to technical differences (e.g. extent of depletion, imaging parameters)
in the often demanding assays being used.
11.1.2.2 Both Too Little and Too Much Kif18A Cause

Cellular Phenotypes
Interestingly, cellular phenotypes are induced by any change in Kif18A abundance,

be it up or down (Fig. 11.2). This is interesting with respect to cancer cells: higher
levels of Kif18A seem to be frequent in cancer [37, 38], and lower levels have been
observed in artificially induced, tetraploid cells (Storchova Z, personal communication),
which are considered one possible intermediate to the development of aneuploid
cancer cells (Fig. 11.4). Hence, both perturbations may play a role in cancer.
healthy, euploid cell

Kif18A
normal
Fig. 11.4 Altered Kif18A

levels in induced tetraploidy
and in cancer cells. Artificial
induction of tetraploidy leads tetraploidization event other routes to aneuploidy
to surviving clones with low
Kif18A abundance. This low
abundance may facilitate the
formation of larger spindles Kif18A
that can accommodate all the normal
chromosomes. Although
tetraploidy is thought to be aneuploid cancer cell
one possible intermediate Kif18A
through which chromosomal high
instability and aneuploidy
can be achieved, aneuploid
cancer cells tend to have high
levels of Kif18A. The high
low Kif18A may help formation
abundance of Kif18A could of larger spindle and viability Kif18A
facilitate chromosomal high
instability or invasive
potential through modulating
microtubule dynamics. When Kif18A
during the process of cancer low
development and how
Kif18A levels are regulated is Kif18A
little understood. One study high
indicated chromatin
modifications that may
underlie Kif18A/Kif18B
expression changes [108]
The effects of Kif18A overexpression are largely opposite to those of Kif18A

depletion. Whereas Kif18A depletion leads to longer spindles with badly aligned
chromosomes that show large oscillations on the metaphase plate [15, 26, 32, 36],
higher levels of Kif18A lead to shorter spindles with suppressed chromosome
oscillations [26]. The effect of Kif18A abundance can even be observed within one
spindle: kinetochore microtubules in the periphery of the spindle have higher
Kif18A abundance and smaller oscillations, whereas those in the center of the spin-
dle have lower Kif18A abundance and stronger oscillations [26]. The long spindles
occurring at reduced Kif18A levels may be beneficial in the larger tetraploid cells
(see below). The potential benefit of higher Kif18A levels in aneuploid cancer cells
is more difficult to understand. The overexpression of Kif18A seems to dampen
microtubule dynamics [18, 26]. A reduced turn-over of kinetochore-microtubules
by different means has been shown to favor chromosomal instability [67, 68].
Whether Kif18A overexpression leads to chromosome missegregation by affecting
microtubule dynamics has not been addressed. In one study, overexpression of
Kif18A was observed to cause multipolar spindles [18], which is a phenotype that
may promote aneuploidy and chromosomal instability [69, 70].
11.1.2.3 Interaction of Kinesin-8 Members with Proteins

Other Than Microtubules
Despite ample evidence that Kif18A and related Kinesin-8 proteins can influence
microtubule dynamics by themselves (based on in vitro assays with purified com-
ponents), there is also evidence that Kif18A binds to other microtubule interactors
or microtubule regulators and therefore may influence microtubule dynamics at
least partially in an indirect fashion. Kif18A has been reported to interact with the
kinesin-7 protein CENP-E and the spindle assembly checkpoint protein and micro-
tubule regulator BubR1, and their levels were slightly reduced when Kif18A was
depleted [23, 71]. Impairment of CENP-E function is known to cause chromosome
congression defects [72, 73]. Although Kif18A depletion causes chromosomes to
show large oscillations from the metaphase plate, which gives the appearance of a
congression defect, most chromosomes nevertheless seem bi-oriented [15, 26],
which is not the case in CENP-E depletion. A considerable reduction of BubR1
would be expected to cause the mitotic checkpoint to malfunction [74]. Kif18A-
depleted cells are delayed in mitosis in a checkpoint-dependent manner [15], and
therefore BubR1 functionality still seems good enough to allow this. Nevertheless,
the reduction of CENP-E and BubR1 may contribute to the Kif18A-depletion phe-
notype to some extent. Reduction of Kif18A furthermore reduces the level of astrin
at metaphase kinetochores [75]. This correlates with increased kinetochore levels
of the Kinesin-13 Kif2b, which promotes kinetochore-microtubule turnover [68].
Thus, some of the effects of Kif18A RNAi might result from elevated levels of
Kif2b at kinetochores. Indeed, co-depletion of Kif2b has been observed to dampen
the effect of Kif18A RNAi on chromosome congression and mitotic progression
[75]. Kif18A also seems to interact with the microtubule-associated protein HURP
[76]. Overexpression of a fragment of HURP leads to a mal-localization of Kif18A

on microtubules with less enrichment of Kif18A at microtubule plus-ends, presum-
ably because HURP impairs Kif18A’s walking ability [76]. As would be expected,
these cells show a Kif18A-RNAi-like phenotype [76]. Knock-down of HURP,
however, affected Kif18A localization only slightly (with a more focused rather
than gradient-like appearance at kinetochore-microtubule plus-ends). The effects
of loss of HURP or loss of its interaction with Kif18A on Kif18A functionality
have not been tested, and it therefore remains unclear if and how HURP contributes
to Kif18A function.
Kif18B was found to interact not only with EB1 [30, 31], but also with the
microtubule-depolymerizing Kinesin-13 MCAK [31]. By engineering kinesin chi-
meras the authors concluded that Kif18B executes its microtubule-depolymerizing
effect through MCAK and uses its own kinesin domain only for targeting. Kinesin-8
molecules from other organisms also have functionally relevant interaction partners:
budding yeast Kip3p seems to transport the microtubule-binding protein Stu2p
(the ortholog of X.l. XMAP215 and H.s. ch-TOG/CKAP5) to microtubule plus-ends,
where it can promote microtubule rescue in order to facilitate capture of chromo-
somes in early mitosis [63]. Interestingly, in anaphase, Kip3p and Stu2p then act as
opponents in spindle elongation [77].
The tail of fission yeast Klp5 and Klp6 was shown to interact with the protein
phosphatase PP1, which helps silencing of the spindle assembly checkpoint [78].
Although a PP1-binding motif also exists in human Kif18A and interaction between
Kif18A and the PP1 alpha catalytic subunit PP1CA has been observed in a yeast
two-hybrid assay [79], the function of this motif has not been tested. Furthermore,
the little characterized fission yeast protein Mcp1 – a coiled-coil protein with no
obvious orthologs outside fission yeast – depends on Klp6 for its microtubule plus-end
localization and its depletion at least partially mimics a Klp5/Klp6 depletion [80].
11.2 Kinesin-8 in Cancer
11.2.1 Expression of Kinesin-8 in Cancer Cells
The analyses of the cancer genome landscape revealed that in common solid tumors
about 33–66 genes have non-silent mutations most of which are single base substi-
tutions [81]. According to their impact on tumor growth, they are classified as driver
gene mutation if they confer a selective growth advantage to the tumor cells or as
passenger gene mutation in case they do not actively promote tumor growth. Driver
gene mutations in Kinesin-8 family members are not known. However, there is an
increasing body of evidence that Kif18A is overexpressed in solid tumors and that
its overexpression correlates with tumor progression. Analyses of 113 surgical spec-
imen of colorectal cancer (CRC) revealed that Kif18A was highly overexpressed in
cancer cells compared to control tissue and its expression level was significantly
correlated with primary tumor stage as well as lymph node metastasis [37]. Further
analyses revealed that expression levels of Kif18A were not a suitable prognostic
marker for relapse-free survival, however, they could be used as a significant prog-
nostic factor for 5-year overall survival. Similar results were obtained by analyzing
the expression profile of Kif18A in human breast cancer cells using the publicly
available ONCOMINE database [38]. Like in CRC, increased Kif18A mRNA
expression strongly correlated with tumor grade, metastasis and poor survival.
Collectively, these findings suggest that Kif18A protein levels are increased in
certain tumor cells.
Since Kif18A accumulates during mitosis [15], the question arises whether
increased levels of Kif18A are a consequence of a higher percentage of dividing
cells within these tumor tissues compared to normal tissues, or whether individual
tumor cells indeed express Kif18A at elevated levels. If the former applies, increased
Kif18A levels would be a simple side effect of the increased proliferative activity of
tumor cells. If the latter is the case, this might indicate that increased Kif18A levels
are actually beneficial for tumor growth and progression, which would make Kif18A
an attractive target for therapy. To address this aspect, Zhang and colleagues per-
formed xenograft assays in nude mice using the breast cancer cell line MDA-MB-231
[38]. These studies revealed that long-term RNAi-mediated suppression of Kif18A
expression reduced tumor formation and tumor growth compared to control cells.
Similarly, compared to wildtype mice, Kif18A knock-out mice seemed to be less
prone to develop dysplastic lesions in chemically induced colonic neoplasia [82].
These observations are not particularly surprising given Kif18A’s critical function
in mitotic divisions [15, 32]. Intriguingly, however, assays with the human breast
cancer cell line MCF-7 (Michigan Cancer Foundation-7) revealed that the depletion
of Kif18A reduced the migration rates of these cells [38]. Similarly, the ability to
migrate and invade was reduced in the colorectal cancer cell line CaR-1 upon
Kif18A depletion, whereas the overexpression of active Kif18A positively affected
both activities [37]. What could be the reason for the reduced migration and inva-
sion ability of cancer cell lines in the absence of Kif18A? Measurements of micro-
tubule plus-end dynamics in MCF-7 cells using GFP-EB1 as marker revealed that
depletion of Kif18A increases the dynamic behavior of interphase microtubules
[38]. Since stabilization of microtubule plus-ends within the leading edge of cells is
important for cell migration and proliferation [83], these observations indicate that
more dynamic microtubules in the absence of Kif18A might account for the
observed migration and invasion phenotypes and, potentially, for the reduced tumor
growth in mice when Kif18A was depleted from MDA-MB-231 cells. The effect of
Kif18A depletion on the dynamic behavior of interphase microtubules is surprising
given that Kif18A seems to localize primarily to the nucleus in interphasic HeLa
cells [15] and, consequently, Du et al. had to engineer a mutant Kif18A with a
defective nuclear localization signal (NLS) to analyze the effect of Kif18A on the
interphasic microtubule cytoskeleton [18]. These findings suggest that the subcel-
lular localization of Kif18A varies depending on the analyzed cancer cell line, for
which there is some evidence [43]. Further research is required to dissect a potential
function of Kif18A in regulating the dynamics of interphase microtubules.
11.2.2 Kif18A as Predictive Marker
The analysis of expression profiles has been a valuable approach to classify cancer
patients into risk groups, predicting the outcome of the cancer and supporting the
decision regarding the appropriate therapeutic strategy. In fact, the so-called
“70-gene”-signature which was assembled based on the correlation of gene expres-
sion profiles with the clinical outcome of breast cancer patients outperforms most
standard clinical parameters in predicting disease outcome [84]. Driven by this find-
ing, immense efforts have been undertaken to define the best possible signature for
a given cancer. By identifying genes essential for error-free mitotic divisions in
Drosophila melanogaster and analyzing the expression profile of their human
orthologues in breast cancer, Damasco et al. generated a novel so-called “Drosophila
mitotic” (DM) signature which shares very few genes with pre-existing signatures
and outperforms most of the well-established signatures, such as the “70-gene”
signature, in its predictive power [85]. Human Kif18A is part of the “DM”-signature
and, importantly, it clearly contributes to its predictive power. There is also some
evidence that Kif18A may serve as plasma biomarker in chemically-induced
cholangiocarcinoma [86]. Similarly, proteomic analysis of serum samples identified
Kif18A as a potential blood biomarker to identify asbestosis patients at risk of
developing lung cancer [87], but a follow-up study calls this result into question [88].
11.2.3 Therapeutic Potential of Interfering

with Kif18A Function
While the number of available diagnostic methods constantly increases, the thera-
peutic options for the treatment of cancer patients are limited and, consequently,
there is an urgent need for novel therapeutic strategies. A promising route in the
treatment of cancer is dendritic cell (DC) based immunotherapy where DCs loaded
with tumor-associated antigens (TAAs) are used to trigger a cancer-specific immune
response. One challenge of cancer immunotherapy is to identify the ideal TAA that
is strongly expressed by cancer cells but hardly present in non-cancer cells and that
contains peptide sequences that bind to major histocompatibility molecules recog-
nized by cytotoxic T lymphocytes (CTLs). The search for novel TAAs identified
Kif18A as a specific tumor-associated antigen [89]. Two Kif18A peptides were able
to induce cancer-reactive CTLs from peripheral blood mononuclear cells of colon
cancer patients (n = 6). Intriguingly, the peptides had no effect on peripheral blood
mononuclear cells derived from healthy donors (n = 5). Based on these observations,
the authors add Kif18A to the growing list of TAAs that might be an appropriate
target for specific immunotherapy in colon cancer patients. Clearly, future studies
are required to investigate if this promise holds true.
Numerical and structural chromosome abnormalities are a hallmark of cancer
and, therefore, represent attractive therapeutic targets. Recent studies addressing the
mechanism underlying numerical aberrations proposed that tetraploidy serves as a

transient intermediate in the development of aneuploid cells [90, 91] and indeed, a
causal link between tetraploidy, chromosomal instability and tumorigenesis has
been experimentally demonstrated in model systems [92, 93]. Intriguingly, tran-
scriptome analyses of artificially created tetraploid HCT-116 cells revealed that the
expression of Kif18A was significantly down-regulated compared to the parental
diploid cell line (Storchova Z, personal communication). The down-regulation of
Kif18A in these tetraploid cells contrasts its up-regulation in many cancer cells and,
therefore, future studies are required to better understand the route from diploid
cells via tetraploidy to aneuploidy and tumorigenesis (Fig. 11.4). One possible
hypothesis could be that reduced Kif18A levels, which result in increased spindle
length [15, 26, 29, 32], are beneficial for the assembly of a mitotic spindle that
has to accommodate the double amount of chromosomes, whereas subsequently
reduced levels of Kif18A might promote aneuploidy by enhancing chromosomal
instability.
11.2.3.1 Targeting Kif18A by Small Molecule Inhibitors
The search for more specific chemotherapeutic agents with fewer side effects on
non-dividing cells, such as the peripheral neuropathy that is often associated with
tubulin-targeting antimitotic agents, drew the attention to mitotic kinesins [94].
In particular, the tetrameric kinesin Eg5 came into the limelight as an attractive
anti-cancer target because of its key function in the assembly of a bipolar spindle.
Currently, multiple clinical trials are testing the efficacy of chemically diverse
Eg5-targeting agents for the treatment of cancer – with somewhat sobering results
so far [94]. To our knowledge, no clinical trials on Kif18A inhibitors are currently
performed. This might be due to the limited availability of appropriate inhibitors.
The only published Kif18A inhibitor, termed BTB-1, was identified by screening a
9,000-member pharmacophore library for small molecules inhibiting the ATPase
activity of the recombinant motor domain of Kif18A [95]. BTB-1 inhibits Kif18A
in a reversible manner with an IC50-value of about 1.7 µM. Since BTB-1 was not
optimized in respect to its inhibitor efficacy, it is likely that further small molecule
screens guided by BTB-1 structure activity relationship analyses will yield more
potent Kif18A inhibitors. BTB-1’s mode of inhibition is intriguing in that it is com-
petitive with ATP and uncompetitive with microtubules, i.e. BTB-1 competes with
ATP for Kif18A binding only when the motor-protein is associated with its pseudo-
substrate microtubules. HeLa cells treated with BTB-1 display a mitotic delay [95],
a phenotype consistent with the mitotic function of Kif18A [15, 26]. Yet, these
mitotically arrested cells do not display elongated spindles with curvy and bent
microtubules as it has been reported for the Kif18A-RNAi phenotype [15]. A trivial
explanation could be that Kif18A is not the relevant target of BTB-1 in cells.
To address this point, a BTB1-resistant cell line would have to be generated followed
by the identification of the mutated gene conferring the resistance. This approach
has been elegantly used to confirm that the kinesin Eg5 is the target of monastrol
and STLC [96] and that the kinase Aurora B is the cellular target of the drug
ZM447439 [97]. Alternatively, it is possible that Kif18A is indeed the major target
of BTB-1 in vivo but that the drug-induced and RNAi-mediated phenotypes do not
match, because the phenotype of removing the protein is different from having the
protein present in an inhibited form [98]. This may well be the case for Kif18A:
depletion of the protein is expected to affect levels and localization of its binding
partners [23, 71, 75], whereas inhibition may have a milder effect in this regard.
Having an inactive Kif18A present on microtubules will also affect microtubules in
a different way than the absence of Kif18A [26]. Thus, it is unlikely that inhibition
of Kif18A and its depletion from the cellular environment result in the same pheno-
type and, therefore, the open question is which cellular phenotype should be
expected from BTB-1. Clearly, the validation of Kif18A as the relevant cellular
target of BTB-1 merits further investigation.
11.3 Outlook
Irrespective of the question surrounding the cellular effect of BTB-1, the concept of
targeting mitosis as a powerful approach to kill cancer cells has recently been ques-
tioned [99]. The basic principle of this concept was that cancer cells that divide
rapidly are highly vulnerable to cell division targeting compounds such as microtu-
bule drugs. Since microtubules are not only fundamental for cell division but also
for a plethora of processes taking place in non-dividing cells, the logic conclusion
of this concept was that compounds that specifically target mitosis might have less
severe side effects than microtubule poisons [100]. While this concept intrigues by
its simplicity and has laid the foundation for enormous research and screening
efforts in traditional pharmaceutical companies as well as in newly founded start-up
companies, the anticancer activity of mitosis-targeting compounds, such as Eg5-
inhibitors, did not meet the high expectations [101–103]. Yet, it is beyond dispute
that paclitaxel and other microtubule drugs are powerful agents for treating certain
types of epithelial cancer, e.g. breast and ovarian cancer [104]. Thus, the success of
microtubule drugs and the limited efficacy of Eg5-inhibitors combined with the fact
that certain cancer types, which are affected by microtubule drugs, proliferate rather
slowly – the median proliferation rates of breast cancer cells seem to be lower than
that of bone marrow or the gut [105–107] – raises the question if mitosis is indeed
the Achilles’ heel by which microtubule drugs kill cancer cells [99, 104]. Instead,
interphasic effects could be more important than mitotic ones. Yet, since there is
evidence that increased levels of Kif18A might be beneficial for two activities
characteristic for metastasis – migration and invasion – the further development of
highly potent and specific Kif18A inhibitors appears worthwhile to pursue.
References
1. Lawrence CJ, Dawe RK, Christie KR, Cleveland DW, Dawson SC, Endow SA, Goldstein LS,
Goodson HV, Hirokawa N, Howard J, Malmberg RL, McIntosh JR, Miki H, Mitchison TJ,
Okada Y, Reddy AS, Saxton WM, Schliwa M, Scholey JM, Vale RD, Walczak CE, Wordeman
L (2004) A standardized kinesin nomenclature. J Cell Biol 167(1):19–22. doi:10.1083/
jcb.200408113
ture and function. Trends Cell Biol 15(9):467–476. doi:10.1016/j.tcb.2005.07.006
3. Su X, Ohi R, Pellman D (2012) Move in for the kill: motile microtubule regulators. Trends
Cell Biol 22(11):567–575. doi:10.1016/j.tcb.2012.08.003
4. Mayr MI, Storch M, Howard J, Mayer TU (2011) A non-motor microtubule binding site is
essential for the high processivity and mitotic function of kinesin-8 Kif18A. PLoS One
6(11):e27471. doi:10.1371/journal.pone.0027471
5. Stumpff J, Du Y, English CA, Maliga Z, Wagenbach M, Asbury CL, Wordeman L, Ohi R
(2011) A tethering mechanism controls the processivity and kinetochore-microtubule
plus-end enrichment of the kinesin-8 Kif18A. Mol Cell 43(5):764–775. doi:10.1016/
j.molcel.2011.07.022
6. Su X, Qiu W, Gupta ML Jr, Pereira-Leal JB, Reck-Peterson SL, Pellman D (2011) Mechanisms
underlying the dual-mode regulation of microtubule dynamics by Kip3/kinesin-8. Mol Cell
43(5):751–763. doi:10.1016/j.molcel.2011.06.027
7. Varga V, Helenius J, Tanaka K, Hyman AA, Tanaka TU, Howard J (2006) Yeast kinesin-8
depolymerizes microtubules in a length-dependent manner. Nat Cell Biol 8(9):957–962.
doi:10.1038/ncb1462
8. Gardner MK, Odde DJ, Bloom K (2008) Kinesin-8 molecular motors: putting the brakes on
chromosome oscillations. Trends Cell Biol 18(7):307–310. doi:10.1016/j.tcb.2008.05.003
9. Niwa S, Nakajima K, Miki H, Minato Y, Wang D, Hirokawa N (2012) KIF19A is a
microtubule-depolymerizing kinesin for ciliary length control. Dev Cell 23(6):1167–1175.
doi:10.1016/j.devcel.2012.10.016
10. Su X, Arellano-Santoyo H, Portran D, Gaillard J, Vantard M, Thery M, Pellman D (2013)
Microtubule-sliding activity of a kinesin-8 promotes spindle assembly and spindle-length
control. Nat Cell Biol 15(8):948–957. doi:10.1038/ncb2801
11. Bormuth V, Nitzsche B, Ruhnow F, Mitra A, Storch M, Rammner B, Howard J, Diez S (2012)
The highly processive kinesin-8, Kip3, switches microtubule protofilaments with a bias
toward the left. Biophys J 103(1):L4–L6. doi:10.1016/j.bpj.2012.05.024
12. Severin F, Habermann B, Huffaker T, Hyman T (2001) Stu2 promotes mitotic spindle elonga-
tion in anaphase. J Cell Biol 153(2):435–442
13. Ems-McClung SC, Walczak CE (2010) Kinesin-13s in mitosis: key players in the spatial and
temporal organization of spindle microtubules. Semin Cell Dev Biol 21(3):276–282.
doi:10.1016/j.semcdb.2010.01.016
14. Gupta ML Jr, Carvalho P, Roof DM, Pellman D (2006) Plus end-specific depolymerase
activity of Kip3, a kinesin-8 protein, explains its role in positioning the yeast mitotic spindle.
Nat Cell Biol 8(9):913–923. doi:10.1038/ncb1457
15. Mayr MI, Hummer S, Bormann J, Gruner T, Adio S, Woehlke G, Mayer TU (2007)
The human kinesin Kif18A is a motile microtubule depolymerase essential for chromosome
congression. Curr Biol 17(6):488–498. doi:10.1016/j.cub.2007.02.036
16. Grissom PM, Fiedler T, Grishchuk EL, Nicastro D, West RR, McIntosh JR (2009) Kinesin-8
from fission yeast: a heterodimeric, plus-end-directed motor that can couple microtubule
depolymerization to cargo movement. Mol Biol Cell 20(3):963–972. doi:10.1091/mbc.
E08-09-0979
17. Peters C, Brejc K, Belmont L, Bodey AJ, Lee Y, Yu M, Guo J, Sakowicz R, Hartman J,
Moores CA (2010) Insight into the molecular mechanism of the multitasking kinesin-8
motor. EMBO J 29(20):3437–3447. doi:10.1038/emboj.2010.220
18. Du Y, English CA, Ohi R (2010) The kinesin-8 Kif18A dampens microtubule plus-end
dynamics. Curr Biol 20(4):374–380. doi:10.1016/j.cub.2009.12.049
19. Erent M, Drummond DR, Cross RA (2012) S. pombe kinesins-8 promote both nucleation and
catastrophe of microtubules. PLoS One 7(2):e30738. doi:10.1371/journal.pone.0030738
20. Varga V, Leduc C, Bormuth V, Diez S, Howard J (2009) Kinesin-8 motors act cooperatively
to mediate length-dependent microtubule depolymerization. Cell 138(6):1174–1183.
doi:10.1016/j.cell.2009.07.032
21. Gardner MK, Zanic M, Gell C, Bormuth V, Howard J (2011) Depolymerizing kinesins Kip3
and MCAK shape cellular microtubule architecture by differential control of catastrophe.
Cell 147(5):1092–1103. doi:10.1016/j.cell.2011.10.037
22. Lee YM, Kim E, Park M, Moon E, Ahn SM, Kim W, Hwang KB, Kim YK, Choi W, Kim W
(2010) Cell cycle-regulated expression and subcellular localization of a kinesin-8 member
human KIF18B. Gene 466(1–2):16–25. doi:10.1016/j.gene.2010.06.007
23. Liu XS, Zhao XD, Wang X, Yao YX, Zhang LL, Shu RZ, Ren WH, Huang Y, Huang L,
Gu MM, Kuang Y, Wang L, Lu SY, Chi J, Fen JS, Wang YF, Fei J, Dai W, Wang ZG
(2010) Germinal cell aplasia in Kif18a mutant male mice due to impaired chromosome
congression and dysregulated BubR1 and CENP-E. Genes Cancer 1(1):26–39.
doi:10.1177/1947601909358184
24. Luboshits G, Benayahu D (2005) MS-KIF18A, new kinesin; structure and cellular expres-
sion. Gene 351:19–28. doi:10.1016/j.gene.2005.02.009
25. Flicek P, Amode MR, Barrell D, Beal K, Billis K, Brent S, Carvalho-Silva D, Clapham P,
Coates G, Fitzgerald S, Gil L, Giron CG, Gordon L, Hourlier T, Hunt S, Johnson N,
Juettemann T, Kahari AK, Keenan S, Kulesha E, Martin FJ, Maurel T, McLaren WM,
Murphy DN, Nag R, Overduin B, Pignatelli M, Pritchard B, Pritchard E, Riat HS, Ruffier M,
Sheppard D, Taylor K, Thormann A, Trevanion SJ, Vullo A, Wilder SP, Wilson M, Zadissa A,
Aken BL, Birney E, Cunningham F, Harrow J, Herrero J, Hubbard TJ, Kinsella R, Muffato
M, Parker A, Spudich G, Yates A, Zerbino DR, Searle SM (2014) Ensembl 2014. Nucleic
Acids Res 42(1):D749–D755. doi:10.1093/nar/gkt1196
26. Stumpff J, von Dassow G, Wagenbach M, Asbury C, Wordeman L (2008) The kinesin-8
motor Kif18A suppresses kinetochore movements to control mitotic chromosome alignment.
Dev Cell 14(2):252–262. doi:10.1016/j.devcel.2007.11.014
27. Weaver LN, Walczak C (2011) Kinesin-8s hang on by a tail. Bioarchitecture 1(5):236–239.
doi:10.4161/bioa.18427
28. Jannasch A, Bormuth V, Storch M, Howard J, Schaffer E (2013) Kinesin-8 is a low-force
motor protein with a weakly bound slip state. Biophys J 104(11):2456–2464. doi:10.1016/
j.bpj.2013.02.040
29. Weaver LN, Ems-McClung SC, Stout JR, LeBlanc C, Shaw SL, Gardner MK, Walczak CE
(2011) Kif18A uses a microtubule binding site in the tail for plus-end localization and spindle
length regulation. Curr Biol 21(17):1500–1506. doi:10.1016/j.cub.2011.08.005
30. Stout JR, Yount AL, Powers JA, Leblanc C, Ems-McClung SC, Walczak CE (2011) Kif18B
interacts with EB1 and controls astral microtubule length during mitosis. Mol Biol Cell
22(17):3070–3080. doi:10.1091/mbc.E11-04-0363
31. Tanenbaum ME, Macurek L, van der Vaart B, Galli M, Akhmanova A, Medema RH (2011)
A complex of Kif18b and MCAK promotes microtubule depolymerization and is negatively
regulated by Aurora kinases. Curr Biol 21(16):1356–1365. doi:10.1016/j.cub.2011.07.017
32. Zhu C, Zhao J, Bibikova M, Leverson JD, Bossy-Wetzel E, Fan JB, Abraham RT, Jiang W
(2005) Functional analysis of human microtubule-based motor proteins, the kinesins and
dyneins, in mitosis/cytokinesis using RNA interference. Mol Biol Cell 16(7):3187–3199.
doi:10.1091/mbc.E05-02-0167
33. Sedgwick GG, Hayward DG, Di Fiore B, Pardo M, Yu L, Pines J, Nilsson J (2013)
Mechanisms controlling the temporal degradation of Nek2A and Kif18A by the APC/
C-Cdc20 complex. EMBO J 32(2):303–314. doi:10.1038/emboj.2012.335
34. Singh SA, Winter D, Kirchner M, Chauhan R, Ahmed S, Ozlu N, Tzur A, Steen JA, Steen H
(2014) Co-regulation proteomics reveals substrates and mechanisms of APC/C-dependent
degradation. EMBO J 33(4):385–399. doi:10.1002/embj.201385876
35. Masuda N, Shimodaira T, Shiu SJ, Tokai-Nishizumi N, Yamamoto T, Ohsugi M (2011)
Microtubule stabilization triggers the plus-end accumulation of Kif18A/kinesin-8. Cell Struct
Funct 36(2):261–267
36. Jaqaman K, King EM, Amaro AC, Winter JR, Dorn JF, Elliott HL, McHedlishvili N,
McClelland SE, Porter IM, Posch M, Toso A, Danuser G, McAinsh AD, Meraldi P, Swedlow
JR (2010) Kinetochore alignment within the metaphase plate is regulated by centromere stiff-
ness and microtubule depolymerases. J Cell Biol 188(5):665–679. doi:10.1083/jcb.200909005
37. Nagahara M, Nishida N, Iwatsuki M, Ishimaru S, Mimori K, Tanaka F, Nakagawa T, Sato T,
Sugihara K, Hoon DS, Mori M (2011) Kinesin 18A expression: clinical relevance to colorec-
tal cancer progression. Int J Cancer 129(11):2543–2552. doi:10.1002/ijc.25916
38. Zhang C, Zhu C, Chen H, Li L, Guo L, Jiang W, Lu SH (2010) Kif18A is involved in human
breast carcinogenesis. Carcinogenesis 31(9):1676–1684. doi:10.1093/carcin/bgq134
39. Kim JJ, Park YM, Baik KH, Choi HY, Yang GS, Koh I, Hwang JA, Lee J, Lee YS, Rhee H,
Kwon TS, Han BG, Heath KE, Inoue H, Yoo HW, Park K, Lee JK (2012) Exome sequencing
and subsequent association studies identify five amino acid-altering variants influencing
human height. Hum Genet 131(3):471–478. doi:10.1007/s00439-011-1096-4
40. FANTOM Consortium and the RIKEN PMI and CLST (DGT) (2014) A promoter-level
mammalian expression atlas. Nature 507(7493):462–470. doi:10.1038/nature13182
41. Tanenbaum ME, Macurek L, Janssen A, Geers EF, Alvarez-Fernandez M, Medema RH
(2009) Kif15 cooperates with eg5 to promote bipolar spindle assembly. Curr Biol
19(20):1703–1711. doi:10.1016/j.cub.2009.08.027
genes in mouse and human. Proc Natl Acad Sci U S A 98(13):7004–7011. doi:10.1073/
pnas.111145398
43. Uhlen M, Oksvold P, Fagerberg L, Lundberg E, Jonasson K, Forsberg M, Zwahlen M, Kampf C,
Wester K, Hober S, Wernerus H, Bjorling L, Ponten F (2010) Towards a knowledge-based
human protein atlas. Nat Biotechnol 28(12):1248–1250. doi:10.1038/nbt1210-1248
44. Wickstead B, Gull K (2006) A “holistic” kinesin phylogeny reveals new kinesin families and
predicts protein functions. Mol Biol Cell 17(4):1734–1743. doi:10.1091/mbc.E05-11-1090
45. Garcia MA, Koonrugsa N, Toda T (2002) Two kinesin-like Kin I family proteins in fission
yeast regulate the establishment of metaphase and the onset of anaphase A. Curr Biol
12(8):610–621
46. Li Y, Chang EC (2003) Schizosaccharomyces pombe Ras1 effector, Scd1, interacts with Klp5
and Klp6 kinesins to mediate cytokinesis. Genetics 165(2):477–488
47. DeZwaan TM, Ellingson E, Pellman D, Roof DM (1997) Kinesin-related KIP3 of
Saccharomyces cerevisiae is required for a distinct step in nuclear migration. J Cell Biol
138(5):1023–1040
48. Goshima G, Vale RD (2003) The roles of microtubule-based motor proteins in mitosis: com-
prehensive RNAi analysis in the Drosophila S2 cell line. J Cell Biol 162(6):1003–1016.
doi:10.1083/jcb.200303022
49. Goshima G, Vale RD (2005) Cell cycle-dependent dynamics and regulation of mitotic kinesins
in Drosophila S2 cells. Mol Biol Cell 16(8):3896–3907. doi:10.1091/mbc.E05-02-0118
50. Miller RK, Heller KK, Frisen L, Wallack DL, Loayza D, Gammie AE, Rose MD (1998) The
kinesin-related proteins, Kip2p and Kip3p, function differently in nuclear migration in yeast.
Mol Biol Cell 9(8):2051–2068
51. Savoian MS, Gatt MK, Riparbelli MG, Callaini G, Glover DM (2004) Drosophila Klp67A is
required for proper chromosome congression and segregation during meiosis I. J Cell Sci
117(Pt 16):3669–3677. doi:10.1242/jcs.01213
52. Tytell JD, Sorger PK (2006) Analysis of kinesin motor function at budding yeast kineto-
chores. J Cell Biol 172(6):861–874. doi:10.1083/jcb.200509101
53. West RR, Malmstrom T, McIntosh JR (2002) Kinesins klp5(+) and klp6(+) are required for
normal chromosome movement in mitosis. J Cell Sci 115(Pt 5):931–940
54. West RR, Malmstrom T, Troxell CL, McIntosh JR (2001) Two related kinesins, klp5+ and
klp6+, foster microtubule disassembly and are required for meiosis in fission yeast. Mol Biol
Cell 12(12):3919–3932
55. Gandhi R, Bonaccorsi S, Wentworth D, Doxsey S, Gatti M, Pereira A (2004) The Drosophila
kinesin-like protein KLP67A is essential for mitotic and male meiotic spindle assembly. Mol
Biol Cell 15(1):121–131. doi:10.1091/mbc.E03-05-0342
56. Wargacki MM, Tay JC, Muller EG, Asbury CL, Davis TN (2010) Kip3, the yeast kinesin-8,
is required for clustering of kinetochores at metaphase. Cell Cycle 9(13):2581–2588
57. Gardner MK, Bouck DC, Paliulis LV, Meehl JB, O’Toole ET, Haase J, Soubry A, Joglekar
AP, Winey M, Salmon ED, Bloom K, Odde DJ (2008) Chromosome congression by Kinesin-5
motor-mediated disassembly of longer kinetochore microtubules. Cell 135(5):894–906.
doi:10.1016/j.cell.2008.09.046
58. Straight AF, Sedat JW, Murray AW (1998) Time-lapse microscopy reveals unique roles for
kinesins during anaphase in budding yeast. J Cell Biol 143(3):687–694
59. Syrovatkina V, Fu C, Tran PT (2013) Antagonistic spindle motors and MAPs regulate meta-
phase spindle length and chromosome segregation. Curr Biol 23(23):2423–2429.
doi:10.1016/j.cub.2013.10.023
60. Cottingham FR, Hoyt MA (1997) Mitotic spindle positioning in Saccharomyces cerevisiae is
accomplished by antagonistically acting microtubule motor proteins. J Cell Biol
138(5):1041–1053
61. Rischitor PE, Konzack S, Fischer R (2004) The Kip3-like kinesin KipB moves along micro-
tubules and determines spindle position during synchronized mitoses in Aspergillus nidulans
hyphae. Eukaryot Cell 3(3):632–645. doi:10.1128/EC.3.3.632-645.2004
destabilizing enzymes. Cell 96(1):69–78
63. Gandhi SR, Gierlinski M, Mino A, Tanaka K, Kitamura E, Clayton L, Tanaka TU (2011)
Kinetochore-dependent microtubule rescue ensures their efficient and sustained interactions
in early mitosis. Dev Cell 21(5):920–933. doi:10.1016/j.devcel.2011.09.006
64. Unsworth A, Masuda H, Dhut S, Toda T (2008) Fission yeast kinesin-8 Klp5 and Klp6 are
interdependent for mitotic nuclear retention and required for proper microtubule dynamics.
Mol Biol Cell 19(12):5104–5115. doi:10.1091/mbc.E08-02-0224
65. Tischer C, Brunner D, Dogterom M (2009) Force- and kinesin-8-dependent effects in the
spatial regulation of fission yeast microtubule dynamics. Mol Syst Biol 5:250. doi:10.1038/
msb.2009.5
66. Stumpff J, Wagenbach M, Franck A, Asbury CL, Wordeman L (2012) Kif18A and chromoki-
nesins confine centromere movements via microtubule growth suppression and spatial con-
trol of kinetochore tension. Dev Cell 22(5):1017–1029. doi:10.1016/j.devcel.2012.02.013
67. Bakhoum SF, Genovese G, Compton DA (2009) Deviant kinetochore microtubule dynamics
underlie chromosomal instability. Curr Biol 19(22):1937–1942. doi:10.1016/j.cub.2009.09.055
68. Bakhoum SF, Thompson SL, Manning AL, Compton DA (2009) Genome stability is ensured
by temporal control of kinetochore-microtubule dynamics. Nat Cell Biol 11(1):27–35.
doi:10.1038/ncb1809
69. Silkworth WT, Nardi IK, Scholl LM, Cimini D (2009) Multipolar spindle pole coalescence is
a major source of kinetochore mis-attachment and chromosome mis-segregation in cancer
cells. PLoS One 4(8):e6564. doi:10.1371/journal.pone.0006564
70. Ganem NJ, Godinho SA, Pellman D (2009) A mechanism linking extra centrosomes to chro-
mosomal instability. Nature 460(7252):278–282. doi:10.1038/nature08136
71. Huang Y, Yao Y, Xu HZ, Wang ZG, Lu L, Dai W (2009) Defects in chromosome congression
and mitotic progression in KIF18A-deficient cells are partly mediated through impaired func-
tions of CENP-E. Cell Cycle 8(16):2643–2649
72. Schaar BT, Chan GK, Maddox P, Salmon ED, Yen TJ (1997) CENP-E function at kinetochores
is essential for chromosome alignment. J Cell Biol 139(6):1373–1382
kinetochore motor required for metaphase chromosome alignment. Cell 91(3):357–366
74. Meraldi P, Draviam VM, Sorger PK (2004) Timing and checkpoints in the regulation of
mitotic progression. Dev Cell 7(1):45–60. doi:10.1016/j.devcel.2004.06.006
75. Manning AL, Bakhoum SF, Maffini S, Correia-Melo C, Maiato H, Compton DA (2010)
CLASP1, astrin and Kif2b form a molecular switch that regulates kinetochore-microtubule
dynamics to promote mitotic progression and fidelity. EMBO J 29(20):3531–3543.
doi:10.1038/emboj.2010.230
76. Ye F, Tan L, Yang Q, Xia Y, Deng LW, Murata-Hori M, Liou YC (2011) HURP regulates
chromosome congression by modulating kinesin Kif18A function. Curr Biol 21(18):
1584–1591. doi:10.1016/j.cub.2011.08.024
77. Rizk RS, Discipio KA, Proudfoot KG, Gupta ML Jr (2014) The kinesin-8 Kip3 scales ana-
phase spindle length by suppression of midzone microtubule polymerization. J Cell Biol
204(6):965–975. doi:10.1083/jcb.201312039
78. Meadows JC, Shepperd LA, Vanoosthuyse V, Lancaster TC, Sochaj AM, Buttrick GJ,
Hardwick KG, Millar JB (2011) Spindle checkpoint silencing requires association of PP1 to
both Spc7 and kinesin-8 motors. Dev Cell 20(6):739–750. doi:10.1016/j.devcel.2011.05.008
79. Colland F, Jacq X, Trouplin V, Mougin C, Groizeleau C, Hamburger A, Meil A, Wojcik J,
Legrain P, Gauthier JM (2004) Functional proteomics mapping of a human signaling path-
way. Genome Res 14(7):1324–1332. doi:10.1101/gr.2334104
80. Zheng F, Li T, Cheung M, Syrovatkina V, Fu C (2014) Mcp1p tracks microtubule plus ends
to destabilize microtubules at cell tips. FEBS Lett. doi:10.1016/j.febslet.2014.01.055
81. Vogelstein B, Papadopoulos N, Velculescu VE, Zhou S, Diaz LA Jr, Kinzler KW (2013)
Cancer genome landscapes. Science 339(6127):1546–1558. doi:10.1126/science.1235122
82. Zhu H, Xu W, Zhang H, Liu J, Xu H, Lu S, Dang S, Kuang Y, Jin X, Wang Z (2013) Targeted
deletion of Kif18a protects from colitis-associated colorectal (CAC) tumors in mice through
impairing Akt phosphorylation. Biochem Biophys Res Commun 438(1):97–102.
doi:10.1016/j.bbrc.2013.07.032
83. Honore S, Pasquier E, Braguer D (2005) Understanding microtubule dynamics for improved
cancer therapy. Cell Mol Life Sci 62(24):3039–3056. doi:10.1007/s00018-005-5330-x
84. van’t Veer LJ, Dai H, van de Vijver MJ, He YD, Hart AA, Mao M, Peterse HL, van der Kooy
K, Marton MJ, Witteveen AT, Schreiber GJ, Kerkhoven RM, Roberts C, Linsley PS, Bernards
R, Friend SH (2002) Gene expression profiling predicts clinical outcome of breast cancer.
Nature 415(6871):530–536. doi:10.1038/415530a
85. Damasco C, Lembo A, Somma MP, Gatti M, Di Cunto F, Provero P (2011) A signature
inferred from Drosophila mitotic genes predicts survival of breast cancer patients. PLoS One
86. Rucksaken R, Khoontawad J, Roytrakul S, Pinlaor P, Hiraku Y, Wongkham C, Pairojkul C,
Boonmars T, Pinlaor S (2012) Proteomic analysis to identify plasma orosomucoid 2 and
kinesin 18A as potential biomarkers of cholangiocarcinoma. Cancer Biomark 12(2):81–95.
doi:10.3233/CBM-130296
87. Tooker BC, Newman LS, Bowler RP, Karjalainen A, Oksa P, Vainio H, Pukkala E,
Brandt-Rauf PW (2011) Proteomic detection of cancer in asbestosis patients using SELDI-
TOF discovered serum protein biomarkers. Biomarkers 16(2):181–191. doi:10.3109/13547
50X.2010.543289
88. Brindley SM, Tooker BC, Pass HI, Newman LS (2012) Attempted validation of surface
enhanced laser desorption ionization-time of flight derived kinesin biomarkers in malignant
mesothelioma. J Data Min Genomics Proteomics S1:1–6. doi:10.4172/2153-0602.S1-002
89. Shichijo S, Ito M, Azuma K, Komatsu N, Maeda Y, Ishihara Y, Nakamura T, Harada M,
Itoh K (2005) A unique gene having homology with the kinesin family member 18A encodes
a tumour-associated antigen recognised by cytotoxic T lymphocytes from HLA-A2+ colon
cancer patients. Eur J Cancer 41(9):1323–1330. doi:10.1016/j.ejca.2005.02.025
90. Shackney SE, Smith CA, Miller BW, Burholt DR, Murtha K, Giles HR, Ketterer DM, Pollice
AA (1989) Model for the genetic evolution of human solid tumors. Cancer Res
49(12):3344–3354
91. Storchova Z, Pellman D (2004) From polyploidy to aneuploidy, genome instability and
cancer. Nat Rev Mol Cell Biol 5(1):45–54. doi:10.1038/nrm1276
92. Fujiwara T, Bandi M, Nitta M, Ivanova EV, Bronson RT, Pellman D (2005) Cytokinesis
failure generating tetraploids promotes tumorigenesis in p53-null cells. Nature
437(7061):1043–1047. doi:10.1038/nature04217
93. Lv L, Zhang T, Yi Q, Huang Y, Wang Z, Hou H, Zhang H, Zheng W, Hao Q, Guo Z, Cooke
HJ, Shi Q (2012) Tetraploid cells from cytokinesis failure induce aneuploidy and spontane-
ous transformation of mouse ovarian surface epithelial cells. Cell Cycle 11(15):2864–2875.
doi:10.4161/cc.21196
94. Rath O, Kozielski F (2012) Kinesins and cancer. Nat Rev Cancer 12(8):527–539. doi:10.1038/
nrc3310
95. Catarinella M, Gruner T, Strittmatter T, Marx A, Mayer TU (2009) BTB-1: a small molecule
inhibitor of the mitotic motor protein Kif18A. Angew Chem Int Ed Engl 48(48):9072–9076.
doi:10.1002/anie.200904510
96. Tcherniuk S, van Lis R, Kozielski F, Skoufias DA (2010) Mutations in the human kinesin Eg5
that confer resistance to monastrol and S-trityl-L-cysteine in tumor derived cell lines.
Biochem Pharmacol 79(6):864–872. doi:10.1016/j.bcp.2009.11.001
97. Girdler F, Sessa F, Patercoli S, Villa F, Musacchio A, Taylor S (2008) Molecular basis of
drug resistance in aurora kinases. Chem Biol 15(6):552–562. doi:10.1016/
j.chembiol.2008.04.013
98. Weiss WA, Taylor SS, Shokat KM (2007) Recognizing and exploiting differences between
RNAi and small-molecule inhibitors. Nat Chem Biol 3(12):739–744. doi:10.1038/
nchembio1207-739
99. Mitchison TJ (2012) The proliferation rate paradox in antimitotic chemotherapy. Mol Biol
Cell 23(1):1–6. doi:10.1091/mbc.E10-04-0335
100. Salmela AL, Kallio MJ (2013) Mitosis as an anti-cancer drug target. Chromosoma
122(5):431–449. doi:10.1007/s00412-013-0419-8
101. Komlodi-Pasztor E, Sackett D, Wilkerson J, Fojo T (2011) Mitosis is not a key target of
microtubule agents in patient tumors. Nat Rev Clin Oncol 8(4):244–250. doi:10.1038/
nrclinonc.2010.228
102. Purcell JW, Davis J, Reddy M, Martin S, Samayoa K, Vo H, Thomsen K, Bean P, Kuo WL,
Ziyad S, Billig J, Feiler HS, Gray JW, Wood KW, Cases S (2010) Activity of the kinesin
spindle protein inhibitor ispinesib (SB-715992) in models of breast cancer. Clin Cancer Res
16(2):566–576. doi:10.1158/1078-0432.CCR-09-1498
103. Chan KS, Koh CG, Li HY (2012) Mitosis-targeted anti-cancer therapies: where they stand.
Cell Death Dis 3:e411. doi:10.1038/cddis.2012.148
104. Jordan MA, Wilson L (2004) Microtubules as a target for anticancer drugs. Nat Rev Cancer
4(4):253–265. doi:10.1038/nrc1317
105. Amadori D, Volpi A, Maltoni R, Nanni O, Amaducci L, Amadori A, Giunchi DC, Vio A,
Saragoni A, Silvestrini R (1997) Cell proliferation as a predictor of response to chemotherapy
in metastatic breast cancer: a prospective study. Breast Cancer Res Treat 43(1):7–14
106. Meyer JS, McDivitt RW, Stone KR, Prey MU, Bauer WC (1984) Practical breast carcinoma
cell kinetics: review and update. Breast Cancer Res Treat 4(2):79–88
107. Skipper HE (1971) Kinetics of mammary tumor cell growth and implications for therapy.
Cancer 28(6):1479–1499
108. Zou JX, Duan Z, Wang J, Sokolov A, Xu J, Chen CZ, Li JJ, Chen HW (2014) Kinesin family
deregulation coordinated by bromodomain protein ANCCA and histone methyltransferase
MLL for breast cancer cell growth, survival and tamoxifen resistance. Mol Cancer Res.
doi:10.1158/1541-7786.MCR-13-0459
Chapter 12
The Kinesin-6 Members MKLP1, MKLP2
and MPP1
Ryan D. Baron and Francis A. Barr
12.1 Introduction to the Kinesin-6 Family
The kinesin-6 family of motor proteins comprises three human proteins MKLP1
(KIF23), MKLP2 (KIF20A) and MPP1 (KIF20B) as well as their orthologues from
other species. The human kinesin-6 family members are all structurally similar pro-
teins with closely related spatio-temporal regulation, yet are each functionally dis-
tinct with critical non-redundant roles involved in the organization and regulation of
the cell cycle central spindle and cytokinesis. Kinesin-6 family members are found
in a wide range of metazoans, protists, and some fungi [1]. These proteins are clas-
sified together as a family based on amino acid sequence conservation within their
motor domains [1–5]. The family number 6 derives from the previous family name
N-6 and the original name of MKLP2, Rabkinesin-6 [1, 2]. The family is further
classified into two subfamilies, the MKLP1 (KIF23) family and the MKLP2/MPP1
(KIF20A/B) family based on evolutionary conservation [2, 3]. There are two iso-
forms of MKLP1, the conventional MKLP1 and its larger CHO1 isoform, which
contains the additional exon 18. Exon 18 encodes an actin-binding domain required
for late cytokinesis [6]. There are no characterized splice variants of MKLP2 or
MPP1 [7].
Like other kinesins, all members of the kinesin-6 family are molecular motor
proteins and comprise four distinct domains, each contributing to their molecular
function. These domains include the motor head domain, the neck linker, a coiled
coil stalk and a globular tail domain. The kinesin-6 family members are all
N-terminal kinesins with the motor domain located at the amino end of the molecule.
R.D. Baron (*)

