Professional Documents
Culture Documents
2010 Book RNATechnologiesAndTheirApplica
2010 Book RNATechnologiesAndTheirApplica
For almost 50 years, our understanding of molecular processes was very much
influenced by a statement formulated by Francis Crick, which is known as the
“central dogma of molecular biology.” According to this dogma, DNA is converted
to an mRNA and the mRNA to a protein. Another way of stating this dogma is that
one gene is converted to one mRNA and the mRNA to one protein with one
function. The present day molecular biology relies basically on genomic data for
creating new hypotheses, which allow the replacement of the term “descriptive
science” by the much more attractive “discovery science.” The discovery science
has revolutionized biology and gave new tools for hypothesis-driven research,
which concerns primarily, but not exclusively, nucleic acids.
The interest to apply RNA structural and functional characteristics in molecular
biology and medicine began in the late 1980s, when catalytic RNAs and in vitro
selection approaches were an exciting new frontier. Now, thirty years after those
discoveries, we begin to understand the novel aspects of RNA biology. Although
the pathways and molecular components involved in RNA-mediated gene regula-
tion are being elucidated very rapidly, the chemical and mechanistic basis still has
to be worked out. The understanding of molecular mechanisms, and the possibi-
lities for employing these processes for therapeutic purposes, falls surely into the
realm of chemical biology.
The causal relationship between sequence, structure, and function significantly
affects the interaction of RNA molecules with proteins, metabolites, and other
nucleic acids, making RNA a malleable and attractive molecule to drive program-
mable function. RNA molecules, which derive sophisticated behavior from an
ability to adopt complex structures, can be generated from potentially all possible
sequence combinations, leading to diverse secondary structures and functions.
These structures can exist in the form of modular domains, which confer specific
and unique functionality. RNA molecules have evolved to regulate gene expression
in a wide variety of ways in cells and viruses. Despite that, we are only beginning to
appreciate how much of known phenotypic variation can be explained by these
novel classes of RNA regulators.
v
vi Preface
The recognition of the biological roles of small molecular weight RNAs has
been one of the most significant discoveries in molecular biology. These RNA
molecules influence the translation of messenger RNAs (mRNAs) in posttranscrip-
tional manner that makes the regulation of RNAs even more complex.
Recent advances in RNA biology and nucleic acid engineering are inspiring the
use of RNA molecules for the construction of different RNA tools. RNA has
become a focus of investigations into novel therapeutic schemes. Oligonucleotide-
based approaches depend on the Watson–Crick base pairing of oligonucleotides to
their corresponding mRNA target. This leads to posttranscriptional gene silencing
by mRNA cleavage or translational inhibition. Oligonucleotide-based therapies
have great potential for the treatment of RNA virus infections and various
diseases. They include antisense oligonucleotides and their derivatives such as
peptide nucleic acids, locked nucleic acids and morpholinos oligonucleotides
(ONs), RNAi, microRNA ribozymes, aptamers, and Spiegelmers. Beyond sequence
conservation, a very important point is the fact that the RNA target must be
accessible for oligonucleotide interaction. Although the effects of antisense
RNAs on the corresponding sense RNAs have not been clearly established, a
number of examples indicate that they may exert control at various levels of gene
expression, such as transcription, mRNA processing, splicing, stability, transport,
and translation.
Multiple challenges, such as optimization of selectivity, stability, delivery, and
long-term safety, have to be addressed in order for RNA drugs to become a
successful therapeutic tool. Not all RNA classes (e.g., ribozymes or RNA decoys)
were so far successfully developed as drugs. The use of RNA-mediated interference
(RNAi) for gene silencing has provided a powerful tool for loss-of-function studies
in a variety of metazoans. siRNA-mediated gene silencing by degradation of target
messenger RNA has been widely used for the functional characterization of genes.
The secondary structure of mRNA target sites has been reported to strongly
influence RNAi activity. Compared with the laborious, time-consuming, and very
costly gene knockout models, siRNA provides an efficient, specific, and economic
solution for inhibiting the expression of target genes. Efficient siRNA delivery is
essential for the success of specific gene silencing and is therefore understandable
that a number of different laboratories are currently working on the problem of
siRNA delivery in living organisms. Because high doses of siRNAs do provoke an
altered expression of many other genes, selection of an optimal condition could be
very helpful to minimize potential side effects. The advantage of the system lies in
the application of short RNAs, which can be synthesized relatively cheap and can
evolve quickly, to regulate a large and complex protein synthesis.
In this volume, 10 papers out of 19 are dealing with various aspects of RNA
interference. They cover basic issues of the technique and its application in biology
and medicine. There are also three contributions on antisense RNA approaches,
which show a high therapeutic potential.
It is becoming clear that microRNAs are essential regulators of many of the key
pathways implicated in tumor pathogenesis. While adding another layer of com-
plexity, the discovery of the role of miRNAs in tumorigenesis has revealed a new
Preface vii
ix
x Contents
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
Contributors
xi
xii Contributors
Anke Detzer Institut für Molekulare Medizin, ZMSZ, Universität zu Lübeck &
Universitätsklinikum Schleswig-Holstein, Ratzeburger Allee 160, 23538 Lübeck,
Germany
Karin von Eije Center for Infection and Immunity Amsterdam (CINIMA),
Department of Medical Microbiology, Laboratory of Experimental Virology,
Academic Medical Center of the University of Amsterdam, Amsterdam, The
Netherlands
Tobias Restle Institut für Molekulare Medizin, ZMSZ, Universität zu Lübeck &
Universitätsklinikum Schleswig-Holstein, Ratzeburger Allee 160, 23538 Lübeck,
Germany
Contributors xv
Georg Sczakiel Institut für Molekulare Medizin, ZMSZ, Universität zu Lübeck &
Universitätsklinikum Schleswig-Holstein, Ratzeburger Allee 160, 23538 Lübeck,
Germany
Nathalie Spruyt Institut de Biology de Lille, CNRS UMR 8161, 1 rue Pr Calmette,
BP 447 59021, Lille Cedex, France
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2 RNA Silencing Effector as a Two-Component System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3 Small RNA Biogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.1 miRNAs and siRNAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.2 Dicer-Independent Pathways . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.3 Endo-siRNAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4 RISC Loading, Sorting, and Target-Sensing of Small RNAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4.1 Sorting by Precursor Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.2 Sorting by the 50 Ends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
4.3 Sorting by Dicer Processing Polarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
4.4 Target-Sensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
5 Safeguards for RNA Silencing Pathways . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
6 Effector Modes of RNA Silencing Pathways . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
7 Regulations of RNA Silencing Pathways . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
7.1 Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
7.2 Modification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
7.3 RISC Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
8 Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 1
RNA Technologies, DOI 10.1007/978-3-642-12168-5_1,
# Springer-Verlag Berlin Heidelberg 2010
2 K. Saito et al.
of complexity to the way cells regulate protein levels. These pathways, which are
collectively referred to as RNA silencing, mediate biological activities that fall
into two broad categories; genomic surveillance and gene regulation. RNA silenc-
ing occurs in a variety of organisms and is evolutionarily conserved. Central to
these processes is small RNA generation by Dicer and inactivation of cognate
RNA targets by small RNA–Argonaute complexes acting in combination with a
multitude of interacting and collaborating proteins. In some systems, silencing
signals are amplified and small RNAs are produced by a Dicer-independent
pathway, which challenges our perception and definition of RNAi. There has
been remarkable progress in our understanding of the mechanisms underlying
RNAi and related silencing processes, which hails the prospect of fully decipher-
ing the RNAi machinery.
1 Introduction
The mechanism underlying the RNA-induced defects appears to be distinct from that of
conventional antisense RNA because both the sense and antisense RNA strands cause
similar defects. Because the mechanism is not known, we will refer to this technique as
RNAi, for RNA-mediated interference (Rocheleau et al. 1997).
Craig Mello chose the term “RNAi” to refer to a mysterious gene silencing process
first observed in 1997. Then, in 1998, the key molecule in the process was found to
be dsRNA (Fire et al. 1998). The discovery of RNAi heralded a new RNA revolu-
tion and led to the discovery of “hidden layers” of gene expression regulation, in
which many previously unidentified families of small RNAs, 20–30 nt in length,
mediate gene silencing. These findings led to the unification of a number of
different RNA-based silencing pathways, including microRNA (miRNA)- and
small interfering RNA (siRNA)-mediated silencing in plants and animals, cosup-
pression and paramutation in plants, quelling in fungi, heterochromatin formation
in fission yeast, and RNA-directed DNA methylation (Stefani and Slack 2008; Ding
and Voinnet 2007; Girard and Hannon 2008; B€ uhler and Moazed 2007). In addi-
tion, RNAi has rapidly become one of the most powerful and indispensable
functional genomics tools and is also considered as a novel and invaluable clinical
therapeutic approach to specifically target genes associated with a variety of dis-
eases (Novina and Sharp 2004). Key steps in the RNAi pathway are shared by a
diverse set of gene regulatory mechanisms, including mechanisms that silence
endogenous genes, particularly genes involved in development and stem cell
maintenance, and mechanisms that restrain the expression of transposons or viruses
and that direct transcriptional gene silencing (Stefani and Slack 2008; Ding and
Voinnet 2007; Girard and Hannon 2008; B€ uhler and Moazed 2007; Siomi and
Siomi 2008).
The Key Features of RNA Silencing 3
Long dsRNA is the trigger molecule in RNAi and can be derived from various
sources, such as simultaneous sense and antisense transcription of specific genomic
loci, foldback-structured transcripts from repetitive sequences, and viral replication
intermediates. The basic biochemical requirements of an RNAi response can be
broken down into three steps (Meister and Tuschl 2004; Tomari and Zamore 2005;
Siomi and Siomi 2009): (a) long dsRNA is processed by a ribonuclease (RNase) III
enzyme called Dicer into small interfering RNA (siRNA) duplexes; (b) these are
subsequently unwound, and one strand, the so-called guide strand (as compared
with the complementary passenger strand), is preferentially loaded onto the RNA-
induced silencing complex (RISC); (c) siRNAs serve as the sequence determinants
of the RNAi pathway by scanning the resident population of mRNAs (possibly any
single-stranded RNA) and directing multiple rounds of cleavage to homologous
mRNAs, via Slicer, a RISC component endonuclease (Fig. 1). Despite using
divergent proteins and mechanisms, organisms employ strikingly convergent stra-
tegies, which comprise a rather simple two-component system; small RNAs act as
specificity factors and Argonaute proteins act as effectors for repression. Depending
on both the nature of the Argonaute protein and the degree of complementarity
between the small RNA and the target sequence, association of RISC with target
RNAs can result in a range of functions. These findings together with structural and
biochemical studies culminated with the revelation that Argonaute is the RNA-
guided Mg2+-dependent RNA endonuclease of RISC that cleaves a single phos-
phodiester bond in the target, otherwise known as the Slicer (Meister et al. 2004;
Liu et al. 2004; Song et al. 2004). Once loaded with a small RNA, Argonaute
proteins exhibit a variety of functions for controlling protein synthesis and RNA
stability, maintaining genome integrity and mediating the production of a specific
set of small RNAs (Peters and Meister 2007; Hutvagner and Simard 2008).
Although the establishment of functional specificity for the different Argonaute
proteins remains to be understood, some of these specialized functions are likely to
be caused by structures and associated proteins specific to each Argonaute.
Emerging data also support Dicer as a RISC component, thus mechanistically
connecting the initiation phase with the effector phase of gene silencing. Refine-
ments of RNAi-related systems include built-in molecular rulers that define the size
of small RNAs, apparatuses that determine small RNA strand selection or polarity,
mechanisms that direct further amplification rounds using RNA-dependent RNA
polymerase (RdRP) activities on additional templates or by forming a cleavage-
mediated cycle, and safeguards for off-target silencing. The analysis of the biogen-
esis and targeting of small RNAs has also benefited from a classic genetic analysis
combined with a more recent approach that uses novel high-throughput cDNA
sequencing technologies with sophisticated bioinformatics tools (Mardis 2008).
The picture emerging from these analyses is of multiple variations on the core
RNA silencing mechanisms. Importantly, it is becoming clear that the activity of
RNA silencing pathways is subject to intense regulation at different levels, from
biogenesis of small RNAs to silencing modes of RISC.
RNAi is of great biomedical interest as it has potential for the study of gene
function, the validation of candidate drug targets, and even the treatment of disease
4 K. Saito et al.
a b
transposon locus flamenco
? Primary ?
Aub Piwi
Drosha/Pasha U U
Dicer2/R2D2
c d
21U-RNA primary siRNA/
Core sequence Argonaute complex
CTGTTTCA YRNT
A/T rich A/T rich Chr. IV
Spacer 3' 5'
Large motif Small motif
(~20nt ) 5' 3'
~34nt 4nt
5'
RdRP
P
PP
3'
U
5' 3'
Fig. 1 Small RNA production and loading onto Argonautes. (a) In Drosophila, dsRNAs from an
endogenous or exogenous source are processed by RNase III domain nucleases, such as Drosha
and Dicer, into small RNA duplexes of about 21 nucleotides. Accessory proteins for RNase III
domain nucleases, such as Loqs and Pasha, help to distinguish the dsRNA precursors depending on
secondary structures or the degree of complementarity. Upon this association, one strand of the
small RNA duplex is selected and loaded onto one of the Argonautes. (b) In Drosophila gonad,
PIWI-subfamily proteins (Piwi, Aub and AGO3) associate with piRNAs and act in transposon
silencing. In the primary processing pathway, piRNAs are produced from single-stranded
precursors transcribed from the genome. In the secondary pathway (also known as the amplifica-
tion loop), piRNAs are produced by a Slicer-dependent mechanism. Aub and AGO3 generate the
50 end of piRNAs associating with AGO3 and Aub, respectively. The amplification loop is
independent of de-novo synthesis of piRNAs. (c) In C. elegans, the 21U-RNAs are predominantly
generated from chromosome IV and characterized by a specific upstream motif. Similar to
Drosophila piRNAs, 21U-RNAs are associated with a PIWI-subfamily protein, PRG-1, and are
produced in a Dicer-independent manner. (d) In C. elegans, Primary siRNAs associate with RDE-
1 and are then guided to the target RNA. Target RNAs are used as a template for producing
secondary siRNAs by RNA-directed RNA polymerase (RdRP). The secondary siRNAs can then
engage in another round of silencing
(Hannon 2002). Here, we review the biogenesis of small guide RNAs and discuss
our understanding of the molecular mechanisms of RNA silencing based on recent
insights into the regulatory circuitry of the silencing state. Although much of the
The Key Features of RNA Silencing 5
Since the discovery of the RNAi pathway, many endogenous small RNA pathways
that share many common features with the siRNA pathway have been identified.
These small RNAs function by degrading mRNAs and effect translational repression
of mRNAs or regulate gene silencing at the transcriptional level by heterochromatin
modification. A common feature of small RNAs is that they function as specificity
determinants for the repressive activities of Argonaute-containing effector com-
plexes. In principal, the RNA silencing system is an RNA-guided enzyme system
that requires only one (nonsequence-specific) protein for its enzymatic activity.
Sequence specificity is achieved by the small RNA component of the RNA-protein
(RNP) complex (H€ uttenhofer and Schattner 2006). In RNA silencing, one protein,
Argonaute, binds to many small guide RNAs that recognize their target by Watson–
Crick base pairing and thereby guide the Argonaute complex to different substrates.
It should be noted that although small guide RNAs might home in on homologous
DNA sequences, to date they have only been shown to target homologous RNAs.
The core protein component of all RISCs is a member of the Argonaute family of
small RNA-binding proteins (Tabara et al. 1999; Faehnle and Joshua-Tor 2007).
Members of this family are defined by the presence of PAZ and PIWI domains and
they normally consist of one variable N-terminal domain and conserved C-terminal
PAZ, MID and PIWI domains (Faehnle and Joshua-Tor 2007). The PAZ domain
recognizes the 30 end of small RNAs in a sequence-independent manner. The PIWI
domain adopts a folded structure similar to that of RNase H enzymes and exhibits
endonuclease or Slicer activity. Three residues (DDH) within the PIWI domain
form a catalytic triad. 50 monophosphate on the guide RNA strand binds to a
conserved binding pocket contained within a cleft bridging the MID and PIWI
domains. Argonaute proteins are divided into three phylogenetic groups: AGO,
from its founding member Arabidopsis Argonaute 1 (Ago1); PIWI, from Drosophila
Piwi (P-element induced wimpy testis); and WAGO/group III, Caenorhabditis
elegans-specific proteins. Many metazoan organisms have multiple Argonautes,
while single celled organisms like the fission yeast Schizosaccharomyces pombe
and the protozoan parasite Trypanosoma Brucei encode only one Argonaute
(Cerutti and Casas-Mollano 2006). This remarkable diversity of Argonautes raised
the possibility that different members of the family have become specialized in
each organism to perform distinct functions (see below).
A hallmark of RNA silencing is the production of short dsRNA molecules
(21–28 nt) by RNase III enzymes. Expansion of Dicer enzymes has also occurred:
Arabidopsis thaliana has four Dicer-like proteins and Drosophila melanogaster has
two Dicers, whereas mammals and yeast encoded a single Dicer. Despite high levels
of functional conservation, the complexity of the RNA silencing machinery varies
6 K. Saito et al.
greatly between different organisms (Cerutti and Casas-Mollano 2006; Kim et al.
2009; Siomi and Siomi 2009). It has been argued that the last common ancestor of
eukaryotes likely had at least one Argonaute, one Dicer, and one RNA-dependent
RNA polymerase (RdRP) (Cerutti and Casas-Mollano 2006), although interestingly,
among animals, only nematodes are reported to contain genes encoding RdRP. Two
main categories of small RNAs, siRNAs, and miRNAs have been defined, differing
in the nature of their precursors (Fig. 1a). Other small RNA species have been
recently identified, including piRNAs and endogenous siRNAs in flies and mam-
mals (Fig. 1a, b). The biogenesis of these small RNAs challenges the definition of
RNAi, since some do not appear to be produced “in response to dsRNA.”
III, Drosha, first defines one end of the miRNA–miRNA* duplex and releases
approximately 65 nt pre-miRNAs (Fig. 1a). The pre-miRNA hairpin is then exported
to the cytoplasm, where the second RNase III, Dicer, completes the processing.
Drosha is part of a large complex, known as the “Microprocessor,” which acts like a
molecular ruler to determine the cleavage site (Han et al. 2006). In this complex,
Drosha interacts with its dsRNA-binding domain (dsRBD) cofactor protein, DGCR8
(also known as Pasha) (Denli et al. 2004; Gregory et al. 2004). A typical metazoan
pri-miRNA contains areas of local snapback structure that consists of a 33 bp stem, a
terminal loop and flanking segments and can be “cropped” by the Drosha complex.
DGCR8 prefers the junction between flexible single-stranded RNA and a double-
stranded stem; indeed, the flanking single-stranded RNA (ssRNA) segments are vital
for binding to DGCR8, and the 33 bp stem is also required for efficient binding.
Drosha may not be in direct contact with RNA at this stage. Drosha may interact
transiently with the stem of this “pre-cleavage” complex, where the processing center
of the enzyme, located about 11 bp from the ssRNA–dsRNA (SD) junction, makes a
staggered pair of breaks in the RNA to create the 65 nt long pre-miRNA. Thus,
DGCR8 may function as the molecular anchor that measures the distance from the
SD junction (Han et al. 2006). The Microprocessor could recognize the terminal loop
as ssRNA and bind to the stem–loop in the opposite orientation. In this case, abortive
cleavage can occur at an alternative site at 11 bp from the terminal loop. However,
most pri-miRNAs contain internal bulges or weakly paired bases 11 bp from the
terminal loop (Han et al. 2006). The computational search for miRNAs may be
facilitated by seeking sequences capable of folding into a structure predicted to be
bound by DGCR8 and to promote Drosha cleavage 11 bp from the junction of a
stem with single-stranded RNA tails.
siRNA and miRNA duplexes derive from the processing of longer duplexes
and pre-miRNA hairpins, respectively. This is performed by Dicer, which forms
21–25-nt duplexes possessing 50 -monophosphates, 30 -hydroxyl groups, and 2 nt 30
overhangs, which are classic hallmarks of RNase III enzymes. Dicer binds dsRNA
and generates RNA products of specific lengths. The prototypical Dicer contains,
from N to C terminus, a PAZ domain, two tandem RNase III domains, and a dsRNA-
binding domain (dsRBD). The distance between the PAZ domain that binds the 30
end of dsRNA and the two catalytic RNase III domains matches the length spanned
by 25 base pairs of RNA (MacRae et al. 2007). Thus Dicer itself is a molecular ruler.
The examination of sequencing data of small RNAs from D. melanogaster led to
the identification of clusters of small RNAs originating from the outer edges of an
annotated small intron (Okamura et al. 2007; Ruby et al. 2007). The 30 end of the
stem–loop precursor structure of these intronic small RNAs coincides with the 30
splice site, and is cleaved by nuclear pre-mRNA splicing rather than by Drosha.
After being debranched, these small intronic RNAs mimic the structural features of
pre-miRNA hairpins and enter the miRNA-processing pathway without Drosha-
mediated cleavage. These pre-miRNAs/introns were termed “mirtrons.”
The imprecision of Drosha or Dicer cleavage can result in the production of an
miRNA:miRNA* duplex with different 50 and 30 ends. This population of miRNA
variants is termed isomiRs. Most miRNAs in animals form imperfect hybrids with
8 K. Saito et al.
sequences in the target mRNA, with the miRNA 50 -proximal “seed” region providing
most of the pairing specificity. Therefore, the imprecise cleavage either alters the seed
sequence or inverts the relative stabilities of the 50 end of the duplex (see next
section). Recent deep-sequencing results of small RNAs reveal that human cells
may take advantage of such imprecise cleavage to create a diverse set of miRNAs
from a single precursor, which could broaden the reach of the miRNA regulatory
network (Landgraf et al. 2007; Azuma-Mukai et al. 2008; Morin et al. 2008).