Department of Molecular and Clinical Cancer Medicine, University of Liverpool,
Daulby Street, Liverpool L69 3GA, UK
e-mail: r.d.baron@liv.ac.uk
F.A. Barr
Department of Biochemistry, University of Oxford, South Parks Road, Oxford OX1 3QU, UK
e-mail: francis.barr@bioch.ox.ac.uk

194 R.D. Baron and F.A. Barr
Although the motor head domain is highly conserved within the family, the
C-terminal domains are much more divergent [8]. Like other kinesins the motor or
head regions of all kinesin-6 family members contain a highly conserved ATPase
and microtubule-binding domain. This family can be distinguished from other kine-
sins by a long insertion within loop-6 of the motor domain, and a long neck linker
region [1, 9–11]. The kinesin-6 family contain the longest loop-6 of all kinesin
families, the longest being the MPP1 loop-6 with 77 amino acids. However, this
long insertion appears to play no role in modulating either the ATPase catalytic
activity or microtubule binding; the two key functions of the motor domain [10].
Loop-6 is located on the exposed cytoplasmic surface of the molecule on the oppo-
site side of the motor head from the microtubule-binding surface [10, 12]. This has
led to the suggestion it may play a role in binding cytosolic regulatory factors. The
neck linker domain is also substantially longer amongst the kinesin-6 family mem-
bers with approximately 70 amino acids between the motor domain and the coiled
coil domain, compared to 13–15 amino acids in other kinesins [9, 13–15]. This
extended neck linker domain is loosely conserved within the kinesin-6 family [1].
Intriguingly, both of the insertions are within loops and interactions between the
neck linker loop and head loops are in regions important for motility [14]. It has
therefore been suggested that the extended loop-6 and neck-linker domains may
have co-evolved [9], and could therefore be important features imparting specific
regulatory or movement properties on kinesin-6 family members. The functional
advantage of the kinesin-6 long neck linker region is not understood, however, it has
been postulated that this may allow kinesin-6 dimers to bridge longer distances than
adjacent tubulin binding sites. This could allow longer steps along the same proto-
filament, allow kinesin-6 dimers to traverse laterally over one or more protofila-
ments, or even cross to an adjacent microtubule [10].
The kinesin-6 family motor proteins dimerise through their C-terminal coiled-
coil domains [16, 17] and this allows cooperativity and processive movement along
microtubules [10]. Like most other N-terminal kinesins, kinesin-6 family members
show directed movement towards the microtubule fast growing or plus end.
However, kinesin-6 family motor proteins are slow plus end directed kinesins with
reduced processivity when compared with the canonical kinesin, Kinesin-1 [10, 18],
the first kinesin discovered, responsible for fast axonal transport in neurons [19].
The direction of kinesin movement is generally determined by the neck-linker
region adjacent to the motor head [14, 20–23]. As already mentioned, this region is
extended in the kinesin-6 family and may therefore impart altered regulation of
movement and reduced processivity relative to conventional kinesin.
12.2 Overview of Kinesin-6 Function in Mitosis

and Cytokinesis
All kinesin-6 family members have been shown to play key roles in the regulation
of mitosis and cytokinesis [9, 24–29]. Further supporting their classification as a
family, most published data demonstrates similar cell cycle regulation and
12 The Kinesin-6 Members MKLP1, MKLP2 and MPP1 195
spatio-temporal localization during mitosis within the family and across species.
Kinesin-6 proteins are expressed from G2 onwards accumulating in the nucleus at
the end of interphase [9, 24, 28–31]. They become diffusely cytoplasmic at nuclear
envelope break down before localizing to the midzone in anaphase and finally to the
midbody in cytokinesis, before their degradation prior to the subsequent cell cycle
[9, 24, 29, 32, 33]. Subtle differences in localization between family members
are however apparent from telophase onwards, MKLP1 localizes to the centre of the
midbody, whereas both MKLP2 and MPP1 flank this region [9, 24, 33]. Interestingly
in drug-induced monopolar cytokinesis [34], or in the anaphase half-spindles aris-
ing from PRC1 depletion [35], the CPC lies distal along the microtubule to MKLP1.
This suggests that the chromosome passenger complex bounding the midbody in
each half-spindle may be positioned by the plus ends of microtubules arising in the
opposite half-spindle [36]. This structural precision is likely to be important for the
collaborative regulation of cytokinesis as functional interaction has been demon-
strated between the different kinesin-6 complexes [37].
Kinesin-6 members spring into action at the metaphase-to-anaphase transition
and are essential for the fidelity of cytokinesis, the final stage of cell division. Their
inactivity earlier in the cell cycle results from their cell cycle dependent expression
[9, 24, 29], nuclear sequestration [9, 24, 29, 30, 32] and inactivating phosphoryla-
tions [18, 38, 39]. At the metaphase to anaphase transition inactivation of Cdk1-
cyclin B allows dephosphorylation and activation of these kinesins.
Once active, kinesin-6 members are able to couple processive movement with
cargo and interaction partner binding. This allows dynamic control of the processes
they regulate. Full-length kinesin-6 family motors have all been shown to crosslink
microtubules in vitro and produce slow plus end directed movement of tubulin poly-
mers in gliding assays [9, 10, 18, 32]. In order to perform this function these motors
have two microtubule binding domains, one within the motor head and a second at
the C-terminus [32, 40]. In addition, Klp9p the kinesin-6 family member in
Schizosaccharomyces pombe has been demonstrated to form both dimers and tetra-
mers and plays a role in Anaphase B spindle elongation that is enhanced by binding
Ase1p/PRC1 [41]. Despite this observation in Schizosaccharomyces Pombe, no
clear direct effect on spindle elongation dynamics have been reported for the mam-
malian kinesin-6 proteins, although interestingly all three kinesin-6 proteins have
been demonstrated to interact with PRC1 [33, 42]. Failure to observe a spindle
elongation phenotype may result from a combination of functional redundancy [43]
and complexity of analysing kinesin depletion phenotypes in higher species [44],
alternatively this role may have been lost in higher eukaryotes. Interestingly, despite
their similarities, all members of the kinesin-6 family are independently required for
cytokinesis [9, 24–27, 29]. This may be explained by specific, non-redundant
molecular interactions between the motors and their interacting signalling com-
plexes. MKLP1 interacts with the Rac1 GAP CYK4 [16, 45] and MKLP2 interacts
with the chromosome passenger complex, which includes the kinase Aurora B [46].
These interactions link the kinesins through down stream signalling pathways with
cytokinesis control. MPP1 has been shown to interact with PRC1 [33] and PIN1 [8],
however no major downstream interacting signalling partner has yet been identified.
It is these specific interactions that allow kinesin-6 proteins to fulfil their distinct
roles, which include organization of the anaphase central spindle, specifying the
ingression furrow position and finally roles in abscission. This high level of speci-
ficity and non-redundancy between cargo-motor interactions along with their essen-
tial roles in the regulation of cytokinesis makes this family of kinesins potentially
attractive anticancer drug targets [47, 48].
12.3 MKLP1 and Formation of the Midbody

During Cytokinesis
MKLP1 was first identified in an immunological screen generating antibodies

against mitotic spindle components from Chinese Hamster Ovary (CHO) cell lines
[31]. The 95/115-kDa characterized antigens appeared in a cell cycle dependent
manner and immunofluorescence microscopy suggested it associated with the cen-
tral spindle microtubules localizing to the interdigitating antiparallel microtubules
of the midzone in anaphase and the midbody in telophase. A nuclear and centro-
somal localization pattern was also seen in some interphase cells with this staining
lost as cells entered mitosis [28, 31]. Microinjection of anti-MKLP1 serum or puri-
fied antibody resulted in a cell cycle arrest with PtK1 cells entering a prolonged
metaphase arrest with misaligned chromosomes [28]. The MKLP1 gene was cloned
in humans using this antibody to probe a HeLa cDNA expression library [32]. The
gene identified encoded a 110-kDa protein containing a 350 amino acid domain in
the amino-terminal half with significant sequence homology to the motor domain of
other known kinesin family proteins [49, 50]. MKLP1 was capable of crosslinking
microtubules in vitro and sliding microtubules in modified gliding assays [32] and
was originally classified as an orphan kinesin [43] until it found its place in the
kinesin-6 family [2, 3]. These initial studies led to the suggestion that MKLP1 might
play a role in chromosome congression [28], modulating microtubule dynamics or
spindle elongation in Anaphase B [31, 32, 51].
Genetic analysis in D. melanogaster and C. elegans clarified the role of MKLP1 in
central spindle assembly and cytokinesis rather than in chromosome congression or
spindle elongation [25–27]. In Drosophila with mutations in Pavarotti, the ortho-
logue of MKLP1, chromosomes segregate normally at anaphase onset, however, a
highly abnormal central spindle forms with defects in antiparallel microtubule bun-
dling. These cells also fail to form a contractile ring and therefore do not undergo
cytokinesis [25]. A similar disorganized central spindle phenotype is also seen in
C. elegans with either a null mutation of Zen-4, the C. elegans orthologue of MKLP1
[27] or with depletion of Zen-4 using siRNA [26]. In C. elegans [26] and mamma-
lian cells [52–54], unlike in D. melanogaster, depletion of Zen-4/MKLP1 does not
prevent furrow formation, however, a stable midbody fails to form resulting in late
stage cytokinesis failure.
MKLP1 dimers interacts with dimers of CYK-4 (also known as RacGAP1 and
MgcRacGAP) a Rho family GTPase activating protein (GAP) [55] forming a high
affinity, 2:2 stoichiometric tetramer, the centralspindlin complex [16, 45]. Only the
centralspindlin complex and not the individual subunits localizes to the central
spindle and has microtubule bundling activity, this complex is essential for central
spindle integrity [25–27, 45, 56, 57]. MKLP1 interacts with the CYK-4 N-terminus
via its characteristic extended neck linker region [16]. Interestingly the sequences
that mediate this interaction are not evolutionarily conserved suggesting a degree of
plasticity in the binding domain [45]. Recent work has demonstrated that although
the Zen-4 neck linker is similarly structured in both the unbound and CYK-4 bound
states, binding dramatically stabilises the relative positions of the neck linker
regions and reduces the rate of microtubule gliding activity in vitro [58].
Centralspindlin fulfils a number of essential roles during the cell cycle including:
promoting central spindle assembly and microtubule bundling, positioning the divi-
sion plane and regulating Rho A activity and recruiting critical regulators of abscis-
sion [15, 59, 60].
At anaphase onset centralspindlin relocates from the cytoplasm to the overlap-
ping plus ends of the central spindle antiparallel microtubules where it is important
for microtubule bundling and organizing central spindle assembly [18, 25–27, 45,
56]. The precise mechanistic details of how centralspindlin stabilises and organizes
the central spindle remain to be elucidated, however, this is subject to a number of
regulatory processes including phosphorylation by Cdk1, Aurora B and Plk1, oligo-
merization and binding partner interactions. Centralspindlin binding to and bun-
dling of microtubules is inhibited by Cdk-cyclin-B phosphorylation of the highly
basic N-terminal extension of the MKLP1 motor [18]. This basic N-terminal exten-
sion of MKLP1 may interact with the acidic tail of tubulin in a manner analogous to
the interaction of the KIF1A K-loop [61]. Supporting this, in the presence of non-
degradable cyclins that stabilize Cdk1 activity, the central spindle fails to assemble
[62, 63]. This mechanism delays the onset of microtubule bundling until after the
metaphase to anaphase transition.
The catalytic activity of the MKLP1 motor is essential for the localization of the
centralspindlin complex to the central spindle [52, 64]. Individual MKLP1 motor
dimers are however non-processive and require oligomerization into higher order
clusters to become processive [65]. This multimerization of MKLP1 complexes is
negatively regulated by binding of the 14-3-3 proteins, that is in turn regulated by
Aurora B phosphorylation [66]. MKLP1 is constitutively phosphorylated on the
S710 residue and this acts as the phosphorylated residue essential for binding the
14-3-3, phospho-serine, phospho-threonine binding complex. Binding of the 14-3-3
proteins inhibits multimerization and therefore processivity and microtubule bun-
dling activity of MKLP1 [66]. MKLP1 is released from this inhibition by Aurora B
phosphorylation of S708 [67] within the 14-3-3 binding domain allowing multimer-
ization and accumulation at, and bundling of the central spindle [66]. In C. elegans
depletion of the Aurora family kinase Air-2, prevents the accumulation of Zen-4 at
the central spindle [68–70]. Acting through a second mechanism the continued
presence of Aurora B, a component of the chromosome passenger complex [71], is
required at the central spindle to maintain MKLP1 localization [37]. The chromo-
some passenger complex is itself relocated from the inner centromere to the central
spindle at the metaphase-to-anaphase transition by the related kinesin-6 family
member MKLP2 [46]. At the central spindle Aurora B phosphorylates MKLP1

S911, this Aurora phosphorylation site overlaps with the C-terminal bipartite
nuclear localization sequence (NLS) and prevents premature sequestration of
MKLP1 into the reforming daughter cell nuclei [37, 54], allowing MKLP1 to fulfil
its late stage roles in the regulation of cytokinesis [6]. The other components of the
chromosome passenger complex have also been shown to be essential to control
MKLP1 function at the central spindle providing further support for this mechanism
[53, 70]. In C. elegans the association between Zen-4 and CDC14 is also important
for Zen-4 localisation to the central spindle and depletion of CDC14 prevents Zen-4
localisation [72]. Cells depleted of MKLP1 maintain the ability to localize MKLP2
and the chromosomal passenger complex to the central spindle [46], however many
other central spindle components fail to be recruited, including CEP55, ECT2 and
FIP3, which are important for furrowing and abscission [73–77].
Once established the central spindle along with the astral microtubules is
important in signalling the cleavage furrow position [25, 78–84]. This occurs
through the localization and activation of Rho, A which in turn activates several
effectors that regulate contractile ring assembly and cytokinesis [60, 85].
Centralspindlin plays a key role in this process by recruiting and activating ECT2,
the critical guanine nucleotide exchange factor (GEF) responsible for activating
Rho A [86–88] at the central spindle [75, 76, 85, 89, 90]. The interaction between
CYK-4 and ECT2 is temporally regulated by phosphorylation [90]. ECT2 is inhib-
ited from binding to CYK-4 before the onset of anaphase by a metaphase specific
phosphorylation of ECT2; this is most likely mediated by Cdk1 [76]. In addition
to this negative regulation, the CYK-4 N-terminus is phosphorylated at the central
spindle by Plk-1 on residue S157, this modification is required to bind the
N-terminal tandem BRCT (BRCA1 C-terminal) domains of ECT2 that form a
phosphopeptide binding module [91–94]. When S157 is mutated to a non-phos-
phorylated residue, ECT2 fails to localize to the central spindle and errors in
cleavage furrowing result [93, 94]. It has been postulated that CYK-4 binding to
ECT2 not only localizes but also activates ECT2 by relieving auto inhibition cre-
ated by interaction of the ECT2 N-terminus with its own C-terminal GEF domain
[95, 96]. This central spindle pool of active ECT2 is assumed to limit local activa-
tion of Rho A to the equatorial cortex [93, 94]. The exact mechanism by which
centralspindlin bound ECT2 activates Rho A is not understood, however, it is
likely that Rho A is activated while membrane bound [97], suggesting either active
ECT2 or the centralspindlin-ECT2 complex travel to the cortex. Despite these
observations, a central spindle localized pool of ECT2 does not appear to be an
absolute requirement for cleavage furrow ingression. Depletion of MKLP1,
although resulting in delocalization of CYK-4 and ECT2, does not prevent activa-
tion of Rho A at the equatorial cortex, although this is spatially impaired with a
wider distribution [52, 76, 81]. Perturbation of Rho A, ECT2 or CYK-4 on the
other hand prevent furrow formation, leading to cytokinesis failure and multinu-
cleated cells [76, 80, 85, 88, 95]. The alternative mechanism proposed for activa-
tion of Rho A involves the pool of astral microtubules that converge at the central
spindle. However the mechanistic details of how astral microtubules localize Rho
A are not understood, and the relative contributions from the central spindle and
astral pools of microtubules appear to be species and cell type specific [79–81].
Interestingly, given its indirect role in the activation of Rho A by recruiting and
activating the RhoGEF ECT2, CYK-4 is in fact a Rho-family GTPase activating
protein (GAP). Although initial biochemical experiments demonstrated that CYK-4
had greater GAP activity towards Rac1 and Cdc42 than to Rho GTPases [55, 56]
and genetic evidence suggesting that CYK-4 regulates Rac proteins [98], a long
standing debate developed within the literature regarding the target GTPase of the
CYK-4 GAP domain [99, 100]. The early observation that Rho and not Rac1 or
Cdc42 depletion led to cytokinesis defects, the positioning of Rho A at the equato-
rial cortex directly adjacent to CYK-4 and the observation of a cycling activity state
of Rho A drove the argument that CYK-4 may be a Rho GAP [45, 56, 86, 89, 90].
Contrasting evidence however suggested some cell lines had no requirement for
CYK-4 GAP activity [101, 102] or that the GAP activity was mediated through Rac
pathways [103]. The suggestion that Aurora B could phosphorylate CYK-4 adjacent
to its catalytic domain switching its activity from Rac to Rho [104] seemed to unify
some of these findings, however, consideration of the three dimensional structure of
Rho and known Rho GAPs suggested that this switch in target was unlikely [42,
100]. Recent work has clarified that the centralspindlin complex only has GAP
activity towards Rac and Cdc42 and not towards Rho GTPases and cytokinesis
defects driven by loss of CYK-4 GAP activity result from excessive cell adhesion
and can be rescued by depletion of Rac1 or its effectors ARGEF7, PAK1 and PAK2
[42]. This work clarifies CYK-4 as a Rac1 GAP, required to inactivate Rac1 at the
central spindle, to release cell adhesions in this region in order to allow the cell to
complete cytokinesis. It also explains why non-adherent cells have a lesser require-
ment for CYK-4 GAP activity [101, 102].
The ingressing cleavage furrow is responsible for compacting the midzone into
the midbody between the two future daughter cells [105]. During midbody forma-
tion three distinct regions have been defined and protein rearrangements from their
midzone positions to their final midbody locations is observed. Centralspindlin
relocates from the overlapping microtubule plus ends to the flanking bulge as a ring
structure around the midbody “dark zone” [105]. In this position it co-localizes with
its known interacting partners CEP55, ARF6 and ECT2 [65, 73, 76, 106]. In its final
position MKLP1 no longer appears associated with microtubules, it is also observed
that MKLP1 C-terminal truncations lacking the motor domain also localize to this
midbody flank [105]. MKLP1 stabilises the midbody structure and in its absence a
stable midbody does not form [26, 52–54]. The structural mechanisms underpinning
this localization are not yet understood, however, MKLP1 may form a bridge from
the midbody to the actomyosin contractile ring or cell cortex either directly or
through its interaction partners [107].
The CHO1 splice variant of MKLP1 is required for stable midbody formation in
mammalian cells [6, 54]. The CHO1 splice variant contains the additional exon 18
that encodes an F-actin binding domain [6]. Deletion of this actin binding domain
[54] or injection of antibodies directed to this region [6] result in the midbody
becoming detached from the cortex with regression of the cleavage furrow in late
cytokinesis, although the midbody structure of bundled microtubules remains intact.

Although this domain may contribute to midbody stability and progression to
abscission a degree of redundancy is likely as a partial rescue was observed even
with the construct lacking the actin-binding domain [54]. Recently it has been dem-
onstrated that the centralspindlin complex can bind to the plasma membrane directly
through the CYK-4 C1 domain that interacts with plasma membrane polyanionic
phospho-inositide lipids [107]. Mutations in the C1 domain prevent binding of
centralspindlin to the plasma membrane and result in defects in cytokinesis.
Alternatively MKLP1 may stabilize the midbody through its binding partner inter-
actions. Arf6 is a GTP binding protein with roles in endocytosis and organization of
the actin cytoskeleton [108] that may function in abscission. Arf6 is membrane
associated, binding through its lipid tail, and also binds to the MKLP1 C-terminus
[106, 109]. Although depletion or knockout of Arf6 does not result in mislocalisa-
tion of MKLP1, it does cause mild cytokinesis defects. It has also been suggested
that Arf6 stabilises the midbody by preventing 14-3-3 protein mediated premature
disassembly [110]. Anillin is another candidate that may play a role in midbody
stabilization through its interactions with CYK-4 (Drosophila)/ECT2 (human and
mouse) as well as with actin, myosin and Rho A [85, 111, 112].
The final role for centralspindlin involves abscission. MKLP1 binds and recruits
the dimeric coiled-coil protein CEP55 to the midbody and the absence of CEP55
leads to midbody defects [73, 113]. CEP55 in turn binds and recruits the endosomal-
sorting complex required for transport I (ESCRT-I) subunit, TSG101 and the ESCRT-I
associated protein ALIX [114, 115]. These proteins recruit the ESCRT-III complex,
which along with vacuolar protein sorting 4 (VPS4) and spastins are required for the
completion of cytokinesis [116, 117]. This process is intricately timed by phosphory-
lation events. Cdk1 phosphorylates Cep55 in metaphase on residues S425/S428;
these phosphorylated residues interact with the polo-box domain of Plk1, leading to
Plk1 phosphorylation of S436 [113, 118]. This phosphorylation negatively regulates
the association of CEP55 and MKLP1 acting as a molecular timer to prevent prema-
ture CEP55 recruitment to the midbody and failure of abscission [118].
Following abscission centralspindlin rings remain in the membrane-associated
midbody derivatives. These are usually retained and degraded through auto phago-
cytic pathways but can also be released as remnants after abscission [119–123].
Disruption of the centralspindlin complex results in premature disassembly of the
midbody derivative [56]. Interestingly midbody derivatives appear to be retained in
stem cells and cancer cells, and in the latter instance be linked to increased
tumorigenicity [120].
12.4 Dysregulation of MKLP1 in Cancer
MKLP1 expression has been shown to be upregulated in both glioma and non-small
cell lung cancer compared with normal tissues [124, 125]. Whole exome sequenc-
ing has also demonstrated non-synonymous missense variations in the KIF23 gene,
encoding MKLP1, in both melanoma cell lines and a colorectal cancer family
without one of the common mutations associated with colorectal cancers (lynch
syndrome, FAP or MUTγH mutations) [126, 127]. The KIF23 gene also falls in a
region previously implicated as a colorectal cancer linkage region (15q22) [127].
These studies suggest MKLP1 may play a role in the development of various can-
cers. Specific immunogenicity against MKLP1 was identified in the sera of mice
inoculated with glioma made immunogenic by the injection of Herpes Simplex
Virus (HSV) compared with control sera suggesting MKLP1 as a potential molecu-
lar target in cancer [125]. Furthermore, down regulation of MKLP1 using siRNA
resulted in reduced growth of glioma cell lines both in culture and in tumour growth
assays in-vivo [125].
Recently the rare autosomal dominant condition, congenital dyserythropoietic
anaemia type III (CDAIII) has been demonstrated to be caused by a heterozygous
single point mutation (c.2747C > G) in the KIF23 gene, resulting in an amino acid
substitution P916R at the C-terminus of the MKLP1 protein [128]. CDAIII is char-
acterized by dyserythropoiesis with large multinucleated erythroblasts (giganto-
blasts) in the bone marrow and intravascular haemolysis [128, 129]. These clinical
findings suggest a failure of cytokinesis. RNA interference experiments in HeLa
cells demonstrated that the wild type but not the P916R mutant could rescue cytoki-
nesis. Following furrowing and midbody formation with normal kinetics, P916R
mutant cells underwent furrow regression becoming multinucleate. Given this
observation it is surprising that the clinical phenotype is limited to erythrogenesis,
however, monoclonal gammopathy of undetermined significance, and myeloma
have also developed in a substantial number of patients in a large Swedish pedigree
affected by CDAIII [128, 129]. This mutation may disrupt MKLP1 function by
interfering with its C-terminal interaction partner binding with either Arf6 or the
14-3-3 proteins, or by prevention of Aurora B mediated phosphorylation of S911
leading to premature nuclear import of MKLP1, the exact underlying mechanism
however awaits clarification.
Transcription of the KIF23 gene is regulated by the cell cycle genes homology
region (CHR) within its promoter region, which is negatively regulated by the
tumour suppressor p53 in a p21WAF1/CIP1 dependant manner via the DREAM/MMB
pathways. Consistent with its tumour suppressive role expression of p53 therefore
reduces the level of MKLP1 expression at both the mRNA and protein levels regu-
lating cell cycle control [130]. The centralspindlin components, MKLP1 and CYK-4
themselves have roles as transcriptional regulators [131, 132].
Although its essential role in cytokinesis suggests MKLP1 as a potential molecu-
lar cancer target, it also has defined functions outside of cell division and the side
effects from disruption of these functions must be balanced against any potential
therapeutic benefit from targeting MKLP1. MKLP1 is highly expressed in the hip-
pocampus during rodent development [133, 134] and continues to be expressed in
cultured post-mitotic differentiated neurons [135, 136]. Depletion of MKLP1 before
dendritic differentiation prevented dendritic development [136, 137] and depletion
after robust dendrite formation resulted in loss of dendrites, and dendrites becoming
axon-like in morphology, microtubule orientation and organelle distribution [138].
Within dendrites microtubules are bidirectionally orientated with both plus-end and
minus-end distally directed microtubules, whereas axons contain only plus-end
distal microtubules [139, 140]. This pattern arises through the transport of minus
end distal microtubules into the dendrites from the cell body by MKLP1 [136–138,
141]. Both the N-terminal motor domain and the C-terminal tail of MKLP1 are
important for its dendritic targeting [142]. The non-mitotic role of MKLP1 extends
beyond neurons, the complex processes of renal glomerular podocytes have also
been demonstrated to contain microtubules of mixed polarity and analogous to den-
dritic organization this has been demonstrated to be dependent on MKLP1 [143].
Along with other midbody associated proteins MKLP1 has also been demonstrated
to localize to the ciliary basal body. A temperature sensitive mutant resulted in
defects in both ciliary morphology and function [144]. Finally, MKLP1 has been
implicated in gametogenesis and fertility, epidermal morphogenesis and in epithe-
lial polarization required for normal foregut epithelium development [145–147].
These non-mitotic functions of MKLP1 must be kept in mind when targeting
MKLP1 and trials of therapeutic agents targeting MKLP1 must consider neurotox-
icity, renal impairment, ciliary dysfunction and teratogenesis.
12.5 MKLP2: The Transport System for Aurora