3.2.1 piRNAs
mainly AGO3 and Aub (Brennecke et al. 2007; Gunawardane et al. 2007), and Piwi
is spatially separated from them at subcellular and cell-type levels (Brennecke et al.
2007; Gunawardane et al. 2007; Nishida et al. 2007). In addition, piRNAs derived
from a particular piRNA cluster (flamenco locus) associate almost exclusively
with Piwi (Brennecke et al. 2007). These observations indicate that some other
mechanisms of piRNA biogenesis must operate. It remains unclear how primary
Piwi-associated antisense piRNAs are produced. Perhaps the most crucial question
is how transposon antisense transcripts, among the sea of the transcriptome, are
distinguished from other cellular transcripts and are specifically recognized by the
piRNA biogenesis machinery as source transcripts for primary piRNAs.
Recently, a new class of abundant 21 nt small RNAs was discovered in
C. elegans (Fig. 1c). They are characterized by a 50 -U bias (and are thus termed
21U-RNAs) and have a characteristic sequence motif about 42 bp upstream of the
start of the small RNA (Ruby et al. 2006). They originate in only a few clusters that
are specific to chromosome IV. 21U-RNAs are not produced by processing dsRNA
precursors but may be derived from thousands of individual, autonomously
expressed loci, broadly scattered in two large regions of chromosome IV; each
21U-RNA could be the product of an individual RNA polymerase transcription
event. 21U-RNAs are expressed in the germline and interact with the PIWI protein
PRG-1 (Batista et al. 2008; Das et al. 2008). They depend on PRG-1 activity for
their accumulation but are independent of Dicer for their production. Mutations in
prg-1 exhibit a reduced brood size and a temperature-sensitive sterile phenotype,
consistent with the notion that Piwi proteins are linked to germline maintenance.
Thus, 21U-RNAs are the piRNAs of C. elegans. Like the abundant pachytene
piRNAs found in mammals, 21U-RNAs encode remarkable sequence diversity
and yet lack obvious targets. In addition, none of the 15,000 individual
21U-RNA sequences are conserved between C. elegans and Caenorhabditis brigg-
sae. Although only one transposon (Tc3)-directed 21U-RNA, out of more than
15,000 different 21U-RNAs encoded in C. elegans, was identified, PRG-1/21U-
RNA complexes still suppress the transposon, probably by the production of
secondary siRNAs by RNA-dependent RNA polymerase (RdRP) activity (see
below). The sequence diversity and lack of obvious targets may suggest that
21U-RNAs act in cis to regulate their own genomic loci.
Scan RNAs (scnRNAs) in the ciliate Tetrahymena thermophila direct elimina-
tion of transposon-like DNA sequences and associate with a Piwi protein, Twi1.
Thus, they are the piRNAs in Tetrahymena. However, they are produced by the
Dicer-dependent pathway (Mochizuki and Gorovsky 2005). These examples indi-
cate that, during evolution, the core Piwi and piRNA machinery may have adopted
different strategies for the production and silencing of targets.
3.3 Endo-siRNAs
pathways that repress transposons. Endo-siRNAs are also present in mouse oocytes
and are derived from a variety of sources, including transposable elements (Tam
et al. 2008; Watanabe et al. 2008). However, a fraction of endo-siRNAs are
processed from overlapping regions of functional genes and from cognate pseudo-
genes suggesting that pseudogenes, previously thought to be nonfunctional protein
fossils, may actually regulate the expression of their founder gene.
Although siRNAs and miRNAs are distinguished from one another, not by their
size or function, but rather by their origin (Tomari and Zamore 2005), with
miRNAs being processed from stem–loop structures and siRNAs arising from the
cleavage of long dsRNA precursors, the discovery of endo-siRNAs makes this
distinction more difficult. This blurring of boundaries among different types of
small RNA has interesting evolutionary implications. Long stem–loop structures
for endo-siRNAs are reminiscent of plant miRNA precursors. One hypothesis for
the evolutionary origin of plant miRNAs is that new plant miRNA loci may evolve
from the inverted duplication of founder loci, producing a hairpin RNA (Chapman
and Carrington 2007). The transcribed hairpin RNAs would exhibit near perfect
self-complementarity and may be processed by DCL enzymes other than DCL1, the
primary plant miRNA processing enzyme, since DCL1 has limited activity on such
substrates. Subsequent acquisition of mutations by genetic drift would produce a
hairpin with imperfect complementarity and channel its processing to DCL1. Thus,
stem–loop structures for endo-siRNAs could be evolutionary intermediates in the
gradual transformation into miRNAs. An adaptive switch from Dicer2- to Dicer1-
mediated processing during the course of miRNA gene evolution might occur in
Drosophila, as suggested for miRNAs in plants.
siRNA duplexes are converted into a single-stranded form as they assemble onto
the RISC, where they provide the sequence specificity or guide for mRNA degra-
dation. Thus, the key steps in converting pre-RISC to mature RISC (holo-RISC) are
siRNA strand unwinding and preferential strand retention or strand selection. The
prevalent view of RISC loading is that thermodynamic asymmetry along the siRNA
or miRNA duplex determines which RNA strand will be retained as the “guide” and
which RNA strand will be discarded as the “passenger.” Specifically, the RNA
strand with its 50 end at the thermodynamically less stable end of the siRNA duplex
is preferentially loaded onto the RISC as the guide strand, a phenomenon referred to
as the asymmetry rule (Khvorova et al. 2003; Schwarz et al. 2003).
Considering the interactions between Dicer and Argonaute proteins (Tabara
et al. 2002), siRNA production and RISC assembly with siRNAs might be physi-
cally coupled. In Drosophila, Dicer-2 does not simply transfer siRNAs to a distinct
RISC but rather assembles onto the RISC along with the siRNAs, indicating that its
role extends beyond the initiation phase. Loading of siRNA duplexes onto AGO2 is
facilitated by the RISC-loading complex (RLC), which includes Dicer-2 and its
dsRBD partner protein, R2D2 (Liu et al. 2003). Which strand of the siRNA duplex
The Key Features of RNA Silencing 13
The emerging picture suggests that miRNAs and siRNAs are initially bound to
the Argonaute-containing complex as duplexes, and are subsequently unwound,
resulting in bound single-stranded RNAs. Subsequently, how is a small RNA
directed to a specific silencing pathway? Small RNAs destined for the different
silencing pathways are often generated by the same Dicer proteins; therefore, the
most obvious solution would be to have different Argonaute proteins for different
silencing pathways, dependent on the source of the dsRNA. siRNA production by
Dicer may be directly coupled to RISC assembly as described. According to this
view, Dicer may pass siRNAs directly to RISC, without siRNAs diffusing freely in
the cytoplasm after their production. This would also aid the discrimination of bona
fide siRNAs from various RNA degradation products within the cell. However,
analyses of how different small RNAs are channeled to different AGO proteins
shows multiple variations.
In Drosophila, AGO1 and AGO2 are responsible for miRNA and siRNA pathways,
respectively (Okamura et al. 2004). Although such restriction of each class of small
RNA to a distinct Argonaute complex could occur because miRNAs and siRNAs
are produced by different Dicer pathways, Dicer1 for miRNAs and Dicer2 for
siRNAs, it appears that in flies small RNA production and small RNA loading
onto Argonaute protein complexes are separate steps. Small RNAs are loaded into
either AGO1 or AGO2 based on the structure of a small intermediate RNA duplex
(Tomari et al. 2007). If the duplex has a bulge in the middle (frequently observed in
miRNA precursors), the RNA is routed onto AGO1. If the duplex is perfectly
matched, the small RNA is channeled onto AGO2. This is because the Dicer-2/
R2D2 heterodimer binds well to highly paired small RNA duplexes but poorly to
duplexes bearing central mismatches. Thus, the Dicer-2/R2D2 heterodimer not only
determines the polarity of siRNA loading based on thermodynamic stability rules
but also acts as a gatekeeper for AGO2 complex assembly, promoting the incor-
poration of siRNAs over miRNAs. These observations suggest that each siRNA
duplex dissociates from the Dicer active site after it is produced and is subsequently
recaptured by the Dicer-2/R2D2 heterodimer; sorting can occur postdicing in
Drosophila, highlighting the importance of central mismatches in precursor
duplexes. Indeed, a single central bulge involving positions 7–11 was sufficient to
redirect an otherwise fully-paired siRNA duplex to AGO1 (Tomari et al. 2007).
Thus, a central unpaired region serves as both an antideterminant for the AGO2-
loading pathway and a preferred binding substrate for the AGO1 pathway.
Although AGO1 favors small RNA duplexes with central mismatches, the large
fraction of an miRNA/miRNA* duplex whose central region is base paired still
associates with the AGO1 RISC (Okamura et al. 2004; Kawamura et al. 2008),
suggesting that the AGO1-loading pathway is inherently selective and not a default
pathway for small RNAs rejected by the AGO2 pathway. The proteins facilitating
The Key Features of RNA Silencing 15
AGO1 loading remain to be identified. Recent studies in C. elegans also support this
“precursor structure” model for small RNA sorting (Jannot et al. 2008).
The 50 nucleotide identity or phosphorylation status at the 50 end of small RNAs has
been shown to influence Argonaute protein associations. Arabidopsis miRNAs and
ta-siRNA begin with a 50 -terminal uridine residue and preferentially associate with
AGO1 (Mi et al. 2008). By contrast, AGO2 associates preferentially with small
RNAs containing 50 -terminal adenosines, whereas AGO5 prefers small RNAs with
50 -terminal cytosines. Interestingly, the opposite strands of miRNAs (miRNA*) with
50 adenosines and 50 cytosines are also bound to AGO2 and AGO5, respectively. These
findings have led to the hypothesis that the binding affinity of Argonaute proteins for
small RNAs is determined by the nucleotide at the 50 end. This implies that some
miRNA variants are loaded more efficiently than others, according to the identity
of their 50 nucleotide. Although these 50 -terminal nucleotide preferences generally
hold true for these Argonautes in plants, miR172 has a 50 -terminal adenosine, but
the majority of miR172 molecules associate with AGO1. Also, AGO7 preferentially
associates with miR390, which possesses a 50 -terminal adenosine (Montgomery et al.
2008). Therefore, it does not appear to be the sole determinant influencing Argonaute
association. In Arabidopsis, Dicer processing may be uncoupled from association with
Argonaute because miRNAs are generated by DCL1.
Secondary siRNAs in C. elegans are specifically loaded onto SAGOs (Yigit et al.
2006). Secondary siRNAs carry a 50 -triphosphate modification (Pak and Fire 2007;
Sijen et al. 2007), the hallmark of RdRP products (Fig. 1d). This might serve as a
recognition element for specific SAGO binding, while excluding binding by pri-
mary Argonautes, such as RDE-1. These findings imply that Argonaute diversifica-
tion is a consequence of which small RNA they recruit. Whether the protein
conformation dictates its small RNA partners awaits the determination of eukary-
otic Argonaute protein structures.
The 50 nucleotide was found to be inserted into a pocket in the MID domain
(Faehnle and Joshua-Tor 2007). It can be envisioned that the residues constituting
the 50 end binding pocket may differ between Argonaute proteins to accommodate
particular 50 terminal nucleotides. It will be interesting and important to solve the
structures of different Argonaute proteins in plants and animals.