B in Anaphase
MKLP2 was first identified using a yeast two-hybrid approach to screen for interact-
ing effectors of active, GTP bound, Rab6; and originally named Rabkinesin-6 [40].
Related to the role of Rab6, MKLP2 was initially proposed to play a role in the
transport of vesicles between the Golgi and the endoplasmic reticulum [40]. The
Rab6 binding domain has been mapped to the coiled-coil stalk domain of MKLP2
between the motor domain and the C-terminal globular cargo-binding domain [24,
40, 148]. Subsequent work has however demonstrated that MKLP2 is regulated in a
cell cycle dependent manner on both the mRNA and protein levels showing a simi-
lar pattern to cyclin B2 regulation. MKLP2 is essentially absent in interphase cells,
accumulating in the nucleus during G2, prior to entry into mitosis [24, 29]. On
nuclear envelope breakdown MKLP2 is released into the cytoplasm until the
metaphase-to-anaphase transition when it relocates to the microtubule midzone and
the midbody during telophase in a manner analogous to MKLP1 with which it
shows significant sequence homology [24, 29]. No co-localisation of MKLP2 was
seen with either the Golgi or Rab6 [24]. Microinjection of antibodies against
MKLP2 results in failure of cytokinesis and the accumulation of binucleate cells
[24, 29]. A direct role of MKLP2 in cytokinesis was confirmed by depletion experi-
ments; after normal alignment and segregation of the chromosomes defects were
observed in the extent and timing of furrow ingression leading to cytokinesis failure
and binucleate cell formation [149]. Consistent with this role in regulating
cytokinesis, overexpression of MKLP2 leads to cell death following mitosis [24].
The major role of MKLP2 in central spindle stabilization and cytokinesis is per-
formed through the recruitment and localization of its interaction partners. It may
however directly contribute to central spindle stabilization as it has been
demonstrated oligomerize [38] and contains a C-terminal microtubule-binding

domain in addition to its motor domain [40, 46, 149]. MKLP2 promotes the local-
ization of Plk1 and Aurora B, the two major anaphase regulatory kinases at the
central spindle [46, 149, 150]. These in turn are crucial for the regulation of cytoki-
nesis. Plk1 primes its own binding site by phosphorylating MKLP2 S528 in the
neck linker region [149]. Within its C-terminus Plk1 contains a polo box domain
(PBD), which forms a phosphopeptide-binding module capable of interacting with
pS528 of MKLP2 [149, 151, 152] and localizing Plk1 to the central spindle. Through
this phosphorylation and interaction Plk1 negatively regulates the microtubule bun-
dling activity of MKLP2 at the central spindle [149]. Inhibition of Plk1 not only
results in loss of Plk1 localisation from the central spindle but also in the failure of
MKLP2 retention at the central spindle and its accumulation at the cortex by telo-
phase [92, 149, 150]. In addition to its regulation of MKLP2, Plk1 regulates several
other proteins at the central spindle including: MKLP1, PRC1, CYK-4, CEP55 and
ROCK2, it therefore plays a crucial role in regulating the early and late stages of
cytokinesis [93, 94, 118, 150, 153–155]. In addition to MKLP2, PRC1 and MKLP1
also have roles in binding and localizing Plk1 to the central spindle [150, 153].
Aurora B is the kinase component of the chromosome passenger complex (CPC),
which is completed by INCENP, survivin and borealin [68, 156–160]. At the
metaphase-to-anaphase transition the chromosome passenger complex relocates
from the inner centromere to the central spindle and this is prevented by the deple-
tion of MKLP2 [46]. Not only does the CPC fail to relocate in the absence of
MKLP2, MKLP2 is also reciprocally dependent on the chromosome passenger
complex for its localization to the central spindle [39]. MKLP2 has been shown to
directly bind to both Aurora B and INCENP components of the CPC [38, 39, 46].
Cdk1-cyclin B phosphorylation negatively regulates the interaction between
MKLP2 and the CPC, controlling the temporal relocation of Aurora B through
phosphorylation of both INCENP and MKLP2 [38, 39]. Additionally the Mad2
spindle assembly checkpoint protein and Aurora B kinase activity have roles in the
regulation of Aurora B relocalisation [161, 162].
MKLP2 localisation of Aurora B to the anaphase central spindle results in a
restricted pool of activation of Aurora B on the spindle specifically in anaphase cells
[37, 42, 163]. Aurora B and Plk1 regulate cytokinesis through independent
mechanisms and inhibition of both kinases results in a more penetrant phenotype of
cytokinesis failure than either alone [92]. The main roles of Aurora B appear to be
in stabilizing the central spindle and the late regulation of cytokinesis [164, 165].
Aurora B regulates the central spindle through its regulatory activity over a number
of other kinesins in addition to MKLP2, including the recruitment [66–70] and
retention [37] of MKLP1 and the recruitment and activation of KIF4 [42] at the
central spindle. In addition to this positive regulation, Aurora B also phosphorylates
and excludes the kinesin KIF2A from the central spindle limiting its microtubule
depolymerase activity to microtubule minus ends [166]. The chromosome passen-
ger complex also plays a role in the regulation of the constriction ring, not only by
acting upstream of centralspindlin recruitment to the central spindle but also through
centralspindlin independent mechanisms [167]. This is evident from the additive
effect of co-depletion of Aurora B and MKLP1 on the rate and extent of constriction
of the contractile ring and the observation that distinct manipulations rescue the
phenotype produced by either Aurora B or MKLP1 depletion [167]. Aurora B
depletion is partially rescued by depletion of septin or partial depletion of anillin
suggesting a role in regulating the maturation of the contractile ring [167]. Aurora
B plays a further role in the regulation of the contractile ring through its inhibitory
phosphorylation of GEF-H1 a microtubule-regulated exchange factor coupling
microtubule dynamics to Rho A activation [168]. This inhibitory phosphorylation is
reversed allowing Rho A activation at the equator at the initiation of cytokinesis
[168]. Targeting of Aurora B to the equatorial cortex appears to be important in
generating the highly focused accumulation of active Rho A [169]. This is achieved
by the ability of MKLP2 to bind myosin-II and actomyosin filaments. An MKLP2
mutant defective in myosin-II targeting, which can localize to the central spindle but
not to the equatorial cortex, fails to support complete furrow ingression [169].
Finally, Aurora B plays a key role in the abscission (NoCut) checkpoint, preventing
abscission in the presence of chromosome bridges to avoid DNA damage and
aneuploidy [170–172]. The Chromosome passenger complex binds the ESCRT-III
component CHMP4C through its borealin subunits and this allows Aurora B
phosphorylation of CHMP4C [173, 174]. Phosphorylated CHMP4C forms a
complex with ANCHR and VPS4 at the midbody preventing the action of VPS4 and
the final stages of abscission until the Aurora B mediated abscission checkpoint has
been silenced [175].
12.6 Dysregulation and Targeting of MKLP2 in Cancer
MKLP2 overexpression has been reported in a multitude of different solid cancers

including: pancreatic ductal adenocarcinoma, gastric cancer, hepatocellular carci-
noma, cholangiocarcinoma, breast cancer, bladder cancer, non-small cell lung can-
cer, small cell lung cancer and malignant melanoma [176–180]. Higher levels of
overexpression were associated with chromosomal instability and the more prolif-
erative and undifferentiated carcinomas characterized by mutation of TP53 and
inactivation of the pRb pathway [179]. MKLP2 is also upregulated in regenerating
hepatocytes and preneoplastic liver demonstrating its role in normal as well as
malignant cell division [179]. The MKLP2 gene maps to the smallest commonly
deleted segment identified in myeloid leukaemia on the long arm of chromosome 5,
del(5q) [7]. Consistent with its essential role in cell division, no inactivating muta-
tions of the remaining MKLP2 gene have been identified in patients with MDS or
AML and abnormalities of chromosome 5 [7]. It may however be possible that
haploinsufficiency results in genomic instability and aneuploidy as a driver of car-
cinogenesis. Lenalidomide a drug used to treat patients with 5q deleted myelodys-
plastic syndrome [181] results in cytokinesis failure and apoptosis [182]. Gene
expression profiling of lenalidomide treated cells demonstrated decreased expres-
sion of M-phase related genes including MKLP2 [182]. Given its position at the 5q
locus it is possible that MKLP2 may be a target of lenalidomide.
Depletion of MKLP2 has been demonstrated to reduce cell viability in a number

of cancers including: pancreatic ductal adenocarcinoma, gastric adenocarcinoma,
hepatocellular carcinoma, breast ductal carcinoma, HeLa cells and osteosarcoma
[176, 179, 183–185]. Although cell death was a common finding, variations were
noted in these studies, some of which can be accounted for by methodological dif-
ferences. Taniuchi et al. demonstrated attenuated cell growth in pancreatic ductal
adenocarcinoma cells, but failed to show multinucleate cell development, this may
have resulted from the extended siRNA treatment of cells for 10 days during which
time multinucleate cells may have undergone cell death [176]. Gasnereau et al. on
the other hand demonstrated a G2/M arrest and polyploidy in hepatoma cells but did
not see apoptosis, however, by analysing the cells after only 36 h of siRNA treat-
ment they may have been too early to see this [179]. Groth-Pedersen et al. identified
MKLP2 in a screen for cytoskeletal-associated proteins whose depletion resulted in
cancer cell death. This study also failed to see multinucleate cells and suggested
MKLP2 depletion led to a G1 arrest in the MCF7 breast ductal carcinoma cell line
before cells underwent a lysosomal cell death, rather than apoptosis [183]. Consistent
with their finding of lysosomal instability, MKLP2 depletion sensitized cells to the
lysosome-disrupting drug siramesine, suggesting MKLP2 may have a role outside
of the regulation of cytokinesis [183]. This study identified three other kinesins:
Eg-5, KIF21A and KIF25 in addition to MKLP2 whose depletions also caused non-
apoptotic cell death and may themselves be potential therapeutic targets [183]. In
the same cell line however, Zou et al. demonstrate MKLP2 depletion leading to
apoptotic cell death characterized by the induction of cleaved caspase and PARP
proteins [185]. They report that a large number of kinesins including MKLP2 are
upregulated in response to oestrogen stimulation in oestrogen receptor (ER)-positive
breast cancer cells. Whether MKLP2 overexpression has a causal role or is purely
consequential in carcinogenesis is not yet understood, however these findings sug-
gest it may have a role as a tumour marker and potential therapeutic target.
The first MKLP2 specific inhibitor Paprotrain (PAssenger PROtein TRAnsport
Inhibitor) has recently been described [186]. Paprotrain is a reversible inhibitor of
MKLP2, uncompetitive with ATP and non-competitive with microtubules.
Paprotrain is cell permeable and treatment of cells resulted in polyploidy and
apoptosis as well as mitotic spindle defects with failure of relocation of the chromo-
somal passenger complex. Paprotrain appears to be specific to MKLP2 and does not
inhibit MKLP1 or MPP1; however the phenotype produced by paprotrain treatment
differs slightly from MKLP2 depletion. After treatment with paprotrain the majority
of metaphase spindles had misaligned chromosomes, this differs from the pheno-
type seen with siRNA [186]. A recent paper using paprotrain to characterize oocyte
maturation demonstrated that, treatment of mouse oocytes with paprotrain resulted
in failure of polar body extrusion consistent with the role of MKLP2 in completing
cytokinesis. Spindle formation and chromosome alignment however appeared to
occur normally in meiosis-I [187]. Further molecular characterization of paprotrain
is required before investigating this compound as a therapeutic drug.
Genistein the predominant isoflavone present in soybeans has been shown to
have anticancer effects in animal models and various cancer cells including breast,
prostate, gastric and ovarian cancer cells [184]. Genistein induces a G2/M arrest and
apoptosis of cancer cells. Quantitative SILAC proteomics analysis of gastric cancer

cells treated with or without genistein identified that a number of the dysregulated
proteins were involved in the regulation of cellular proliferation, mitotic transition,
mitotic spindle, microtubule dynamics and chromosome segregation [184]. These
included five kinesins: Eg-5, MKLP1, MKLP2, KID, and CENPF. MKLP2 was the
most down regulated protein after Genistein treatment. Genistein treatment results
in delocalization of Rho A from the equatorial cortex, cleavage furrow regression,
failure of cytokinesis and multinucleate cells. Aberrant chromosome segregation,
chromosome bridges and lagging chromosomes are also seen [188]. Overexpression
of MKLP2 was able to partly rescue viability of cells treated with Genistein sug-
gesting that MKLP2 may play a potential role as a target of Genistein [188].
In recent years immunotherapy has emerged as a potential alternative way to
target cancers and MKLP2 has been identified as potential tumour associated anti-
gen (TAA) target. As discussed MKLP2 is frequently overexpressed in many can-
cers, however, amongst normal tissues is only highly expressed in testis with lower
levels in thymus and bone marrow and virtually no observed expression in all other
tissues as detected by northern blot, PCR, western blotting and immunohistochem-
istry [7, 176, 189]. Although MKLP2 is not the perfect cancer-testis antigen its
non-redundant role makes it a promising target, as cancer cells are known to escape
from the immune response through the loss of the tumour-associated antigens.
Consistent with its essential role KIF20A expression was highly conserved in both
advanced malignant melanoma and samples of metastatic lesions [180]. Short pep-
tide KIF20A epitopes have been identified that can stimulate the generation of
human cytotoxic T-lymphocytes (CTL’s) that show KIF20A-specific cytotoxicity in
both HLA-A2 and HLA-A24 restricted manners [189, 190]. These peptide-specific
CTL’s exerted potent IFN-γ production and cytotoxic activity in vitro, specifically
responding to HLA restricted cancer cells expressing KIF20A. Promisingly no
autoimmunity was observed when the peptide was tested in a mouse model [189].
Four phase-I clinical trials have now been reported investigating the same HLA-
A24 restricted KIF20A short peptide vaccine either as a single peptide [191, 192] or
in combination with other TAA’s [193, 194]. Three trails have investigated the role
of the peptide immunotherapy in pancreatic ductal adenocarcinoma [191–193] and
one looked at cholangiocarcinoma [194]. These non-randomized, open-label, single
centre trials have shown promising initial results with no serious adverse events
reported, two patients were withdrawn in the largest trial due to autoimmune hepa-
titis and interstitial pneumonia both patients recovered from these complications
however it could not be ruled out if the adverse events were related to the vaccines
or not [191]. All trials used similar protocols with 3 or 4 vaccinations during a 28
day cycle. During the first cycle stable disease was identified in 72 % [191], 44 %
[193], 55 % [194] and 44 % [192] of enrolled patients. Progression free survival
(PFS) and overall survival (OS) were similar to or better than best medical care,
despite the trial population having failed standard therapy previously [191–194].
Further phase-I clinical trials are on going and phase-II randomized trials are
planned. Developing this strategy further Tomita et al. have identified long peptides
with predicted HLA class II-binding epitopes adjacent to the CTL epitope, which
induce tumour-specific T-helper type 1 (TH1) cells in addition to CTLs [195]. This
work demonstrated the additive role that CD4+ helper cells can exert in promoting
the activity of CTL’s [195, 196].
As with MKLP1, MKLP2 is also regulated by the cell cycle-dependant element
(CDE) and cell cycle homology region (CHR) [29]. TP53 acting via the DREAM/
MMB pathways is therefore also important in regulating MKLP2 expression [197].
Supporting this notion, high levels of MKLP2 mRNA are found in HCC’s with
mutated TP53 [179]. The transcription factor Forkhead box M1 (FoxM1) can also
upregulate MKLP2 amongst other genes involved in chromosome segregation and
mitosis [198]. Elevated levels of FoxM1 have been identified in basal cell carci-
noma of the skin and in hepatocellular carcinoma [47, 198].
Despite the initial suggestion that MKLP2 may have a role as a Rab6 effector in
membrane trafficking pathways no clear role for MKLP2 has emerged outside of its
regulation of mitosis. This along with its low levels of expression in most vital tis-
sues makes MKLP2 a potentially good therapeutic choice for targeting cancer. The
major potential side effects would relate to its role in lymphopoiesis and haemopoi-
esis [7], however, adverse events related to these roles were not observed in the
vaccine trials conducted to date.
12.7 MPP1 (M-Phase Phosphoprotein 1)
MPP1 C-terminal peptide was originally identified in a screen for proteins phos-
phorylated in metaphase and Cdk1 was identified as the kinase responsible for its
phosphorylation [199]. This peptide was discovered to be the C-terminal peptide of
the kinesin MPP1 after screening of a HeLa cDNA library using autoimmune serum
from a scleroderma patient identified its cDNA [8]. Analysis of the sequence dem-
onstrated similarities to the MKLP1 and MKLP2 proteins within the putative motor
domain and a 566 amino acid sequence at the carboxy terminus was found to be
identical to the previously characterized MPP1 sequence [8, 199]. Biochemically
the protein was characterized as a slow plus end directed kinesin with microtubule
stimulated ATPase activity and the ability to bundle and slide microtubules in glid-
ing assays [9]. Unlike MKLP2, the MPP1 protein is expressed in both interphase
and mitosis [8, 200], however the protein level was demonstrated to increase two to
threefold as cells passed from G1 to G2/M phases [9]. In interphase MPP1 localises
to the nucleus and punctate cytoplasmic foci unrelated to microtubules, accumulat-
ing in the nucleus during S phase [9, 33]. In anaphase MPP1 localises as a band at
the spindle midzone between the dividing sister chromatids and in telophase local-
ises to two bands flanking the midbody analogous to MKLP2 [9, 33, 201]. The
C-terminal Cdk1 phosphorylation was mapped to T1604 and confirmed to be phos-
phorylated in mitosis [8]. MPP1 depleted cells progressed normally through mito-
sis, forming a midbody, but failed to complete cytokinesis with either regression of
the furrow or persistence of the intercellular bridge [9, 33]. This suggests a late role
for MPP1 in regulating cytokinesis and abscission. Both cells that became
multinucleate and those with persistent intercellular bridges underwent apoptosis

either immediately or delayed on the second round of cell division [9]. In contrast
to these findings Zhu et al. reported MPP1 depletion had no obvious effect on either
mitosis or cytokinesis in HeLa cells, whereas both MKLP1 and MKLP2 depletions
led to cytokinesis defects [202]. This finding is however most likely due to depletion
efficiency as MPP1 levels were only depleted by 78 % compared with over 90 %
depletion for both MKLP1 and MKLP2 [202]. Interestingly a forward genetic
screen for the mouse microcephaly, magoo, phenotype characterized by a small
cerebral cortex, with reduced production of progenitors and neurons, but preserved
layering has identified a loss of function mutation in the MPP1 gene [201]. This
mutation specifically affects the polarized neuroepithelial stem cell divisions with
changes in the apical midbody number, shape and position relative to the apical
membrane. Interestingly the mutation appears to result in apoptosis rather than
generation of multinucleate cells. The fact that many of the mouse tissues develop
normally in this mutant suggests MPP1 may not be absolutely required for
cytokinesis [201].
Transient overexpression of MPP1 in Cos-7 cells has been reported to lead to cell
cycle arrest at the G2/M boundary and subsequent cell death [8], however, Kanehira
et al. found expressing MPP1 in NIH3T3 cells promoted increased cell growth both
in vitro and in vivo [33]. The C-terminus of MPP1 has been shown to interact with
the WW phospho Ser/Thr-Pro binding module of peptidyl-propyl isomerase 1
(Pin1) [8]. Pin1 catalyses phosphorylation-dependent propyl isomerization which
leads to conformational changes regulating the function of post-phosphorylated
proteins [203]. Pin1 knockout or depletion results in prolonged cytokinesis and
increased frequency of cytokinetic defects leading to multinucleated cells [204]. It
has been proposed that this defect arises through a Cep55 mediated mechanism,
where Pin1 binds to and promotes the phosphorylation of Cep55 S436 by Plk1 dur-
ing mitosis [204]. This phosphorylation of Cep55 is known to be a key timer of
abscission [118]. In common with MKLP1 and MKLP2, MPP1 co-
immunoprecipitates with PRC1 and this interaction is likely important in stabilizing
and organizing the spindle midzone and midbody [33, 42].
12.8 MPP1 Dysregulation and Targeting in Cancer
MPP1 has been demonstrated to be upregulated in bladder cancer [33] and large B
cell lymphoma [205], as well as in hepatic, colon, cervical, non-small cell lung, and
bladder cancer cell lines [206]. Further supporting its role in oncogenesis, mouse
fibroblasts stably transfected with MPP1 grew much faster and lost contact inhibi-
tion in vitro compared with mock-transfected cells [33]. Additionally these stably
transfected cells were tumour forming when transplanted subcutaneously into
BALB/cA Jcl-nu mice suggesting MPP1 is an oncogene [33]. The MPP1 interacting
protein Pin1 is also overexpressed in many cancers including: prostate, liver, colon,
breast, cervical, lung, brain and melanoma [207]. Depletion of MPP1 significantly
suppressed the growth of J82 and UM-UC-3 bladder cancer cell lines suggesting it
may also be a potential therapeutic cancer target.
Although MPP1 is expressed in cancer cells its expression in normal tissues has
been variably reported; Kanehira et al. found MPP1 was only expressed in testis and
hardly expressed in other tissues. Abaza et al. however found MPP1 to be expressed
in brain, kidney, ovary and testis [9]. Its classification as a cancer testis antigen must
therefore be viewed cautiously. Based on HLA epitope prediction a short peptide of
MPP1 was found that potently induced CTL responses in an HLA-A24 restricted
manner [208]. The established CTL clones produced IFN-γ when mixed with HLA-
A24 positive HT1376 bladder cancer cells endogenously expressing MPP1. Clinical
trial of this MPP1 peptide vaccine in patients with pathologically confirmed bladder
cancer unresponsive to standard therapy was conducted. Vaccines were well toler-
ated without any major adverse events. Positive cytotoxic T lymphocyte (CTL)
responses against the MPHOS1-278 peptide were observed in three of the four vac-
cinated patients. The three patients with CTL responses showed either stable dis-
ease or objective response on either CT or MRI imaging whereas the patient without
CTL response demonstrated progressive disease [208]. A phase I/II study is now
underway.
MPP1 has also been targeted using Cancer Targeting Gene-Viro-Therapy
(CTGVT). This strategy aims to strengthen the anti-tumour effect of an oncolytic
virus by expressing inserted foreign anti-tumour genes [206]. Liu et al. used a novel
adenoviral vector controlled by the tumour-specific survivin promoter to deliver and
express IL-24 and short hairpin RNA, shMPP1, in tumour cells [206]. They reported
promising anti-tumour effects both in vitro and in vivo [206]. Tumour cells (SW620,
A549) infected with the virus demonstrated multinuclear cells after 24 h infection
and began to show apoptosis at 36 h after infection. Video microscopy demonstrated
multinuclear cells attempted a further cycle of mitosis, entered a prolonged mitotic
arrest before developing mitotic defects and progressing into apoptosis [206]. The
in vivo oncolytic potential of the virus was also evaluated using an SW620 tumour
xenograft model. Growth of the virus treated tumours was significantly suppressed
compared with control treated tumours after 2 months. IHC demonstrated virus
treated tumours had markedly lower MPP1 expression levels compared with control
treated tumours. The anti-tumour effect may be exerted by upregulation of IL-24,
down-regulation of MPP1, or the oncolytic activity of the virus [206]. High multi-
plicity of infection (MOI) resulted in apoptosis; low MOI resulted in cellular senes-
cence [206].
MPP1 is found in brain, kidney and ovary in addition testis [9]. As MPP1 is not
a perfect cancer-testis antigen the therapeutic benefit of targeting MPP1 must be
balanced against potential side effects in normal tissues. As discussed already kine-
sin-6 family members can have non-mitotic functions, and MPP1 has a role in
development of the normal cerebral cortex [201]. It is also highly expressed in the
developing mouse brain localizing to the tip of the growing axon and through inter-
action with its potential cargo Shootin1 plays a role in neuronal polarization and
migration [209]. Although no definite role has been demonstrated in post mitotic
neurons, autoantibodies to MPP1 are found in patients with idiopathic ataxia [210]
and the MPP1 gene localizes to an Alzheimer’s disease linkage region 10q22–24
[211]. These observations may suggest an as yet uncharacterized role for MPP1
[210, 212].
12.9 Outlook for Targeting the Kinesin-6 Family in Cancer
As a result of their essential and non-redundant roles in regulating cytokinesis kine-

sin-6 family members are attracting increasing attention as potential therapeutic
targets. Although they have conserved structures and similar spatio-temporal pat-
terns of regulation they each have distinct roles in the regulation of the central spin-
dle organization and in cytokinesis. The unique differences between kinesins results
from their specific cargo and binding partner interactions linking them to distinct
down stream regulatory pathways. Their essential role in cytokinesis makes them
attractive therapeutic targets, as cells are unable to escape this through antigen inac-
tivation, which would result in cell death. Despite their independent roles in regulat-
ing cytokinesis the outcome of targeting the kinesin-6 family members is common,
resulting in the failure of cytokinesis. Cytokinesis failure generates multinucleated
tetraploid cells. The fate of these tetraploid cells is critically important in evaluating
the therapeutic potential of targeting kinesin-6 family members. It has been well
established that tetraploid cells can undergo apoptosis either during mitosis, during
a subsequent G1 arrest or in the next mitotic cycle, through both TP53 dependent
and independent mechanisms [206, 213–217]. Tetraploidy has however also been
implicated as an intermediary state in the development of aneuploidy and tumori-
genesis [218, 219]. Following failed cytokinesis and development of tetraploidy,
cells can enter the next mitotic cycle with centrosomal duplication leading to multi-
polar or pseudo-bipolar spindles [48, 59, 215–218]. These states promote genomic
instability, chromosomal rearrangements and development of aneuploidy and these
cells are tumorigenic in mice models [218, 219]. Targeting kinesin-6 family mem-
bers may therefore prove a double-edged sword with a limited therapeutic window
[48] and further work is required to establish if cells can be pushed down the apop-
totic pathway rather than toward further genomic instability. It may be that targeting
kinesin-6 family members is useful in only a subset of cancer cell types and genetic
contexts.
References
ture and function. Trends Cell Biol 15(9):467–476. doi:10.1016/j.tcb.2005.07.006
2. Lawrence CJ, Dawe RK, Christie KR, Cleveland DW, Dawson SC, Endow SA, Goldstein LS,
Goodson HV, Hirokawa N, Howard J, Malmberg RL, McIntosh JR, Miki H, Mitchison TJ, Okada
Y, Reddy AS, Saxton WM, Schliwa M, Scholey JM, Vale RD, Walczak CE, Wordeman L (2004)
A standardized kinesin nomenclature. J Cell Biol 167(1):19–22. doi:10.1083/jcb.200408113
genes in mouse and human. Proc Natl Acad Sci U S A 98(13):7004–7011. doi:10.1073/
pnas.111145398
4. Goodson HV, Kang SJ, Endow SA (1994) Molecular phylogeny of the kinesin family of
microtubule motor proteins. J Cell Sci 107(Pt 7):1875–1884
5. Lawrence CJ, Malmberg RL, Muszynski MG, Dawe RK (2002) Maximum likelihood meth-
ods reveal conservation of function among closely related kinesin families. J Mol Evol
54(1):42–53. doi:10.1007/s00239-001-0016-y
6. Kuriyama R, Gustus C, Terada Y, Uetake Y, Matuliene J (2002) CHO1, a mammalian kinesin-
like protein, interacts with F-actin and is involved in the terminal phase of cytokinesis. J Cell
Biol 156(5):783–790. doi:10.1083/jcb.200109090
7. Lai F, Fernald AA, Zhao N, Le Beau MM (2000) cDNA cloning, expression pattern, genomic
structure and chromosomal location of RAB6KIFL, a human kinesin-like gene. Gene
248(1–2):117–125. doi:10.1016/S0378-1119(00)00135-9
8. Kamimoto T, Zama T, Aoki R, Muro Y, Hagiwara M (2001) Identification of a novel kinesin-
related protein, KRMP1, as a target for mitotic peptidyl-prolyl isomerase Pin1. J Biol Chem
276(40):37520–37528. doi:10.1074/jbc.M106207200
9. Abaza A, Soleilhac JM, Westendorf J, Piel M, Crevel I, Roux A, Pirollet F (2003) M phase
phosphoprotein 1 is a human plus-end-directed kinesin-related protein required for cytokine-
sis. J Biol Chem 278(30):27844–27852. doi:10.1074/jbc.M304522200
10. Hizlan D, Mishima M, Tittmann P, Gross H, Glotzer M, Hoenger A (2006) Structural analysis
of the ZEN-4/CeMKLP1 motor domain and its interaction with microtubules. J Struct Biol
153(1):73–84. doi:10.1016/j.jsb.2005.10.007
11. Wade RH (2002) Sequence landmark patterns identify and characterize protein families.
Structure 10(10):1329–1336. doi:10.1016/S0969-2126(02)00854-7
12. Kikkawa M, Okada Y, Hirokawa N (2000) 15 A resolution model of the monomeric kinesin
motor, KIF1A. Cell 100(2):241–252. doi:10.1016/S0092-8674(00)81562-7
13. Vale RD, Fletterick RJ (1997) The design plan of kinesin motors. Annu Rev Cell Dev Biol
13:745–777. doi:10.1146/annurev.cellbio.13.1.745
14. Wade RH, Kozielski F (2000) Structural links to kinesin directionality and movement. Nat
Struct Biol 7(6):456–460. doi:10.1038/75850
15. White EA, Glotzer M (2012) Centralspindlin: at the heart of cytokinesis. Cytoskeleton
(Hoboken) 69(11):882–892. doi:10.1002/cm.21065
16. Pavicic-Kaltenbrunner V, Mishima M, Glotzer M (2007) Cooperative assembly of CYK-4/
MgcRacGAP and ZEN-4/MKLP1 to form the centralspindlin complex. Mol Biol Cell
18(12):4992–5003. doi:10.1091/mbc.E07-05-0468
17. Kuriyama R, Dragas-Granoic S, Maekawa T, Vassilev A, Khodjakov A, Kobayashi H (1994)
Heterogeneity and microtubule interaction of the CHO1 antigen, a mitosis-specific
kinesin-like protein. Analysis of subdomains expressed in insect sf9 cells. J Cell Sci
107(Pt 12):3485–3499
18. Mishima M, Pavicic V, Gruneberg U, Nigg EA, Glotzer M (2004) Cell cycle regulation of
central spindle assembly. Nature 430(7002):908–913. doi:10.1038/nature02767
kinesin, involved in microtubule-based motility. Cell 42(1):39–50. doi:10.1016/
S0092-8674(85)80099-4
20. Case RB, Pierce DW, Hom-Booher N, Hart CL, Vale RD (1997) The directional preference
of kinesin motors is specified by an element outside of the motor catalytic domain. Cell
90(5):959–966. doi:10.1016/S0092-8674(00)80360-8
21. Endow SA, Waligora KW (1998) Determinants of kinesin motor polarity. Science
281(5380):1200–1202. doi:10.1126/science.281.5380.1200
22. Henningsen U, Schliwa M (1997) Reversal in the direction of movement of a molecular
motor. Nature 389(6646):93–96. doi:10.1038/38022
23. Rice S, Lin AW, Safer D, Hart CL, Naber N, Carragher BO, Cain SM, Pechatnikova E,
Wilson-Kubalek EM, Whittaker M, Pate E, Cooke R, Taylor EW, Milligan RA, Vale RD
(1999) A structural change in the kinesin motor protein that drives motility. Nature
402(6763):778–784. doi:10.1038/45483
24. Hill E, Clarke M, Barr FA (2000) The Rab6-binding kinesin, Rab6-KIFL, is required for
cytokinesis. EMBO J 19(21):5711–5719. doi:10.1093/emboj/19.21.5711
25. Adams RR, Tavares AA, Salzberg A, Bellen HJ, Glover DM (1998) Pavarotti encodes a
kinesin-like protein required to organize the central spindle and contractile ring for cytokine-
sis. Genes Dev 12(10):1483–1494
26. Powers J, Bossinger O, Rose D, Strome S, Saxton W (1998) A nematode kinesin required for
cleavage furrow advancement. Curr Biol 8(20):1133–1136. doi:10.1016/
S0960-9822(98)70470-1
27. Raich WB, Moran AN, Rothman JH, Hardin J (1998) Cytokinesis and midzone microtubule
organization in Caenorhabditis elegans require the kinesin-like protein ZEN-4. Mol Biol Cell
9(8):2037–2049. doi:10.1091/mbc.9.8.2037
28. Nislow C, Sellitto C, Kuriyama R, McIntosh JR (1990) A monoclonal antibody to a mitotic
microtubule-associated protein blocks mitotic progression. J Cell Biol 111(2):511–522.
doi:10.1083/jcb.111.2.511
29. Fontijn RD, Goud B, Echard A, Jollivet F, van Marle J, Pannekoek H, Horrevoets AJ (2001)
The human kinesin-like protein RB6K is under tight cell cycle control and is essential for
cytokinesis. Mol Cell Biol 21(8):2944–2955. doi:10.1128/MCB.21.8.2944-2955.2001
30. Deavours BE, Walker RA (1999) Nuclear localization of C-terminal domains of the kinesin-
like protein MKLP-1. Biochem Biophys Res Commun 260(3):605–608. doi:10.1006/
bbrc.1999.0952
31. Sellitto C, Kuriyama R (1988) Distribution of a matrix component of the midbody during the
cell cycle in Chinese hamster ovary cells. J Cell Biol 106(2):431–439. doi:10.1083/
jcb.106.2.431
32. Nislow C, Lombillo VA, Kuriyama R, McIntosh JR (1992) A plus-end-directed motor
enzyme that moves antiparallel microtubules in vitro localizes to the interzone of mitotic
spindles. Nature 359(6395):543–547. doi:10.1038/359543a0
33. Kanehira M, Katagiri T, Shimo A, Takata R, Shuin T, Miki T, Fujioka T, Nakamura Y (2007)
Oncogenic role of MPHOSPH1, a cancer-testis antigen specific to human bladder cancer.
Cancer Res 67(7):3276–3285. doi:10.1158/0008-5472.CAN-06-3748
34. Hu CK, Coughlin M, Field CM, Mitchison TJ (2008) Cell polarization during monopolar
cytokinesis. J Cell Biol 181(2):195–202. doi:10.1083/jcb.200711105
KIF4 and its binding partner PRC1 in organized central spindle midzone formation. EMBO
J 23(16):3237–3248. doi:10.1038/sj.emboj.7600347
36. Glotzer M (2009) The 3Ms of central spindle assembly: microtubules, motors and MAPs. Nat
Rev Mol Cell Biol 10(1):9–20. doi:10.1038/nrm2609
37. Neef R, Klein UR, Kopajtich R, Barr FA (2006) Cooperation between mitotic kinesins controls
the late stages of cytokinesis. Curr Biol 16(3):301–307. doi:10.1016/j.cub.2005.12.030
38. Kitagawa M, Fung SY, Hameed UF, Goto H, Inagaki M, Lee SH (2014) Cdk1 coordinates
timely activation of MKLP2 kinesin with relocation of the chromosome passenger complex
for cytokinesis. Cell Rep 7(1):166–179. doi:10.1016/j.celrep.2014.02.034
39. Hummer S, Mayer TU (2009) Cdk1 negatively regulates midzone localization of the mitotic
kinesin MKLP2 and the chromosomal passenger complex. Curr Biol 19(7):607–612.
doi:10.1016/j.cub.2009.02.046
40. Echard A, Jollivet F, Martinez O, Lacapere JJ, Rousselet A, Janoueix-Lerosey I, Goud B
(1998) Interaction of a Golgi-associated kinesin-like protein with Rab6. Science
279(5350):580–585. doi:10.1126/science.279.5350.580
41. Fu C, Ward JJ, Loiodice I, Velve-Casquillas G, Nedelec FJ, Tran PT (2009) Phospho-
regulated interaction between kinesin-6 Klp9p and microtubule bundler Ase1p promotes
spindle elongation. Dev Cell 17(2):257–267. doi:10.1016/j.devcel.2009.06.012
42. Nunes Bastos R, Gandhi SR, Baron RD, Gruneberg U, Nigg EA, Barr FA (2013) Aurora B
suppresses microtubule dynamics and limits central spindle size by locally activating
KIF4A. J Cell Biol 202(4):605–621. doi:10.1083/jcb.201301094
43. Barton NR, Goldstein LS (1996) Going mobile: microtubule motors and chromosome
segregation. Proc Natl Acad Sci U S A 93(5):1735–1742
molecular assays. Semin Cell Dev Biol 21(3):260–268. doi:10.1016/j.semcdb.2010.01.018
45. Mishima M, Kaitna S, Glotzer M (2002) Central spindle assembly and cytokinesis require a
kinesin-like protein/RhoGAP complex with microtubule bundling activity. Dev Cell 2(1):41–
54. doi:10.1016/S1534-5807(01)00110-1
46. Gruneberg U, Neef R, Honda R, Nigg EA, Barr FA (2004) Relocation of Aurora B from
centromeres to the central spindle at the metaphase to anaphase transition requires MKLP2.
J Cell Biol 166(2):167–172. doi:10.1083/jcb.200403084
47. Yu Y, Feng YM (2010) The role of kinesin family proteins in tumorigenesis and progression:
potential biomarkers and molecular targets for cancer therapy. Cancer 116(22):5150–5160.
doi:10.1002/cncr.25461
48. Rath O, Kozielski F (2012) Kinesins and cancer. Nat Rev Cancer 12(8):527–539. doi:10.1038/
nrc3310
49. Goldstein LS (1991) The kinesin superfamily: tails of functional redundancy. Trends Cell
Biol 1(4):93–98. doi:10.1016/0962-8924(91)90036-9
50. Endow SA (1991) The emerging kinesin family of microtubule motor proteins. Trends
Biochem Sci 16(6):221–225. doi:10.1016/0968-0004(91)90089-E
51. Kuriyama R, Nislow C (1992) Molecular components of the mitotic spindle. Bioessays
14(2):81–88. doi:10.1002/bies.950140203
52. Matuliene J, Kuriyama R (2002) Kinesin-like protein CHO1 is required for the formation of
midbody matrix and the completion of cytokinesis in mammalian cells. Mol Biol Cell
13(6):1832–1845. doi:10.1091/mbc.01-10-0504
53. Zhu C, Bossy-Wetzel E, Jiang W (2005) Recruitment of MKLP1 to the spindle midzone/
midbody by INCENP is essential for midbody formation and completion of cytokinesis in
human cells. Biochem J 389(Pt 2):373–381. doi:10.1042/BJ20050097
54. Matuliene J, Kuriyama R (2004) Role of the midbody matrix in cytokinesis: RNAi
and genetic rescue analysis of the mammalian motor protein CHO1. Mol Biol Cell
15(7):3083–3094. doi:10.1091/mbc.E03-12-0888
55. Toure A, Dorseuil O, Morin L, Timmons P, Jegou B, Reibel L, Gacon G (1998) MgcRacGAP,
a new human GTPase-activating protein for Rac and Cdc42 similar to Drosophila rotundRac-
GAP gene product, is expressed in male germ cells. J Biol Chem 273(11):6019–6023
56. Jantsch-Plunger V, Gonczy P, Romano A, Schnabel H, Hamill D, Schnabel R, Hyman AA,
Glotzer M (2000) CYK-4: a Rho family gtpase activating protein (GAP) required for central
spindle formation and cytokinesis. J Cell Biol 149(7):1391–1404. doi:10.1083/jcb.149.7.1391
57. Miyauchi K, Zhu X, Foong C, Hosoya H, Murata-Hori M (2007) Aurora B kinase activity is
required to prevent polar cortical ingression during cytokinesis. Cell Cycle 6(20):2549–2553.
doi:10.4161/cc.6.20.4817
58. White EA, Raghuraman H, Perozo E, Glotzer M (2013) Binding of the CYK-4 subunit of the
centralspindlin complex induces a large scale conformational change in the kinesin subunit.
J Biol Chem 288(27):19785–19795. doi:10.1074/jbc.M113.463695
59. Barr FA, Gruneberg U (2007) Cytokinesis: placing and making the final cut. Cell
131(5):847–860. doi:10.1016/j.cell.2007.11.011
60. Glotzer M (2005) The molecular requirements for cytokinesis. Science 307(5716):1735–1739.
doi:10.1126/science.1096896
61. Kikkawa M, Sablin EP, Okada Y, Yajima H, Fletterick RJ, Hirokawa N (2001) Switch-based
mechanism of kinesin motors. Nature 411(6836):439–445. doi:10.1038/35078000
62. Parry DH, O’Farrell PH (2001) The schedule of destruction of three mitotic cyclins can dic-
tate the timing of events during exit from mitosis. Curr Biol 11(9):671–683. doi:10.1016/
S0960-9822(01)00204-4
63. Wheatley SP, Hinchcliffe EH, Glotzer M, Hyman AA, Sluder G, Wang Y (1997) CDK1 inac-
tivation regulates anaphase spindle dynamics and cytokinesis in vivo. J Cell Biol 138(2):
385–393. doi:10.1083/jcb.138.2.385
64. Minestrini G, Harley AS, Glover DM (2003) Localization of Pavarotti-KLP in living