RdRPs produce small RNAs directly from the target mRNA, in a primer-independent
manner. Therefore, all secondary siRNAs are of negative polarity and serve to
reinforce silencing of the RNA target (Yigit et al. 2006; Pak and Fire 2007; Sijen
et al. 2007). In fission yeast, the physical association of Dicer with a RdRP
complex, termed RNA-dependent RNA polymerase complex (RDRC), and an
Argonaute complex, named RNA-induced transcriptional silencing complex
(RITS), may facilitate loading of siRNAs onto Argonaute in a directional manner
as Dicer moves along and cuts the dsRNA products of RdRP, giving rise to
antisense strand bias (B€ uhler and Moazed 2007). Thus, Dicer processing polarity
defines siRNA strand polarity. RITS and RDRC are proposed to act in a self-
reinforcing loop in which DNA-interacting proteins and the siRNA in RITS guide
H3K9 methylation and heterochromatin formation, and also the RdRP-mediated
conversion of transcripts into siRNA precursors (Moazed 2009). Thus, active
transcription may be a prerequisite for the assembly of heterochromatin in fission
yeast. Interestingly, all of these processing events appear to occur in the nucleus of
fission yeast. By contrast, nuclear functions for mammalian AGO and Dicer
proteins are still a matter of debate, although nuclear functions of siRNAs have
been reported in humans (Meister 2008).
4.4 Target-Sensing
Once loaded with single-strand guide small RNA, how does RISC find its target
RNA? It is clear that most of the binding energy that tethers RISC to a target RNA
comes from bases in the 50 half (the “seed” region) of the small RNA (Haley and
Zamore 2004), which has led to the hypothesis of the “two-state” model for RISC
target binding (Filipowicz 2005; Tomari and Zamore 2005). In this model, the 50
part of the small RNA within RISC has a favorable structure for base pairing,
whereas the arrangement of the 30 part antagonizes base pairing with the target
RNA. After a base-pairing between the seed and the target is formed, the confor-
mational change of an Argonaute occurs so that the 30 half of the small RNA can
base-pair with the target RNA. However, it should be noted that some miRNA
sequences, including let-7 and miR-34, are perfectly conserved across a vast
evolutionary distance. And about 10% of the miRNA families identified in inverte-
brates are completely conserved in mammals. Therefore, the principle that base-
pairing to the 50 seed part of the miRNA is a dominant factor in miRNA target
recognition alone does not account for the perfect conservation of the full sequences
of these miRNAs. It appears that accessibility of the target site can be sensed by an
intrinsic, nonspecific affinity of RISC for single-stranded RNA, which follows the
initial specific RISC-target association via the 50 “seed” part of the siRNA (Ameres
et al. 2007). On the other hand, the accessibility of the target site correlates directly
with the efficiency of cleavage, indicating that RISC is unable to unfold structured
RNA. Since mRNA targets exist in the cell as RNPs (Dreyfuss et al. 2002),
accessibility can, therefore, also be controlled by a number of RNA-binding
The Key Features of RNA Silencing 17
proteins that either mask the target site or facilitate unfolding. Binding site accessi-
bility provides an additional layer of regulatory control.
By analogy with classic immunity, there are also mechanisms in RNA silencing that
manage the population of effector molecules involved in surveillance, both by
subtracting out specific components that are not engaging targets and in some
cases by amplifying specific components that are engaging their targets. In RNA
silencing, one nonsequence-specific RNA-binding protein, Argonaute, binds many
different small guide RNAs, resulting in effector RISC complexes. Because target
recognition uses complementary RNA sequences, once a particular element or gene
is recognized by the RNAi system, all copies within a cell will be targets for
inactivation. This system, therefore, requires gatekeepers to ensure that Argonaute
can bind small guide RNAs but cannot degrade small RNAs.
Small RNAs produced by the RNase III enzymes, Drosha and Dicer, character-
istically leave 50 -monophosphates and 2 nt 30 overhangs on the processed product.
Therefore, the PAZ domain of Argonautes can initially distinguish these small
RNAs, by binding to their characteristic 30 -overhangs, from degraded RNAs that
are derived from nonrelated pathways. In addition, to become incorporated into
RISC and mediate target cleavage, the guide strand of siRNA needs to display a
phosphate group at the 50 end (Pham and Sontheimer 2005). In humans, the enzyme
capable of phosphorylating the 50 end of siRNAs is hClp1 (Weitzer and Martinez
2007), which also has roles in the splicing of transfer RNAs (tRNAs) and in the
formation of 30 ends of mRNAs, suggesting a functional link between these fun-
damental processes of RNA metabolism. Interestingly, both tRNA splicing and
mRNA 30 -end formation occur in the nucleus (Paushkin et al. 2004; Danckwardt
et al. 2008), suggesting that synthetic siRNA duplexes with a 50 OH group and
dephosphorylated siRNA duplexes are transported (or diffuse) into the nucleus, and
after phosphorylation by hClp1, they are exported to the cytoplasm and become
assembled onto RISC.
The production of dsRNA must be carefully controlled to prevent inappropriate
silencing. While amplification of the silencing signal would have obvious benefits
for suppressing the expression of repetitive elements and viruses, this should be
balanced against the danger of amplifying off-target silencing. Although the Slicer-
mediated ping-pong mechanism for piRNA production does not lead to “transitive”
RNAi, but rather to conservative amplification of functional primary piRNA
sequences, any off-target events mediated by RdRPs could, conceivably, lead to a
chain reaction or transitive effect of silencing with deleterious consequences. Thus,
safeguards must exist to prevent pervasive use of RdRPs. One such safeguard is
built into the pathway itself. In C. elegans, the processing of trigger dsRNA and
loading of the primary siRNAs onto the RDE-1 complex appear to be inherently
inefficient to limit the initial round of target recognition by RDE-1 and thus to
18 K. Saito et al.
minimize the risk of amplifying off-target silencing reactions (Yigit et al. 2006). In
addition, each secondary siRNA appears to be made by RdRP in a nonprocessive,
self-termination manner, which restricts transitive effects (Aoki et al. 2007; Pak and
Fire 2007; Sijen et al. 2007). Furthermore, secondary siRNAs associate with worm-
specific Argonautes (SAGOs), which lack catalytic residues for mRNA cleavage,
suggesting that they may be unable to generate cleaved substrates for further
amplification, thereby preventing them from inducing the exponential generation
of secondary siRNAs (Yigit et al. 2006; but see also Aoki et al. 2007). SAGOs are
also present in limited supply and thus provide limited capacity to support multiple
simultaneous silencing reactions.
mRNA quality control will also be important. For instance, in fission yeast, the
RNAi machinery is negatively regulated by the conserved siRNA nuclease, Eri1
(Iida et al. 2006), and links transgene silencing to a protein complex resembling the
Trf4-Air1/Air2-Mtr4 polyadenylation (TRAMP) complex of budding yeast, a
nuclear surveillance mechanism that degrades aberrant transcripts via the exosome
(B€uhler et al. 2007). Thus, RNAi in fission yeast is actively restricted from exerting
its effects throughout the genome and appears to be subject to competition from an
RNA quality control machinery.
mRNA targets (or nearly so) direct endonucleolytic cleavage within the base-paired
region (Tomari and Zamore 2005). However, mRNA decay by partially comple-
mentary miRNAs in animal cells may not occur via endonucleolytic cleavage but
rather by removal of the 30 poly(A) tail from mRNAs (Bagga et al. 2005; Lim et al.
2005; Giraldez et al. 2006). Deadenylation and the consequent loss of poly(A)-
binding protein triggers 50 decapping, thereby exposing the message to the general
mRNA degradation machinery. miRNA-mediated mRNA decay requires an Argo-
naute protein, GW182, the CCR4–CAF1–NOT deadenylase complex, the decap-
ping enzyme DCP2, and several decapping activators (Behm-Ansmant et al. 2006;
Eulalio et al. 2007b). However, the contribution to gene silencing of translational
repression or mRNA degradation appears to differ for each miRNA-target pair and
is likely to depend on the particular set of proteins bound to the 30 UTR of the
mRNA or on proteins that bind to miRNA RISC (miRISC) complexes (Kedde et al.
2007). In some cases, RNA-binding proteins may physically block access to an
miRNA target site. In other cases, RNA-binding proteins may change the subcellu-
lar localization of an mRNA, taking it out of reach of miRNAs. mRNA structure
could also restrict miRISC accessibility to the miRNA target site. Thus, the final
outcome of miRNA regulation is influenced by other proteins interacting with the
target mRNA or with RISC, thereby counteracting the effects of the miRNA, or by
mRNA structure influencing miRISC accessibility, resulting in differential, tissue-
specific regulation (Bhattacharyya and Filipowicz 2007; Brodersen and Voinnet
2009). These findings also predict that if the miRNA target site is close to a binding
site for a strong translational repressor, RISC might compete with the translational
repressor for access to the mRNA, resulting in translational activation by miRNAs
(Brodersen and Voinnet 2009).
Many plant and animal viruses encode suppressor proteins that block host RNA
silencing at various stages of the siRNA and miRNA pathways (Mlotshwa et al.
2008). Some virus suppressors sequester siRNA duplexes and prevent them from
entering RISC. Other suppressors inhibit DCLs or Argonaute activities. Cellular
proteins can also regulate the RNA silencing pathway.
7.1 Processing
a b
Control of miRNA homeostatic control of
biogenesis miRNA biogenesis
pri-miRNA
Drosha DGCR8
cap AAAA
pre-miRNA DGCR8 mRNA
export protein
stabilization
translation
Dicer control of
Lin28 recognizes miRISC activity
miRNA loop
containinig
GGAG Argonaute
Lin28 recruits cap ORF AAAA
Argonaute TUT4
UUUU Terminal
uridylation
Argonaute
cap ORF AAAA
TRIM-NHL
Degradation
miRNA biogenesis is
repressed in stem cells
Fig. 2 Control of the miRNA-mediated silencing pathway. (a) A subset of miRNA biogenesis is
controlled posttranscriptionally. Lin28, mainly expressed in undifferentiated stem cells, recog-
nizes the miRNA loop containing GGAG. Lin28 recruits TUT-4, a noncanonical poly (A)
polymerase, to pre-miRNAs, which adds a uridine tail to the 30 end of pre-miRNAs. Uridylation
prevents processing by Dicer, resulting in degradation of the miRNA. (b) DGCR8 mRNA is
cleaved and downregulated by the Drosha/DGCR8 complex. Drosha is stabilized by DGCR8 via
protein–protein interaction. This feedback circuit may help maintain the homeostatic control of
miRNA production. (c) TRIM-NHL proteins increase miRNA activity by modulating the interac-
tion of miRNP with downstream effectors. The molecular mechanisms for this enhancement are
largely unknown. The expression of TRIM-NHL is regulated during development and differentia-
tion, indicating that miRISC activity might be controlled during development
7.2 Modification
8 Perspective
Biogenesis pathways of small guide RNAs and the ways in which small
RNA–Argonaute complexes regulate gene expression are surprisingly diverse.