Drosophila embryos suggests roles in reorganizing the cortical cytoskeleton during the
mitotic cycle. Mol Biol Cell 14(10):4028–4038. doi:10.1091/mbc.E03-04-0214
65. Hutterer A, Glotzer M, Mishima M (2009) Clustering of centralspindlin is essential
for its accumulation to the central spindle and the midbody. Curr Biol 19(23):2043–2049.
doi:10.1016/j.cub.2009.10.050
66. Douglas ME, Davies T, Joseph N, Mishima M (2010) Aurora B and 14-3-3 coordinately
regulate clustering of centralspindlin during cytokinesis. Curr Biol 20(10):927–933.
doi:10.1016/j.cub.2010.03.055
67. Guse A, Mishima M, Glotzer M (2005) Phosphorylation of ZEN-4/MKLP1 by aurora B regu-
lates completion of cytokinesis. Curr Biol 15(8):778–786. doi:10.1016/j.cub.2005.03.041
68. Schumacher JM, Golden A, Donovan PJ (1998) AIR-2: an Aurora/Ipl1-related protein kinase
associated with chromosomes and midbody microtubules is required for polar body extrusion
and cytokinesis in Caenorhabditis elegans embryos. J Cell Biol 143(6):1635–1646.
doi:10.1083/jcb.143.6.1635
69. Severson AF, Hamill DR, Carter JC, Schumacher J, Bowerman B (2000) The aurora-related
kinase AIR-2 recruits ZEN-4/CeMKLP1 to the mitotic spindle at metaphase and is required
for cytokinesis. Curr Biol 10(19):1162–1171. doi:10.1016/S0960-9822(00)00715-6
70. Kaitna S, Mendoza M, Jantsch-Plunger V, Glotzer M (2000) INCENP and an aurora-like
kinase form a complex essential for chromosome segregation and efficient completion of
cytokinesis. Curr Biol 10(19):1172–1181. doi:10.1016/S0960-9822(00)00721-1
71. Vagnarelli P, Earnshaw WC (2004) Chromosomal passengers: the four-dimensional regula-
tion of mitotic events. Chromosoma 113(5):211–222. doi:10.1007/s00412-004-0307-3
72. Gruneberg U, Glotzer M, Gartner A, Nigg EA (2002) The CeCDC-14 phosphatase is required
for cytokinesis in the Caenorhabditis elegans embryo. J Cell Biol 158(5):901–914.
doi:10.1083/jcb.200202054
73. Zhao WM, Seki A, Fang G (2006) Cep55, a microtubule-bundling protein, associates with
centralspindlin to control the midbody integrity and cell abscission during cytokinesis. Mol
Biol Cell 17(9):3881–3896. doi:10.1091/mbc.E06-01-0015
74. Kamijo K, Ohara N, Abe M, Uchimura T, Hosoya H, Lee JS, Miki T (2006) Dissecting the
role of Rho-mediated signaling in contractile ring formation. Mol Biol Cell 17(1):43–55.
doi:10.1091/mbc.E05-06-0569
75. Nishimura Y, Yonemura S (2006) Centralspindlin regulates ECT2 and RhoA accumulation at
the equatorial cortex during cytokinesis. J Cell Sci 119(Pt 1):104–114. doi:10.1242/ jcs.02737
76. Yuce O, Piekny A, Glotzer M (2005) An ECT2-centralspindlin complex regulates the local-
ization and function of RhoA. J Cell Biol 170(4):571–582. doi:10.1083/jcb.200501097
77. Simon GC, Schonteich E, Wu CC, Piekny A, Ekiert D, Yu X, Gould GW, Glotzer M, Prekeris
R (2008) Sequential Cyk-4 binding to ECT2 and FIP3 regulates cleavage furrow ingression
and abscission during cytokinesis. EMBO J 27(13):1791–1803. doi:10.1038/emboj.2008.112
78. Alsop GB, Zhang D (2003) Microtubules are the only structural constituent of the spindle
apparatus required for induction of cell cleavage. J Cell Biol 162(3):383–390. doi:10.1083/
jcb.200301073
79. Bringmann H, Hyman AA (2005) A cytokinesis furrow is positioned by two consecutive
signals. Nature 436(7051):731–734. doi:10.1038/nature03823
80. Dechant R, Glotzer M (2003) Centrosome separation and central spindle assembly act in
redundant pathways that regulate microtubule density and trigger cleavage furrow formation.
Dev Cell 4(3):333–344. doi:10.1016/S1534-5807(03)00057-1
81. Werner M, Munro E, Glotzer M (2007) Astral signals spatially bias cortical myosin recruit-
ment to break symmetry and promote cytokinesis. Curr Biol 17(15):1286–1297. doi:10.1016/j.
cub.2007.06.070
82. Rappaport R (1985) Repeated furrow formation from a single mitotic apparatus in cylindrical
sand dollar eggs. J Exp Zool 234(1):167–171. doi:10.1002/jez.1402340120
83. Bastos RN, Penate X, Bates M, Hammond D, Barr FA (2012) CYK4 inhibits Rac1-dependent
PAK1 and ARHGEF7 effector pathways during cytokinesis. J Cell Biol 198(5):865–880.
doi:10.1083/jcb.201204107
84. Bement WM, Benink HA, von Dassow G (2005) A microtubule-dependent zone of active
RhoA during cleavage plane specification. J Cell Biol 170(1):91–101. doi:10.1083/
jcb.200501131
85. Piekny A, Werner M, Glotzer M (2005) Cytokinesis: welcome to the Rho zone. Trends Cell
Biol 15(12):651–658. doi:10.1016/j.tcb.2005.10.006
86. Miller AL, Bement WM (2009) Regulation of cytokinesis by Rho GTPase flux. Nat Cell Biol
11(1):71–77. doi:10.1038/ncb1814
87. Prokopenko SN, Brumby A, O’Keefe L, Prior L, He Y, Saint R, Bellen HJ (1999) A putative
exchange factor for Rho1 GTPase is required for initiation of cytokinesis in Drosophila.
Genes Dev 13(17):2301–2314
88. Tatsumoto T, Xie X, Blumenthal R, Okamoto I, Miki T (1999) Human ECT2 is an exchange
factor for Rho GTPases, phosphorylated in G2/M phases, and involved in cytokinesis. J Cell
Biol 147(5):921–928. doi:10.1083/jcb.147.5.921
89. Zavortink M, Contreras N, Addy T, Bejsovec A, Saint R (2005) Tum/RacGAP50C provides a
critical link between anaphase microtubules and the assembly of the contractile ring in
Drosophila melanogaster. J Cell Sci 118(Pt 22):5381–5392. doi:10.1242/jcs.02652
90. Somers WG, Saint R (2003) A RhoGEF and Rho family GTPase-activating protein complex
links the contractile ring to cortical microtubules at the onset of cytokinesis. Dev Cell
4(1):29–39. doi:10.1016/S1534-5807(02)00402-1
91. Chalamalasetty RB, Hummer S, Nigg EA, Sillje HH (2006) Influence of human Ect2
depletion and overexpression on cleavage furrow formation and abscission. J Cell Sci
119(Pt 14):3008–3019. doi:10.1242/ jcs.03032
92. Petronczki M, Glotzer M, Kraut N, Peters JM (2007) Polo-like kinase 1 triggers the initiation
of cytokinesis in human cells by promoting recruitment of the RhoGEF Ect2 to the central
spindle. Dev Cell 12(5):713–725. doi:10.1016/j.devcel.2007.03.013
93. Burkard ME, Maciejowski J, Rodriguez-Bravo V, Repka M, Lowery DM, Clauser KR, Zhang
C, Shokat KM, Carr SA, Yaffe MB, Jallepalli PV (2009) Plk1 self-organization and priming
phosphorylation of HsCYK-4 at the spindle midzone regulate the onset of division in human
cells. PLoS Biol 7(5):e1000111. doi:10.1371/journal.pbio.1000111
94. Wolfe BA, Takaki T, Petronczki M, Glotzer M (2009) Polo-like kinase 1 directs assembly of
the HsCyk-4 RhoGAP/Ect2 RhoGEF complex to initiate cleavage furrow formation. PLoS
Biol 7(5):e1000110. doi:10.1371/journal.pbio.1000110
95. Saito S, Liu XF, Kamijo K, Raziuddin R, Tatsumoto T, Okamoto I, Chen X, Lee CC, Lorenzi
MV, Ohara N, Miki T (2004) Deregulation and mislocalization of the cytokinesis regulator
ECT2 activate the Rho signaling pathways leading to malignant transformation. J Biol Chem
279(8):7169–7179. doi:10.1074/jbc.M306725200
96. Kim JE, Billadeau DD, Chen J (2005) The tandem BRCT domains of Ect2 are required for
both negative and positive regulation of Ect2 in cytokinesis. J Biol Chem 280(7):5733–5739.
doi:10.1074/jbc.M409298200
97. Dvorsky R, Ahmadian MR (2004) Always look on the bright site of Rho: structural
implications for a conserved intermolecular interface. EMBO Rep 5(12):1130–1136.
doi:10.1038/sj.embor.7400293
98. D’Avino PP, Savoian MS, Glover DM (2004) Mutations in sticky lead to defective organiza-
tion of the contractile ring during cytokinesis and are enhanced by Rho and suppressed by
Rac. J Cell Biol 166(1):61–71. doi:10.1083/jcb.200402157
99. Mishima M, Glotzer M (2003) Cytokinesis: a logical GAP. Curr Biol 13(15):R589–R591.
doi:10.1016/S0960-9822(03)00521-9
100. Glotzer M (2009) Cytokinesis: GAP gap. Curr Biol 19(4):R162–R165. doi:10.1016/j.
cub.2008.12.028
101. Goldstein AY, Jan YN, Luo L (2005) Function and regulation of Tumbleweed (RacGAP50C)
in neuroblast proliferation and neuronal morphogenesis. Proc Natl Acad Sci U S A
102(10):3834–3839. doi:10.1073/pnas.0500748102
102. Yamada T, Hikida M, Kurosaki T (2006) Regulation of cytokinesis by mgcRacGAP in B
lymphocytes is independent of GAP activity. Exp Cell Res 312(18):3517–3525. doi:10.1016/j.
yexcr.2006.07.026
103. Canman JC, Lewellyn L, Laband K, Smerdon SJ, Desai A, Bowerman B, Oegema K (2008)
Inhibition of Rac by the GAP activity of centralspindlin is essential for cytokinesis. Science
322(5907):1543–1546. doi:10.1126/science.1163086
104. Minoshima Y, Kawashima T, Hirose K, Tonozuka Y, Kawajiri A, Bao YC, Deng X, Tatsuka
M, Narumiya S, May WS Jr, Nosaka T, Semba K, Inoue T, Satoh T, Inagaki M, Kitamura T
(2003) Phosphorylation by aurora B converts MgcRacGAP to a RhoGAP during cytokinesis.
Dev Cell 4(4):549–560. doi:10.1016/S1534-5807(03)00089-3
105. Hu CK, Coughlin M, Mitchison TJ (2012) Midbody assembly and its regulation during cyto-
kinesis. Mol Biol Cell 23(6):1024–1034. doi:10.1091/mbc.E11-08-0721
106. Boman AL, Kuai J, Zhu X, Chen J, Kuriyama R, Kahn RA (1999) Arf proteins bind to mitotic
kinesin-like protein 1 (MKLP1) in a GTP-dependent fashion. Cell Motil Cytoskeleton
44(2):119–132. doi:10.1002/(SICI)1097-0169(199910)44:2<119::AID-CM4>3.0.CO;2-C
107. Lekomtsev S, Su KC, Pye VE, Blight K, Sundaramoorthy S, Takaki T, Collinson LM,
Cherepanov P, Divecha N, Petronczki M (2012) Centralspindlin links the mitotic spindle to
the plasma membrane during cytokinesis. Nature 492(7428):276–279. doi:10.1038/
nature11773
108. D’Souza-Schorey C, Chavrier P (2006) ARF proteins: roles in membrane traffic and beyond.
Nat Rev Mol Cell Biol 7(5):347–358. doi:10.1038/nrm1910
109. Makyio H, Ohgi M, Takei T, Takahashi S, Takatsu H, Katoh Y, Hanai A, Ueda T, Kanaho Y,
Xie Y, Shin HW, Kamikubo H, Kataoka M, Kawasaki M, Kato R, Wakatsuki S, Nakayama K
(2012) Structural basis for Arf6-MKLP1 complex formation on the Flemming body respon-
sible for cytokinesis. EMBO J 31(11):2590–2603. doi:10.1038/emboj.2012.89
110. Joseph N, Hutterer A, Poser I, Mishima M (2012) ARF6 GTPase protects the post-mitotic
midbody from 14-3-3-mediated disintegration. EMBO J 31(11):2604–2614. doi: 10.1038/
emboj.2012.139
111. D’Avino PP, Takeda T, Capalbo L, Zhang W, Lilley KS, Laue ED, Glover DM (2008)
Interaction between Anillin and RacGAP50C connects the actomyosin contractile ring
with spindle microtubules at the cell division site. J Cell Sci 121(Pt 8):1151–1158.
doi:10.1242/ jcs.026716
112. Frenette P, Haines E, Loloyan M, Kinal M, Pakarian P, Piekny A (2012) An anillin-Ect2
complex stabilizes central spindle microtubules at the cortex during cytokinesis. PLoS One
113. Fabbro M, Zhou BB, Takahashi M, Sarcevic B, Lal P, Graham ME, Gabrielli BG, Robinson
PJ, Nigg EA, Ono Y, Khanna KK (2005) Cdk1/Erk2- and Plk1-dependent phosphorylation of
a centrosome protein, Cep55, is required for its recruitment to midbody and cytokinesis. Dev
Cell 9(4):477–488. doi:10.1016/j.devcel.2005.09.003
114. Carlton JG, Martin-Serrano J (2007) Parallels between cytokinesis and retroviral budding: a
role for the ESCRT machinery. Science 316(5833):1908–1912. doi:10.1126/science.1143422
115. Morita E, Sandrin V, Chung HY, Morham SG, Gygi SP, Rodesch CK, Sundquist WI (2007)
Human ESCRT and ALIX proteins interact with proteins of the midbody and function in
cytokinesis. EMBO J 26(19):4215–4227. doi:10.1038/sj.emboj.7601850
116. Elia N, Sougrat R, Spurlin TA, Hurley JH, Lippincott-Schwartz J (2011) Dynamics of endo-
somal sorting complex required for transport (ESCRT) machinery during cytokinesis and its
role in abscission. Proc Natl Acad Sci U S A 108(12):4846–4851. doi:10.1073/pnas.1102714108
117. Guizetti J, Schermelleh L, Mantler J, Maar S, Poser I, Leonhardt H, Muller-Reichert T,
Gerlich DW (2011) Cortical constriction during abscission involves helices of ESCRT-III-
dependent filaments. Science 331(6024):1616–1620. doi:10.1126/science.1201847
118. Bastos RN, Barr FA (2010) Plk1 negatively regulates Cep55 recruitment to the midbody to
ensure orderly abscission. J Cell Biol 191(4):751–760. doi:10.1083/jcb.201008108
119. Mullins JM, Biesele JJ (1977) Terminal phase of cytokinesis in D-98 s cells. J Cell Biol
73(3):672–684. doi:10.1083/jcb.73.3.672
120. Kuo TC, Chen CT, Baron D, Onder TT, Loewer S, Almeida S, Weismann CM, Xu P,
Houghton JM, Gao FB, Daley GQ, Doxsey S (2011) Midbody accumulation through evasion
of autophagy contributes to cellular reprogramming and tumorigenicity. Nat Cell Biol
13(10):1214–1223. doi:10.1038/ncb2332
121. Gromley A, Yeaman C, Rosa J, Redick S, Chen CT, Mirabelle S, Guha M, Sillibourne J,
Doxsey SJ (2005) Centriolin anchoring of exocyst and SNARE complexes at the midbody is
required for secretory-vesicle-mediated abscission. Cell 123(1):75–87. doi:10.1016/j.
cell.2005.07.027
122. Pohl C, Jentsch S (2009) Midbody ring disposal by autophagy is a post-abscission event of
cytokinesis. Nat Cell Biol 11(1):65–70. doi:10.1038/ncb1813
123. Goss JW, Toomre DK (2008) Both daughter cells traffic and exocytose membrane at the
cleavage furrow during mammalian cytokinesis. J Cell Biol 181(7):1047–1054. doi:10.1083/
jcb.200712137
124. Valk K, Vooder T, Kolde R, Reintam MA, Petzold C, Vilo J, Metspalu A (2010) Gene expres-
sion profiles of non-small cell lung cancer: survival prediction and new biomarkers. Oncology
79(3–4):283–292. doi:10.1159/000322116
125. Takahashi S, Fusaki N, Ohta S, Iwahori Y, Iizuka Y, Inagawa K, Kawakami Y, Yoshida K,
Toda M (2012) Downregulation of KIF23 suppresses glioma proliferation. J Neurooncol
106(3):519–529. doi:10.1007/s11060-011-0706-2
126. Cifola I, Pietrelli A, Consolandi C, Severgnini M, Mangano E, Russo V, De Bellis G, Battaglia
C (2013) Comprehensive genomic characterization of cutaneous malignant melanoma cell
lines derived from metastatic lesions by whole-exome sequencing and SNP array profiling.
PLoS One 8(5):e63597. doi:10.1371/journal.pone.0063597
127. DeRycke MS, Gunawardena SR, Middha S, Asmann YW, Schaid DJ, McDonnell SK, Riska
SM, Eckloff BW, Cunningham JM, Fridley BL, Serie DJ, Bamlet WR, Cicek MS, Jenkins
MA, Duggan DJ, Buchanan D, Clendenning M, Haile RW, Woods MO, Gallinger SN, Casey
G, Potter JD, Newcomb PA, Le Marchand L, Lindor NM, Thibodeau SN, Goode EL (2013)
Identification of novel variants in colorectal cancer families by high-throughput exome
sequencing. Cancer Epidemiol Biomarkers Prev 22(7):1239–1251. doi:10.1158/1055-9965.
EPI-12-1226
128. Liljeholm M, Irvine AF, Vikberg AL, Norberg A, Month S, Sandstrom H, Wahlin A,
Mishima M, Golovleva I (2013) Congenital dyserythropoietic anemia type III (CDA III) is
caused by a mutation in kinesin family member, KIF23. Blood 121(23):4791–4799.
doi:10.1182/blood-2012-10-461392
129. Sandstrom H, Wahlin A, Eriksson M, Bergstrom I, Wickramasinghe SN (1994) Intravascular
haemolysis and increased prevalence of myeloma and monoclonal gammopathy in
congenital dyserythropoietic anaemia, type III. Eur J Haematol 52(1):42–46.
doi:10.1111/j.1600-0609.1994.tb01283.x
130. Fischer M, Grundke I, Sohr S, Quaas M, Hoffmann S, Knorck A, Gumhold C, Rother K
(2013) p53 and cell cycle dependent transcription of kinesin family member 23 (KIF23) is
controlled via a CHR promoter element bound by DREAM and MMB complexes. PLoS One
131. Lyberopoulou A, Venieris E, Mylonis I, Chachami G, Pappas I, Simos G, Bonanou S,
Georgatsou E (2007) MgcRacGAP interacts with HIF-1alpha and regulates its transcriptional
activity. Cell Physiol Biochem 20(6):995–1006. doi:10.1159/000110460
132. Jones WM, Chao AT, Zavortink M, Saint R, Bejsovec A (2010) Cytokinesis proteins Tum and
Pav have a nuclear role in Wnt regulation. J Cell Sci 123(Pt 13):2179–2189. doi:10.1242/
jcs.067868
133. Ferhat L, Cook C, Chauviere M, Harper M, Kress M, Lyons GE, Baas PW (1998) Expression
of the mitotic motor protein Eg5 in postmitotic neurons: implications for neuronal develop-
ment. J Neurosci 18(19):7822–7835
134. Van de Putte T, Zwijsen A, Lonnoy O, Rybin V, Cozijnsen M, Francis A, Baekelandt V, Kozak
CA, Zerial M, Huylebroeck D (2001) Mice with a homozygous gene trap vector insertion in
mgcRacGAP die during pre-implantation development. Mech Dev 102(1–2):33–44.
doi:10.1016/S0925-4773(01)00279-9
135. Ferhat L, Kuriyama R, Lyons GE, Micales B, Baas PW (1998) Expression of the mitotic
motor protein CHO1/MKLP1 in postmitotic neurons. Eur J Neurosci 10(4):1383–1393.
doi:10.1046/j.1460-9568.1998.00159.x
136. Sharp DJ, Yu W, Ferhat L, Kuriyama R, Rueger DC, Baas PW (1997) Identification of a
microtubule-associated motor protein essential for dendritic differentiation. J Cell Biol
138(4):833–843. doi:10.1083/jcb.138.4.833
137. Yu W, Sharp DJ, Kuriyama R, Mallik P, Baas PW (1997) Inhibition of a mitotic motor
compromises the formation of dendrite-like processes from neuroblastoma cells. J Cell Biol
136(3):659–668. doi:10.1083/jcb.136.3.659
138. Yu W, Cook C, Sauter C, Kuriyama R, Kaplan PL, Baas PW (2000) Depletion of a
microtubule-associated motor protein induces the loss of dendritic identity. J Neurosci
20(15):5782–5791.
139. Baas PW, Black MM, Banker GA (1989) Changes in microtubule polarity orientation during
the development of hippocampal neurons in culture. J Cell Biol 109(6 Pt 1):3085–3094.
doi:10.1083/jcb.109.6.3085
140. Baas PW, Deitch JS, Black MM, Banker GA (1988) Polarity orientation of microtubules in
hippocampal neurons: uniformity in the axon and nonuniformity in the dendrite. Proc Natl
Acad Sci U S A 85(21):8335–8339
141. Sharp DJ, Kuriyama R, Essner R, Baas PW (1997) Expression of a minus-end-directed motor
protein induces Sf9 cells to form axon-like processes with uniform microtubule polarity
orientation. J Cell Sci 110(Pt 19):2373–2380
142. Xu X, He C, Zhang Z, Chen Y (2006) MKLP1 requires specific domains for its dendritic
targeting. J Cell Sci 119(Pt 3):452–458. doi:10.1242/ jcs.02750
143. Kobayashi N, Reiser J, Kriz W, Kuriyama R, Mundel P (1998) Nonuniform microtubular
polarity established by CHO1/MKLP1 motor protein is necessary for process formation of
podocytes. J Cell Biol 143(7):1961–1970. doi:10.1083/jcb.143.7.1961
144. Smith KR, Kieserman EK, Wang PI, Basten SG, Giles RH, Marcotte EM, Wallingford JB
(2011) A role for central spindle proteins in cilia structure and function. Cytoskeleton
(Hoboken) 68(2):112–124. doi:10.1002/cm.20498
145. Portereiko MF, Saam J, Mango SE (2004) ZEN-4/MKLP1 is required to polarize the foregut
epithelium. Curr Biol 14(11):932–941. doi:10.1016/j.cub.2004.05.052
146. Hardin J, King R, Thomas-Virnig C, Raich WB (2008) Zygotic loss of ZEN-4/MKLP1
results in disruption of epidermal morphogenesis in the C. elegans embryo. Dev Dyn
237(3):830–836. doi:10.1002/dvdy.21455
147. Haglund K, Nezis IP, Stenmark H (2011) Structure and functions of stable intercellular
bridges formed by incomplete cytokinesis during development. Commun Integr Biol 4(1):1–9.
doi:10.4161/cib.4.1.13550
148. Echard A, Opdam FJ, de Leeuw HJ, Jollivet F, Savelkoul P, Hendriks W, Voorberg J, Goud B,
Fransen JA (2000) Alternative splicing of the human Rab6A gene generates two close but
functionally different isoforms. Mol Biol Cell 11(11):3819–3833. doi:10.1091/
mbc.11.11.3819
149. Neef R, Preisinger C, Sutcliffe J, Kopajtich R, Nigg EA, Mayer TU, Barr FA (2003)
Phosphorylation of mitotic kinesin-like protein 2 by polo-like kinase 1 is required for cytoki-
nesis. J Cell Biol 162(5):863–875. doi:10.1083/jcb.200306009
150. Neef R, Gruneberg U, Kopajtich R, Li X, Nigg EA, Sillje H, Barr FA (2007) Choice of Plk1
docking partners during mitosis and cytokinesis is controlled by the activation state of Cdk1.
Nat Cell Biol 9(4):436–444. doi:10.1038/ncb1557
151. Elia AE, Cantley LC, Yaffe MB (2003) Proteomic screen finds pSer/pThr-binding domain
localizing Plk1 to mitotic substrates. Science 299(5610):1228–1231. doi:10.1126/
science.1079079
152. Elia AE, Rellos P, Haire LF, Chao JW, Ivins FJ, Hoepker K, Mohammad D, Cantley LC,
Smerdon SJ, Yaffe MB (2003) The molecular basis for phosphodependent substrate
targeting and regulation of Plks by the polo-box domain. Cell 115(1):83–95. doi:10.1016/
S0092-8674(03)00725-6
153. Liu X, Zhou T, Kuriyama R, Erikson RL (2004) Molecular interactions of polo-like-kinase 1
with the mitotic kinesin-like protein CHO1/MKLP-1. J Cell Sci 117(Pt 15):3233–3246.
doi:10.1242/jcs.01173
154. Li J, Wang J, Jiao H, Liao J, Xu X (2010) Cytokinesis and cancer: polo loves ROCK‘n’
Rho(A). J Genet Genomics 37(3):159–172. doi:10.1016/S1673-8527(09)60034-5
155. Lowery DM, Clauser KR, Hjerrild M, Lim D, Alexander J, Kishi K, Ong SE, Gammeltoft S,
Carr SA, Yaffe MB (2007) Proteomic screen defines the polo-box domain interactome and
identifies Rock2 as a Plk1 substrate. EMBO J 26(9):2262–2273. doi:10.1038/sj.
emboj.7601683
156. Wheatley SP, Carvalho A, Vagnarelli P, Earnshaw WC (2001) INCENP is required for proper
targeting of Survivin to the centromeres and the anaphase spindle during mitosis. Curr Biol
11(11):886–890. doi:10.1016/S0960-9822(01)00238-X
157. Adams RR, Maiato H, Earnshaw WC, Carmena M (2001) Essential roles of Drosophila inner
centromere protein (INCENP) and aurora B in histone H3 phosphorylation, metaphase chro-
mosome alignment, kinetochore disjunction, and chromosome segregation. J Cell Biol
153(4):865–880. doi:10.1083/jcb.153.4.865
158. Bolton MA, Lan W, Powers SE, McCleland ML, Kuang J, Stukenberg PT (2002) Aurora B
kinase exists in a complex with survivin and INCENP and its kinase activity is stimulated by
survivin binding and phosphorylation. Mol Biol Cell 13(9):3064–3077. doi:10.1091/mbc.
E02-02-0092
159. Sessa F, Mapelli M, Ciferri C, Tarricone C, Areces LB, Schneider TR, Stukenberg PT,
Musacchio A (2005) Mechanism of Aurora B activation by INCENP and inhibition by hes-
peradin. Mol Cell 18(3):379–391. doi:10.1016/j.molcel.2005.03.031
160. Jeyaprakash AA, Klein UR, Lindner D, Ebert J, Nigg EA, Conti E (2007) Structure of a
Survivin-Borealin-INCENP core complex reveals how chromosomal passengers travel
together. Cell 131(2):271–285. doi:10.1016/j.cell.2007.07.045
161. Xu Z, Ogawa H, Vagnarelli P, Bergmann JH, Hudson DF, Ruchaud S, Fukagawa T, Earnshaw
WC, Samejima K (2009) INCENP-aurora B interactions modulate kinase activity and chromo-
some passenger complex localization. J Cell Biol 187(5):637–653. doi:10.1083/jcb.200906053
162. Lee SH, McCormick F, Saya H (2010) Mad2 inhibits the mitotic kinesin MKLP2. J Cell Biol
191(6):1069–1077. doi:10.1083/jcb.201003095
163. Fuller BG, Lampson MA, Foley EA, Rosasco-Nitcher S, Le KV, Tobelmann P, Brautigan DL,
Stukenberg PT, Kapoor TM (2008) Midzone activation of aurora B in anaphase produces an
intracellular phosphorylation gradient. Nature 453(7198):1132–1136. doi:10.1038/
nature06923
164. Carmena M, Wheelock M, Funabiki H, Earnshaw WC (2012) The chromosomal passenger
complex (CPC): from easy rider to the godfather of mitosis. Nat Rev Mol Cell Biol
13(12):789–803. doi:10.1038/nrm3474
165. van der Horst A, Lens SM (2014) Cell division: control of the chromosomal passenger com-
plex in time and space. Chromosoma 123(1–2):25–42. doi:10.1007/s00412-013-0437-6
166. Uehara R, Tsukada Y, Kamasaki T, Poser I, Yoda K, Gerlich DW, Goshima G (2013) Aurora
B and Kif2A control microtubule length for assembly of a functional central spindle during
anaphase. J Cell Biol 202(4):623–636. doi:10.1083/jcb.201302123
167. Lewellyn L, Carvalho A, Desai A, Maddox AS, Oegema K (2011) The chromosomal pas-
senger complex and centralspindlin independently contribute to contractile ring assembly.
J Cell Biol 193(1):155–169. doi:10.1083/jcb.201008138
168. Birkenfeld J, Nalbant P, Bohl BP, Pertz O, Hahn KM, Bokoch GM (2007) GEF-H1 modulates
localized RhoA activation during cytokinesis under the control of mitotic kinases. Dev Cell
12(5):699–712. doi:10.1016/j.devcel.2007.03.014
169. Kitagawa M, Fung SY, Onishi N, Saya H, Lee SH (2013) Targeting Aurora B to the equatorial
cortex by MKLP2 is required for cytokinesis. PLoS One 8(6):e64826. doi:10.1371/journal.
pone.0064826
170. Norden C, Mendoza M, Dobbelaere J, Kotwaliwale CV, Biggins S, Barral Y (2006) The
NoCut pathway links completion of cytokinesis to spindle midzone function to prevent chro-
mosome breakage. Cell 125(1):85–98. doi:10.1016/j.cell.2006.01.045
171. Mendoza M, Norden C, Durrer K, Rauter H, Uhlmann F, Barral Y (2009) A mechanism for
chromosome segregation sensing by the NoCut checkpoint. Nat Cell Biol 11(4):477–483.
doi:10.1038/ncb1855
172. Steigemann P, Wurzenberger C, Schmitz MH, Held M, Guizetti J, Maar S, Gerlich DW

(2009) Aurora B-mediated abscission checkpoint protects against tetraploidization. Cell
136(3):473–484. doi:10.1016/j.cell.2008.12.020
173. Capalbo L, Montembault E, Takeda T, Bassi ZI, Glover DM, D’Avino PP (2012) The chro-
mosomal passenger complex controls the function of endosomal sorting complex required for
transport-III Snf7 proteins during cytokinesis. Open Biol 2(5):120070. doi:10.1098/
rsob.120070
174. Carlton JG, Caballe A, Agromayor M, Kloc M, Martin-Serrano J (2012) ESCRT-III
governs the Aurora B-mediated abscission checkpoint through CHMP4C. Science
336(6078):220–225. doi:10.1126/science.1217180
175. Thoresen SB, Campsteijn C, Vietri M, Schink KO, Liestol K, Andersen JS, Raiborg C,
Stenmark H (2014) ANCHR mediates Aurora-B-dependent abscission checkpoint control
through retention of VPS4. Nat Cell Biol 16(6):550–560. doi:10.1038/ncb2959
176. Taniuchi K, Nakagawa H, Nakamura T, Eguchi H, Ohigashi H, Ishikawa O, Katagiri T,
Nakamura Y (2005) Down-regulation of RAB6KIFL/KIF20A, a kinesin involved with mem-
brane trafficking of discs large homologue 5, can attenuate growth of pancreatic cancer cell.
Cancer Res 65(1):105–112
177. Lu Y, Liu P, Wen W, Grubbs CJ, Townsend RR, Malone JP, Lubet RA, You M (2010)
Cross-species comparison of orthologous gene expression in human bladder cancer and
carcinogen-induced rodent models. Am J Transl Res 3(1):8–27
178. Claerhout S, Lim JY, Choi W, Park YY, Kim K, Kim SB, Lee JS, Mills GB, Cho JY (2011)
Gene expression signature analysis identifies vorinostat as a candidate therapy for gastric
cancer. PLoS One 6(9):e24662. doi:10.1371/journal.pone.0024662
179. Gasnereau I, Boissan M, Margall-Ducos G, Couchy G, Wendum D, Bourgain-Guglielmetti F,
Desdouets C, Lacombe ML, Zucman-Rossi J, Sobczak-Thepot J (2012) KIF20A mRNA and
its product MKLP2 are increased during hepatocyte proliferation and hepatocarcinogenesis.
Am J Pathol 180(1):131–140. doi:10.1016/j.ajpath.2011.09.040
180. Yamashita J, Fukushima S, Jinnin M, Honda N, Makino K, Sakai K, Masuguchi S, Inoue Y,
Ihn H (2012) Kinesin family member 20A is a novel melanoma-associated antigen. Acta
Derm Venereol 92(6):593–597. doi:10.2340/00015555-1416
181. List A, Kurtin S, Roe DJ, Buresh A, Mahadevan D, Fuchs D, Rimsza L, Heaton R, Knight R,
Zeldis JB (2005) Efficacy of lenalidomide in myelodysplastic syndromes. N Engl J Med
352(6):549–557. doi:10.1056/NEJMoa041668
182. Matsuoka A, Tochigi A, Kishimoto M, Nakahara T, Kondo T, Tsujioka T, Tasaka T, Tohyama Y,
Tohyama K (2010) Lenalidomide induces cell death in an MDS-derived cell line with deletion
of chromosome 5q by inhibition of cytokinesis. Leukemia 24(4):748–755. doi:10.1038/
leu.2009.296
183. Groth-Pedersen L, Aits S, Corcelle-Termeau E, Petersen NH, Nylandsted J, Jaattela M (2012)
Identification of cytoskeleton-associated proteins essential for lysosomal stability and sur-
vival of human cancer cells. PLoS One 7(10):e45381. doi:10.1371/journal.pone.0045381
184. Yan GR, Zou FY, Dang BL, Zhang Y, Yu G, Liu X, He QY (2012) Genistein-induced mitotic
arrest of gastric cancer cells by downregulating KIF20A, a proteomics study. Proteomics
12(14):2391–2399. doi:10.1002/pmic.201100652
185. Zou JX, Duan Z, Wang J, Sokolov A, Xu J, Chen CZ, Li JJ, Chen HW (2014) Kinesin family
deregulation coordinated by bromodomain protein ANCCA and histone methyltransferase
MLL for breast cancer cell growth, survival, and tamoxifen resistance. Mol Cancer Res
12(4):539–549. doi:10.1158/1541-7786.MCR-13-0459
186. Tcherniuk S, Skoufias DA, Labriere C, Rath O, Gueritte F, Guillou C, Kozielski F (2010)
Relocation of Aurora B and survivin from centromeres to the central spindle impaired by a
kinesin-specific MKLP-2 inhibitor. Angew Chem Int Ed Engl 49(44):8228–8231.
doi:10.1002/anie.201003254
187. Liu J, Wang QC, Cui XS, Wang ZB, Kim NH, Sun SC (2013) MKLP2 inhibitior paprotrain
affects polar body extrusion during mouse oocyte maturation. Reprod Biol Endocrinol
11:117. doi:10.1186/1477-7827-11-117
188. Nakayama Y, Saito Y, Soeda S, Iwamoto E, Ogawa S, Yamagishi N, Kuga T, Yamaguchi N

(2014) Genistein induces cytokinesis failure through RhoA delocalization and anaphase
chromosome bridging. J Cell Biochem 115(4):763–771. doi:10.1002/jcb.24720
189. Imai K, Hirata S, Irie A, Senju S, Ikuta Y, Yokomine K, Harao M, Inoue M, Tomita Y, Tsunoda
T, Nakagawa H, Nakamura Y, Baba H, Nishimura Y (2011) Identification of HLA-A2-
restricted CTL epitopes of a novel tumour-associated antigen, KIF20A, overexpressed in
pancreatic cancer. Br J Cancer 104(2):300–307. doi:10.1038/sj.bjc.6606052
190. Osawa R, Tsunoda T, Yoshimura S, Watanabe T, Miyazawa M, Tani M, Takeda K, Nakagawa
H, Nakamura Y, Yamaue H (2012) Identification of HLA-A24-restricted novel T Cell epitope
peptides derived from P-cadherin and kinesin family member 20A. J Biomed Biotechnol
2012:848042. doi:10.1155/2012/848042
191. Asahara S, Takeda K, Yamao K, Maguchi H, Yamaue H (2013) Phase I/II clinical trial using
HLA-A24-restricted peptide vaccine derived from KIF20A for patients with advanced pan-
creatic cancer. J Transl Med 11(1):291. doi:10.1186/1479-5876-11-291
192. Suzuki N, Hazama S, Ueno T, Matsui H, Shindo Y, Iida M, Yoshimura K, Yoshino S, Takeda
K, Oka M (2014) A phase I clinical trial of vaccination with KIF20A-derived peptide in
combination with gemcitabine for patients with advanced pancreatic cancer. J Immunother
37(1):36–42. doi:10.1097/CJI.0000000000000012
193. Okuyama R, Aruga A, Hatori T, Takeda K, Yamamoto M (2013) Immunological responses to
a multi-peptide vaccine targeting cancer-testis antigens and VEGFRs in advanced pancreatic
cancer patients. Oncoimmunology 2(11):e27010. doi:10.4161/onci.27010
194. Aruga A, Takeshita N, Kotera Y, Okuyama R, Matsushita N, Ohta T, Takeda K, Yamamoto M
(2014) Phase I clinical trial of multiple-peptide vaccination for patients with advanced biliary
tract cancer. J Transl Med 12:61. doi:10.1186/1479-5876-12-61
195. Tomita Y, Yuno A, Tsukamoto H, Senju S, Kuroda Y, Hirayama M, Irie A, Kawahara K, Yatsuda
J, Hamada A, Jono H, Yoshida K, Tsunoda T, Kohrogi H, Yoshitake Y, Nakamura Y, Shinohara
M, Nishimura Y (2013) Identification of promiscuous KIF20A long peptides bearing both
CD4+ and CD8+ T-cell epitopes: KIF20A-specific CD4+ T-cell immunity in patients with
malignant tumor. Clin Cancer Res 19(16):4508–4520. doi:10.1158/1078-0432.CCR-13-0197
196. Melief CJ, van der Burg SH (2008) Immunotherapy of established (pre)malignant disease by
synthetic long peptide vaccines. Nat Rev Cancer 8(5):351–360. doi:10.1038/nrc2373
197. Reichert N, Wurster S, Ulrich T, Schmitt K, Hauser S, Probst L, Gotz R, Ceteci F, Moll R,
Rapp U, Gaubatz S (2010) Lin9, a subunit of the mammalian DREAM complex, is essential
for embryonic development, for survival of adult mice, and for tumor suppression. Mol Cell
Biol 30(12):2896–2908. doi:10.1128/MCB.00028-10
198. Wonsey DR, Follettie MT (2005) Loss of the forkhead transcription factor FoxM1 causes
centrosome amplification and mitotic catastrophe. Cancer Res 65(12):5181–5189.
doi:10.1158/0008-5472.CAN-04-4059
199. Westendorf JM, Rao PN, Gerace L (1994) Cloning of cDNAs for M-phase phosphoproteins
recognized by the MPM2 monoclonal antibody and determination of the phosphorylated
epitope. Proc Natl Acad Sci U S A 91(2):714–718
200. Matsumoto-Taniura N, Pirollet F, Monroe R, Gerace L, Westendorf JM (1996) Identification
of novel M phase phosphoproteins by expression cloning. Mol Biol Cell 7(9):1455–1469.
doi:10.1091/mbc.7.9.1455
201. Janisch KM, Vock VM, Fleming MS, Shrestha A, Grimsley-Myers CM, Rasoul BA, Neale SA,
Cupp TD, Kinchen JM, Liem KF Jr, Dwyer ND (2013) The vertebrate-specific Kinesin-6,
Kif20b, is required for normal cytokinesis of polarized cortical stem cells and cerebral cortex
size. Development 140(23):4672–4682. doi:10.1242/dev.093286
202. Zhu C, Zhao J, Bibikova M, Leverson JD, Bossy-Wetzel E, Fan JB, Abraham RT, Jiang W
(2005) Functional analysis of human microtubule-based motor proteins, the kinesins and
dyneins, in mitosis/cytokinesis using RNA interference. Mol Biol Cell 16(7):3187–3199.
doi:10.1091/mbc.E05-02-0167
203. Ryo A, Liou YC, Lu KP, Wulf G (2003) Prolyl isomerase Pin1: a catalyst for oncogenesis and
a potential therapeutic target in cancer. J Cell Sci 116(Pt 5):773–783. doi:10.1242/jcs.00276
204. van der Horst A, Khanna KK (2009) The peptidyl-prolyl isomerase Pin1 regulates cytokinesis
through Cep55. Cancer Res 69(16):6651–6659. doi:0.1158/0008-5472.CAN-09-0825
205. Nishiu M, Yanagawa R, Nakatsuka S, Yao M, Tsunoda T, Nakamura Y, Aozasa K (2002)
Microarray analysis of gene-expression profiles in diffuse large B-cell lymphoma: identifica-
tion of genes related to disease progression. Jpn J Cancer Res 93(8):894–901.
doi:10.1111/j.1349-7006.2002.tb01335.x
206. Liu XR, Cai Y, Cao X, Wei RC, Li HL, Zhou XM, Zhang KJ, Wu S, Qian QJ, Cheng B,
Huang K, Liu XY (2012) A new oncolytic adenoviral vector carrying dual tumour suppressor
genes shows potent anti-tumour effect. J Cell Mol Med 16(6):1298–1309.
doi:10.1111/j.1582-4934.2011.01396.x
207. Bao L, Kimzey A, Sauter G, Sowadski JM, Lu KP, Wang DG (2004) Prevalent overexpres-
sion of prolyl isomerase Pin1 in human cancers. Am J Pathol 164(5):1727–1737. doi:10.1016/
S0002-9440(10)63731-5
208. Obara W, Ohsawa R, Kanehira M, Takata R, Tsunoda T, Yoshida K, Takeda K, Katagiri T,
Nakamura Y, Fujioka T (2012) Cancer peptide vaccine therapy developed from oncoantigens
identified through genome-wide expression profile analysis for bladder cancer. Jpn J Clin
Oncol 42(7):591–600. doi:10.1093/jjco/hys069
209. Sapir T, Levy T, Sakakibara A, Rabinkov A, Miyata T, Reiner O (2013) Shootin1 acts
in concert with KIF20B to promote polarization of migrating neurons. J Neurosci
33(29):11932–11948. doi:10.1523/JNEUROSCI.5425-12.2013
210. Fritzler MJ, Kerfoot SM, Feasby TE, Zochodne DW, Westendorf JM, Dalmau JO, Chan EK
(2000) Autoantibodies from patients with idiopathic ataxia bind to M-phase phosphoprotein-
1 (MPP1). J Investig Med 48(1):28–39
211. Liu F, Arias-Vasquez A, Sleegers K, Aulchenko YS, Kayser M, Sanchez-Juan P, Feng BJ,
Bertoli-Avella AM, van Swieten J, Axenovich TI, Heutink P, van Broeckhoven C, Oostra BA,
van Duijn CM (2007) A genomewide screen for late-onset alzheimer disease in a genetically
isolated Dutch population. Am J Hum Genet 81(1):17–31. doi:10.1086/518720
212. Zochodne DW, Auer R, Fritzler MJ (2003) Longstanding ataxic demyelinating polyneu-
ronopathy with a novel autoantibody. Neurology 60(1):127–129. doi:10.1212/01.
WNL.0000040660.76868.3C
213. Blagosklonny MV (2007) Mitotic arrest and cell fate: why and how mitotic inhibition of
transcription drives mutually exclusive events. Cell Cycle 6(1):70–74. doi:10.4161/
cc.6.1.3682
214. Casenghi M, Mangiacasale R, Tuynder M, Caillet-Fauquet P, Elhajouji A, Lavia P, Mousset
S, Kirsch-Volders M, Cundari E (1999) p53-independent apoptosis and p53-dependent block
of DNA rereplication following mitotic spindle inhibition in human cells. Exp Cell Res
250(2):339–350. doi:10.1006/excr.1999.4554
215. Sagona AP, Stenmark H (2010) Cytokinesis and cancer. FEBS Lett 584(12):2652–2661.
doi:10.1016/j.febslet.2010.03.044
216. Hayashi MT, Karlseder J (2013) DNA damage associated with mitosis and cytokinesis
failure. Oncogene 32(39):4593–4601. doi:10.1038/onc.2012.615
217. Lacroix B, Maddox AS (2012) Cytokinesis, ploidy and aneuploidy. J Pathol 226(2):338–351.
doi:10.1002/path.3013
218. Fujiwara T, Bandi M, Nitta M, Ivanova EV, Bronson RT, Pellman D (2005) Cytokinesis
failure generating tetraploids promotes tumorigenesis in p53-null cells. Nature
437(7061):1043–1047. doi:10.1038/nature04217
219. Lv L, Zhang T, Yi Q, Huang Y, Wang Z, Hou H, Zhang H, Zheng W, Hao Q, Guo Z, Cooke
HJ, Shi Q (2012) Tetraploid cells from cytokinesis failure induce aneuploidy and spontane-
ous transformation of mouse ovarian surface epithelial cells. Cell Cycle 11(15):2864–2875.
doi:10.4161/cc.21196
Chapter 13
Non-motor Spindle Proteins as Cancer
Chemotherapy Targets
Robert L. Margolis and Mythili Yenjerla
13.1 Introduction
The mitotic spindle is a fusiform bipolar array of microtubules that assures the proper
attachment, metaphase alignment and precise segregation of chromosomes. The spin-
dle is an elaborate and highly regulated machine that performs its functions through
microtubule assembly at specific sites and through changes in the stability and dynam-
ics of the microtubule polymers so that they first align chromosomes and then segre-
gate them toward the spindle poles. Throughout mitosis there is an intimate interplay
between microtubule motors and non-motor proteins to assure controlled spindle
assembly and dynamics. Many of the non-motor spindle proteins interact with micro-
tubules as cofactors and regulators of motors, but many have functions entirely inde-
pendent of motor activity. The non-motor spindle proteins have essential roles in the
control of spindle assembly and function [1], and in linkage of microtubules to kineto-
chores and to centrosomes [2, 3], as well as in bundling of anti-parallel arrays of
microtubules to sustain the spindle midzone [4–6] (Table 13.1). During mitosis, the
interplay of microtubule stability and dynamics assures that microtubules will per-
form as reliable platforms for the motor proteins, and as the essential dynamic core of
the spindle structure. Interaction of the non-motor proteins with motor proteins is
required for several purposes; to maintain a continuous poleward flux, or treadmilling,
of microtubules, to create and sustain spindle bipolarity, to enable metaphase chromo-
some alignment and to initiate poleward migration of chromatid arms in anaphase.
Throughout mitosis microtubule dynamics and spindle function are controlled
by local assembly of molecular machines at kinetochores, spindle poles and the
anaphase spindle midzone; and by regulation of the spindle components through
posttranslational modification, most prominently by spindle associated protein
R.L. Margolis (*) • M. Yenjerla

Tumor Initiation and Maintenance Program, Sanford-Burnham Medical Research Institute,
10901 N Torrey Pines Rd, La Jolla, CA 92037, USA
e-mail: rmargolis@sanfordburnham.org

224
Table 13.1 Non-motor spindle proteins and protein kinases upregulated in tumors
Protein Localization in cells Mitotic function Expression in tumors Reference
Aurora A Spindle poles Recruitment of γ-tubulin, centrosome maturation, ↑ Oncogenic, tumor [3, 43, 152–154]
chromosomal alignment and segregation, cytokinesis progression
Aurora B Kinetochore, spindle Chromosomal alignment, recruitment of kinetochore ↑ Non-small cell lung [27–31, 53, 134, 136]
midzone and checkpoint proteins, microtubule stabilization carcinoma, pancreatic cancer
Plk1 Kinetochore, spindle Metaphase alignment of kinetochores, centrosome ↑ Variety of tumors [67, 138, 174]
poles, spindle midzone maturation, cytokinesis,
Ndc80/Hec1 Kinetochore Kinetochore binding to microtubule, ↑ Breast cancer [114, 116, 220, 221]
complex Chromosomal alignment, recruitment of Mps1
kinase, Mad1/Mad2
Stathmin Binds to tubulin Destabilizes microtubules ↑ Broad range of tumors [91, 208]
Survivin Kinetochore, spindle Kinetochore binding to microtubule, chromosomal ↑ Broad range of tumors [222]
midzone alignment
TD-60 Kinetochore, spindle Aurora activation, localization of Aurora B, ↑ Broad range of tumors [129, 131, RLM,
midzone INCENP, Survivin unpublished data]
NuMA Spindle poles Spindle pole organization ↑ Oral, liver adenocarcinoma [142, 223]
TPX2 Spindle microtubules Targets aurora A, Xklp2, activates Aurora A, Bipolar ↑ Esophageal squamous cell [145–148, 224, 225]
spindle formation carcinoma, cervical cancer
PRC1 Microtubules, spindle Microtubule bundling, formation of central spindle, ↑ Breast and bladder cancer [163, 164, 167, 226,
midzone localization of midzone proteins cells 227]
R.L. Margolis and M. Yenjerla
13 Non-motor Spindle Proteins as Cancer Chemotherapy Targets 225
kinases. Among the protein kinases, the most important to the control of spindle
function throughout mitosis are Aurora A, Aurora B, Plk1 and Cdk1.
To assemble the spindle and create its function, spindle proteins exploit the
intrinsic polarity of microtubules. At one end of each microtubule, designated the
plus end, net assembly occurs, while the minus end either permits net disassembly
of the polymer, or is capped to suppress dynamics. The mitotic spindle is always
organized so that all microtubule minus ends associate with spindle poles, and all
plus ends are either kinetochore associated or freely interact in antiparallel fashion
with microtubules that emanate from the other spindle pole, or form astral arrays
with free microtubule ends, that emanate from centrosomes [7]. Where the protein
interacts is important to its function. Motors, protein kinases and crosslinking pro-
teins may function at microtubule ends, or can act on lateral microtubule surfaces.
In contrast, non-motor proteins that control dynamics and linkage of microtubules
to subcellular structures, such as the plus-tip proteins, perform their functions
uniquely at microtubule ends, where tubulin subunits actively exchange.
13.2 Formation and Function of the Dynamic Spindle

Through Microtubule Dynamics
Formation of a bipolar spindle is an essential step in mitotic progression. The bipo-

lar spindle is both formed and sustained by multiple protein factors. In this process,
motor and non-motor protein interactions that establish the functioning spindle can
either be cooperative or counterbalancing.
There are three distinct phases of mitosis with respect to spindle dynamics and
function. In prometaphase, the formation of the spindle involves induction of bipolar-
ity and outgrowth of microtubules both from kinetochores and from spindle poles. In
this phase, spindle microtubules are primarily nucleated at the spindle poles, which
are normally demarcated by centrosomes. The spindle poles create an astral array of
microtubules that exhibits a search and capture mechanism, relying on dynamic
instability, involving a constant flux in microtubule length through rapid gain and loss
of tubulin from microtubule tips at the periphery of the aster. Spindle pole microtu-
bules may thus “fish” to capture kinetochores [8]. At the same time microtubule
growth also clearly initiates at the kinetochores [2, 9] and in the vicinity of chromo-
somes [9], and this may be the predominant mechanism of kinetochore integration
into the spindle in mammalian cells [2, 9]. This mechanism would entail growth of
microtubules that are anchored at kinetochores through net-assembly, a process that
implies extraordinary stabilization of microtubule minus ends to achieve net growth.
Centrosomes normally serve as the foci of the spindle poles in mitosis, but two
spindle poles can also form in the absence of centrosomes [10, 11]. Indeed, under
appropriate conditions a fusiform spindle can form without chromosomes or centro-
somes, and can exhibit robust poleward treadmilling dynamics [12]. Thus cells have
an intrinsic capacity to generate a focused fusiform spindle with appropriate dynam-
ics in early mitosis, independent of the normal spindle organizing centers.
226 R.L. Margolis and M. Yenjerla
Based on work on Xenopus oocyte spindles, RCC1 (a Ran guanine nucleotide

exchange factor) associated with chromosomes has been shown to create a chromo-
some centric fusiform spindle through production of a gradient of RanGTP [13].
RanGTP diffuses from the chromosomes to activate two key spindle assembly pro-
teins, TPX2 and NuMA [14]. However, what is true of Xenopus oocytes may not be
true of mammalian somatic cells. Recent evidence in mammalian cells suggests that
a Ran gradient has no influence on mammalian bipolar spindle assembly [15].
Importantly, Xenopus RCC1, which is distributed throughout the chromatin, controls
spindle assembly, even from chromatin without kinetochores [16, 17], but in mam-
malian somatic cells, the kinetochore is the essential element in forming the bipolar
spindle. When kinetochores are physically separated from bulk chromatin a bipolar
spindle is established from kinetochores and not bulk chromatin (Fig. 13.1), and thus
forms at right angles to the RanGTP gradient that emanates from the RCC1 associ-
ated with bulk chromatin. Interestingly, a bipolar spindle can form in Xenopus oocyte
extracts in the absence of a RanGTP gradient, but in this case the CPC, provided by
sperm nuclei, is required to establish a kinetochore linked bipolar spindle [18].
As mitosis continues, the mature bipolar spindle consists of both kinetochore
associated microtubules and interpolar microtubules that radiate from the spindle
poles and form antiparallel linkages with microtubules from the opposite pole. At
this stage, central spindle microtubules continuously flux, or treadmill, from their
plus ends toward the spindle poles, while maintaining spindle length, and associa-
tions with kinetochores and the opposing half spindle [19, 20].
When cells enter anaphase and the chromatids separate, the third phase of mito-
sis involves stabilization of interpolar microtubules and shortening of the kineto-
chore fibers to enable poleward chromatid migration. In this phase, both motor and
Fig. 13.1 Distribution of Aurora A and Aurora B during mitosis. Aurora A first associates with
centrosomes, and then with the whole spindle. Aurora B associates first with kinetochores and then
with the overlap region of the midzone microtubules in late mitosis
non-motor microtubule proteins concentrate at the midzone of the spindle to create

the conditions for cell cleavage. The spindle asters are of importance for anchorage
of the spindle to the cell cortex throughout mitosis, but are of especial importance
during anaphase and telophase. Aster association with the cortex is controlled by a
ternary complex of cortex associated proteins, G-protein alpha, LGN, and NuMA
[21, 22], that couples with dynein [23, 24] to create microtubule linkage. This com-
plex then enables a dynamic microtubule plus end that is tethered to the cortex,
permitting the intrinsic dynamics of the spindle to play out while the spindle posi-
tion in the cell remains constant. The tethering permits chromosomes to move
toward poles near the cell cortex, and a well-defined spindle center enables cell
cleavage to occur equidistant from the poles.
From the onset of anaphase until the end of mitosis, a profound change occurs,
suppressing microtubule dynamics. Stable interpolar polymers bundle and the two
antiparallel half spindle microtubule bundles slide apart in the midzone, separating
the spindle poles, while at the same time kinetochore fibers shorten by controlled
disassembly to reel in the chromatids toward the poles.
Throughout mitosis, microtubule dynamics are essential to spindle function.
Without an inherent capacity to flux, microtubules cannot maintain a stable length
while remaining attached to kinetochores. Freely fluxing microtubules cannot be
tethered by end capping proteins, but rather must associate with proteins that permit
their flowing displacement from points of attachment. End capping proteins then
play an important role in later stages of mitosis, when spindle elongation requires a
nexus of stable microtubules.
The dynamics of the mitotic spindle dictate its structure. The spindle is fusiform
because of an intrinsic coalescence of microtubules at the spindle poles, and because
the turnover of microtubules counterbalances stable linkages at their termini. The
polymer termini, being the sites of subunit addition and loss, dictate both spindle
dynamics and the form that results from tension between dynamics and linkage.
Further, the elongation of interpolar microtubules in anaphase results from an intrin-
sic change in polymer dynamics and stability consequent to a decrease in Cdk1
kinase activity in anaphase, a stabilization that occurs even when Cdk1 suppression
is induced in cells that contain a monastral spindle [25, 26].
13.3 Protein Kinases, Posttranslational Modification

and Global Control of Spindle Function
Mitotic spindle assembly and function are regulated by a number of mitosis specific
protein kinases, each with multiple spindle functions. The principal protein kinases
involved in mitotic control are Aurora A, Aurora B, polo-like kinase (Plk1) and
Cdk1, the protein kinase that activates the mitotic state.
Between them, these protein kinases regulate almost all aspects of spindle assem-
bly and dynamics as mitosis progresses. Because of their obvious importance to the
mitotic process, and because each of the Aurora kinases [27–31], and Plk1 [32, 33],
are dysregulated in tumor cells, they have generated a great deal of attention as
potential targets for chemotherapy [33–35], and small drug inhibitors of each of
these kinases have entered clinical trials.
Aurora A and Aurora B are paralogue mitotic serine-threonine kinases. Both are
important to mitotic progression, but both their localization and their functions are
distinct [36–39]. Aurora A associates with centrosomes and with centrosome proxi-
mal microtubules [27, 40] (Fig. 13.1), and both binds to and is activated by TPX2,
a microtubule binding protein that localizes Aurora A to the mitotic spindle [40].
Aurora A is required for mitotic entry, and for this it must be activated by associa-
tion with Ajuba at the centrosome [41]. It is then required for centrosome matura-
tion and formation of a bipolar mitotic spindle [42], for mitotic checkpoint function
[43, 44] and for accurate segregation of both centrosomes and chromosomes into
daughter cells during mitotic exit [36, 45, 46].
In contrast, Aurora B first binds to a complex of proteins at the inner centromere,
the chromosome passenger complex (CPC), and like the other passenger proteins, it
first associates with the inner centromere, then separates from centromeres to relo-
cate to the spindle midzone during anaphase and cell cleavage (Fig. 13.1). It binds
to and is activated by INCENP, another passenger protein [47–51]. Aurora B is
absolutely required for spindle assembly checkpoint function, correct chromosome
segregation and for cell cleavage [52–57].
Despite their structural similarities, the two proteins have unique spectra of bind-
ing partners [58] and of phosphorylation substrates [59, 60], perhaps because of their
different localizations within the cell. Aurora A substrates include p53, CENP-A,
Lats2, RalA, HURP, BRCA1, Eg5, TACC3/Maskin, Protein Phosphatase 1, CDC25B,
MBP3, Nuf2, NDEL1 and Plk1 [59–62]. Of these, the motor protein Eg5, as well as
TACC3/Maskin, CENP-A, Nuf2 and Plk1 are of particular interest with respect to
spindle function. In addition to histone H3, Aurora B substrates include Survivin,
INCENP, Borealin, CENP-A and Hec1/Ndc80 at the centromere [59–61], and
MCAK, MgcRacGAP, vimentin and ZEN4/MKLP1 at the spindle midzone [60, 61]
in late mitosis. CENP-A uniquely appears to be a substrate first of Aurora A, and then
of Aurora B during mitotic progression [63]. Among their substrates, it is noteworthy
that each Aurora kinase phosphorylates and regulates several mitotic motor proteins.
Structural analysis indicates the interaction of Aurora B with INCENP occurs in
a region of the catalytic domain that overlaps with that required for TPX2 binding
to Aurora A [51]. The localization, dictated by its binding partner, appears to be the
most distinctive feature of each Aurora kinase, as a single amino acid substitution in
Aurora A changes its binding partner from TPX2 to INCENP, and it then can com-
pletely substitute for ablated Aurora B, both with respect to localization and activity
[64]. Both Aurora A and Aurora B are specifically upregulated in G2, perform their
mitotic functions, and then are destroyed by APC/Ccdh1 dependent ubiquitination
and subsequent proteolysis at the end of mitosis [65, 66].
Polo-like kinase 1 (Plk1) is another mitosis specific protein kinase that pro-
foundly affects spindle assembly and function throughout mitosis [67]. Just as the
two Aurora kinases are localized to distinct subcellular sites in mitosis, and migrate
in a distinctive manner from one locus to another, Plk1 also localizes to different
sites as mitosis progresses. Plk1 first associates with centrosomes in prophase [68,
69], and then transfers to the kinetochores during prometaphase and metaphase
[68]. Like Aurora B and the other CPC proteins, Plk1 then relocates to the spindle
midzone during anaphase and to the midbody during cell cleavage [67, 68]. Plk1
activity is essential for the maturation of the centrosome, for formation of the bipo-
lar spindle, for metaphase alignment of kinetochores, for spindle elongation in ana-
phase, and for the process of cell cleavage. The best understood of Plk1 activities is
its regulation of anaphase spindle elongation and bundling, and initiation of cell
cleavage by regulating the anaphase microtubule bundling protein PRC1 [70, 71],
and the RhoA-GTPase GEF Ect2 [72] at the central spindle, allowing Ect2 to acti-
vate RhoA dependent initiation of cell cleavage.
Cdk1-cyclin B protein kinase drives mitotic entry and has a large number of
mitotic substrates. Its role in spindle assembly is difficult to ascertain, as its essen-
tial role in mitotic entry is epistatic to any specific functions that might be revealed
by its suppression. Cdk1 substrates include the microtubule proteins, tubulin, stath-
min and MAP7 [73, 74], suggesting that important aspects of microtubule assembly
and dynamics are directly regulated by Cdk1 activity.
At the metaphase to anaphase transition, Cdk1 activity is abruptly suppressed by
cyclin B degradation. If Cdk1 activity persists after metaphase checkpoint satisfac-
tion, cells enter anaphase, but the CPC proteins do not separate from centromeres
and the cells do not cleave [75]. Further, loss of Cdk1 activity in cells blocked with
a monastral spindle in prometaphase causes spindle elongation and a distal budding
or “cleavage” event at the end of the elongated prometaphase spindle (Fig. 13.2).
Global mitotic protein kinase interaction and phospho-protein substrate net-
works for these principal mitotic kinases have been determined by mass spectrom-
etry methods. The key mitotic interacting and substrate proteins for Aurora A,
Aurora B, Plk1 and Cdk1, as determined by this exhaustive method, are listed in
Dephoure et al. [76]. Further, there are many mitotic substrates of Cdk1, Aurora A,
Aurora B and Plk1 that are specific to the mitotic spindle, including many microtu-
bule motor and non-motor proteins [77].
Another important posttranslational modification in mitosis involves the addition
and removal of C-terminal tyrosine on alpha-tubulin. The tubulin C-terminus is
modified continuously by addition and removal of the terminal tyrosine, using tubu-
lin tyrosine ligase [78] and a tubulin carboxypeptidase, provisionally identified as
AGBL2 [79]. The presence of C-terminal tyrosine produces an EEY motif that is
commonly recognized by plus-tip proteins, and tyrosinated tubulin is important for
recruitment of CAP-Gly motif plus-tip proteins such as CLIP-170, CLIP-115 and
p150glued, that track dynamic microtubule plus-ends in association with EB1 [80],
and link microtubules to kinetochores and to the cell cortex [81, 82]. Tubulin tyro-
sine ligase (TTL) is often suppressed in tumor cells, presumably because its loss
confers a growth advantage [83, 84]. Assuming recruitment of plus-tip proteins such
as CLIP-170, which are important to spindle function, requires TTL activity [80], it
is of interest that the spindle core retains tyrosinated tubulin even in cells that are
devoid of TTL [85]. The mechanism by which this occurs is unknown, but must
involve either selective recruitment of newly translated tyrosinated tubulin to the
Fig. 13.2 Mitotic HeLa cells

with monastral spindles
created by an Eg5 inhibitor
(STLC) undergo cleavage by
producing a microtubule
bundle extension from the
aster after suppression of
Cdk1 activity. The
chromosomes remain
clustered at the centrosome,
and cleavage produces an
empty sac (Reproduced from
Skoufias et al. [26])
central spindle (the mRNA codes for C-terminal tyrosine), or a specific spindle
associated TTL-like activity. The absence of tyrosinated tubulin in asters in the
same cell suggests that tyrosination correlates with flux dynamics in the central
spindle, as opposed to dynamic instability in asters [86]. This system of tubulin
posttranslational modification is of potential interest as a chemotherapy target.
13.3.1 Protein Regulation of Microtubule Dynamics
The coordinated activity of both stabilizing and destabilizing microtubule associated

proteins (MAPs) regulates spindle dynamics. Although most stabilizing MAPs are
restricted to non-cycling neuronal cells, stabilizing MAPs, including XMAP215/
hTOG and MAP4 [87–89] also function in mitosis. Additionally, there are destabiliz-
ing MAPs including stathmin [90, 91], the kinesin related protein XKCM1/MCAK
[92] and the microtubule severing protein katanin [93]. The Xenopus homologues of
MCAK and hTOG act in opposition to regulate spindle dynamics in Xenopus extracts
[94]. To some extent this activity is conserved with the human hTOG and MCAK,
which regulate spindle assembly [95] and morphology [96–98].
While MAP4 and hTOG promote microtubule polymerization by increasing
microtubule length and turnover rates [99], stathmin destabilizes microtubules by
enhancing disassembly at the polymer minus ends. Stathmin has been shown to
increase treadmilling rates in vitro, suggesting that it favors poleward flux by induc-
ing depolymerization of minus ends at the spindle pole [91]. Both MAP4 and stath-
min are suppressed by Cdk1 phosphorylation in mitosis, with consequences for
spindle dynamics [100, 101].
Polymer dynamics, as well as binding to kinetochores, centrosomes and the cell
cortex, all occur at the microtubule ends. A diverse group of proteins interact as a
complex at the microtubule net assembly ends, and they play a major role in regulat-
ing microtubule dynamics and spindle function in mitosis. These proteins, collec-
tively designated tracking proteins (plus-tip proteins), are involved both in
directional cell motility and in control of spindle dynamics. Several of the plus-tip
proteins, including CLIP-170 and EB1, have the fascinating capacity to “surf” along
the polymerizing end [102], while permitting plus-end assembly to occur [103,
104]. The capacity to surf or slide, first demonstrated for neuronal MAP6 [105], is
posited to be due to lateral association with the polymer, allowing free exchange of
subunits at the microtubule terminus [106, 107].
EB1 is required to maintain proper spindle length, and is a required component
for astral microtubules to link to the cell cortex [108] thus positioning the spindle
poles and orienting the mitotic apparatus. CLASP1 (CLIP associated protein) is
another stabilizing MAP [109], and the Drosophila orthologue of CLASP (Mast/
Orbit) acts in opposition to the motor protein dynein to maintain spindle bipolarity
and microtubule-kinetochore attachment [110].
CLASP1 forms a complex at the microtubule end with CLIP-170, with EB1
related proteins, and EB1 binds to APC (Adenomatous Polyposis Coli). The
complex of plus-tip proteins localizes at the kinetochore and guides its attachment
to microtubules [104, 111, 112]. The association of plus-tip proteins is in turn regu-
lated by tubulin tyrosination, as noted above.
13.3.2 Regulation of Spindle at the Kinetochore
The kinetochore is a highly complex mitosis specific organelle that forms at the
primary constriction of each chromosome, at the site demarcated by deposition of
the centromere specific histone CENP-A [113]. The kinetochore is composed of
multiple protein complexes, each with a specific role [114]. Their individual roles
include creating a stable microtubule attachment while permitting microtubule flux
outward toward the spindle poles, maintaining the spindle assembly checkpoint
until the chromosomes are correctly aligned at metaphase [115], continuously cor-
recting improper microtubule linkages to prevent aneuploid chromosome segrega-
tion, and finally enabling the controlled shortening of the kinetochore microtubule
fiber during anaphase.
The core microtubule binding complex of the kinetochore, referred to as the
KMN network, consists of three protein complexes (KNL-1/Mis12 complex/
Ndc80) [116]. Kinetochore proteins that directly link to microtubules include the
Ndc80/Hec1 complex, containing four proteins, Ndc80/Hec1, Nuf2, Spc24 and
Spc25 [117] of which Ndc80 and Nuf2 bind to the microtubule lattice while the
Spc24-Spc25 dimer binds the kinetochore. The Ndc80 complex forms rod like pro-
jections on the kinetochore proximal microtubule lattice [116] and regulates
kinetochore-microtubule attachment while permitting constant poleward flux of
microtubules [118]. Depletion of any of the Ndc80 complex proteins, including
Hec1 itself, leads to chromosomal misalignment [114]. Working in conjunction
with this linkage, microtubule associated proteins associated with plus tip function,
including CLIP170, CLASP, EB1, hSgo1 and APC, as well as the microtubule
motors dynein/dynactin and CENP-E regulate microtubule motility at the kineto-
chore [104, 115, 119, 120].
The outer plate of the kinetochore contains the microtubule linking proteins, as
well as checkpoint proteins that read correct attachment and permit entry into ana-
phase after the spindle assembly checkpoint (SAC) is satisfied. The satisfaction of
the SAC appears to depend on KMN dependent changes in protein architecture at
the kinetochore [121]. The proteins implicated in activating the SAC until correct
microtubule attachment to kinetochores include the Rod-Zw10-Zwilch complex
[122], acting with dynein and Spindly [123, 124], and the checkpoint readers Bub1,
BubR1, Bub3, Mad1, Mad2 and Mps1 [120].
In yeast cells, Ndc80 is required for loading Dam1/DASH complexes [125],
which tether the kinetochores to the microtubule plus ends [126] as an encircling
ring [127]. The ring has been proposed to follow shortening microtubules so that the
kinetochore remains attached to the microtubule while the energy from microtubule
depolymerization is used to generate processive chromosome movement [128]. The
mammalian orthologues of Dam1/DASH have not been identified and the signifi-
cance of the ring formation in humans is yet to be established. Some possible func-
tional counterparts of Dam1 such as MCAK, Ska1/2, Cep57 and Bod1 have been
suggested in humans, but their role in modulating the kinetochore-microtubule
attachment remains to be clarified [117]. The CPC proteins Aurora B, INCENP,
Borealin, TD-60 and Survivin promote binding of kinetochores to microtubules in
early mitosis [129, 130]. One of the CPC proteins, TD-60 binds directly to microtu-
bules, and is required to activate both Aurora B and the histone kinase haspin, whose
activity recruits Aurora B to the kinetochore [129, 131, 132], generating correct
microtubule attachments that satisfy the spindle assembly checkpoint.
Most interestingly, the CPC proteins displace from centromere localization in
early mitosis to the central spindle during anaphase and telophase [47] where they
function to induce cell cleavage. Figure 13.3 shows the typical displacement of
TD-60 as mitosis proceeds. The displacement is as shown in Fig. 13.1 for Aurora
B. The CPC proteins are required for the stabilization of kinetochore microtubules
and spindle formation [133], for recruitment of other proteins to the kinetochore,
and to assure proper chromosomal alignment. Aurora B is required to recruit the
outer kinetochore proteins, including the microtubule regulatory proteins Ndc80,
CLIP-170 and p150Glued (a dynein regulating protein), and the checkpoint proteins
Mps1, Mad2, Bub1 and BubR1 [134], which transiently associate with the kineto-
chores to prevent induction of anaphase until chromosomes are properly aligned
and under tension [135]. Aurora B phosphorylates and suppresses the microtubule
Fig. 13.3 Distribution of the passenger protein TD-60 during mitosis. Co-stain with DNA shows
that it localizes to kinetochores in early mitosis (upper panels). Followed through the course of
mitosis, it behaves like Aurora B, first associating with kinetochores and then displacing to the
spindle midzone as the chromatids separate in anaphase (lower panels)
destabilization activity of MCAK (mitotic centromere-associated kinesin) at the

centromeres [53]. It also senses tension at the kinetochore and favors bipolar attach-
ments [136], correcting merotelically attached (misaligned) chromosomes by
phosphorylation of MCAK [137].
Among the kinetochore proteins, several are frequently upregulated [31, 61, 138]
or mutated [139] in tumor cells, with consequences for chromosome missegregation
that can drive tumor aneuploidy. For example, Aurora B overexpression is common
in tumors, and will interfere with Aurora B dependent correction of faulty merotelic
chromosome attachments to the spindle [140]. Interestingly, Aurora A overexpres-
sion has been shown to override the spindle assembly checkpoint in tumor cells
[43], ultimately leading to tumor aneuploidy [46]. Over 100 kinetochore proteins
have been described to date, and this review only touches the surface of their various
functions. For more detail, see [141]. Our purpose is to show the complexity of the
structure and the possibility of its dysregulation in tumors. Of the many spindle
proteins, the protein kinases have been the most appealing targets for chemotherapy
trials, both because of their dysregulation and because the kinome (the full comple-
ment of protein kinases) has a history of success in drug development.
In addition to protein kinases, other non-motor proteins involved in kinetochore
function can easily be envisioned as valid targets for chemotherapy, particularly if
routinely dysregulated or mutated in specific tumors. For proteins without enzy-
matic activity, effective screens would entail disruption of the function of the pro-
tein that is regulated by the target. Examples would include suppression of INCENP
binding to Aurora B or TPX2 binding to Aurora A, with consequent suppression of
kinase activity. The best candidates for intervention would include those proteins
that exclusively modulate mitotic events.
13.3.3 Regulation of Spindle at the Centrosome
As described above, the centrosomes are not absolutely required to form mitotic
spindle poles, but when present they dictate the foci of the bipolar spindle, and the
destination of segregating chromatids. To a large extent, the normal mitotic state is
dictated by centrosome-associated proteins, which play important roles in the for-
mation of the dynamic mitotic spindle. While many proteins regulate centrosome
control of microtubules throughout the cell cycle, there are many regulatory pro-
teins that are specific to mitosis. For example, the MAPs NuMA and TPX2 are
intranuclear in interphase but associate with spindle poles and microtubules during
mitosis, and are important to control of microtubule dynamics [3]. NuMA organizes
the spindle pole by stabilizing and cross-linking microtubule minus-ends in con-
junction with dynein and dynactin [142, 143], and TPX2 is required for the forma-
tion of normal bipolar spindles [144]. TPX2 targets Xklp2 to microtubule minus
ends [145], and is the essential Aurora A binding partner that both activates the
kinase [146, 147] and localizes it to interpolar microtubules [148]. As a result,
Aurora A targets the TACC3-XMAP215 complex to the centrosome proximal spin-
dle [149], which counterbalances the destabilizing activity of MCAK at centro-
somes [149, 150]. TPX2 is in turn activated through its displacement from Importin-β
by Ran-GTP [151].
Aurora A and Plk1 are both important to centrosome maturation in mitosis [3,
152]. Both kinases are required for recruitment of γ-tubulin to the centrosome [153,
154], and for proper metaphase alignment of chromosomes. At the onset of mitosis,
Aurora A phosphorylates TACC3 which induces growth of microtubules at both
ends and activates kinesin-like motors that drive centrosome maturation and separa-
tion [149]. The potential of Aurora A to establish the spindle pole, and create a
bipolar spindle, has been dramatically demonstrated in Xenopus extracts, where
Aurora A coated beads can create bipolar spindles in the absence of centrosomes
and chromatin [155], in a manner strikingly parallel to the ability of chromatin
beads to generate a bipolar spindle in Xenopus extracts in the absence of kineto-
chores and centrosomes. When the spindle is created by Aurora coated beads,
Aurora A is present at the spindle poles, and it recruits an Aurora centered microtu-
bule stabilizing, bundling and regulatory complex, comprised of γTURC,
XMAP215, HURP, TPX2 and Eg5 [156].
The processes that preserve the role of the centrosome in creating a bipolar spin-
dle in nontransformed cells are often aberrant in tumor cells [157]. In addition to the
dysregulation of Aurora A and Plk1, another important centrosome associated pro-
tein kinase, Nek2, is often highly upregulated in tumors [158–160]. Active mostly
in G2 [161], Nek2 functions to regulate the centrosome cohesion proteins, C-Nap1
and Rootletin, prior to mitosis [162]. Its dysregulation disturbs centrosome number
control, thus creating multipolar spindles and consequent aneuploidy in tumors.
Because its dysregulation is believed to play a pivotal role in tumor growth and
development of aneuploidy, characteristic of the most aggressive tumors, Nek2 is an
attractive therapeutic target.
13.3.4 Regulation of Spindle at the Midzone
Antiparallel microtubules are cross-linked by microtubule bundling proteins such as