The future challenges are obvious. How many new classes of silencing processes
remain to be discovered? How are all of these pathways regulated? Does the
number of Argonaute family genes among species set an upper limit on the number
of classes of small regulatory RNAs that remain to be identified and on the number
of small RNA-guided regulatory processes? Given that many classes of small
RNAs are modified at their ends (Farazi et al. 2008), it is still uncertain whether
current cloning technologies (including ligation of linkers to 30 and 50 ends with
potential PCR amplification bias) are really sampling the entire spectrum of small
RNAs in cells. As the power of our observational tools, such as novel deep-
sequencing technology, increases (Mardis 2008), we need to develop new cloning
methods that are insensitive to specific modifications of small RNAs to reveal the
entire spectrum of small RNA molecules in cells.
It is important to note that small guide RNAs, their precursors, and their target
RNAs are not naked, but rather they are mostly associated with multiple proteins
that regulate many aspects of gene expression. A major challenge for future studies
will be to identify how specific RNA-binding proteins influence the final outcome
of small RNA regulation. For example, genome-wide in vivo approaches, with a
combination of immunoprecipitation methodologies and high-throughput sequenc-
ing, will be required to establish protein–RNA interactions or RNP occupancy at
certain regions of RNA transcripts in vivo (Chi et al. 2009). It is also important to
identify cellular activators and repressors of RNA silencing pathways. Identifica-
tion of such modifiers of the pathways will, in turn, help to identify compounds that
regulate RNAi for potential therapeutic use.
Finally, it is becoming apparent that changes in the activity of RNA silencing
pathways could create quantitative genetic variation in gene expression. Have such
changes contributed to human biology, such as the domestication of Homo sapiens?
There is relatively little correlation between morphological complexity and the
number and diversity of protein-coding genes. However, the number of miRNAs
correlates well with an organism’s total number of neurons (Grimson et al. 2008).
This suggests that miRNA diversity in higher eukaryotes may correlate with an
increase in their relative morphological complexity. Since miRNA seed matches
are often necessary and sufficient for target regulation, a single mutation in a
cellular gene or in the seed sequence of an miRNA gene is presumably enough to
inactivate an miRNA target sequence or to create a new miRNA target site and
might be a rapid way for evolution to fine tune gene expression. Polymorphic
The Key Features of RNA Silencing 23
Acknowledgments We apologize to all colleagues whose relevant primary publications were not
included because of space limitations and the focus on recent results. The authors thank all
members of the Siomi laboratory for their comments and critical reading of the manuscript.
References
Ameres SL, Martinez J, Schroeder R (2007) Molecular basis for target RNA recognition and
cleavage by human RISC. Cell 130:101–112
Aoki K, Moriguchi H, Yoshioka T et al (2007) In vitro analyses of the production and activity of
secondary small interfering RNAs in C. elegans. EMBO J 26:5007–5019
Aravin AA, Sachidanandam R, Girard A et al (2007) Developmentally regulated piRNA clusters
implicate MILI in transposon control. Science 316:744–747
Azuma-Mukai A, Oguri H, Mituyama T et al (2008) Characterization of endogenous human
Argonautes and their miRNA partners in RNA silencing. Proc Natl Acad Sci USA
105:7964–7969
Bagga S, Bracht J, Hunter S et al (2005) Regulation by let-7 and lin-4 miRNAs results in target
mRNA degradation. Cell 122:553–563
Bartel DP (2009) MicroRNAs: target recognition and regulatory functions. Cell 136:215–233
Batista PJ, Ruby JG, Claycomb JM et al (2008) PRG-1 and 21U-RNAs interact to form the piRNA
complex required for fertility in C. elegans. Mol Cell 31:67–78
Behm-Ansmant I, Rehwinkel J, Doerks T et al (2006) mRNA degradation by miRNAs and GW182
requires both CCR4:NOT deadenylase and DCP1:DCP2 decapping complexes. Genes Dev
20:1885–1898
Bhattacharyya SN, Filipowicz W (2007) Argonautes and company: sailing against the wind. Cell
128:1027–1028
Bhattacharyya SN, Habermacher R, Martine U et al (2006) Relief of microRNA-mediated
translational repression in human cells subjected to stress. Cell 125:1111–1124
Brennecke J, Stark A, Russell RB et al (2005) Principles of microRNA-target recognition. PLoS
Biol 3:e85
Brennecke J, Aravin AA, Stark A et al (2007) Discrete small RNA-generating loci as master
regulators of transposon activity in Drosophila. Cell 128:1089–1103
Brennecke J, Malone CD, Aravin AA et al (2008) An epigenetic role for maternally inherited
piRNAs in transposon silencing. Science 322:1387–1392
Brodersen P, Voinnet O (2009) Revisiting the principles of microRNA target recognition and
mode of action. Nat Rev Mol Cell Biol 10:141–148
B€uhler M, Moazed D (2007) Transcription and RNAi in heterochromatic gene silencing. Nat
Struct Mol Biol 14:1041–1048
B€uhler M, Haas W, Gygi SP et al (2007) RNAi-dependent and -independent RNA turnover
mechanisms contribute to heterochromatic gene silencing. Cell 129:707–721
Castanotto D, Sakurai K, Lingeman R et al (2007) Combinatorial delivery of small interfering
RNAs reduces RNAi efficacy by selective incorporation into RISC. Nucleic Acids Res
35:5154–5164
Cerutti H, Casas-Mollano JA (2006) On the origin and functions of RNA-mediated silencing: from
protists to man. Curr Genet 50:81–99
Chapman EJ, Carrington JC (2007) Specialization and evolution of endogenous small RNA
pathways. Nat Rev Genet 8:884–896
24 K. Saito et al.
Chendrimada TP, Gregory RI, Kumaraswamy E et al (2005) TRBP recruits the Dicer complex to
Ago2 for microRNA processing and gene silencing. Nature 436:740–744
Chendrimada TP, Finn KJ, Ji X et al (2007) MicroRNA silencing through RISC recruitment of
eIF6. Nature 447:823–828
Chi SW, Zang JB, Mele A et al (2009) Argonaute HITS-CLIP decodes microRNA–mRNA
interaction maps. Nature 460:479–486
Czech B, Malone CD, Zhou R et al (2008) An endogenous small interfering RNA pathway in
Drosophila. Nature 453:798–802
Danckwardt S, Hentze MW, Kulozik AE (2008) 30 end mRNA processing: molecular mechanisms
and implications for health and disease. EMBO J 27:482–498
Das PP, Bagijn MP, Goldstein LD et al (2008) Piwi and piRNAs act upstream of an endogenous
siRNA pathway to suppress Tc3 transposon mobility in the Caenorhabditis elegans germline.
Mol Cell 31:79–90
Davis BN, Hilyard AC, Lagna G et al (2008) SMAD proteins control DROSHA-mediated
microRNA maturation. Nature 454:56–61
Denli AM, Tops BB, Plasterk RH et al (2004) Processing of primary microRNAs by the
Microprocessor complex. Nature 432:231–235
Diederichs S, Haber DA (2007) Dual role for argonautes in microRNA processing and posttran-
scriptional regulation of microRNA expression. Cell 131:1097–1108
Ding SW, Voinnet O (2007) Antiviral immunity directed by small RNAs. Cell 130:413–426
Dreyfuss G, Kim VN, Kataoka N (2002) Messenger-RNA-binding proteins and the messages they
carry. Nat Rev Mol Cell Biol 3:195–205
ENCODE Project Consortium (2007) Identification and analysis of functional elements in 1% of
the human genome by the ENCODE pilot project. Nature 447:799–816
Eulalio A, Behm-Ansmant I, Izaurralde E (2007a) P bodies: at the crossroads of post-
transcriptional pathways. Nat Rev Mol Cell Biol 8:9–22
Eulalio A, Rehwinkel J, Stricker M et al (2007b) Target-specific requirements for enhancers of
decapping in miRNA-mediated gene silencing. Genes Dev 21:2558–2570
Faehnle CR, Joshua-Tor L (2007) Argonautes confront new small RNAs. Curr Opin Chem Biol
11:569–577
Farazi TA, Juranek SA, Tuschl T (2008) The growing catalog of small RNAs and their association
with distinct Argonaute/Piwi family members. Development 135:1201–1214
Filipowicz W (2005) RNAi: the nuts and bolts of the RISC machine. Cell 122:17–20
Filipowicz W, Bhattacharyya SN, Sonenberg N (2008) Mechanisms of post-transcriptional
regulation by microRNAs: are the answers in sight? Nat Rev Genet 9:102–114
Fire A, Xu S, Montgomery MK et al (1998) Potent and specific genetic interference by double-
stranded RNA in Caenorhabditis elegans. Nature 391:806–811
Flynt A, Liu N, Martin R et al (2009) Dicing of viral replication intermediates during silencing of
latent Drosophila viruses. Proc Natl Acad Sci USA 106:5270–5275
Förstemann K, Tomari Y, Du T et al (2005) Normal microRNA maturation and germ-line stem cell
maintenance requires Loquacious, a double-stranded RNA-binding domain protein. PLoS Biol
3:e236
Franco-Zorrilla JM, Valli A, Todesco M et al (2007) Target mimicry provides a new mechanism
for regulation of microRNA activity. Nat Genet 39:1033–1037
Friedman RC, Farh KK, Burge CB et al (2009) Most mammalian mRNAs are conserved targets of
microRNAs. Genome Res 19:92–105
Giraldez AJ, Mishima Y, Rihel J et al (2006) Zebrafish MiR-430 promotes deadenylation and
clearance of maternal mRNAs. Science 312:75–79
Girard A, Hannon GJ (2008) Conserved themes in small-RNA-mediated transposon control.