PRC1 and Centralspindlin. Centralspindlin is a tetramer composed of the Rho-GAP
Cyk4/MgcRacGAP and the kinesin ZEN4/MKLP1, which act together to form the
central spindle in anaphase [163, 164]. PRC1, a protein with direct microtubule
bundling activity [4], localizes to the spindle midzone in late mitosis (Fig. 13.4)
where it binds and colocalizes with kinesin KIF4A [165]. Both the Centralspindlin
and PRC1 protein complexes are required for anaphase central spindle bundling,
and to mark and activate the site of cell cleavage at the point of overlap in the bundle
of antiparallel microtubules that forms as the spindle elongates in anaphase.
Aurora B, localizing to the central spindle in anaphase with other components of
the CPC, suppresses microtubule dynamics and regulates the midzone by activating
KIF4A [166]. Knockdown of PRC1 causes loss of microtubule bundles in the spin-
dle midzone and therefore the CPC proteins such as Aurora B and Survivin improp-
erly localize as a subcortical ring at the spindle midzone instead of to a central
spindle bundle [167]. Although cells can still undergo furrowing in anaphase with-
out PRC1 dependent localization of the CPC complex, they do not undergo
abscission to separate the two daughter cells. PRC1 is thus required for final telo-
phase abscission [167]. PRC1 is itself regulated by Cdk1 phosphorylation in early
mitosis [4, 70], and Plk1 phosphorylation in late mitosis [70], although Plk1
phosphorylation of PRC1 has also been reported to control bipolar spindle forma-
tion in early mitosis [71]. Interestingly, the APC/C dependent degradation of Aurora
A is required to produce a robust spindle midzone [168]. In contrast, Aurora B activ-
ity during the abscission step of cleavage promotes a checkpoint (NoCut) that pre-
vents abscission from trapping strands of DNA [169].
Fig. 13.4 PRC1 associates

with the spindle midzone
during late mitosis and cell
cleavage. The micrograph
shows a cleaving cell. PRC1
completely overlaps with
tubulin in the midzone, but
does not associate with
microtubules elsewhere in the
cell. Note the interphase
microtubule bundling induced
by ectopic PRC1 expression
in the adjacent cell
13.4 Spindle Proteins as Chemotherapy Targets
Among the anti-mitotic drugs that block the growth of rapidly dividing cancer cells,
microtubule targeted drugs have been most successful (Table 13.2). However, sev-
eral drugs that inhibit the activity of motors and non-motor proteins have shown
promise and are under development. We focus on development of drugs targeting
non-motor proteins.
Table 13.2 Cancer drugs that target non-motor spindle proteins and protein kinases
Target/mechanism of
Drug class action Use Reference
Vinblastine, Vincristine, Depolymerize Clinical use [228]
Vinorelbine microtubules at plus ends
Paclitaxel, Ixabepilone Stabilize microtubules Clinical use [170, 172]
2-Methoxyestradiol Depolymerize Phase II [170]
(Panzem) microtubules
Aurora kinase inhibitors Aurora kinase Clinical trials [183, 229]
MLN8237 (Alisertib) Aurora A Phase I, Phase
II
GSK1070916A Aurora B and C Phase I
AT9283 Aurora A and B Phase I
PF-03814735 Aurora A and B Phase I
CYC116 Aurora A, B and C Phase I
AMG 900 Aurora A, B and C Phase I
SNS-314 Aurora A, B and C Phase I
Polo-like kinase Plk1 Clinical trials [183, 229]
inhibitors
TAK-960 Plk1 Phase I
BI 6727 (Volasertib) Plk1 Phase II
GSK461364 Plk1 Phase I
TKM-080301 siRNA against Plk1 Phase I
NMS-1286937 Plk1 Phase I
Hec1 inhibitors Inhibits Hec1-Nek2 Preclinical trials [221]
interation
Bifunctional shRNA Stathmin Phase I www.gradalisinc.com
Stathmin 1-knockdown
Lipoplex
YM155 Survivin Phase I, II [230, 193]
S12 Survivin Preclinical
trials
Clinical trial details from http://www.clinicaltrials.gov/
13.4.1 Microtubule Targeting Agents
Drugs that bind to tubulin alter the polymerization and depolymerization dynamics
of spindle microtubules (for a detailed review see [170]) and arrest cells in mitosis.
Based on their binding site on tubulin, these drugs fall into three classes, the vinca
alkaloids that bind at the tubulin vinca domain, drugs that bind to the colchicine site
on tubulin, and drugs that interact with the tubulin paclitaxel binding site. The fam-
ily of vinca domain binding drugs include Vinblastine, Vincristine, Vinflunine,
Dolastatins and Halichondrins, which all act to destabilize microtubule assembly.
They have been used to treat lymphomas, as well as solid tumors such as breast
cancer and non-small-cell lung cancer. The second class of drugs, those that bind to
the colchicine site, include Combretastatin A-4 phosphate and 2-methoxyestradiol.
These drugs also act by destabilizing microtubule assembly. The third class, com-
prising the taxanes, stabilizes the microtubule polymer in its assembled state. This
class includes paclitaxel and docetaxel, which have been widely used to treat
patients with breast and ovarian cancer. Frequently, the taxanes are used in combi-
nation therapy with drugs that induce DNA damage, such as carboplatin and doxo-
rubicin. Novel taxane congeners and formulations are being developed, and some
have been approved by the FDA [171].
Regardless of whether they stabilize or destabilize microtubules, these drugs kill
dividing tumor cells by suppressing satisfaction of the metaphase spindle assembly
checkpoint that blocks completion of mitosis until after proper chromosome align-
ment at the metaphase plate. Cells blocked in abnormal metaphase and eventually
undergo apoptosis [170].
Another nontaxane microtubule-targeting agents approved for clinical use
include Ixabepilone, a novel analogue of epothilone B for use in taxane resistant
tumors [172]. The epothilones act by stabilizing microtubules in a manner similar to
the taxanes. Additionally, Eribulin, a synthetic analogue of a marine sponge derived
Halichondrin B is used for breast cancer. It destabilizes microtubules and blocks
cells in prometaphase with misaligned chromosomes [173].
13.4.2 Spindle Protein Inhibitors
The protein kinases that are essential to mitotic progression, including Aurora A,
Aurora B and Plk1, are increasingly the targets of an emerging class of chemother-
apy compounds [138, 174, 175]. These protein kinases, as described above, are
essential for many aspects of mitosis, they are frequently overexpressed in tumors,
and their overexpression has been linked to promotion of tumorigenesis [175, 176]
(Table 13.1).
The Aurora inhibitors either block interaction between Aurora A and its cofac-
tors or substrates, or can block the ATP binding site of the kinase [177]. Inhibition
or depletion of Aurora A can induce monopolar mitotic spindles [178], or cause

misalignment of chromosomes, activating the spindle assembly checkpoint [179,
180]. Inhibition of Aurora B suppresses the spindle assembly checkpoint, leading to
improper mitotic exit. It also causes failure in cytokinesis, driving both polyploidy
and aneuploidy [180–182].
Many of the Plk1 inhibitors compete with ATP, block the substrate binding site
[183], or bind the essential Polo box domain of Plk1 [184]. Inhibition of Plk1 cre-
ates monopolar spindles, and induces mitotic arrest [185, 186] ultimately activating
apoptosis [187]. Aurora B and Plk1 inhibitors have shown efficacy in suppressing
tumor growth in human xenograft models [188, 189]. Several Aurora and Plk1
inhibitors are in clinical trials (Table 13.2) as a single agent or in combination with
other drugs.
The passenger proteins, which regulate both the spindle assembly checkpoint
and cell cleavage, are attractive targets for chemotherapy. Disruption of Survivin
causes impaired chromosome movements, spindle checkpoint override and poly-
ploidy [190], suppressing cell proliferation [191]. YM155, a drug that interferes
with induction of the Survivin promoter, has entered clinical trials, with inconclu-
sive results thus far in non-small-cell lung cancer [192]. Another drug, S12, that
allosterically blocks Survivin interactions with CPC binding partners, shows prom-
ise as an inhibitor of tumor growth in vivo [193]. Although Survivin has been
targeted because of a presumptive anti-apoptotic function, it appears that its princi-
pal function in cells is as a passenger protein in mitosis [194], and it is thus appro-
priate to discuss Survivin solely as a mitotic chemotherapy target.
Several mitotic spindle inhibitors have been clinically successful, but the reason
for their success when other spindle inhibitors are without effect remains a mystery.
It is further difficult to assess why only certain tumors are susceptible to a particular
spindle agent. The outcome perhaps depends on different drug induced mecha-
nisms. Drugs such as paclitaxel cause apoptosis and the activity increases by addi-
tional proapoptotic changes induced by the drug when compared to nocodazole and
kinesin-5 inhibitor [195]. Similarly the pharmacokinetic response in vivo varies
with different drugs as they induce different genetic responses [196]. Another factor
is a difference in drug concentrations that reach tumors.
Efficacy can vary due to tumor type, relative gene expression of the target and
differences in drug induced apoptotic changes. For example, p53 deficient tumors or
tumors with mutated Ras are more sensitive to Plk1 inhibitors [197, 198].
Overexpressed Aurora A phosphorylates, inactivates and degrades p53 [199, 200],
while overexpression of Aurora B increases the level of phosphorylated Survivin
[201], which can be countered by using Aurora inhibitors. Similarly, expression of
BRCA1 mediates sensitivity to apoptosis caused by microtubule targeting drugs
when compared to tumor response to etoposide or bleomycin [202]. Drug resistance
through changes in gene expression poses another problem.
Efficacy of inhibitors can also be highly variable in different cells of the same
tumor cell line. Death following exposure to spindle inhibitors can either occur dur-
ing prolonged mitotic arrest or after inappropriate mitotic exit [203–205]. Prolonged
mitotic arrest is rapidly lethal, while inappropriate exit is lethal in the longer term,
but it is not clear that prolonged arrest is actually achieved during chemotherapy,
which, for paclitaxel, typically involves brief drug infusions on a weekly basis
[206]. Which mechanism is of greatest consequence in actual tumor therapy in situ
is at present unknown, but will be of importance to the design of therapeutic
approaches that either induce prolonged mitotic arrest or suppress the SAC and
drive premature mitotic exit [203, 207].
The non-motor spindle proteins are often overexpressed in tumors, and overex-
pression perhaps determines the tumor response to microtubule-targeted drugs. For
example, overexpression of stathmin and MAP4 in tumor cells correlates with
reduced sensitivity to paclitaxel and vinblastine respectively [208–210], while over-
expression of Aurora A causes resistance to paclitaxel [43]. In a related manner,
decrease in expression of Survivin and Plk1 increases sensitivity to paclitaxel
[211–214].
13.5 Summary
The focus of this review is to highlight potential molecular targets for chemotherapy
among the non-motor spindle proteins. The therapeutic success of drugs that have
targeted microtubule dynamics is substantial, but limited by several factors. Perhaps
the most important of these limiting factors is the non-specificity of tubulin as a
target. Microtubules are required by every cell both for mitosis and for interphase
functions such as regulating cell motility and organelle transport. The broad spec-
trum effects of these drugs thus limit therapeutic applications.
It therefore would be most welcome to identify proteins that are both essential to
mitosis and specific to mitosis, and that are dysregulated in tumors. Many of the
proteins described in this review meet these criteria and a few, certainly not the
majority, should be promising candidates for drug development. For drug screens,
the most fruitful approach should be to target proteins with well-identified activity,
or that act as well identified modifiers of activity. The protein kinases such as Aurora
A and B, and Plk1 are already being targeted in drug screens and in clinical trials.
Nek2 is another obvious candidate for chemotherapy. However, drug screens may
not always require activity readouts, but can be based on molecular analysis of the
protein structure. Rational drug design based on protein structure is an alternative,
as exemplified by S12, a drug that interferes with a specific Survivin motif [193],
which was developed using an in-house allosteric modeling drug discovery (CIAM)
algorithm.
To date, drugs that disrupt spindle function have been used in combination ther-
apy with DNA damaging agents, and appear to be more successful when used in
combination, as witnessed by many clinical studies [215–220]. Such combination
approaches may play on common aberrations in tumor cells that interfere with the
cell cycle arrest following DNA damage during interphase. These aberrations
include suppression or mutation of p53 and the CDKN2A complex [139]. Needless
to say, combination approaches of this nature may be of greater value for chemo-
therapy when they target both aberrant DNA repair and aberrant expression of
mitotic protein kinases in the same tumor cell.
As noted above, the dysregulation of multiple spindle components in tumor cells
may actually suppress the response to agents that target tubulin in chemotherapy
[43, 208–214]. Drugs that instead target other spindle components, especially those
that are dysregulated, could therefore be highly effective. In a selected target
approach, exploiting personalized medicine, one can anticipate that inhibitors of
dysregulated mitotic kinases and other specific dysregulated mitotic targets may
ultimately be much more successful than the use of tubulin inhibitors in
chemotherapy.
Acknowledgements Supported by NIH grants R01GM068107 and R01GM088716 (R.L.M.)
References
1. Compton DA (2000) Spindle assembly in animal cells. Annu Rev Biochem 69:95–114
2. Maiato H, Rieder CL, Khodjakov A (2004) Kinetochore-driven formation of kinetochore
fibers contributes to spindle assembly during animal mitosis. J Cell Biol 167(5):831–840
3. Fant X, Merdes A, Haren L (2004) Cell and molecular biology of spindle poles and NuMA. Int
Rev Cytol 238:1–57
4. Mollinari C et al (2002) PRC1 is a microtubule binding and bundling protein essential to
maintain the mitotic spindle midzone. J Cell Biol 157(7):1175–1186
5. Wong J et al (2007) A protein interaction map of the mitotic spindle. Mol Biol Cell
18(10):3800–3809
6. Rojas AM et al (2012) Uncovering the molecular machinery of the human spindle–an integra-
tion of wet and dry systems biology. PLoS One 7(3):e31813
7. Margolis RL, Wilson L, Keifer BI (1978) Mitotic mechanism based on intrinsic microtubule
behaviour. Nature 272(5652):450–452
8. Witt PL, Ris H, Borisy GG (1980) Origin of kinetochore microtubules in Chinese hamster
ovary cells. Chromosoma 81(3):483–505
9. Khodjakov A, Copenagle L, Gordon MB, Compton DA, Kapoor TM (2003) Minus-end cap-
ture of preformed kinetochore fibers contributes to spindle morphogenesis. J Cell Biol
160(5):671–683
10. Khodjakov A, Cole RW, Oakley BR, Rieder CL (2000) Centrosome-independent mitotic
spindle formation in vertebrates. Curr Biol 10(2):59–67
11. Hinchcliffe EH, Miller FJ, Cham M, Khodjakov A, Sluder G (2001) Requirement of a centro-
somal activity for cell cycle progression through G1 into S phase. Science
291(5508):1547–1550
12. Sawin KE, Mitchison TJ (1994) Microtubule flux in mitosis is independent of chromosomes,
centrosomes, and antiparallel microtubules. Mol Biol Cell 5(2):217–226
13. Carazo-Salas RE et al (1999) Generation of GTP-bound Ran by RCC1 is required for
chromatin-induced mitotic spindle formation. Nature 400(6740):178–181
14. Tsai MY et al (2003) A Ran signalling pathway mediated by the mitotic kinase Aurora A in
spindle assembly. Nat Cell Biol 5(3):242–248
15. O’Connell CB, Loncarek J, Kalab P, Khodjakov A (2009) Relative contributions of chroma-
tin and kinetochores to mitotic spindle assembly. J Cell Biol 187(1):43–51
16. Kalab P, Weis K, Heald R (2002) Visualization of a Ran-GTP gradient in interphase and
mitotic Xenopus egg extracts. Science 295(5564):2452–2456
17. Heald R et al (1996) Self-organization of microtubules into bipolar spindles around artificial
chromosomes in Xenopus egg extracts. Nature 382(6590):420–425
18. Maresca TJ et al (2009) Spindle assembly in the absence of a RanGTP gradient requires
localized CPC activity. Curr Biol 19(14):1210–1215
19. Mitchison TJ, Salmon ED (1992) Poleward kinetochore fiber movement occurs during both
metaphase and anaphase-A in newt lung cell mitosis. J Cell Biol 119(3):569–582
20. Maddox P, Desai A, Oegema K, Mitchison TJ, Salmon ED (2002) Poleward microtubule flux
is a major component of spindle dynamics and anaphase a in mitotic Drosophila embryos.
Curr Biol 12(19):1670–1674
21. Grill SW, Howard J, Schaffer E, Stelzer EH, Hyman AA (2003) The distribution of active
force generators controls mitotic spindle position. Science 301(5632):518–521
22. Kotak S, Gonczy P (2013) Mechanisms of spindle positioning: cortical force generators in
the limelight. Curr Opin Cell Biol 25(6):741–748
23. Woodard GE et al (2010) Ric-8A and Gi alpha recruit LGN, NuMA, and dynein to the cell
cortex to help orient the mitotic spindle. Mol Cell Biol 30(14):3519–3530
24. Laan L et al (2012) Cortical dynein controls microtubule dynamics to generate pulling forces
that position microtubule asters. Cell 148(3):502–514
25. Canman JC et al (2003) Determining the position of the cell division plane. Nature
424(6952):1074–1078
26. Skoufias DA, Indorato RL, Lacroix F, Panopoulos A, Margolis RL (2007) Mitosis persists in
the absence of Cdk1 activity when proteolysis or protein phosphatase activity is suppressed.
J Cell Biol 179(4):671–685
27. Zhou H et al (1998) Tumour amplified kinase STK15/BTAK induces centrosome amplifica-
tion, aneuploidy and transformation. Nat Genet 20(2):189–193
28. Tanaka T et al (1999) Centrosomal kinase AIK1 is overexpressed in invasive ductal carci-
noma of the breast. Cancer Res 59(9):2041–2044
29. Warner SL et al (2006) Comparing Aurora A and Aurora B as molecular targets for growth
inhibition of pancreatic cancer cells. Mol Cancer Ther 5(10):2450–2458
30. Araki K, Nozaki K, Ueba T, Tatsuka M, Hashimoto N (2004) High expression of Aurora-B/
Aurora and Ipll-like midbody-associated protein (AIM-1) in astrocytomas. J Neurooncol
67(1–2):53–64
31. Smith SL et al (2005) Overexpression of aurora B kinase (AURKB) in primary non-small cell
lung carcinoma is frequent, generally driven from one allele, and correlates with the level of
genetic instability. Br J Cancer 93(6):719–729
32. Eckerdt F et al (2005) Polo-like kinase 1-mediated phosphorylation stabilizes Pin1 by inhibit-
ing its ubiquitination in human cells. J Biol Chem 280(44):36575–36583
33. Strebhardt K, Ullrich A (2006) Targeting polo-like kinase 1 for cancer therapy. Nat Rev
Cancer 6(4):321–330
34. Keen N, Taylor S (2004) Aurora-kinase inhibitors as anticancer agents. Nat Rev Cancer
4(12):927–936
35. Malumbres M, Barbacid M (2009) Cell cycle, CDKs and cancer: a changing paradigm. Nat
Rev Cancer 9(3):153–166
36. Bischoff JR, Plowman GD (1999) The Aurora/Ipl1p kinase family: regulators of chromo-
some segregation and cytokinesis. Trends Cell Biol 9(11):454–459
37. Giet R, Prigent C (1999) Aurora/Ipl1p-related kinases, a new oncogenic family of mitotic
serine-threonine kinases. J Cell Sci 112(Pt 21):3591–3601
38. Giet R, Petretti C, Prigent C (2005) Aurora kinases, aneuploidy and cancer, a coincidence or
a real link? Trends Cell Biol 15(5):241–250
39. Ruchaud S, Carmena M, Earnshaw WC (2007) Chromosomal passengers: conducting cell
division. Nature reviews. Mol Cell Biol 8(10):798–812
40. Kufer TA et al (2002) Human TPX2 is required for targeting Aurora-A kinase to the spindle.
J Cell Biol 158(4):617–623
41. Hirota T et al (2003) Aurora-A and an interacting activator, the LIM protein Ajuba, are
required for mitotic commitment in human cells. Cell 114(5):585–598
42. Hannak E, Kirkham M, Hyman AA, Oegema K (2001) Aurora-A kinase is required for cen-
trosome maturation in Caenorhabditis elegans. J Cell Biol 155(7):1109–1116
43. Anand S, Penrhyn-Lowe S, Venkitaraman AR (2003) AURORA-A amplification overrides
the mitotic spindle assembly checkpoint, inducing resistance to Taxol. Cancer Cell
3(1):51–62
44. Macurek L et al (2008) Regulation of microtubule nucleation from membranes by complexes
of membrane-bound gamma-tubulin with Fyn kinase and phosphoinositide 3-kinase.
Biochem J 416(3):421–430
45. Dutertre S, Descamps S, Prigent C (2002) On the role of aurora-A in centrosome function.
Oncogene 21(40):6175–6183
46. Meraldi P, Honda R, Nigg EA (2002) Aurora-A overexpression reveals tetraploidization as a
major route to centrosome amplification in p53-/- cells. EMBO J 21(4):483–492
47. Cooke CA, Heck MM, Earnshaw WC (1987) The inner centromere protein (INCENP) anti-
gens: movement from inner centromere to midbody during mitosis. J Cell Biol
105(5):2053–2067
48. Adams RR et al (2000) INCENP binds the Aurora-related kinase AIRK2 and is required to
target it to chromosomes, the central spindle and cleavage furrow. Curr Biol
10(17):1075–1078
49. Bishop JD, Schumacher JM (2002) Phosphorylation of the carboxyl terminus of inner centro-
mere protein (INCENP) by the Aurora B kinase stimulates Aurora B kinase activity. J Biol
Chem 277(31):27577–27580
50. Honda R, Korner R, Nigg EA (2003) Exploring the functional interactions between Aurora
B, INCENP, and survivin in mitosis. Mol Biol Cell 14(8):3325–3341
51. Sessa F et al (2005) Mechanism of Aurora B activation by INCENP and inhibition by hesper-
adin. Mol Cell 18(3):379–391
52. Carmena M, Earnshaw WC (2003) The cellular geography of aurora kinases. Nat Rev Mol
Cell Biol 4(11):842–854
6(2):253–268
54. Jeyaprakash AA et al (2007) Structure of a Survivin-Borealin-INCENP core complex reveals
how chromosomal passengers travel together. Cell 131(2):271–285
55. Jelluma N et al (2008) Mps1 phosphorylates Borealin to control Aurora B activity and chro-
mosome alignment. Cell 132(2):233–246
56. Ciferri C et al (2008) Implications for kinetochore-microtubule attachment from the structure
of an engineered Ndc80 complex. Cell 133(3):427–439
57. Fuller BG et al (2008) Midzone activation of aurora B in anaphase produces an intracellular
phosphorylation gradient. Nature 453(7198):1132–1136
58. Tien AC et al (2004) Identification of the substrates and interaction proteins of aurora kinases
from a protein-protein interaction model. Mol Cell Proteomics 3(1):93–104
59. Meraldi P, Honda R, Nigg EA (2004) Aurora kinases link chromosome segregation and cell
division to cancer susceptibility. Curr Opin Genet Dev 14(1):29–36
60. Ohashi S et al (2006) Phospho-regulation of human protein kinase Aurora-A: analysis using
anti-phospho-Thr288 monoclonal antibodies. Oncogene 25(59):7691–7702
61. Fu J, Bian M, Jiang Q, Zhang C (2007) Roles of Aurora kinases in mitosis and tumorigenesis.
Mol Cancer Res 5(1):1–10
62. Macurek L et al (2008) Polo-like kinase-1 is activated by aurora A to promote checkpoint
recovery. Nature 455(7209):119–123
63. Kunitoku N et al (2003) CENP-A phosphorylation by Aurora-A in prophase is required for
enrichment of Aurora-B at inner centromeres and for kinetochore function. Dev Cell
5(6):853–864
64. Hans F, Skoufias DA, Dimitrov S, Margolis RL (2009) Molecular distinctions between
Aurora A and B: a single residue change transforms Aurora A into correctly localized and
functional Aurora B. Mol Biol Cell 20(15):3491–3502
65. Lindon C, Pines J (2004) Ordered proteolysis in anaphase inactivates Plk1 to contribute to
proper mitotic exit in human cells. J Cell Biol 164(2):233–241
66. Crane R, Kloepfer A, Ruderman JV (2004) Requirements for the destruction of human
Aurora-A. J Cell Sci 117(Pt 25):5975–5983
67. Petronczki M, Lenart P, Peters JM (2008) Polo on the rise-from mitotic entry to cytokinesis
with Plk1. Dev Cell 14(5):646–659
68. van Vugt MA, Medema RH (2005) Getting in and out of mitosis with Polo-like kinase-1.
Oncogene 24(17):2844–2859
69. Schatten H (2008) The mammalian centrosome and its functional significance. Histochem
Cell Biol 129(6):667–686
70. Neef R et al (2007) Choice of Plk1 docking partners during mitosis and cytokinesis is con-
trolled by the activation state of Cdk1. Nat Cell Biol 9(4):436–444
71. Hu CK, Ozlu N, Coughlin M, Steen JJ, Mitchison TJ (2012) Plk1 negatively regulates PRC1
to prevent premature midzone formation before cytokinesis. Mol Biol Cell
23(14):2702–2711
72. Petronczki M, Glotzer M, Kraut N, Peters JM (2007) Polo-like kinase 1 triggers the initiation
of cytokinesis in human cells by promoting recruitment of the RhoGEF Ect2 to the central
spindle. Dev Cell 12(5):713–725
73. Fourest-Lieuvin A et al (2006) Microtubule regulation in mitosis: tubulin phosphorylation by
the cyclin-dependent kinase Cdk1. Mol Biol Cell 17(3):1041–1050
74. Blethrow JD, Glavy JS, Morgan DO, Shokat KM (2008) Covalent capture of kinase-specific
phosphopeptides reveals Cdk1-cyclin B substrates. Proc Natl Acad Sci U S A 105(5):1442–1447
75. Wheatley SP et al (1997) CDK1 inactivation regulates anaphase spindle dynamics and cyto-
kinesis in vivo. J Cell Biol 138(2):385–393
76. Dephoure N et al (2008) A quantitative atlas of mitotic phosphorylation. Proc Natl Acad Sci
U S A 105(31):10762–10767
77. Nousiainen M, Sillje HH, Sauer G, Nigg EA, Korner R (2006) Phosphoproteome analysis of
the human mitotic spindle. Proc Natl Acad Sci U S A 103(14):5391–5396
78. Barra HS, Arce CA, Argarana CE (1988) Posttranslational tyrosination/detyrosination of
tubulin. Mol Neurobiol 2(2):133–153
79. Sahab ZJ et al (2011) Tumor suppressor RARRES1 interacts with cytoplasmic carboxypepti-
dase AGBL2 to regulate the alpha-tubulin tyrosination cycle. Cancer Res 71(4):1219–1228
80. Bieling P et al (2008) CLIP-170 tracks growing microtubule ends by dynamically recogniz-
ing composite EB1/tubulin-binding sites. J Cell Biol 183(7):1223–1233
81. Lansbergen G, Akhmanova A (2006) Microtubule plus end: a hub of cellular activities.
Traffic 7(5):499–507
82. Akhmanova A, Steinmetz MO (2008) Tracking the ends: a dynamic protein network controls
the fate of microtubule tips. Nat Rev Mol Cell Biol 9(4):309–322
83. Lafanechere L et al (1998) Suppression of tubulin tyrosine ligase during tumor growth. J Cell
Sci 111(Pt 2):171–181
84. Mialhe A et al (2001) Tubulin detyrosination is a frequent occurrence in breast cancers of
poor prognosis. Cancer Res 61(13):5024–5027
85. Peris L et al (2006) Tubulin tyrosination is a major factor affecting the recruitment of CAP-
Gly proteins at microtubule plus ends. J Cell Biol 174(6):839–849
86. Mitchison TJ et al (2004) Bipolarization and poleward flux correlate during Xenopus extract
spindle assembly. Mol Biol Cell 15(12):5603–5615
87. Pedrotti B, Ulloa L, Avila J, Islam K (1996) Characterization of microtubule-associated pro-
tein MAP1B: phosphorylation state, light chains, and binding to microtubules. Biochemistry
35(9):3016–3023
88. Valiron O, Caudron N, Job D (2001) Microtubule dynamics. Cell Mol Life Sci
58(14):2069–2084
89. Chien CL, Lu KS, Lin YS, Hsieh CJ, Hirokawa N (2005) The functional cooperation of
MAP1A heavy chain and light chain 2 in the binding of microtubules. Exp Cell Res
308(2):446–458
90. Curmi PA et al (1999) Stathmin and its phosphoprotein family: general properties, biochemi-
cal and functional interaction with tubulin. Cell Struct Funct 24(5):345–357
91. Manna T, Thrower D, Miller HP, Curmi P, Wilson L (2006) Stathmin strongly increases the
minus end catastrophe frequency and induces rapid treadmilling of bovine brain microtubules
at steady state in vitro. J Biol Chem 281(4):2071–2078
92. Walczak CE, Mitchison TJ, Desai A (1996) XKCM1: a Xenopus kinesin-related protein that
regulates microtubule dynamics during mitotic spindle assembly. Cell 84(1):37–47
93. McNally FJ, Thomas S (1998) Katanin is responsible for the M-phase microtubule-severing
activity in Xenopus eggs. Mol Biol Cell 9(7):1847–1861
94. Tournebize R et al (2000) Control of microtubule dynamics by the antagonistic activities of
XMAP215 and XKCM1 in Xenopus egg extracts. Nat Cell Biol 2(1):13–19
95. Dionne MA, Sanchez A, Compton DA (2000) ch-TOGp is required for microtubule aster
formation in a mammalian mitotic extract. J Biol Chem 275(16):12346–12352
96. Cassimeris L, Morabito J (2004) TOGp, the human homolog of XMAP215/Dis1, is required
for centrosome integrity, spindle pole organization, and bipolar spindle assembly. Mol Biol
Cell 15(4):1580–1590
97. Holmfeldt P, Stenmark S, Gullberg M (2004) Differential functional interplay of TOGp/
XMAP215 and the KinI kinesin MCAK during interphase and mitosis. EMBO J
23(3):627–637
98. Kinoshita K, Noetzel TL, Arnal I, Drechsel DN, Hyman AA (2006) Global and local control
of microtubule destabilization promoted by a catastrophe kinesin MCAK/XKCM1. J Muscle
Res Cell Motil 27(2):107–114
99. Kinoshita K, Habermann B, Hyman AA (2002) XMAP215: a key component of the dynamic
microtubule cytoskeleton. Trends Cell Biol 12(6):267–273
100. Ookata K et al (1995) Cyclin B interaction with microtubule-associated protein 4 (MAP4)
targets p34cdc2 kinase to microtubules and is a potential regulator of M-phase microtubule
dynamics. J Cell Biol 128(5):849–862
101. Cassimeris L (1999) Accessory protein regulation of microtubule dynamics throughout the
cell cycle. Curr Opin Cell Biol 11(1):134–141
102. Wu X, Xiang X, Hammer JA 3rd (2006) Motor proteins at the microtubule plus-end. Trends
Cell Biol 16(3):135–143
103. Tirnauer JS, Canman JC, Salmon ED, Mitchison TJ (2002) EB1 targets to kinetochores with
attached, polymerizing microtubules. Mol Biol Cell 13(12):4308–4316
104. Galjart N (2005) CLIPs and CLASPs and cellular dynamics. Nat Rev Mol Cell Biol
6(6):487–498
105. Pabion M, Job D, Margolis RL (1984) Sliding of STOP proteins on microtubules.
Biochemistry 23(26):6642–6648
106. Margolis RL, Wilson L (1981) Microtubule treadmills–possible molecular machinery. Nature
293(5835):705–711
107. Garel JR, Job D, Margolis RL (1987) Model of anaphase chromosome movement based on
polymer-guided diffusion. Proc Natl Acad Sci U S A 84(11):3599–3603
108. Rogers SL, Rogers GC, Sharp DJ, Vale RD (2002) Drosophila EB1 is important for proper
assembly, dynamics, and positioning of the mitotic spindle. J Cell Biol 158(5):873–884
109. Maiato H et al (2003) Human CLASP1 is an outer kinetochore component that regulates
spindle microtubule dynamics. Cell 113(7):891–904
110. Reis R et al (2009) Dynein and mast/orbit/CLASP have antagonistic roles in regulating
kinetochore-microtubule plus-end dynamics. J Cell Sci 122(Pt 14):2543–2553
111. Schuyler SC, Pellman D (2001) Microtubule “plus-end-tracking proteins”: the end is just the
beginning. Cell 105(4):421–424
112. Maiato H, Sampaio P, Sunkel CE (2004) Microtubule-associated proteins and their essential
roles during mitosis. Int Rev Cytol 241:53–153
113. Palmer DK, O’Day K, Trong HL, Charbonneau H, Margolis RL (1991) Purification of the
centromere-specific protein CENP-A and demonstration that it is a distinctive histone. Proc
Natl Acad Sci U S A 88(9):3734–3738
114. Kline-Smith SL, Sandall S, Desai A (2005) Kinetochore-spindle microtubule interactions
during mitosis. Curr Opin Cell Biol 17(1):35–46
115. Logarinho E, Bousbaa H (2008) Kinetochore-microtubule interactions “in check” by Bub1,
Bub3 and BubR1: the dual task of attaching and signalling. Cell Cycle 7(12):1763–1768
116. Cheeseman IM, Chappie JS, Wilson-Kubalek EM, Desai A (2006) The conserved KMN net-
work constitutes the core microtubule-binding site of the kinetochore. Cell 127(5):983–997
117. Tanaka TU, Desai A (2008) Kinetochore-microtubule interactions: the means to the end. Curr
Opin Cell Biol 20(1):53–63
118. DeLuca JG et al (2006) Kinetochore microtubule dynamics and attachment stability are regu-
lated by Hec1. Cell 127(5):969–982
119. Maiato H, DeLuca J, Salmon ED, Earnshaw WC (2004) The dynamic kinetochore-
microtubule interface. J Cell Sci 117(Pt 23):5461–5477
120. Chan GK, Liu ST, Yen TJ (2005) Kinetochore structure and function. Trends Cell Biol
15(11):589–598
121. Wan X et al (2009) Protein architecture of the human kinetochore microtubule attachment
site. Cell 137(4):672–684
122. Karess R (2005) Rod-Zw10-Zwilch: a key player in the spindle checkpoint. Trends Cell Biol
15(7):386–392
123. Griffis ER, Stuurman N, Vale RD (2007) Spindly, a novel protein essential for silencing the
spindle assembly checkpoint, recruits dynein to the kinetochore. J Cell Biol
177(6):1005–1015
124. Chan YW et al (2009) Mitotic control of kinetochore-associated dynein and spindle orienta-
tion by human spindly. J Cell Biol 185(5):859–874
125. Maure JF et al (2011) The Ndc80 loop region facilitates formation of kinetochore attachment
to the dynamic microtubule plus end. Curr Biol 21(3):207–213
126. Tanaka K, Kitamura E, Kitamura Y, Tanaka TU (2007) Molecular mechanisms of microtubule-
dependent kinetochore transport toward spindle poles. J Cell Biol 178(2):269–281
127. Miranda JJ, De Wulf P, Sorger PK, Harrison SC (2005) The yeast DASH complex forms
closed rings on microtubules. Nat Struct Mol Biol 12(2):138–143
128. Nogales E, Ramey VH (2009) Structure-function insights into the yeast Dam1 kinetochore
complex. J Cell Sci 122(Pt 21):3831–3836
129. Mollinari C et al (2003) The mammalian passenger protein TD-60 is an RCC1 family
member with an essential role in prometaphase to metaphase progression. Dev Cell
5(2):295–307
130. Kotwaliwale C, Biggins S (2006) Microtubule capture: a concerted effort. Cell
127(6):1105–1108
131. Rosasco-Nitcher SE, Lan W, Khorasanizadeh S, Stukenberg PT (2008) Centromeric Aurora-B
activation requires TD-60, microtubules, and substrate priming phosphorylation. Science
319(5862):469–472
132. Wang F et al (2010) Histone H3 Thr-3 phosphorylation by Haspin positions Aurora B at
centromeres in mitosis. Science 330(6001):231–235
133. Sampath SC et al (2004) The chromosomal passenger complex is required for chromatin-
induced microtubule stabilization and spindle assembly. Cell 118(2):187–202
134. Emanuele MJ et al (2008) Aurora B kinase and protein phosphatase 1 have opposing roles in
modulating kinetochore assembly. J Cell Biol 181(2):241–254
135. Skoufias DA, Andreassen PR, Lacroix FB, Wilson L, Margolis RL (2001) Mammalian mad2
and bub1/bubR1 recognize distinct spindle-attachment and kinetochore-tension checkpoints.
Proc Natl Acad Sci U S A 98(8):4492–4497
136. Tanaka TU et al (2002) Evidence that the Ipl1-Sli15 (Aurora kinase-INCENP) complex pro-
motes chromosome bi-orientation by altering kinetochore-spindle pole connections. Cell
108(3):317–329
137. Knowlton AL, Lan W, Stukenberg PT (2006) Aurora B is enriched at merotelic attachment
sites, where it regulates MCAK. Curr Biol 16(17):1705–1710
138. Takai N, Hamanaka R, Yoshimatsu J, Miyakawa I (2005) Polo-like kinases (Plks) and cancer.
Oncogene 24(2):287–291
139. Negrini S, Gorgoulis VG, Halazonetis TD (2010) Genomic instability–an evolving hallmark
of cancer. Nat Rev Mol Cell Biol 11(3):220–228
140. Cimini D, Wan X, Hirel CB, Salmon ED (2006) Aurora kinase promotes turnover of kineto-
chore microtubules to reduce chromosome segregation errors. Curr Biol 16(17):1711–1718
141. Gascoigne KE, Cheeseman IM (2011) Kinetochore assembly: if you build it, they will come.
Curr Opin Cell Biol 23(1):102–108
142. Merdes A, Heald R, Samejima K, Earnshaw WC, Cleveland DW (2000) Formation of spindle
poles by dynein/dynactin-dependent transport of NuMA. J Cell Biol 149(4):851–862
143. Merdes A, Ramyar K, Vechio JD, Cleveland DW (1996) A complex of NuMA and cytoplas-
mic dynein is essential for mitotic spindle assembly. Cell 87(3):447–458
144. Garrett S, Auer K, Compton DA, Kapoor TM (2002) hTPX2 is required for normal spindle
morphology and centrosome integrity during vertebrate cell division. Curr Biol
12(23):2055–2059
145. Wittmann T, Wilm M, Karsenti E, Vernos I (2000) TPX2, A novel Xenopus MAP involved in
spindle pole organization. J Cell Biol 149(7):1405–1418
146. Eyers PA, Maller JL (2004) Regulation of Xenopus Aurora A activation by TPX2. J Biol
Chem 279(10):9008–9015
147. Gruss OJ, Vernos I (2004) The mechanism of spindle assembly: functions of Ran and its
target TPX2. J Cell Biol 166(7):949–955
148. Kufer TA, Nigg EA, Sillje HH (2003) Regulation of Aurora-A kinase on the mitotic spindle.
Chromosoma 112(4):159–163
149. Hood FE, Royle SJ (2011) Pulling it together: the mitotic function of TACC3. Bioarchitecture
1(3):105–109
150. Peset I et al (2005) Function and regulation of Maskin, a TACC family protein, in microtu-
bule growth during mitosis. J Cell Biol 170(7):1057–1066
151. Nachury MV et al (2001) Importin beta is a mitotic target of the small GTPase Ran in spindle
assembly. Cell 104(1):95–106
152. Taylor S, Peters JM (2008) Polo and Aurora kinases: lessons derived from chemical biology.
Curr Opin Cell Biol 20(1):77–84
153. Lane HA, Nigg EA (1996) Antibody microinjection reveals an essential role for human polo-
like kinase 1 (Plk1) in the functional maturation of mitotic centrosomes. J Cell Biol 135(6 Pt
2):1701–1713
154. Berdnik D, Knoblich JA (2002) Drosophila Aurora-A is required for centrosome maturation
and actin-dependent asymmetric protein localization during mitosis. Curr Biol 12(8):640–647
155. Tsai MY, Zheng Y (2005) Aurora A kinase-coated beads function as microtubule-organizing
centers and enhance RanGTP-induced spindle assembly. Curr Biol 15(23):2156–2163
156. Barr AR, Gergely F (2007) Aurora-A: the maker and breaker of spindle poles. J Cell Sci
120(Pt 17):2987–2996
157. Pihan GA (2013) Centrosome dysfunction contributes to chromosome instability, chromoa-
nagenesis, and genome reprograming in cancer. Front Oncol 3:277
158. Hayward DG et al (2004) The centrosomal kinase Nek2 displays elevated levels of protein
expression in human breast cancer. Cancer Res 64(20):7370–7376
159. Cappello P et al (2014) Role of Nek2 on centrosome duplication and aneuploidy in breast
cancer cells. Oncogene 33:2375–2384
160. Takahashi Y et al (2014) Up-regulation of NEK2 by microRNA-128 methylation is associ-
ated with poor prognosis in colorectal cancer. Ann Surg Oncol 21:205–212
161. Fry AM, Schultz SJ, Bartek J, Nigg EA (1995) Substrate specificity and cell cycle regulation
of the Nek2 protein kinase, a potential human homolog of the mitotic regulator NIMA of
Aspergillus nidulans. J Biol Chem 270(21):12899–12905
162. Mi J, Guo C, Brautigan DL, Larner JM (2007) Protein phosphatase-1alpha regulates centro-
some splitting through Nek2. Cancer Res 67(3):1082–1089
and length of antiparallel microtubule overlaps. Cell 142(3):420–432
164. Subramanian R et al (2010) Insights into antiparallel microtubule crosslinking by PRC1, a
conserved nonmotor microtubule binding protein. Cell 142(3):433–443
KIF4 and its binding partner PRC1 in organized central spindle midzone formation. EMBO
J 23(16):3237–3248
166. Shrestha S, Wilmeth LJ, Eyer J, Shuster CB (2012) PRC1 controls spindle polarization and
recruitment of cytokinetic factors during monopolar cytokinesis. Mol Biol Cell
23(7):1196–1207
167. Mollinari C et al (2005) Ablation of PRC1 by small interfering RNA demonstrates that cyto-
kinetic abscission requires a central spindle bundle in mammalian cells, whereas completion
of furrowing does not. Mol Biol Cell 16(3):1043–1055
168. Floyd S, Pines J, Lindon C (2008) APC/C Cdh1 targets aurora kinase to control reorganiza-
tion of the mitotic spindle at anaphase. Curr Biol 18(21):1649–1658
169. Agromayor M, Martin-Serrano J (2013) Knowing when to cut and run: mechanisms that
control cytokinetic abscission. Trends Cell Biol 23(9):433–441
170. Jordan MA, Wilson L (2004) Microtubules as a target for anticancer drugs. Nat Rev Cancer
4(4):253–265
171. Yared JA, Tkaczuk KH (2012) Update on taxane development: new analogs and new formu-
lations. Drug Des Devel Ther 6:371–384
172. Dumontet C, Jordan MA, Lee FF (2009) Ixabepilone: targeting betaIII-tubulin expression in
taxane-resistant malignancies. Mol Cancer Ther 8(1):17–25
173. Okouneva T, Azarenko O, Wilson L, Littlefield BA, Jordan MA (2008) Inhibition of centro-
mere dynamics by eribulin (E7389) during mitotic metaphase. Mol Cancer Ther
7(7):2003–2011
174. McInnes C, Mezna M, Fischer PM (2005) Progress in the discovery of polo-like kinase inhib-
itors. Curr Top Med Chem 5(2):181–197
175. Vader G, Lens SM (2008) The Aurora kinase family in cell division and cancer. Biochim
Biophys Acta 1786(1):60–72
176. Smith MR et al (1997) Malignant transformation of mammalian cells initiated by constitutive
expression of the polo-like kinase. Biochem Biophys Res Commun 234(2):397–405
177. Agnese V et al (2007) The role of Aurora-A inhibitors in cancer therapy. Ann Oncol 18(Suppl
6):vi47–vi52
178. Marumoto T, Zhang D, Saya H (2005) Aurora-A – a guardian of poles. Nat Rev Cancer
5(1):42–50
179. Yang H et al (2005) Mitotic requirement for aurora A kinase is bypassed in the absence of
aurora B kinase. FEBS Lett 579(16):3385–3391
180. Shimomura T et al (2010) MK-5108, a highly selective Aurora-A kinase inhibitor, shows
antitumor activity alone and in combination with docetaxel. Mol Cancer Ther 9(1):157–166
181. Hauf S et al (2003) The small molecule Hesperadin reveals a role for Aurora B in correcting
kinetochore-microtubule attachment and in maintaining the spindle assembly checkpoint. J
Cell Biol 161(2):281–294
182. Ditchfield C et al (2003) Aurora B couples chromosome alignment with anaphase by target-
ing BubR1, Mad2, and Cenp-E to kinetochores. J Cell Biol 161(2):267–280
183. Marzo I, Naval J (2013) Antimitotic drugs in cancer chemotherapy: promises and pitfalls.
Biochem Pharmacol 86(6):703–710
184. Medema RH, Lin CC, Yang JC (2011) Polo-like kinase 1 inhibitors and their potential role in
anticancer therapy, with a focus on NSCLC. Clin Cancer Res 17(20):6459–6466
185. Sumara I et al (2004) Roles of polo-like kinase 1 in the assembly of functional mitotic spin-
dles. Curr Biol 14(19):1712–1722
186. Kaestner P, Bastians H (2010) Mitotic drug targets. J Cell Biochem 111(2):258–265
187. Gleixner KV et al (2010) Polo-like kinase 1 (Plk1) as a novel drug target in chronic myeloid
leukemia: overriding imatinib resistance with the Plk1 inhibitor BI 2536. Cancer Res
70(4):1513–1523
188. Wilkinson RW et al (2007) AZD1152, a selective inhibitor of Aurora B kinase, inhibits
human tumor xenograft growth by inducing apoptosis. Clin Cancer Res 13(12):3682–3688
189. Hikichi Y et al (2012) TAK-960, a novel, orally available, selective inhibitor of polo-like
kinase 1, shows broad-spectrum preclinical antitumor activity in multiple dosing regimens.
Mol Cancer Ther 11(3):700–709
190. Wheatley SP, McNeish IA (2005) Survivin: a protein with dual roles in mitosis and apoptosis.
Int Rev Cytol 247:35–88
191. Tao YF et al (2012) Survivin selective inhibitor YM155 induce apoptosis in SK-NEP-1 Wilms
tumor cells. BMC Cancer 12:619
192. Kelly RJ et al (2013) A phase I/II study of sepantronium bromide (YM155, survivin suppres-
sor) with paclitaxel and carboplatin in patients with advanced non-small-cell lung cancer.
Ann Oncol 24(10):2601–2606
193. Berezov A et al (2012) Disabling the mitotic spindle and tumor growth by targeting a cavity-
induced allosteric site of survivin. Oncogene 31(15):1938–1948
194. Yue Z et al (2008) Deconstructing Survivin: comprehensive genetic analysis of Survivin
function by conditional knockout in a vertebrate cell line. J Cell Biol 183(2):279–296
195. Shi J, Orth JD, Mitchison T (2008) Cell type variation in responses to antimitotic drugs that
target microtubules and kinesin-5. Cancer Res 68(9):3269–3276
196. Synold TW, Dussault I, Forman BM (2001) The orphan nuclear receptor SXR coordinately
regulates drug metabolism and efflux. Nat Med 7(5):584–590
197. Degenhardt Y et al (2010) Sensitivity of cancer cells to Plk1 inhibitor GSK461364A is associ-
ated with loss of p53 function and chromosome instability. Mol Cancer Ther 9(7):2079–2089
198. Luo J et al (2009) A genome-wide RNAi screen identifies multiple synthetic lethal interac-
tions with the Ras oncogene. Cell 137(5):835–848
199. Katayama H et al (2004) Phosphorylation by aurora kinase A induces Mdm2-mediated desta-
bilization and inhibition of p53. Nat Genet 36(1):55–62
200. Liu Q et al (2004) Aurora-A abrogation of p53 DNA binding and transactivation activity by
phosphorylation of serine 215. J Biol Chem 279(50):52175–52182
201. Yoon MJ et al (2012) Aurora B confers cancer cell resistance to TRAIL-induced apoptosis via
phosphorylation of survivin. Carcinogenesis 33(3):492–500
202. Quinn JE et al (2003) BRCA1 functions as a differential modulator of chemotherapy-induced
apoptosis. Cancer Res 63(19):6221–6228
203. Weaver BA, Cleveland DW (2005) Decoding the links between mitosis, cancer, and chemo-
therapy: the mitotic checkpoint, adaptation, and cell death. Cancer Cell 8(1):7–12
204. Gascoigne KE, Taylor SS (2008) Cancer cells display profound intra- and interline variation
following prolonged exposure to antimitotic drugs. Cancer Cell 14(2):111–122
205. Bekier ME, Fischbach R, Lee J, Taylor WR (2009) Length of mitotic arrest induced by
microtubule-stabilizing drugs determines cell death after mitotic exit. Mol Cancer Ther
8(6):1646–1654
206. Goldspiel BR (1997) Clinical overview of the taxanes. Pharmacotherapy 17(5 Pt
2):110S–125S
207. Huang HC, Shi J, Orth JD, Mitchison TJ (2009) Evidence that mitotic exit is a better cancer
therapeutic target than spindle assembly. Cancer Cell 16(4):347–358
208. Bhat KM, Setaluri V (2007) Microtubule-associated proteins as targets in cancer chemo-
therapy. Clin Cancer Res 13(10):2849–2854
209. Kavallaris M et al (2001) Multiple microtubule alterations are associated with Vinca alkaloid
resistance in human leukemia cells. Cancer Res 61(15):5803–5809
210. Alli E, Yang JM, Ford JM, Hait WN (2007) Reversal of stathmin-mediated resistance to pacli-
taxel and vinblastine in human breast carcinoma cells. Mol Pharmacol 71(5):1233–1240
211. Spankuch B et al (2006) Down-regulation of Polo-like kinase 1 elevates drug sensitivity of
breast cancer cells in vitro and in vivo. Cancer Res 66(11):5836–5846
212. Spankuch B, Kurunci-Csacsko E, Kaufmann M, Strebhardt K (2007) Rational combinations
of siRNAs targeting Plk1 with breast cancer drugs. Oncogene 26(39):5793–5807
213. Promkan M, Liu G, Patmasiriwat P, Chakrabarty S (2011) BRCA1 suppresses the expression
of survivin and promotes sensitivity to paclitaxel through the calcium sensing receptor
(CaSR) in human breast cancer cells. Cell Calcium 49(2):79–88
214. Chen L et al (2013) Survivin status affects prognosis and chemosensitivity in epithelial ovar-
ian cancer. Int J Gynecol Cancer 23(2):256–263
215. Chello PL, Sirotnak FM (1981) Increased schedule-dependent synergism of vindesine versus
vincristine in combination with methotrexate against L1210 leukemia. Cancer Treat Rep
65(11–12):1049–1053
216. Kano Y, Ohnuma T, Okano T, Holland JF (1988) Effects of vincristine in combination with
methotrexate and other antitumor agents in human acute lymphoblastic leukemia cells in
culture. Cancer Res 48(2):351–356
217. Cardoso F et al (2009) International guidelines for management of metastatic breast cancer:
combination vs sequential single-agent chemotherapy. J Natl Cancer Inst 101(17):1174–1181
218. Gobbi PG, Ferreri AJ, Ponzoni M, Levis A (2013) Hodgkin lymphoma. Crit Rev Oncol
Hematol 85(2):216–237
219. Von Hoff DD et al (2013) Increased survival in pancreatic cancer with nab-paclitaxel plus
gemcitabine. N Engl J Med 369(18):1691–1703
220. Martin-Lluesma S, Stucke VM, Nigg EA (2002) Role of Hec1 in spindle checkpoint signal-
ing and kinetochore recruitment of Mad1/Mad2. Science 297(5590):2267–2270
221. Wu G et al (2008) Small molecule targeting the Hec1/Nek2 mitotic pathway suppresses
tumor cell growth in culture and in animal. Cancer Res 68(20):8393–8399
222. Altieri DC (2003) Survivin, versatile modulation of cell division and apoptosis in cancer.
Oncogene 22(53):8581–8589
223. Quintyne NJ, Reing JE, Hoffelder DR, Gollin SM, Saunders WS (2005) Spindle multipolar-
ity is prevented by centrosomal clustering. Science 307(5706):127–129
224. Chang H et al (2012) The TPX2 gene is a promising diagnostic and therapeutic target for
cervical cancer. Oncol Rep 27(5):1353–1359
225. Hsu PK et al (2014) TPX2 expression is associated with cell proliferation and patient out-
come in esophageal squamous cell carcinoma. J Gastroenterol 49:1231–1240
226. Kanehira M et al (2007) Oncogenic role of MPHOSPH1, a cancer-testis antigen specific to
human bladder cancer. Cancer Res 67(7):3276–3285
227. Shimo A et al (2007) Elevated expression of protein regulator of cytokinesis 1, involved in the
growth of breast cancer cells. Cancer Sci 98(2):174–181
228. Ngan VK et al (2000) Novel actions of the antitumor drugs vinflunine and vinorelbine on
microtubules. Cancer Res 60(18):5045–5051
229. Salmela AL, Kallio MJ (2013) Mitosis as an anti-cancer drug target. Chromosoma
122(5):431–449
230. Nakahara T et al (2011) Broad spectrum and potent antitumor activities of YM155, a novel
small-molecule survivin suppressant, in a wide variety of human cancer cell lines and xeno-
graft models. Cancer Sci 102(3):614–621
Chapter 14
Inhibitors of Mitotic Kinesins for Cancer
Treatment: Consequences for Neurons
Olga I. Kahn and Peter W. Baas
14.1 Introduction
Cancer is among the greatest and most expensive medical challenges in the world,
and potentially escalating in prevalence as more carcinogens are introduced into the
environment. Not surprisingly, there are ongoing efforts to develop novel therapies,
as effective treatments would positively impact the landscape for society. Cancer
involves abnormal cells that rapidly proliferate, move, and invade tissues of the
body [1]. Cell proliferation, movement and invasion all rely on the cytoskeletal
elements known as microtubules, which have historically been a primary target
cancer therapy. Microtubules are dynamic polymers composed of tubulin subunits,
each of which is a dimer of alpha and beta tubulin. Quintessential and crucial for
most of the work done by microtubules are their dynamic properties, which are
governed by a mechanism known as dynamic instability [2]. A portion of the micro-
tubule mass becomes stabilized in living cells, but a portion remains highly dynamic,
undergoing rapid bouts of assembly and disassembly. When their dynamics are
frozen, microtubules are no longer able to do many of their assigned tasks, which is
why drugs that stabilize microtubules inhibit mitosis. Drugs that depolymerize
microtubules are also effective at inhibiting mitosis, as they pare away the mitotic
apparatus itself. The problem with any of these therapies is that virtually all cells of
the body rely on microtubules for a variety of tasks – such as serving as architectural
elements and acting as railways for organelle transport. Cells that are most depen-
dent on microtubules are likely to display the most notable ill effects when patients
are systemically treated with drugs that either stabilize or depolymerize microtu-
bules. Consistent with this, neuropathy is among the most common problems with
microtubule-altering chemotherapy [3, 4].
O.I. Kahn • P.W. Baas (*)