Trends Cell Biol 18:136–148
Gregory RI, Yan KP, Amuthan G et al (2004) The Microprocessor complex mediates the genesis
of microRNAs. Nature 432:235–240
The Key Features of RNA Silencing 25
Gregory RI, Chendrimada TP, Cooch N et al (2005) Human RISC couples microRNA biogenesis
and posttranscriptional gene silencing. Cell 123:631–640
Grimson A, Srivastava M, Fahey B et al (2008) Early origins and evolution of microRNAs and
Piwi-interacting RNAs in animals. Nature 455:1193–1197
Gunawardane LS, Saito K, Nishida KM et al (2007) A slicer-mediated mechanism for repeat-
associated siRNA 50 end formation in Drosophila. Science 315:1587–1590
Haley B, Zamore PD (2004) Kinetic analysis of the RNAi enzyme complex. Nat Struct Mol Biol
11:599–606
Hamilton AJ, Baulcombe DC (1999) A species of small antisense RNA in posttranscriptional gene
silencing in plants. Science 286:950–952
Hammell CM, Lubin I, Boag PR et al (2009) nhl-2 Modulates microRNA activity in Caenorhab-
ditis elegans. Cell 136:926–938
Han J, Lee Y, Yeom KH et al (2006) Molecular basis for the recognition of primary microRNAs by
the Drosha-DGCR8 complex. Cell 125:887–901
Han J, Pedersen JS, Kwon SC et al (2009) Posttranscriptional crossregulation between Drosha and
DGCR8. Cell 136:75–84
Hannon GJ (2002) RNA interference. Nature 418:244–251
Harris AN, MacDonald PM (2001) Aubergine encodes a Drosophila polar granule component
required for pole cell formation and related to eIF2C. Development 128:2823–2832
Heo I, Joo C, Kim YK et al (2009) TUT4 in concert with Lin28 suppresses microRNA biogenesis
through pre-microRNA uridylation. Cell 138:696–708
Horwich MD, Li C, Matranga C et al (2007) The Drosophila RNA methyltransferase, DmHen1,
modifies germline piRNAs and single-stranded siRNAs in RISC. Curr Biol 17:1265–1272
Houwing S, Kamminga LM, Berezikov E et al (2007) A role for Piwi and piRNAs in germ cell
maintenance and transposon silencing in Zebrafish. Cell 129:69–82
H€uttenhofer A, Schattner P (2006) The principles of guiding by RNA: chimeric RNA-protein
enzymes. Nat Rev Genet 7:475–482
Hutvagner G, Simard MJ (2008) Argonaute proteins: key players in RNA silencing. Nat Rev Mol
Cell Biol 9:22–32
Iida T, Kawaguchi R, Nakayama J (2006) Conserved ribonuclease, Eri1, negatively regulates
heterochromatin assembly in fission yeast. Curr Biol 16:1459–1464
Jannot G, Boisvert ME, Banville IH et al (2008) Two molecular features contribute to the
Argonaute specificity for the microRNA and RNAi pathways in C. elegans. RNA 14:829–835
Kawamura Y, Saito K, Kin T et al (2008) Drosophila endogenous small RNAs bind to Argonaute 2
in somatic cells. Nature 453:793–797
Kedde M, Strasser MJ, Boldajipour B et al (2007) RNA-binding protein Dnd1 inhibits microRNA
access to target mRNA. Cell 131:1273–1286
Khvorova A, Reynolds A, Jayasena SD (2003) Functional siRNAs and miRNAs exhibit strand
bias. Cell 115:209–216
Kim VN (2005) MicroRNA biogenesis: coordinated cropping and dicing. Nat Rev Mol Cell Biol
6:376–385
Kim VN, Han J, Siomi MC (2009) Biogenesis of small RNAs in animals. Nat Rev Mol Cell Biol
10:126–139
Kiriakidou M, Tan GS, Lamprinaki S et al (2007) An mRNA m7G cap binding-like motif within
human Ago2 represses translation. Cell 129:1141–1151
Kirino Y, Mourelatos Z (2007) Mouse Piwi-interacting RNAs are 20 -O-methylated at their 30
termini. Nat Struct Mol Biol 14:347–348
Kuramochi-Miyagawa S, Watanabe T, Gotoh K et al (2008) DNA methylation of retrotransposon
genes is regulated by Piwi family members MILI and MIWI2 in murine fetal testes. Genes Dev
22:908–917
Landgraf P, Rusu M, Sheridan R et al (2007) A mammalian microRNA expression atlas based on
small RNA library sequencing. Cell 129:1401–1414
26 K. Saito et al.
Lee RC, Feinbaum RL, Ambros V (1993) The C. elegans heterochronic gene lin-4 encodes small
RNAs with antisense complementarity to lin-14. Cell 75:843–854
Lee Y, Hur I, Park SY et al (2006) The role of PACT in the RNA silencing pathway. EMBO J
25:522–532
Lewis BP, Burge CB, Bartel DP (2005) Conserved seed pairing, often flanked by adenosines,
indicates that thousands of human genes are microRNA targets. Cell 120:15–20
Lim LP, Lau NC, Garrett-Engele P et al (2005) Microarray analysis shows that some microRNAs
downregulate large numbers of target mRNAs. Nature 433:769–773
Liu Q, Rand TA, Kalidas S et al (2003) R2D2, a bridge between the initiation and effector steps of
the Drosophila RNAi pathway. Science 301:1921–1925
Liu J, Carmell MA, Rivas FV et al (2004) Argonaute2 is the catalytic engine of mammalian RNAi.
Science 305:1437–1441
Liu Y, Ye X, Jiang F et al (2009) C3PO, an endoribonuclease that promotes RNAi by facilitating
RISC activation. Science 325:750–753
MacRae IJ, Zhou K, Doudna JA (2007) Structural determinants of RNA recognition and cleavage
by Dicer. Nat Struct Mol Biol 14:934–940
Maniataki E, Mourelatos Z (2005) A human, ATP-independent, RISC assembly machine fueled by
pre-miRNA. Genes Dev 19:2979–2990
Mardis ER (2008) The impact of next-generation sequencing technology on genetics. Trends
Genet 24:133–141
Matranga C, Tomari Y, Shin C et al (2005) Passenger-strand cleavage facilitates assembly of
siRNA into Ago2-containing RNAi enzyme complexes. Cell 123:607–620
Megosh HB, Cox DN, Campbell C et al (2006) The role of PIWI and the miRNA machinery in
Drosophila germline determination. Curr Biol 16:1884–1894
Meister G (2008) Molecular biology. RNA interference in the nucleus. Science 321:496–497
Meister G, Tuschl T (2004) Mechanisms of gene silencing by double-stranded RNA. Nature
431:343–349
Meister G, Landthaler M, Patkaniowska A et al (2004) Human Argonaute2 mediates RNA
cleavage targeted by miRNAs and siRNAs. Mol Cell 15:185–197
Mi S, Cai T, Hu Y et al (2008) Sorting of small RNAs into Arabidopsis argonaute complexes is
directed by the 50 terminal nucleotide. Cell 133:116–127
Miyoshi K, Tsukumo H, Nagami T et al (2005) Slicer function of Drosophila Argonautes and its
involvement in RISC formation. Genes Dev 19:2837–2848
Mlotshwa S, Pruss GJ, Vance V (2008) Small RNAs in viral infection and host defense. Trends
Plant Sci 13:375–382
Moazed D (2009) Small RNAs in transcriptional gene silencing and genome defence. Nature
457:413–420
Mochizuki K, Gorovsky MA (2005) A Dicer-like protein in Tetrahymena has distinct functions in
genome rearrangement, chromosome segregation, and meiotic prophase. Genes Dev 19:77–89
Montgomery TA, Howell MD, Cuperus JT et al (2008) Specificity of ARGONAUTE7-miR390
interaction and dual functionality in TAS3 trans-acting siRNA formation. Cell 133:128–141
Morin RD, O’Connor MD, Griffith M et al (2008) Application of massively parallel sequencing to
microRNA profiling and discovery in human embryonic stem cells. Genome Res 18:610–621
Nishida KM, Saito K, Mori T et al (2007) Gene silencing mechanisms mediated by Aubergine
piRNA complexes in Drosophila male gonad. RNA 13:1911–1922
Novina CD, Sharp PA (2004) The RNAi revolution. Nature 430:161–164
Obernosterer G, Leuschner PJ, Alenius M et al (2006) Post-transcriptional regulation of micro-
RNA expression. RNA 12:1161–1167
O’Carroll D, Mecklenbrauker I, Das PP et al (2007) A Slicer-independent role for Argonaute 2 in
hematopoiesis and the microRNA pathway. Genes Dev 21:1999–2004
Ohara T, Sakaguchi Y, Suzuki T et al (2007) The 30 termini of mouse Piwi-interacting RNAs are
20 -O-methylated. Nat Struct Mol Biol 14:349–350
The Key Features of RNA Silencing 27
Okamura K, Ishizuka A, Siomi H et al (2004) Distinct roles for Argonaute proteins in small RNA-
directed RNA cleavage pathways. Genes Dev 18:1655–1666
Okamura K, Hagen JW, Duan H et al (2007) The mirtron pathway generates microRNA-class
regulatory RNAs in Drosophila. Cell 130:89–100
Okamura K, Chung WJ, Ruby JG et al (2008) The Drosophila hairpin RNA pathway generates
endogenous short interfering RNAs. Nature 453:803–806
Pak J, Fire A (2007) Distinct populations of primary and secondary effectors during RNAi in
C. elegans. Science 315:241–244
Pane A, Wehr K, Sch€ upbach T (2007) Zucchini and squash encode two putative nucleases required
for rasiRNA production in the Drosophila germline. Dev Cell 12:851–862
Pasquinelli AE, Reinhart BJ, Slack F et al (2000) Conservation of the sequence and temporal
expression of let-7 heterochronic regulatory RNA. Nature 408:86–89
Paushkin SV, Patel M, Furia BS et al (2004) Identification of a human endonuclease complex
reveals a link between tRNA splicing and pre-mRNA 30 end formation. Cell 117:311–321
Peters L, Meister G (2007) Argonaute proteins: mediators of RNA silencing. Mol Cell 26:611–623
Pham JW, Sontheimer EJ (2005) Molecular requirements for RNA-induced silencing complex
assembly in the Drosophila RNA interference pathway. J Biol Chem 280:39278–39283
Qi HH, Ongusaha PP, Myllyharju J et al (2008) Prolyl 4-hydroxylation regulates Argonaute 2
stability. Nature 455:421–424
Rand TA, Petersen S, Du F et al (2005) Argonaute2 cleaves the anti-guide strand of siRNA during
RISC activation. Cell 123:621–629
Reinhart BJ, Slack FJ, Basson M et al (2000) (2000) The 21-nucleotide let-7 RNA regulates
developmental timing in Caenorhabditis elegans. Nature 403:901–906
Rocheleau CE, Downs WD, Lin R et al (1997) Wnt signaling and an APC-related gene specify
endoderm in early C. elegans embryos. Cell 90:707–716
Ruby JG, Jan C, Player C et al (2006) Large-scale sequencing reveals 21U-RNAs and additional
microRNAs and endogenous siRNAs in C. elegans. Cell 127:1193–1207
Ruby JG, Jan CH, Bartel DP (2007) Intronic microRNA precursors that bypass Drosha processing.