Department of Neurobiology and Anatomy, Drexel University College of Medicine,
2900 Queen Lane, Philadelphia, PA 19129, USA
e-mail: pbaas@drexelmed.edu

252 O.I. Kahn and P.W. Baas
Taxol, its derivatives and related drugs are currently the most commonly used
microtubule-based drugs for cancer chemotherapy. Taxol binds to beta tubulin in a
pocket on the luminal surface of the microtubule and suppresses the disassembly of
the polymer, thus promoting its stabilization [5, 6]. Taxol does not cross the blood–
brain barrier, but related drugs such as Epothilone D readily cross and actually accu-
mulate in the central nervous system [7, 8]. Painful, debilitating and sometimes
permanent peripheral neuropathies are common in patients treated with taxol and
there is concern about parallel effects on the brain with drugs that cross the blood–
brain barrier.
About 15 years ago, microtubule researchers began to develop a new microtubule-
based strategy to target cancer cells with the goal of prohibiting cell division with
no negative consequences on non-diving cells, such as neurons. Neurons are termi-
nally post-mitotic cells that no longer divide, but rely on microtubules for their
complex polar morphologies and for the efficient and orderly transport of proteins
and organelles within their axons and dendrites. Hyper-stabilization of microtubules
leads to a variety of changes in the organization, composition and behavior of the
microtubule arrays of the neuron [9]. Therefore, the idea was to develop drugs that
would specifically inhibit certain members of the kinesin family of molecular motor
proteins believed to be expressed only in cells that divide. These so-called mitotic
kinesins were shown to be important for organizing and driving the mitotic appara-
tus, and hence would presumably not be expressed in terminally post-mitotic neu-
rons. This was considered a breakthrough for the cancer research community and
the absence of mitotic kinesins from neurons was enthusiastically touted by the
pharmaceutical industry.
A decade and a half later, one might have expected innumerable cancer patients
to owe their lives to a rich kit of drugs inhibiting mitotic kinesins. One might also
have expected taxol to be long since abandoned in favor of such drugs. In fact, thou-
sands of patients continue to rely on taxol to impede the growth of tumors and
continue to suffer the effects of peripheral neuropathies. There are various reasons
why the anticipated success targeting mitotic kinesins has not materialized, at least
not yet. Here, we discuss a potential concern that should be considered far more
than it has been to date. Mitotic kinesins, as it turns out, are expressed not only dur-
ing mitosis in dividing cells, but also in terminally post-mitotic cells [10–16].
Neurons rely on these kinesins for many of their most important developmental
challenges. While the levels of these kinesins diminish in adult neurons, even low
levels of these motor proteins may have important work to do throughout the life of
a neuron (Fig. 14.1). In addition, there is recurring neurogenesis in the adult brain,
which recent research suggests is important for memory and cognition [17]. Drugs
that inhibit the so-called mitotic kinesins would, without a doubt, affect
the differentiation of these newly born neurons if the drugs were to cross the
blood–brain barrier.
Given that chemotherapy is given to patients for discrete periods of time, it may
well be that the central and/or peripheral nervous systems can tolerate prohibited
function of one or more of the mitotic kinesins in a typical therapeutic timeframe.
In fact, inhibiting these kinesins in cultured neurons, while dramatically affecting
14 Kinesins, Cancer and Neurons 253
Fig. 14.1 Effects of mitotic kinesin inhibition on mitosis and neurons. In the left panels,
mitotic kinesins normally generate forces important for the organization of the mitotic spindle in
the dividing cell and for the distinct microtubule arrays in axon and dendrites of the neuron. Arrows
represent forces generated by different mitotic kinesins In the right panels, inhibition of these
kinesins yields a failed mitotic spindle and the forces responsible for organizing the microtubule
arrays in the neuron are imbalanced, leading to axonal and dendritic architectural and functional
abnormalities
neuronal morphology, has no apparent negative consequences on the overall vitality

of the neuron [16]. Even so, the concern lies in the mystery of what might poten-
tially happen to patients taking such drugs. Ill effects may be subtle compared to
those caused by taxol and may even escape notice in clinical trials, but could never-
theless be insidious. For example, minor changes in dendritic arborization or spines
could affect cognition, memory and personality [18–23]. The purpose of this chap-
ter is not to be alarmist about the use of this approach, but to dissuade any tempta-
tion for scientists, clinicians and drug developers to be cavalier about ignoring the
potential impact of such drugs on the nervous system.
14.2 Neuronal Microtubules
Vertebrate neurons in their post-migratory configuration generally consist of a sin-

gle axon and multiple dendrites [24, 25]. These are elongated processes (often com-
plex in their branching patterns) that require microtubules as shape-sustaining
architectural struts. In addition, the microtubules of axons and dendrites serve as
direction-specific railways for organelles and other cargo transport. In a typical ver-
tebrate neuron, nearly all of the microtubules in the axon are oriented with their plus
ends directed away from the cell body [26], whereas the dendrites have a mixed
orientation of microtubules [27]. We have recently discussed in detail how these
patterns were discovered and the contributions of these microtubule polarity pat-
terns to the distinct morphological and compositional features that define the iden-
tity of the axon and the dendrites [28]. Because different types of organelles engage
either plus-end-directed or minus-end-directed molecular motor proteins, the polar-
ity pattern of the microtubule array is a major determinant of which organelles/car-
goes are transported into each type of process from the cell body, as well as the
efficiency and character of anterograde and retrograde organelle movements within
the axon and dendrites. This simple scenario can explain, for example, why Golgi
outposts appear in dendrites, but not the axon [29]. In addition, the polarity patterns
offer an attractive explanation for the differences in morphology and growth proper-
ties between axons and dendrites [28].
Contemporary research on neuronal microtubules has yielded a great deal of
knowledge on the behaviors of microtubules that underlie the growth and mainte-
nance of the axon, the development and plasticity of dendritic arbors, and the migra-
tion of developing neurons to their destinations [24]. Microtubules undergo a variety
of behaviors such as dynamic assembly and disassembly, bundling and splaying,
and severing [30]. They also integrate at many levels with other cytoskeletal
elements, most notably actin filaments [31]. Shorter microtubules undergo rapid
transport, and the same forces that transport short microtubules also impinge upon
longer stationary microtubules in functionally important ways [32]. These various
behaviors are regulated by signaling pathways that impact proteins such as molecu-
lar motor proteins, microtubule-severing enzymes, microtubule-depolymerizing
enzymes, as well as a variety of other proteins that influence the assembly, stabiliza-
tion and bundling of microtubules.
While a fraction of the microtubule array of the axon is quite stable, a fraction is
not, with dendritic microtubules having an even smaller fraction of stable microtu-
bule polymer than axons [33]. Younger neurons also have more labile microtubules
in general than mature neurons, especially in their axonal growth cones, in which
highly dynamic microtubules are important for navigation [34, 35]. The stable and
labile microtubule fractions in the neuron are not separate microtubules, but rather
each microtubule has a stable region toward its minus end with most of these micro-
tubules having a labile region toward its plus end [36, 37]. The stable region has
been likened to a microtubule nucleating structure that controls the distribution and
polarity orientation of the dynamic microtubule polymer [38]. Without that level of
control to restrict where new assembly arises, new microtubules would arise with
haphazard organization, thus corrupting the all-important microtubule polarity pat-
terns of axons and dendrites [39].
The structure and function of microtubules explain why microtubule-stabilizing
drugs such as taxol have such negative and often permanent effects on neurons.
Neurons treated with taxol can display abnormal accumulation and bundling of
microtubules, loss of microtubule domain structure, and corruption of normal
microtubule polarity patterns (Fig. 14.2). Such effects can cause traffic jams in
organelle transport as well as mis-localization of organelles and proteins that should
Fig. 14.2 Effects of taxol on microtubule organization in the neuron. In control neurons,
microtubules are nearly uniformly plus end distal in the axon. In the dendrite, microtubules have a
mixed orientation (left panel). Each microtubule consists of a stable domain toward the minus end
of the microtubule, with most microtubules also consisting of a dynamic (labile) domain toward
the plus end. Dendritic microtubules are less stable than axonal microtubules. In neurons treated
with taxol (right panel), the density of microtubules increases, the normal domain structure of
individual microtubules is lost because the microtubules are stabilized all along their lengths, and
flaws arise in the normal polarity patterns of the microtubules. Such abnormalities can lead to
degeneration of axons and dendrites
be enriched or exclusive to one type of process or the other. In fact, major

abnormalities in axonal microtubule polarity orientation have been documented in
cultured neurons treated with taxol and, of particular concern, the abnormalities
seem to persist even after removing the drug [40]. Even subtle alterations in the
microtubule polarity patterns of axons and dendrites could have profoundly nega-
tive consequences over time [41, 42].
Accumulation of certain tubulin post-translational modifications on microtu-
bules is another relevant aspect of microtubule regulation [43–47]. The best studied
of these are detyrosination and acetylation, both of which occur on the alpha tubulin
component of the tubulin heterodimer and both of which occur only after a tubulin
subunit is incorporated into microtubule polymer. Detyrosination is the removal of
the C-terminal tubulin residue from alpha tubulin, while acetylation is the addition
of an acetyl moiety to lysine 40 of alpha tubulin, which lies at the luminal surface
of the microtubule polymer [45, 48]. Some motor proteins, including kinesins, that
use microtubules to transport organelles, vesicles or even other shorter microtu-
bules, have been found to have a preference for certain microtubule post transla-
tional modifications over others. For example, the depolymerizing kinesins
(kinesin-13 family) interact better with tyrosinated microtubules [49], while

conventional kinesin (kinesin-1) interacts better with detyrosinated microtubules
[50–52]. Because these microtubule modifications only occur on the polymer (not
free subunits of tubulin), and because the modifications accumulate on the polymer
over time, the older the microtubule is, the more modified it becomes. The levels of
the modified subunits are therefore considered a good indication of the stability of
the microtubule, although they do not cause stability, at least not directly. Hyper-
stabilization of microtubules with taxol dramatically increases the levels of these
modifications on the microtubules, which can be consequential given the role of
these modifications in regulating the interaction of the microtubule with such a vast
array of proteins.
These various considerations illustrate why drugs that stabilize microtubules
have such profound impact on neurons, with any of these effects potentially contrib-
uting to the neuropathies experienced by patients using taxol and its derivatives as
chemotherapy. As for drugs that inhibit mitotic kinesins, the question becomes
whether the potential effects of such drugs on neurons are non-existent (as has been
claimed by early advocates for their use), minimal or potentially as bad or worse
than the effects of taxol. Due consideration seems warranted, especially in light of
the expectation of many drug companies that there will be no effects due to their
assumption that mitotic motor proteins are not expressed in neurons.
14.3 Mitotic Kinesins: Function in Neurons
In an ideal world, chemotherapy would be minimally invasive, able to target and

eliminate cancer cells, and avoid other cell types. When mitosis was discovered to
rely on specialized kinesins, these proteins became an appealing target for drug
development. Kinesins were originally discovered as motors that transport membra-
nous organelles, but these specialized mitotic kinesins are different in the sense that
they generate forces between microtubules and neighboring microtubules, between
microtubules and actin filaments, and can also integrate with other components of
the mitotic apparatus in various signaling pathways relevant to cell division [53]. In
this way, the so-called mitotic kinesins contribute to the formation of the bipolar
spindle, the separation of the half-spindles and cytokinesis. But are these mitotic
kinesins really mitosis-specific, as originally believed?
In the early 1990s, a great deal of attention, mainly from our laboratory, focused
on identifying the molecular motor proteins that transport and set up microtubule
arrays in neuronal axons and dendrites. We figured that at least one molecular motor
was needed to transport microtubules with plus ends leading from the cell body into
the axon and developing dendrites, and at least one other motor was needed to trans-
port microtubules with minus ends leading specifically into developing dendrites
(but not the axon). This was at a time prior to the genome projects and prior to the
kinesin tree being established, so we initially thought that entirely new motor pro-
teins were awaiting discovery. It was instructive for us to learn that the mitotic
motors actually have the kinds of properties that we had in mind for such neuronal
motors: they do not transport membranous organelles along microtubules, but rather
generate forces that move and configure the microtubules themselves [54].
Our first hint that mitotic motors could be repurposed for transporting microtu-
bules into cellular processes came from our collaborative studies with Dr. Ryoko
Kuriyama, in which we expressed two different mitotic motors separately into insect
ovarian Sf9 cells, which normally do not bear processes. One of these, a motor
termed CHO1/MKLP1 (also called Kif23, now known as kinesin-6), had the
interesting property of transporting microtubules with minus ends leading toward
the plus ends of other microtubules. Notably, expression of a large fragment of this
motor in Sf9 cells resulted in the outgrowth of microtubule-rich processes and the
concomitant depletion of microtubules from the cell body. CHO1/MKLP1 produced
dendrite-like processes (short and tapering) with mixed orientation of microtubules
[55–57]. We subsequently showed that a motor protein then called CHO2 (now
known as kinesin-14) when expressed as a fragment in Sf9 cells produced axon-like
processes with uniformly plus-end-distal microtubules. Kinesin-14 family members
are unusual among kinesins in that they move toward minus ends of microtubules,
whereas the other kinesins move toward plus ends of microtubules [58]. Thus, like
cytoplasmic dynein, kinesin-14 would transport microtubules against a stationary
substrate with plus ends leading. These experiments established precedent for the
properties of mitotic motor proteins to generate the characteristic microtubule polar-
ity patterns of axons and dendrites by transporting microtubules generated in the
cell body into the processes.
We originally considered the work on Sf9 cells as proof-of-principle, not that
mitotic motors were actually repurposed in developing neurons. However, when we
assayed developing neurons for three different mitotic motors, kinesin-6 (CHO1/
MKLP1), kinesin-5 (also called Eg5, KSP, Kif11), and kinesin-12 (also called
Kif15), we found each one to be strongly expressed. In mitosis, kinesin-5 generates
forces on anti-parallel microtubules during prophase to drive the duplicated centro-
somes apart, and then later in mitosis contributes to the separation of the half-
spindles [59–61]. Kinesin-12 does not appear to be crucial within itself, but it has
overlapping function with kinesin-5 and can compensate for its loss [10, 62–64].
Kinesin-6 generates forces on antiparallel microtubules in the spindle midzone, but
also signals through RhoA to the actin cytoskeleton to cause cortical actin filaments
to concentrate in the region corresponding to the midzone of the spindle to form the
cleavage furrow that ultimately pinches off the two daughter cells [65–67]. In cul-
tured neurons, when we knocked down each of these motors using siRNA, we
observed dramatic morphological phenotypes and pronounced effects on microtu-
bule organization and behavior [16].
Studies on these motors in neurons have produced several papers from our labo-
ratory, often as collaborative projects with scientists prominent in the mitosis field.
We began our work on kinesin-6, but our work on kinesin-5 has been the most thor-
ough, in part because of availability of drugs that inhibit its function. Such drugs,
prototypes for the more sophisticated medications subsequently used for clinical
trials on cancer patients, were used to fortify our knockdown approach and enabled
Fig. 14.3 Model for co-regulation of microtubule polarity in axons and dendrites by mitotic
kinesins. Figure adapted from previous publication [16]. During axonal differentiation, forces
generated by cytoplasmic dynein drive plus-end-distal microtubules into the axon and nascent
dendrites (not shown). (a) Forces generated by kinesin-5 act as a brake on the transport of micro-
tubules into the axon during axonogenesis. Kinesin-5 also acts as a brake on the transport of plus
and minus-end-distal microtubules into dendrites (b) Forces generated by kinesin-12 behave
a variety of experiments, such as ones in which we observed the effects on growth

cones of a gradient of such a drug [68]. The work with drugs or siRNA produced
consistent results, suggesting that such an approach will be useful in exploring the
properties of future drugs that become available to inhibit mitotic motors.
Our studies indicate that knocking down any of the three motors causes the axon
to grow more rapidly and to increase its content of short mobile microtubules mov-
ing in both directions [16]. Knockdown of either kinesin-5 or kinesin-12 causes
growth cones to fail to turn in response to environmental cues [13, 68], with no such
effect observed with knockdown of kinesin-6 [16]. Knockdown of kinesin-5 results
in a more branched axon [12], while knockdown of kinesin-12 results in a less
branched axon [14], with this difference presumably owing to the fact that kinesin-
12 interacts with actin filaments while kinesin-5 does not [10]. Kinesin-6 knock-
down has no effect on axonal branching [16]. Both kinesin-5 and kinesin-6 are
present in the axon and the growth cone, but kinesin-6 is undetectable in the axon,
suggesting that its effects on axonal growth are probably due to an effect at the level
of the cell body that enables more short microtubules to transit into the axon [12, 14,
16]. All three motors are present in the cell body and concentrate in dendrites as
they begin to develop, with the knockdown of kinesin-6 or kinesin-12 proving
detrimental to the appearance of minus-end-distal microtubules in dendrites [16].
Conversely, knockdown of kinesin-5 increases the proportion of minus-end-distal
microtubules in dendrites [69]. Knockdown of any of the three motors causes
dendrites to become thinner, with knockdown of kinesin-6 or kinesin-12 causing
dendrites to become more axonal in character (Fig. 14.3).
We have also observed notable effects of knocking down these motor proteins on
neuronal migration, which is the movement of newly born neurons from their sites of
origin near the ventricles through the developing brain, to form its laminar structure.
Depletion of either kinesin-5 or kinesin-12 increases the rate of neuronal migration,
with either of these motors potentially playing a role in the cessation of migration
once the neuron has reached its appropriate location [14, 15]. Knockdown of kinesin-6
prevents newborn neurons from transitioning from multi-polar to bipolar morphol-
ogy, and prevents the neuron from designating any one process as the leading process
[70]. Such neurons are either stationary or are futile in their movements.
Studies on messenger RNA indicate that these three mitotic motors diminish in
expression in mature neurons, in some cases to almost undetectable levels [71–73].
At first glance, this seems very good news for the potential use of cancer drugs
against mitotic motors, as the developmental events in which the motors are utilized
Fig. 14.3 (continued) similarly to kinesin-6 with regard to introducing minus-end-distal microtu-
bules into the dendrite, but kinesin-12 is also present in the axon and growth cone, pushing plus-
end-distal microtubules back toward the cell body. (c) Forces generated by kinesin-6 in the cell
body oppose the forces generated by cytoplasmic dynein, restricting the transport of plus-end-
distal microtubules into the axon. As the neuron matures, kinesin-6 fuels the transport of short
microtubules with their minus-end distal into all of the processes except the one designated to
remain the axon, thus causing the other processes to differentiate into dendrites. As a result, kine-
sin-12 behaves like kinesin-6 with regard to dendrites but produces effects more like kinesin-5 with
regard to the axon. For more details, see Ref. [16]
will have been long since completed. However, the messenger RNA results do not
necessarily mean that protein levels are down by corresponding levels, as our
Western blot analyses indicate [14, 74]. Moreover, we have some preliminary indi-
cations that expression of certain of these motors may increase when neurons are
challenged, for example by injury leading to regeneration (in the case of kinesin-12,
according to our unpublished data, Liu, M. and Baas, P.W), or perhaps even in
response to normal plastic events such as dendritic remodeling associated with
learning. For example, adult neurons may synthesize “gulps” of mitotic motors that
could be functionally important, especially during bouts of plasticity. Perhaps more
importantly, one of the most provocative areas of neuroscience in recent years has
been the discovery that in the adult brain, there is ongoing neurogenesis and the new
neurons that are produced are functionally important for cognition and memory
[17, 75–77]. When neurogenesis is impeded in the adult brain, rodents have been
shown to undergo notable decline in their ability to perform in cognitive and mem-
ory tests [78, 79]. Thus, even if adult neurons rely minimally or not at all on mitotic
motors, the newly differentiated neurons in the adult brain undoubtedly do, and
during the period of chemotherapy, if the drugs cross the blood–brain barrier (or if
the drugs are actually used to treat brain cancer), there would presumably be strong
effects on these young neurons.
What about other kinesins essential for mitosis? The kinesin-4 proteins (also
called Kif4), the so-called chromokinesins, are important in mitosis for condensing
and moving chromosomes. Kif4A and kif4B are often redundant in function and
sometimes not distinguished in studies, but have essential roles in regulating ana-
phase spindle dynamics and completion of cytokinesis [80–82]. Depending on the
type of cancer, kinesin-4 has been shown to be either overexpressed or downregu-
lated, and has proven to be a valuable prognostic tool in several independent cancer
type cases. Loss of kinesin-4 in some studies had been found to induce aneuploidy
(abnormal number of chromosomes), which potentially contributes to tumor
formation [83], while overexpression of the protein in cervical and lung cancer has
potential to become a biomarker and a target for drugs [84, 85]. In neurons, avail-
able evidence suggests that these kinesins mediate anterograde translocation and
positioning of the cell adhesion molecule L1, ribosomal constituents to axons [86],
and are involved in neuronal survival [87]. Kinesin-13 (Kif2A) is a microtubule-
depolymerizing motor that is required for bipolar spindle formation [88–91].
Overexpression of this protein has been seen to contribute to metastasis of certain
cancers and to play a role in resistance to microtubule-stabilizing drugs [92, 93].
Kif2A has been shown to be present in the neuron, and when inhibited causes
impairment of axonal branch formation [94]. Mutations in the Kif2A gene have
recently been found to contribute to malformations of cortical development and
microcephaly emphasizing the importance of centrosomal and microtubule-related
proteins in cortical development [95]. Alterations in levels of these kinesins
and other mitotic kinesins are a hallmark of some types of cancers, while certain
experimental manipulations of these mitotic motors have been seen to actually
cause cancer [96, 97] or have an effect in resistance to therapy [98]. Potential impact
on neurons of inhibiting these kinesins further complicates the landscape.
14.4 Mitotic Kinesins as Targets for Cancer

Drugs: Where Do We Stand?
Kinesin-5 was the first mitotic kinesin chosen for cancer therapy, because the
knockdown phenotype of a mono-astral spindle was so dramatic and so inconsistent
with cell division. Dr. Mitchisons’s laboratory developed monastrol as the prototype
drug for kinesin-5 inhibition and since then more sophisticated drugs have been
developed by companies and put through clinical trials. Some proof-of-principle
that anti-kinesin-5 drugs could be affective against cancer comes from gene electro-
transfer therapy studies wherein kinesin-5 siRNA drastically reduces outgrowth of
subcutaneous melanoma and ovarian cancer lesions in mice [99] and from lipid
nanoparticle-driven delivery of kinesin-5 siRNA, which has been shown to cause
regression of liver metastases in endometrial cancer in humans [100]. Anti-kinesin-5
drugs such as Ispinesib and ARRY-520 appear to be moderately well tolerated by
patients in clinical trials, with common ill effects of mild to moderate neutropenia,
fatigue, anemia, nausea, leukopenia, thrombocytopenia, and diarrhea [101–103].
Surprisingly, though, anti-kinesin-5 drugs have elicited mixed results in terms of
curtailing various types of cancer, and collectively these results have been rather
disappointing (see also, clinicaltrials.gov).
Newer efforts are shifting toward combinatorial therapy with other anti-cancer
drugs. One possibility is that in cancer cells, excess expression of kinesin-12 can
compensate for inhibition of kinesin-5, so perhaps developing anti-kinesin-12 drugs
to be used in combination with anti-kinesin-5 drugs makes sense. Drug companies
are also putting attention now on other mitotic kinesins with the hope that the impact
on cancer will be more consistent and profound. In terms of nervous system defects,
the good news is that we have been unable to find any reports in available literature
or clinical trial summaries of the kinds of disorders or defects that concerned us
most, given our work on developing neurons. However, it is not clear just how thor-
ough the trials were in terms of seeking potential effects on neurons; and ill effects
that might have escaped notice could be exacerbated as the treatment regimes are
fortified to have greater effects on cancer.
We will conclude by reiterating a need not for alarm, but for due caution in the
use of drugs that inhibit mitotic kinesins. Nature has developed a system of re-
purposing proteins that while beautiful in its economy presents potential obstacles
for cancer therapy that merit consideration. The neuroscience community will hope-
fully delve deeper into whether mitotic motors serve key roles in adult neurons, and
also whether putting adult neurogenesis on hold for periods of chemotherapy is
acceptable to the health of the patient.
14.5 Funding
The work in the Baas laboratory relevant to the topic of this chapter has been funded
over the years by grants from the National Institutes of Health, the National Science
Foundation, the Department of Defense, the Craig H. Neilsen Foundation, and the
Christopher and Dana Reeve Foundation. Olga I. Kahn is the recipient a pre-doctoral
NRSA from the NIH/NIMH.
Acknowledgments We are thankful to past and present members of our laboratory for their con-
tributions to the work presented here. The authors declare no financial or conflicting interests.
References
1. Hanahan D, Weinberg RA (2011) Hallmarks of cancer: the next generation. Cell 144:646–
674. doi:10.1016/j.cell.2011.02.013, S0092-8674(11)00127-9 [pii]
2. Kirschner M, Mitchison T (1986) Beyond self-assembly: from microtubules to morphogen-
esis. Cell 45:329–342. doi:10.1016/0092-8674(86)90318-1 [pii]
3. Ferrier J, Pereira V, Busserolles J, Authier N, Balayssac D (2013) Emerging trends in under-
standing chemotherapy-induced peripheral neuropathy. Curr Pain Headache Rep 17:364.
doi:10.1007/s11916-013-0364-5
4. Jaggi AS, Singh N (2012) Mechanisms in cancer-chemotherapeutic drugs-induced peripheral
neuropathy. Toxicology 291:1–9. doi:10.1016/j.tox.2011.10.019, S0300-483X(11)00458-6
[pii]
5. Amos LA, Lowe J (1999) How taxol stabilises microtubule structure. Chem Biol 6:R65–R69.
doi:S1074-5521(99)89002-4 [pii]
6. Prota AE et al (2013) Molecular mechanism of action of microtubule-stabilizing anticancer
agents. Science 339:587–590. doi:10.1126/science.1230582, science.1230582 [pii]
7. Brunden KR et al (2011) The characterization of microtubule-stabilizing drugs as possible
therapeutic agents for Alzheimer’s disease and related tauopathies. Pharmacol Res 63:341–
351. doi:10.1016/j.phrs.2010.12.002, S1043-6618(10)00227-6 [pii]
8. Zhang B et al (2012) The microtubule-stabilizing agent, epothilone D, reduces axonal dys-
function, neurotoxicity, cognitive deficits, and Alzheimer-like pathology in an interventional
study with aged tau transgenic mice. J Neurosci 32:3601–3611. doi:10.1523/
JNEUROSCI.4922-11.2012, 32/11/3601 [pii]
9. Baas PW, Ahmad FJ (2013) Beyond taxol: microtubule-based treatment of disease and injury
of the nervous system. Brain. doi:10.1093/brain/awt153, awt153 [pii]
10. Buster DW et al (2003) Expression of the mitotic kinesin Kif15 in postmitotic neurons: impli-
cations for neuronal migration and development. J Neurocytol 32:79–96. doi:5146300 [pii]
11. Haque SA, Hasaka TP, Brooks AD, Lobanov PV, Baas PW (2004) Monastrol, a prototype
anti-cancer drug that inhibits a mitotic kinesin, induces rapid bursts of axonal outgrowth from
cultured postmitotic neurons. Cell Motil Cytoskeleton 58:10–16. doi:10.1002/cm.10176
12. Myers KA, Baas PW (2007) Kinesin-5 regulates the growth of the axon by acting as a brake
on its microtubule array. J Cell Biol 178:1081–1091. doi:10.1083/jcb.200702074,
jcb.200702074 [pii]
13. Nadar VC, Ketschek A, Myers KA, Gallo G, Baas PW (2008) Kinesin-5 is essential for
growth-cone turning. Curr Biol 18:1972–1977. doi:10.1016/j.cub.2008.11.021,
S0960-9822(08)01501-7 [pii]
14. Liu M et al (2010) Kinesin-12, a mitotic microtubule-associated motor protein, impacts axo-
nal growth, navigation, and branching. J Neurosci 30:14896–14906. doi:10.1523/
JNEUROSCI.3739-10.2010, 30/44/14896 [pii]
15. Falnikar A, Tole S, Baas PW (2011) Kinesin-5, a mitotic microtubule-associated motor pro-
tein, modulates neuronal migration. Mol Biol Cell 22:1561–1574. doi:10.1091/mbc.E10-
11-0905, mbc.E10-11-0905 [pii]
16. Lin S, Liu M, Mozgova OI, Yu W, Baas PW (2012) Mitotic motors coregulate microtubule
patterns in axons and dendrites. J Neurosci 32:14033–14049. doi:10.1523/JNEUROSCI.
3070-12.2012, 32/40/14033 [pii]
17. Couillard-Despres S, Iglseder B, Aigner L (2011) Neurogenesis, cellular plasticity and cogni-
tion: the impact of stem cells in the adult and aging brain–a mini-review. Gerontology
57:559–564. doi:10.1159/000323481, 000323481 [pii]
18. Verpelli C, Montani C, Vicidomini C, Heise C, Sala C (2013) Mutations of the synapse genes
and intellectual disability syndromes. Eur J Pharmacol. doi:10.1016/j.ejphar.2013.07.023,
S0014-2999(13)00545-1 [pii]
19. Holtmaat A, Randall J, Cane M (2013) Optical imaging of structural and functional synaptic
plasticity in vivo. Eur J Pharmacol. doi:10.1016/j.ejphar.2013.07.020, S0014-2999(13)00542-6
[pii]
20. Shirao T, Gonzalez-Billault C (2013) Actin filaments and microtubules in dendritic spines. J
Neurochem 126:155–164. doi:10.1111/jnc.12313
21. Chang PK, Boridy S, McKinney RA, Maysinger D (2013) Letrozole potentiates mitochon-
drial and dendritic spine impairments induced by beta amyloid. J Aging Res 2013:538979.
doi:10.1155/2013/538979
22. Meng C, He Z, Xing D (2013) Low-level laser therapy rescues dendrite atrophy via upregu-
lating BDNF expression: implications for Alzheimer’s disease. J Neurosci 33:13505–13517.
doi:10.1523/JNEUROSCI. 0918-13.2013, 33/33/13505 [pii]
23. Murmu RP, Li W, Holtmaat A, Li JY (2013) Dendritic spine instability leads to progressive
neocortical spine loss in a mouse model of Huntington’s disease. J Neurosci 33:12997–
13009. doi:10.1523/JNEUROSCI. 5284-12.2013, 33/32/12997 [pii]
24. Conde C, Caceres A (2009) Microtubule assembly, organization and dynamics in axons and
dendrites. Nat Rev Neurosci 10:319–332. doi:10.1038/nrn2631, nrn2631 [pii]
25. Bradke F, Dotti CG (2000) Establishment of neuronal polarity: lessons from cultured hippo-
campal neurons. Curr Opin Neurobiol 10:574–581. doi:S0959-4388(00)00124-0 [pii]
26. Heidemann SR, Landers JM, Hamborg MA (1981) Polarity orientation of axonal microtubules.
J Cell Biol 91:661–665
27. Baas PW, Deitch JS, Black MM, Banker GA (1988) Polarity orientation of microtubules in
hippocampal neurons: uniformity in the axon and nonuniformity in the dendrite. Proc Natl
Acad Sci U S A 85:8335–8339
28. Baas PW, Lin S (2011) Hooks and comets: the story of microtubule polarity orientation in the
neuron. Dev Neurobiol 71:403–418. doi:10.1002/dneu.20818
29. Horton AC, Ehlers MD (2003) Dual modes of endoplasmic reticulum-to-Golgi transport in
dendrites revealed by live-cell imaging. J Neurosci 23:6188–6199. doi:23/15/6188 [pii]
30. Baas PW, Buster DW (2004) Slow axonal transport and the genesis of neuronal morphology.
J Neurobiol 58:3–17. doi:10.1002/neu.10281
31. Myers KA, Baas PW (2011) Microtubule-actin interactions during neuronal development.
Adv Neurobiol 5:73–96
32. Baas PW, Vidya Nadar C, Myers KA (2006) Axonal transport of microtubules: the long and
short of it. Traffic 7:490–498. doi:10.1111/j.1600-0854.2006.00392.x, TRA392 [pii]
33. Baas PW, Slaughter T, Brown A, Black MM (1991) Microtubule dynamics in axons and
dendrites. J Neurosci Res 30:134–153. doi:10.1002/jnr.490300115
34. Challacombe JF, Snow DM, Letourneau PC (1997) Dynamic microtubule ends are required
for growth cone turning to avoid an inhibitory guidance cue. J Neurosci 17:3085–3095
35. Geraldo S, Gordon-Weeks PR (2009) Cytoskeletal dynamics in growth-cone steering. J Cell

Sci 122:3595–3604. doi:10.1242/jcs.042309, 122/20/3595 [pii]
36. Baas PW, Black MM (1990) Individual microtubules in the axon consist of domains that dif-
fer in both composition and stability. J Cell Biol 111:495–509
37. Brown A, Li Y, Slaughter T, Black MM (1993) Composite microtubules of the axon: quantita-
tive analysis of tyrosinated and acetylated tubulin along individual axonal microtubules. J
Cell Sci 104(Pt 2):339–352
38. Baas PW, Ahmad FJ (1992) The plus ends of stable microtubules are the exclusive nucleating
structures for microtubules in the axon. J Cell Biol 116:1231–1241
39. Baas PW, Black MM, Banker GA (1989) Changes in microtubule polarity orientation during
the development of hippocampal neurons in culture. J Cell Biol 109:3085–3094
40. Shemesh OA, Spira ME (2010) Paclitaxel induces axonal microtubules polar reconfiguration
and impaired organelle transport: implications for the pathogenesis of paclitaxel-induced
polyneuropathy. Acta Neuropathol 119:235–248. doi:10.1007/s00401-009-0586-0
41. Kuznetsov AV (2010) Effect of the degree of polar mismatching on traffic jam formation in
fast axonal transport. Comput Methods Biomech Biomed Engin 13:711–722.
doi:10.1080/10255840903505154, 920291996 [pii]
42. Baas PW, Mozgova OI (2012) A novel role for retrograde transport of microtubules in the
axon. Cytoskeleton (Hoboken) 69:416–425. doi:10.1002/cm.21013
43. Janke C, Kneussel M (2010) Tubulin post-translational modifications: encoding functions on
the neuronal microtubule cytoskeleton. Trends Neurosci 33:362–372. doi:10.1016/j.
tins.2010.05.001, S0166-2236(10)00065-2 [pii]
44. Janke C, Bulinski JC (2011) Post-translational regulation of the microtubule cytoskeleton: mech-
anisms and functions. Nat Rev Mol Cell Biol 12:773–786. doi:10.1038/nrm3227, nrm3227 [pii]
45. Wloga D, Gaertig J (2010) Post-translational modifications of microtubules. J Cell Sci
123:3447–3455. doi:10.1242/jcs.063727, 123/20/3447 [pii]
46. Wloga D et al (2010) Hyperglutamylation of tubulin can either stabilize or destabilize microtu-
bules in the same cell. Eukaryot Cell 9:184–193. doi:10.1128/EC.00176-09, EC.00176-09 [pii]
47. Garnham CP, Roll-Mecak A (2012) The chemical complexity of cellular microtubules: tubu-
lin post-translational modification enzymes and their roles in tuning microtubule functions.
Cytoskeleton (Hoboken) 69:442–463. doi:10.1002/cm.21027
48. Westermann S, Weber K (2003) Post-translational modifications regulate microtubule func-
tion. Nat Rev Mol Cell Biol 4:938–947. doi:10.1038/nrm1260, nrm1260 [pii]
49. Peris L et al (2009) Motor-dependent microtubule disassembly driven by tubulin tyrosina-
tion. J Cell Biol 185:1159–1166. doi:10.1083/jcb.200902142, jcb.200902142 [pii]
50. Dunn S et al (2008) Differential trafficking of Kif5c on tyrosinated and detyrosinated micro-
tubules in live cells. J Cell Sci 121:1085–1095. doi:10.1242/jcs.026492, jcs.026492 [pii]
51. Konishi Y, Setou M (2009) Tubulin tyrosination navigates the kinesin-1 motor domain to
axons. Nat Neurosci 12:559–567. doi:10.1038/nn.2314, nn.2314 [pii]
52. Hammond JW et al (2010) Posttranslational modifications of tubulin and the polarized trans-
port of kinesin-1 in neurons. Mol Biol Cell 21:572–583. doi:10.1091/mbc.E09-01-0044,
E09-01-0044 [pii]
53. Verhey KJ, Hammond JW (2009) Traffic control: regulation of kinesin motors. Nat Rev Mol
Cell Biol 10:765–777. doi:10.1038/nrm2782, nrm2782 [pii]
54. Sharp DJ, Rogers GC, Scholey JM (2000) Microtubule motors in mitosis. Nature 407:41–47.
doi:10.1038/35024000
55. Sharp DJ, Kuriyama R, Baas PW (1996) Expression of a kinesin-related motor protein
induces Sf9 cells to form dendrite-like processes with nonuniform microtubule polarity ori-
entation. J Neurosci 16:4370–4375
56. Sharp DJ, Kuriyama R, Essner R, Baas PW (1997) Expression of a minus-end-directed motor
protein induces Sf9 cells to form axon-like processes with uniform microtubule polarity ori-
entation. J Cell Sci 110(Pt 19):2373–2380
57. Sharp DJ et al (1997) Identification of a microtubule-associated motor protein essential for
dendritic differentiation. J Cell Biol 138:833–843
58. Rank KC et al (2012) Kar3Vik1, a member of the kinesin-14 superfamily, shows a novel
kinesin microtubule binding pattern. J Cell Biol 197:957–970. doi:10.1083/jcb.201201132,
jcb.201201132 [pii]
59. Ferenz NP, Gable A, Wadsworth P (2010) Mitotic functions of kinesin-5. Semin Cell Dev
Biol 21:255–259. doi:10.1016/j.semcdb.2010.01.019, S1084-9521(10)00020-0 [pii]
60. Saunders AM, Powers J, Strome S, Saxton WM (2007) Kinesin-5 acts as a brake in ana-
phase spindle elongation. Curr Biol 17:R453–R454. doi:10.1016/j.cub.2007.05.001,
S0960-9822(07)01333-4 [pii]
61. Kashina AS et al (1996) A bipolar kinesin. Nature 379:270–272. doi:10.1038/379270a0
62. Wittmann T, Boleti H, Antony C, Karsenti E, Vernos I (1998) Localization of the kinesin-like
protein Xklp2 to spindle poles requires a leucine zipper, a microtubule-associated protein,
and dynein. J Cell Biol 143:673–685
63. Rogers GC et al (2000) A kinesin-related protein, KRP(180), positions prometaphase spindle
poles during early sea urchin embryonic cell division. J Cell Biol 150:499–512
microtubule substrate. Curr Biol 23:1280–1290. doi:10.1016/j.cub.2013.05.043,
S0960-9822(13)00638-6 [pii]
65. Vale RD, Spudich JA, Griffis ER (2009) Dynamics of myosin, microtubules, and Kinesin-6
at the cortex during cytokinesis in Drosophila S2 cells. J Cell Biol 186:727–738. doi:10.1083/
jcb.200902083, jcb.200902083 [pii]
66. White EA, Raghuraman H, Perozo E, Glotzer M (2013) Binding of the CYK-4 subunit of the
centralspindlin complex induces a large scale conformational change in the kinesin subunit.
J Biol Chem 288:19785–19795. doi:10.1074/jbc.M113.463695, M113.463695 [pii]
67. Douglas ME, Davies T, Joseph N, Mishima M (2010) Aurora B and 14-3-3 coordinately regu-
late clustering of centralspindlin during cytokinesis. Curr Biol 20:927–933. doi:10.1016/j.
cub.2010.03.055, S0960-9822(10)00381-7 [pii]
68. Nadar VC, Lin S, Baas PW (2012) Microtubule redistribution in growth cones elicited by
focal inactivation of kinesin-5. J Neurosci 32:5783–5794. doi:10.1523/
JNEUROSCI.0144-12.2012, 32/17/5783 [pii]
69. Kahn OI, Sharma V, González-Billault C, Baas PW (2015) Effects of kinesin-5 inhibition on
dendritic architecture and microtubule organization. Mol Biol Cell 26(1):66–77. doi:1091/
mbc.E14-08-1313
70. Falnikar A, Tole S, Liu M, Liu JS, Baas PW (2013) Polarity in migrating neurons is related to
a mechanism analogous to cytokinesis. Curr Biol 23:1215–1220. doi:10.1091/mbc.E14-08-
1313, S0960-9822(13)00622-2 [pii]
71. Ferhat L, Kuriyama R, Lyons GE, Micales B, Baas PW (1998) Expression of the mitotic
motor protein CHO1/MKLP1 in postmitotic neurons. Eur J Neurosci 10:1383–1393
72. Ferhat L et al (1998) Expression of the mitotic motor protein Eg5 in postmitotic neurons:
implications for neuronal development. J Neurosci 18:7822–7835
73. Silverman MA et al (2010) Expression of kinesin superfamily genes in cultured hippocampal
neurons. Cytoskeleton (Hoboken) 67:784–795. doi:10.1002/cm.20487
74. Lin S et al (2011) Inhibition of kinesin-5, a microtubule-based motor protein, as a strategy for
enhancingregenerationofadultaxons.Traffic12:269–286.doi:10.1111/j.1600-0854.2010.01152.x
75. Gould E, Beylin A, Tanapat P, Reeves A, Shors TJ (1999) Learning enhances adult neurogen-
esis in the hippocampal formation. Nat Neurosci 2:260–265. doi:10.1038/6365
76. Leuner B, Gould E, Shors TJ (2006) Is there a link between adult neurogenesis and learning?
Hippocampus 16:216–224. doi:10.1002/hipo.20153
77. Monje M, Dietrich J (2012) Cognitive side effects of cancer therapy demonstrate a functional
role for adult neurogenesis. Behav Brain Res 227:376–379. doi:10.1016/j.bbr.2011.05.012,
S0166-4328(11)00401-3 [pii]
78. Pan YW, Chan GC, Kuo CT, Storm DR, Xia Z (2012) Inhibition of adult neurogenesis by
inducible and targeted deletion of ERK5 mitogen-activated protein kinase specifically in
adult neurogenic regions impairs contextual fear extinction and remote fear memory. J
Neurosci 32:6444–6455. doi:10.1523/JNEUROSCI. 6076-11.2012, 32/19/6444 [pii]
79. Mirescu C, Peters JD, Noiman L, Gould E (2006) Sleep deprivation inhibits adult neurogen-
esis in the hippocampus by elevating glucocorticoids. Proc Natl Acad Sci U S A 103:19170–
19175. doi:10.1073/pnas.0608644103, 0608644103 [pii]
80. Hu CK, Coughlin M, Field CM, Mitchison TJ (2011) KIF4 regulates midzone length dur-
ing cytokinesis. Curr Biol 21:815–824. doi:10.1016/j.cub.2011.04.019, S0960-
9822(11)00435-0 [pii]
sins and dyneins, in mitosis/cytokinesis using RNA interference. Mol Biol Cell 16:3187–
3199. doi:10.1091/mbc.E05-02-0167, E05-02-0167 [pii]
82. Mazumdar M, Misteli T (2005) Chromokinesins: multitalented players in mitosis. Trends
Cell Biol 15:349–355. doi:10.1016/j.tcb.2005.05.006, S0962-8924(05)00131-5 [pii]
16:1559–1564. doi:10.1016/j.cub.2006.06.029, S0960-9822(06)01755-6 [pii]
84. Taniwaki M et al (2007) Activation of KIF4A as a prognostic biomarker and therapeutic tar-
get for lung cancer. Clin Cancer Res 13(6624–6631):2007. doi:10.1158/1078-0432.CCR-
07-1328, 13/22/6624 [pii]
85. Narayan G et al (2007) Gene dosage alterations revealed by cDNA microarray analysis in
cervical cancer: identification of candidate amplified and overexpressed genes. Genes
Chromosomes Cancer 46:373–384. doi:10.1002/gcc.20418
86. Bisbal M et al (2009) KIF4 mediates anterograde translocation and positioning of ribosomal
constituents to axons. J Biol Chem 284:9489–9497. doi:10.1074/jbc.M808586200,
M808586200 [pii]
87. Kaplan DR, Miller FD (2006) When a motor goes bad: a kinesin regulates neuronal survival.
Cell 125:224–226. doi:10.1016/j.cell.2006.03.031, S0092-8674(06)00432-6 [pii]
88. Moores CA, Milligan RA (2006) Lucky 13-microtubule depolymerisation by kinesin-13
motors. J Cell Sci 119:3905–3913. doi:10.1242/jcs.03224, 119/19/3905 [pii]
89. Dawson SC et al (2007) Kinesin-13 regulates flagellar, interphase, and mitotic microtubule
dynamics in Giardia intestinalis. Eukaryot Cell 6:2354–2364. doi:10.1128/EC.00128-07,
EC.00128-07 [pii]
90. Ems-McClung SC, Walczak CE (2010) Kinesin-13 s in mitosis: key players in the spatial and
temporal organization of spindle microtubules. Semin Cell Dev Biol 21:276–282.
doi:10.1016/j.semcdb.2010.01.016, S1084-9521(10)00017-0 [pii]
91. Hood EA, Kettenbach AN, Gerber SA, Compton DA (2012) Plk1 regulates the kinesin-13
protein Kif2b to promote faithful chromosome segregation. Mol Biol Cell 23:2264–2274.
doi:10.1091/mbc.E11-12-1013, mbc.E11-12-1013 [pii]
92. Wang CQ et al (2010) Overexpression of Kif2a promotes the progression and metastasis of
squamous cell carcinoma of the oral tongue. Oral Oncol 46:65–69. doi:10.1016/j.oraloncol-
ogy.2009.11.003, S1368-8375(09)00967-1 [pii]
93. Stevens KN et al (2011) Evaluation of associations between common variation in mitotic
regulatory pathways and risk of overall and high grade breast cancer. Breast Cancer Res Treat
129:617–622. doi:10.1007/s10549-011-1587-y
94. Homma N et al (2003) Kinesin superfamily protein 2A (KIF2A) functions in suppression of
collateral branch extension. Cell 114:229–239. doi:S0092867403005221 [pii]
95. Poirier K et al (2013) Mutations in TUBG1, DYNC1H1, KIF5C and KIF2A cause malforma-
tions of cortical development and microcephaly. Nat Genet 45:962. doi:10.1038/ng0813-
962b, ng0813-962b [pii]
96. Liu M et al (2010) Ectopic expression of the microtubule-dependent motor protein Eg5 pro-
motes pancreatic tumourigenesis. J Pathol 221:221–228. doi:10.1002/path.2706
97. Castillo A, Morse HC 3rd, Godfrey VL, Naeem R, Justice MJ (2007) Overexpression of Eg5
causes genomic instability and tumor formation in mice. Cancer Res 67:10138–10147.
doi:10.1158/0008-5472.CAN-07-0326, 67/21/10138 [pii]
98. Ganguly A, Yang H, Cabral F (2011) Overexpression of mitotic centromere-associated

Kinesin stimulates microtubule detachment and confers resistance to paclitaxel. Mol Cancer
Ther 10:929–937. doi:10.1158/1535-7163.MCT-10-1109, 1535-7163.MCT-10-1109 [pii]
99. Marra E, Palombo F, Ciliberto G, Aurisicchio L (2013) Kinesin spindle protein SiRNA slows
tumor progression. J Cell Physiol 228:58–64. doi:10.1002/jcp.24103
100. Tabernero J et al (2013) First-in-humans trial of an RNA interference therapeutic targeting
VEGF and KSP in cancer patients with liver involvement. Cancer Discov 3:406–417.
doi:10.1158/2159-8290.CD-12-0429, 2159-8290.CD-12-0429 [pii]
101. Gomez HL et al (2012) Phase I dose-escalation and pharmacokinetic study of ispinesib, a
kinesin spindle protein inhibitor, administered on days 1 and 15 of a 28-day schedule in
patients with no prior treatment for advanced breast cancer. Anticancer Drugs 23:335–341.
doi:10.1097/CAD.0b013e32834e74d6
102. Khoury HJ et al (2012) A phase 1 dose-escalation study of ARRY-520, a kinesin spindle
protein inhibitor, in patients with advanced myeloid leukemias. Cancer 118:3556–3564.
doi:10.1002/cncr.26664
103. Woessner R et al (2009) ARRY-520, a novel KSP inhibitor with potent activity in hematologi-
cal and taxane-resistant tumor models. Anticancer Res 29:4373–4380. doi:29/11/4373 [pii]
Index
A Chromosome architecture, 138

Allosteric inhibitors, 29, 53, 69, 87, 110, 111 Chromosome congression, 11, 87, 137, 151,
Aneuploidy, 92–93, 108, 120–122, 127, 152, 174, 176, 180, 196
139–141, 143, 179, 180, 184, 204, 210, Clinical trials, 11, 12, 29, 30, 34, 40, 42,
233, 234, 238, 260 44, 47, 48, 63–73, 94, 111, 126,
Antimitotic, 28, 68, 70, 93, 95, 164, 184, 236 184, 206, 209, 228, 236, 238, 239,
ARRY-520, 41–43, 48, 68–72, 261 253, 257, 261
ATPase, 4–6, 8, 13, 15, 46, 53–54, 91, 94, CW069, 110–112
110, 126, 137, 149–151, 166, 184, Cytokinesis, 13–15, 80, 92, 93, 103, 135–137,
194, 207 139, 144, 149–153, 162–164, 166,
Axon, 153, 201, 209, 252–260 193–208, 210, 224, 238, 256, 260
AZ82, 110, 112
D
B Declustering, 110
Bipolar spindle assembly, 77–85, 119, 123, Dendrite, 201, 202, 252–259
166, 226 Drug discovery, 47–49, 167, 239
BTB-1, 14, 184, 185 Drug resistance, 84, 107, 109, 118, 238
C E
Cancer, 12, 28, 63, 83, 87, 105, 117–129, Eg5, 7–11, 27–49, 53–61, 63, 64, 67–72, 77,
135–144, 149–167, 174, 200, 223–240, 78, 80–85, 101, 103, 126, 166, 184,
251–262 185, 228, 230, 234, 257
Cancer signature, 159, 183 Eg5-independent cells (EICs), 81, 83, 84
Cell nucleus, 104
Centrosome(s), 102–105, 118, 120, 123, 124,
139, 223–226, 228–231, 234, 257 G
clustering, 12, 106–111 Genome stability, 139, 140, 142, 143
separation, 77–84
Chemosensitivity, 161, 163
Chemotherapy, 29, 43, 70, 95–96, 107, 122, H
126, 143, 161–164, 223–240, 251, 252, HSET inhibitor, 109–112
256, 260, 261 Human spleen, embryo, and testes protein
Chromokinesin, 13, 87, 135–144, 260 (HSET), 12, 101–112

F. Kozielski (ed.), Kinesins and Cancer, DOI 10.1007/978-94-017-9732-0
270 Index
I Mitosis, 1, 3, 8, 13–15, 29, 43, 48, 63, 70, 71,

Inhibitor, 6, 10–14, 27–49, 53–61, 63–73, 84, 77, 81, 82, 84, 85, 92, 93, 95, 102, 103,
85, 91, 92, 94–96, 107, 110, 111, 117, 105, 106, 108, 110, 117–124, 127,
124–129, 143, 163, 166, 167, 175, 135–140, 142–144, 151, 173, 174, 176,
184–185, 205, 228, 230, 236–240, 177, 180–182, 185, 194–196, 202,
251–262 207–210, 223, 225–235, 237–239,
Ispinesib, 29–32, 34, 46, 48, 67–70, 261 251–253, 256, 257, 260
Mitotic centromere associated kinesin
(MCAK), 8, 12–13, 29, 46, 117–129,
K 166, 181, 228, 230, 232–234
Kif2a, 8, 12, 46, 117–121, 123–126, 129, Mitotic kinesin, 1, 3, 6, 8–10, 28, 63–73, 94,
203, 260 135, 141, 143, 149, 159, 164, 184,
Kif2b, 12, 117–120, 122, 123, 125, 126, 180 251–262
Kif2c, 8, 12, 13, 117, 118, 120 Mitotic kinesin-like protein 1 (MKLP1), 7, 14,
KifC1, 12, 46, 101–112 153, 193–210, 228, 235, 257
KIF14, 15, 149–167 Mitotic kinesin-like protein 2 (MKLP2), 14,
Kif15, 11, 48, 77–85, 141, 166, 257 193–210
Kif19, 14, 172, 173, 175, 176, 178 Mitotic spindle, 29, 30, 46, 63, 70, 89–91,
Kif18A, 14, 89, 172–176, 178–185 102–103, 105, 106, 118, 124, 128,
Kif18B, 14, 172–176, 179, 181 137, 138, 143, 151, 153, 174, 176,
Kinesin-5, 4, 6–12, 59, 77, 80, 81, 101, 238, 184, 196, 205, 206, 223, 225,
257–259, 261 227–229, 234, 238, 253
Kinesin-8, 6–8, 14, 171–185 MKLP1. See Mitotic kinesin-like protein 1
Kinesin-12, 7–9, 11, 78–80, 85, 257–261 (MKLP1)
Kinesin(s), 1–15, 28, 29, 53, 63–73, 77–85, 87, MKLP2. See Mitotic kinesin-like protein 2
101–112, 117–129, 135–144, 149–167, (MKLP2)
171–185, 193–210, 230, 251–262 Monastrol, 185
Kinesin-6 family, 14, 193–198, 209, 210 mechanism, 28, 29, 46, 59, 63, 261
Kinesin motor domain, 1, 3–6, 15, 79, 136, site, 28, 29, 36, 43
137, 149–151, 166, 173 Motor, 1, 53, 54, 63, 77, 78, 87, 88, 101–112,
Kinetochore(s), 9, 11, 82–83, 87–92, 102, 105, 117, 135, 149, 173, 193, 223, 252
118–125, 127, 143, 173, 176, 178, 180, M-phase phosphoprotein 1 (MPP1), 14,
181, 223–227, 229, 231–234 193–210
Myeloma, 42, 48, 68, 69, 71, 72, 94, 95,
140, 201
L
Lead optimisation, 29, 34–36, 40–44, 47, 48
N
Neck linker, 4, 5, 7, 54, 55, 58, 126, 172, 193,
M 194, 197, 203
MAPs. See Microtubule-associated-proteins Neuron, 63, 72, 153, 194, 201, 202, 208, 209,
(MAPs) 251–262
MCAK. See Mitotic centromere associated Non-motor protein, 7, 78, 223–240
kinesin (MCAK)
Mechanochemistry, 3, 4, 6, 54
Microtubule O
depolymeriser, 6, 88, 93, 117–129, 178, Oncogene, 64, 65, 70, 141–142, 155–162,
181, 232, 254, 260 164–165, 208
detachment, 120, 123–125, 128 addiction, 64, 67, 71
dynamics, 13, 14, 82–83, 89–91, 93, 123,
129, 171–173, 178–180, 196, 204, 206,
223, 225–227, 230–231, 234, 235, 239 P
fragments, 124, 127, 128 Paclitaxel, 29, 40, 123–125, 127, 128, 185,
Microtubule-associated-proteins (MAPs), 9, 236–239
77, 78, 84, 105, 180, 230–232, 234 Passenger protein, 205, 228, 233, 238
Index 271
Pharmacokinetics (PK), 34, 36, 39, 40, 43, 48, Switch 1 (Sw1), 4–6, 54, 55, 57–60
65, 67, 68, 111, 238 Switch 2 (Sw2), 4–6, 54, 55, 57
Processive, 3, 4, 6, 7, 88, 171, 177, 194, 195, Synthetic lethality, 65
197, 232
Prognostic factor, 160, 182
Protein kinase, 224, 225, 227–237, 239, 240 T
Targeting protein for Xklp2 (TPX2), 8, 78, 79,
81, 82, 85, 224, 226, 228, 233, 234
R Taxol, 88, 178, 252–256
Replication, 48, 140, 142, 143, 165 Theranostic target, 111
Retinoblastoma, 15, 150, 154–158, 160, 164, Therapeutics, 6, 11–15, 29, 32, 36, 39, 48, 49,
166, 176 63–65, 70, 73, 108–112, 124, 143, 144,
155, 160, 162–167, 171–185, 201, 202,
205, 207, 209, 210, 234, 239, 252
S Toxicity, 32, 36, 40, 63, 65–70, 72, 111,
Spindle 143, 166
assembly, 9, 11–13, 48, 78–80, 85, 91, 93, TPX2. See Targeting protein for Xklp2
102, 105, 109, 120, 122, 129, 137, 180, (TPX2)
181, 196, 197, 203, 223, 226–230
pole, 9, 78, 80, 87, 102, 104, 105, 108,
109, 118–120, 123–125, 127, 128, 135, X
137, 223–227, 231, 234 Xklp2, 8, 78–80, 224, 234

Kinesins and Cancer

Uploaded by

Copyright:

Available Formats

You might also like

Kinesins and Cancer

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Kinesins and Cancer

Uploaded by

Copyright:

Available Formats

Frank

Kinesins and Cancer

ISBN 978-94-017-9731-3 ISBN 978-94-017-9732-0 (eBook)

Library of Congress Control Number: 2015932649

Springer Dordrecht Heidelberg New York London

Printed on acid-free paper

Springer Science+Business Media B.V. Dordrecht is part of Springer Science+Business Media

The validation of proteins as disease targets is a notoriously difficult and complex

Finally, the last chapter – on neuronal kinesins – attempts to focus attention on

London, UK Frank Kozielski

1 The Kinesin Superfamily........................................................................ 1

10 Kif14: A Clinically Relevant Kinesin and Potential

Index ................................................................................................................. 269

Ariane Abrieu Université Montpellier, CRBM, Montpellier, France

Anutosh Ganguly Department of Microbiology and Infectious Diseases, Snyder

Kristen J. Verhey, Jared C. Cochran, and Claire E. Walczak

Kinesin motors play important roles in cellular processes such as intracellular

K.J. Verhey (*)

© Springer Science+Business Media Dordrecht 2015 1

kinesin-1 (HsKIF5A, HsKIF5B, HsKIF5C)

kinesin-2 (HsKIF3A, HsKIF3B, HsKIF3C, family also contains HsKIF17)

kinesin-3 (HsKIF14, family also contains HsKIF1, HsKIF13, HsKIF16)

kinesin-4 (HsKIF4, family also contains HsKIF7, HsKIF21, HsKIF27)

kinesin-6 (HsKIF20A/MKLP2, HsKIF20B/MPP1, HsKIF23/MKLP1)

kinesin-8 (HsKIF18A, HsKIF18B, HsKIF19)

kinesin-12 (HsKIF15/HKlp2, family also contains HsKIF12)

kinesin-13 (HsKIF2A, HsKIF2B, HsKIF2C/MCAK, HsKIF24)

kinesin-14 (HsKIFC1/HSET, HsKIFC2, HsKIFC3)

kinesin motor domain forkhead associated (FHA) domain

ATP-indep. microtubule binding DNA binding, e.g. helix-hairpin-helix (HHH)

1.2 General Principles of Kinesin Motor Proteins

1.2.1 Mechanics of the Kinesin Motor Domain

P-loop Sw2 P-loop

The functional unit of a kinesin motor protein is a dimer, usually a homodimer, as

microtubule localization [154]. Members of the kinesin-10, kinesin-13 and kinesin-14

1.3 Kinesin Families with Functional Roles

1.3.1 Kinesin-5 Family: HsKIF11 (Also Known as Eg5)

Kinesin Activity Localization Function Inhibitor

HsKif11/Eg5 + end motor Spindle pole Numerous

HsKif10/Cenp-E Congression GSK923295

HsKifC1/HSET - end motor Centrosome

1.3.2 The Kinesin-12 Family: HsKIF15

1.3.3 The Kinesin-7 Family: HsCENP-E

up the complexities associated with mitotic protein inhibition as potential therapeutics.

1.3.4 The Kinesin-14 Family: HsKIFC1

The kinesin-14 proteins were first described in S. cerevisiae [191–193] and in

1.3.5 The Kinesin-13 Family: HsKIF2A, HsKIF2B, HsKIF2C

a centrally located kinesin-like catalytic domain, and a C-terminal tail involved in

1.3.6 Chromokinesins (The Kinesin-4

1.3.7 The Kinesin-8 Family: HsKIF18A, HsKIF18B, HsKIF19

1.3.8 The Kinesin-6 Family: HsKIF23 (Also Known

1.3.9 The Kinesin-3 Family: HsKIF14

1.4 Summary and Future Perspectives

7. Rayment I et al (1993) Three-dimensional structure of myosin subfragment-1: a molecular

59. Varga V et al (2006) Yeast kinesin-8 depolymerizes microtubules in a length-dependent

105. Pavicic-Kaltenbrunner V, Mishima M, Glotzer M (2007) Cooperative assembly of CYK-4/

James A.D. Good, Giacomo Berretta, Nahoum G. Anthony,

Abbreviations and Definitions