Nature 448:83–86
Rybak A, Fuchs H, Smirnova L et al (2008) A feedback loop comprising lin-28 and let-7 controls
pre-let-7 maturation during neural stem-cell commitment. Nat Cell Biol 10:987–993
Saito K, Ishizuka A, Siomi H et al (2005) Processing of pre-microRNAs by the Dicer-1-
Loquacious complex in Drosophila cells. PLoS Biol 3:e235
Saito K, Nishida KM, Mori T et al (2006) Specific association of Piwi with rasiRNAs derived from
retrotransposon and heterochromatic regions in the Drosophila genome. Genes Dev
20:2214–2222
Saito K, Sakaguchi Y, Suzuki T et al (2007) Pimet, the Drosophila homolog of HEN1, mediates
20 -O-methylation of Piwi- interacting RNAs at their 30 ends. Genes Dev 21:1603–1608
Schwamborn JC, Berezikov E, Knoblich JA (2009) The TRIM-NHL protein TRIM32 activates
microRNAs and prevents self-renewal in mouse neural progenitors. Cell 136:913–925
Schwarz DS, Hutvágner G, Du T et al (2003) Asymmetry in the assembly of the RNAi enzyme
complex. Cell 115:199–208
Seto AG, Kingston RE, Lau NC (2007) The coming of age for Piwi proteins. Mol Cell 26:603–609
Sijen T, Steiner FA, Thijssen KL et al (2007) Secondary siRNAs result from unprimed RNA
synthesis and form a distinct class. Science 315:244–247
Siomi MC, Kuramochi-Miyagawa S (2009) RNA silencing in germlines–exquisite collaboration
of Argonaute proteins with small RNAs for germline survival. Curr Opin Cell Biol 21:426–434
Siomi H, Siomi MC (2008) Interactions between transposable elements and Argonautes have
(probably) been shaping the Drosophila genome throughout evolution. Curr Opin Genet Dev
18:181–187
Siomi H, Siomi MC (2009) On the road to reading the RNA-interference code. Nature
457:396–404
28 K. Saito et al.
Song JJ, Smith SK, Hannon GJ, Joshua-Tor L (2004) Crystal structure of Argonaute and its
implications for RISC slicer activity. Science 305:1434–1437
Stefani G, Slack FJ (2008) Small non-coding RNAs in animal development. Nat Rev Mol Cell Biol
9:219–230
Tabara H, Sarkissian M, Kelly WG et al (1999) The rde-1 gene, RNA interference, and transposon
silencing in C. elegans. Cell 99:123–132
Tabara H, Yigit E, Siomi H et al (2002) The dsRNA binding protein RDE-4 interacts with RDE-1,
DCR-1, and a DExH-box helicase to direct RNAi in C. elegans. Cell 109:861–871
Tam OH, Aravin AA, Stein P et al (2008) Pseudogene-derived small interfering RNAs regulate
gene expression in mouse oocytes. Nature 453:534–538
Thomson JM, Newman M, Parker JS et al (2006) Extensive post-transcriptional regulation of
microRNAs and its implications for cancer. Genes Dev 20:2202–2207
Tomari Y, Zamore PD (2005) Perspective: machines for RNAi. Genes Dev 19:517–529
Tomari Y, Matranga C, Haley B et al (2004) A protein sensor for siRNA asymmetry. Science
306:1377–1380
Tomari Y, Du T, Zamore PD (2007) Sorting of Drosophila small silencing RNAs. Cell
130:299–308
Vagin VV, Sigova A, Li C et al (2006) A distinct small RNA pathway silences selfish genetic
elements in the germline. Science 313:320–324
Viswanathan SR, Daley GQ, Gregory RI (2008) Selective blockade of microRNA processing by
Lin28. Science 320:97–100
Watanabe T, Totoki Y, Toyoda A et al (2008) Endogenous siRNAs from naturally formed dsRNAs
regulate transcripts in mouse oocytes. Nature 453:539–543
Weitzer S, Martinez J (2007) The human RNA kinase hClp1 is active on 30 transfer RNA exons
and short interfering RNAs. Nature 447:222–226
Yigit E, Batista PJ, Bei Y et al (2006) Analysis of the C. elegans Argonaute family reveals that
distinct Argonautes act sequentially during RNAi. Cell 127:747–757
Yin H, Lin H (2007) An epigenetic activation role of Piwi and a Piwi-associated piRNA in
Drosophila melanogaster. Nature 450:304–308
Selected Strategies for the Delivery of siRNA
In Vitro and In Vivo
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2 Mechanism of RNA Interference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3 Naked Delivery of siRNA In Vitro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.1 Cellular Uptake of Naked Nucleic Acids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.2 The Phosphorothioate-Stimulated Cellular Delivery of siRNA . . . . . . . . . . . . . . . . . . . . . . 34
3.3 The siRNA-Peptide Conjugate Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.4 Intracellular Release of siRNA: A Major Hurdle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4 CPP-Mediated siRNA Delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.1 Cell-Penetrating Peptides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.2 Selected Examples of CPP-Mediated siRNA Delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
5 Selected Examples of siRNA Delivery In Vivo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
6 Conclusions and Future Prospects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 29
RNA Technologies, DOI 10.1007/978-3-642-12168-5_2,
# Springer-Verlag Berlin Heidelberg 2010
30 S.D. Laufer et al.
uptake. Moreover, we will present some of the most promising recent approaches
for siRNA delivery in vivo, which may help to pave the road to drugs of the future.
Keywords Argonaute 2 Caveosomal endocytosis pathway Cell-penetrating
peptides Cellular uptake Clinical trials Delivery Endocytosis Endoplasmic
reticulum Endosomal escape Extracellular RNA Golgi apparatus Ilimaquinone
Microinjection Nanoparticles Non-covalent Nonviral delivery systems Oligo-
nucleotide-based drugs Phosphorothioate-stimulated uptake Polycations Protein
transduction domains RNAi Signal peptides siRNAs siRNA-peptide conjugate
Abbreviations
Ago2 Argonaute 2
AMD age-related macular degeneration
CPP cell-penetrating peptide
dsRNA double-stranded RNA
ER endoplasmic reticulum
exNA extracellular nucleic acids
exRNA extracellular RNA
GFP green fluorescent protein
gp41 glycoprotein 41
HA hemagglutinin
HIV human immunodeficiency virus
IL interleukin
IFN interferon
JEV Japanese encephalitis virus
LF2000 Lipofectamine™ 2000
MEND multifunctional envelope-type nano device
miRNA microRNA
NLS nuclear localization sequence
PCI photochemical internalization
PCR polymerase chain reaction
PEG polyethylene glycol
PEI polyethyleneimine
PLL Poly-L-Lysine
PS-ON phosphorothioate-modified oligonucleotides
PTD protein transduction domain
PTGS posttranscriptional gene silencing
R8/R9 oligoarginines
RBD RNA-binding domain
RISC RNA-induced silencing complex
RNAi RNA interference
Selected Strategies for the Delivery of siRNA In Vitro and In Vivo 31
1 Introduction
In recent years, RNA interference (RNAi) has gained a lot of interest as a tool for
functional genomics and probably equally important as a promising therapeutic
approach for the treatment of various diseases (Bumcrot et al. 2006; Castanotto and
Rossi 2009; de Fougerolles et al. 2007). However, despite these bright prospects, a
major impediment to the development of siRNA-based strategies for treatment and
prevention of diseases is the relatively inefficient means to effectively deliver these
macromolecules into the desired target cells or tissues. Although viral vectors have
been widely used to transfer genetic material into cells (Kootstra and Verma 2003;
Verma and Weitzman 2005), they bear an inherent risk for the patient to encounter
severe immunological responses or even develop cancer (Check 2005; Hacein-Bey-
Abina et al. 2003; Raper et al. 2002, 2003). As a result of these problems, much
attention has been paid in recent years to the delivery of naked RNA into a target
organ such as the lung or eye and the development of nonviral delivery systems.
Accumulating experimental evidence suggests that naked oligonucleotide-based
drugs including siRNA may be taken up by specific cell types in cell culture and
in vivo where they exert suppression of their target gene expression. Those findings
warrant more detailed analyses of this mode of delivery. The conception of nonviral
delivery includes an assortment of fairly unrelated approaches yielding various
degrees of enhanced cellular uptake of nucleic acids. Currently, liposomes and
cationic polymers are used as a standard tool to transfect cells in vitro. However,
these procedures are characterized by a significant lack of efficiency accompanied by
a high level of toxicity rendering them mostly inadequate for in vivo applications. In
this context, cell-penetrating peptides (CPPs) represent an interesting alternative as
they generally are less toxic than liposomes or cationic polymers. Moreover, they are
commonly better suited to transfer cargo into different cell types such as nonadherent
cells and primary cells, which are hard to transfect using commercially available
standard protocols. The most advanced approaches in the field are complex carrier
systems combining vantages of assorted strategies to generate nanoparticles with
better defined properties aimed toward enhanced uptake as well as intracellular
trafficking in combination with cell-specific functionalities.
32 S.D. Laufer et al.
In this chapter, we will report about particular aspects of siRNA delivery in vitro
and in vivo, with special emphasis on naked and CPP-mediated cellular delivery of
these macromolecules. Additionally, we will present and briefly discuss selected
recent examples of promising siRNA delivery approaches in vivo.
In tissue cell culture of mammalian cells and in vivo, nucleic acids including single-
stranded RNA and double-stranded DNA can be isolated from the extracellular
Selected Strategies for the Delivery of siRNA In Vitro and In Vivo 33
Fig. 1 Mechanistic principles of RNAi and structure of Thermus thermophilus Argonaute protein.
(a) Details are given in the text. (b) X-ray structure of the ternary complex of T. thermophilus
Argonaute bound to a 21-nucleotide guide DNA and a 20-nucleotide target RNA (pdb file: 3F73).
The protein contains four defined domains, N-terminal, PAZ, Mid, and PIWI, which are color
coded blue, magenta, gold, and green, respectively. Additionally, two linker regions are shown in
grey. Guide DNA is shown in red and target RNA in blue. The coordinated Mg2+ within the active
site (amino acids: D478, D546 and D660) is shown in cyan
environment. In higher mammals, this includes different body fluids such as blood,
serum, and urine.
Extracellular nucleic acids (exNA) were shown to be released from normal cells
and also from tumor cells, which means that one could hypothesize on tumor
34 S.D. Laufer et al.
cell-specific DNA and RNA in blood. In fact, non-invasive methods of early tumor
diagnostics are increasingly based on the analysis of circulating DNA. Even though
RNA is thought to be highly instable in blood, when compared to DNA, it was
found that extracellular RNA (exRNA) circulates in humans at amounts and
integrity that allow isolation, reverse transcription, and quantification by polymer-
ase chain reaction (PCR). The existence of amplifiable tumor-specific RNA in the
plasma of melanoma patients (Kopreski et al. 1999) and breast cancer patients
(Chen et al. 2000) was discovered despite the fact that the activity of blood RNases
is increased in patients with malignancies (Reddi and Holland 1976). Possible
sources of cell-free DNA and RNA are apoptotic bodies resulting from somatic
cell death [summarized by Garcia-Olmo et al. (2000)] and nutrition (Doerfler et al.
2001). Another endogenous source of cell-free nucleic acids to which cells and
organs of mammals are exposed is blood. The circulating blood system contains
significant concentrations of cell-free DNA and RNA (Anker and Stroun 2002; Ng
et al. 2002). Uptake of exNA by individual cells seems to be possible and may be of
biological relevance (Garcia-Olmo et al. 2000).Thus, it is warranted to speculate on
a biological role of exNA, which implies their recognition by cell surface molecules
and it might even include their cellular uptake.
Little is known about the internalization of cell-free nucleic acids by cells.
However, over the past years, Doerfler and colleagues have shown that mice fed
with naked DNA may incorporate this DNA in specific subsets of mononuclear cell
populations in the bloodstream (Doerfler 1995; Schubbert et al. 1994, 1997).
Surprisingly, such DNA is not completely degraded or metabolized, as fragments
of 200–400 bp in length of exogenously introduced DNA could be unequivocally
detected (Schubbert et al. 1994). For single-stranded DNA (ssDNA), much more is
known about the pathways of their cellular uptake (de Diesbach et al. 2000;
Laktionov et al. 1999).
Conversely, almost nothing is known about the conceivable cellular uptake of
short-chain RNA. By co-incubating a mammalian cell culture set-up with various
classes of nucleic acids and short double-stranded DNA competition of uptake was
measured quantitatively (Lehmann and Sczakiel 2005). Firstly, these studies sug-
gest that higher mammalian cells do take up nucleic acids measurably. Secondly,
cells distinguish between DNA and RNA as well as their characteristics regarding
chain length, global structure, and single- versus double-stranded forms. Recent
studies have shown that simple co-incubation of certain human cell types with
naked siRNA at micromolar and sub-micromolar concentrations leads to their
spontaneous cellular uptake within a few hours (Overhoff et al. 2004). Under
certain conditions, this is related to siRNA-specific target suppression indicating
that critical amounts of siRNA are internalized by the exposed cells in a biologi-
cally functional fashion.
Fig. 2 Model of the PS-stimulated cellular uptake of naked siRNA. Upon stimulation of a yet
unknown cell surface molecule by PS-ON, siRNA is taken up via caveolae into caveosomes and
transported to the perinuclearly located smooth ER. Since RNAi is thought to be a cytoplasmic
process, internalized siRNA needs to be released from the perinuclear compartments to be able to
interact with the RISC machinery. This figure was produced using Servier Medical Art
Sczakiel 2005). This means that siRNA and PS-ON are neither complexed nor is
there any measurable co-uptake of the PS-ON by the cells. A schematic depiction of
the underlying model is shown in Fig. 2. Essentially, one hypothesizes that PS-ON
recognize an unknown cell surface molecule that induces a kind of cellular stimu-
lation cascade giving rise to increased apparent uptake of coincubated extracellular
naked siRNA. This process is critically dependent on a number of characteristics
including the chemistry of the stimulating nucleic acid, its chain length, and its
concentration (Fig. 3). More specifically, one hypothesizes that two characteristics
of the stimulating nucleic acid are important for its activity, the phosphorothioate
internucleotide phosphate and a certain structure of the sugar.
The cellular uptake pathway of the cargo, i.e., siRNA, seems to make use of the
caveosomal endocytosis pathway, which is supported by experimental constraints
using pathway-specific activators or inhibitors and by fluorescence microscopy
(Overhoff and Sczakiel 2005). Present experimental data suggest that siRNA
migrates via caveosomes to the smooth endoplasmic reticulum (ER) where it is
trapped and only small amounts of siRNA seem to be released, thereby giving rise
to suppression of target gene expression (Detzer et al. 2009).
Fig. 3 Key characteristics of the PS-stimulated uptake of siRNA by mammalian cells. (a) Only a
fully PS-modified DNA backbone is active. (b) There is a sharp dependency of the amount of
internalized siRNA on the length of the PS-ON. The dotted line represents the detection limit of
the used nuclease protection assay. (c) The PS-stimulated uptake of siRNA reaches a plateau above
a concentration of PS-modified 24 mer of 500 nM
order to become biologically active as a suppressor of target RNA via RNAi. This
includes microscopic studies in the use of fluorescently labeled siRNA after its PS-
stimulated delivery and the discrepancy between large amounts of intracellular
siRNA and surprisingly low effectiveness, i.e., target suppression (Overhoff and
Sczakiel 2005). This view is compatible with the finding that ilimaquinone, a
substance that transiently disrupts the Golgi apparatus and at higher concentrations
also the ER (Takizawa et al. 1993; Wang et al. 1997), is related to increased target
suppression (Fig. 4). In particular, the concentration-dependent disruption of these
two cellular compartments strongly indicates that capturing of siRNA mainly
occurs in the smooth ER (Detzer et al. 2008).
In case of intracellular sorting of proteins, signal peptides serve as promoters of
intracellular transport, a process that may include transmembrane translocation
steps. Similar transport signals on the level of nucleic acids are not known;
however, one might think of a covalent attachment of signal peptides derived
from intracellular protein sorting to siRNA in order to facilitate the intracellular
release and, hence, to enhance the biological activity of siRNA (Fig. 5). For this
reason, the signal peptide TQIENLKEKG, which is thought to facilitate transloca-
tion of the catalytic domains of several bacterial protein toxins from transport
vesicles into the cytoplasm, was used as a tool to be covalently conjugated to
siRNA, thereby bypassing its presumed capturing in the ER (Detzer et al. 2009).
This study showed increased RNAi, i.e., siRNA-mediated target suppression.
Selected Strategies for the Delivery of siRNA In Vitro and In Vivo 37
Fig. 4 The progressive disruption of the ER and the Golgi apparatus by ilimaquinone is related to
increased intracellular release and biological activity of siRNA. This figure was produced using
Servier Medical Art
Fig. 5 Signal peptides steer a transmembrane translocation step of polypeptides (upper panel).
The concept of covalently attaching signal peptides to siRNA in order to enhance its release from
capturing in intracellular compartments is depicted in the lower panel. This figure was produced
using Servier Medical Art
38 S.D. Laufer et al.
In past years, cellular delivery of siRNA was regarded as one of the major technical
problems for the successful application of oligonucleotide-based drugs in therapeu-
tic settings including antisense oligonucleotides and siRNA. Progressively, it
became obvious that physical delivery to mammalian cells could be substantially
improved with regard to the percentage of transfected cells as well as the total
amounts of internalized oligonucleotide-based drugs. However, in many cases,
improved delivery was not reflected by the extent of target suppression. For exam-
ple, limited biological activity of siRNA was observed in the use of a number of
delivery peptides as well as in the use of the PS-stimulated pathway. Hence, it seems
to be reasonable to assume that a block of “functional delivery” exists intracellu-
larly. Further, those findings suggest that intracellular transport and intracellular
release are crucial for the effectiveness of a variety of delivery modes of siRNA
(Fig. 6). A comparison of the mode of delivery of siRNA, the amount of intracellular
The idea of using peptides as carriers for a controlled cellular delivery of siRNA
represents a promising concept to bypass the problem of poor bioavailability and
clinical efficacy of these nucleic acids. Twenty years ago, it was discovered that the
HIV-1 transactivating protein Tat is taken up by mammalian cells (Frankel and
Pabo 1988; Green and Loewenstein 1988), and a few years later, the Antennapedia
homeodomain of Drosophila melanogaster was shown to act similarly (Joliot et al.
1991). Later on, it could be shown that peptides derived from Tat and Antennape-
dia, i.e., Tat48–60 and penetratin, as well as other proteins are capable of transporting
macromolecular cargo molecules into cells (Allinquant et al. 1995; Fawell et al.
1994; Schwarze et al. 1999). Based on such promising results, a rapidly expanding
field focusing on the so-called cell-penetrating peptides (CPPs), also referred to as
protein transduction domains (PTDs), began to develop. Since the first reports about
Tat, a large number of naturally occurring as well as engineered CPPs have been
described (Foged and Nielsen 2008; Heitz et al. 2009; Langel 2006; Lindgren et al.
2000; Morris et al. 2008; Patel et al. 2007; Veldhoen et al. 2008; Zorko and Langel
2005). In addition to Tat and penetratin, well-known examples include transportan,
a chimeric peptide composed of galanin and mastoparan (Pooga et al. 1998), and
oligoarginines (Futaki et al. 2001; Futaki 2006). Generally, CPPs are short poly-
cationic sequences of less than 30 amino acids that are able to translocate different
cargoes (e.g., nucleic acids, peptides, and even entire proteins) into cells. The only
common characteristic of these peptides appears to be that they are net positively
charged at physiological pH. Table 1 gives an overview of selected “classical”
CPPs. In the majority of cases, the cargo is covalently attached to the CPP, which
can be achieved by expression as a fusion construct or by chemical coupling [for a
review see, Zatsepin et al. (2005)]. In particular cases, cargo and carrier bind each
other non-covalently through mainly ionic interactions (Crombez et al. 2008;
Deshayes et al. 2008; Laufer and Restle 2008; Morris et al. 2008). Depending on
the nature of both binding partners, the assembly of nanoparticles may occur.
40 S.D. Laufer et al.
Fig. 7 Principles of peptide-based nucleic acid delivery systems. Interaction of CPP and cargo is
either achieved by covalent attachment or by non-covalent complexation through mainly ionic
interactions. In case of non-covalent complex formation, a further assembly of cargo/carrier
complexes occurs, leading to the formation of large nan-oparticles. In case of covalently joined
molecules, a similar scenario is less likely, yet cannot be excluded. Prior to the translocation
process, the particles attach to the cell surface by ionic interactions of positively charged CPP
residues with negatively charged membrane components. Subsequently, complexes are taken up
via an endocytotic pathway. Although less likely, direct penetration cannot be excluded and may
occur simultaneously. Once inside the cell, the cargo has to escape from vesicular compartments;
otherwise, it eventually gets degraded in the lysosome. This figure was produced using Servier
Medical Art
Simeoni et al. (2003) were the first who non-covalently complexed siRNA with
the peptide MPG. MPG is a 27 amino acid peptide composed of a hydrophobic
domain derived from the N-terminal fusion sequence of the HIV-1 glycoprotein 41
and a hydrophilic domain derived from the nuclear localization sequence (NLS) of
the SV40 large T-antigen, which are linked by a 3 amino acid spacer (Morris et al.
1997). At a 1:10 ratio of negative nucleic acid to positive peptide charges, a
decrease in luciferase activity of about 80% was detectable in HeLa or Cos-7
cells. This effect was further enhanced to about 90% downregulation by a mutation
in the NLS sequence of the carrier peptide (MPGDNLS), presumably due to an
increased delivery to the cytoplasm, where RISC is localized.
In the following, Veldhoen et al. (2006) used a derivative of the MPG peptide for
the delivery of siRNA. This variant, termed MPGa, differs from MPG by five
amino acids in the hydrophobic part. These changes result in an alteration of the
overall structure of the peptide towards a higher tendency of adopting a helical
conformation (Deshayes et al. 2004). MPGa forms highly stable non-covalent
complexes with nucleic acids through ionic interactions of the positively charged
NLS sequence and negative charges of the cargo. Furthermore, hydrophobic pep-
tide/peptide interactions lead to the formation of nanoparticles. Using a luciferase-
targeted siRNA as cargo, reporter gene activity could be inhibited up to 90% with
an IC50 value in the subnanomolar range. Confocal microscopy studies as well as
transfections in the presence of inhibitors of different endocytotic pathways
strongly indicate that endocytosis is involved in the cellular uptake of peptide/
siRNA complexes. As a key issue, the authors quantified the intracellular number of
siRNA molecules after MPGa-mediated transfection and compared it to the amount
of extracellularly applied RNA. Together with data from microinjection experi-
ments (Laufer and Restle 2008), this comparison yields the percentage of inter-
nalized molecules that are biologically active. In the case of MPGa-mediated
siRNA delivery, only 0.1% of internalized oligonucleotides are biologically active
whereas more than 99% are probably retained in endosomes (confer Fig. 6).
Recently, Crombez et al. (2009a) designed a similar secondary amphipathic
peptide, called CADY, which adopts a helical conformation within cell membranes,
exposing cationic arginine residues on one side and aromatic tryptophan groups on
the other. CADY forms stable complexes with siRNAs already at a molar ratio of
5:1–10:1 (peptide:siRNA), whereas for protection from serum nucleases, optimal
cellular uptake and significant target knockdown higher molar ratios (>20:1) are
required. Cellular uptake and the associated biological response were hardly
affected in the presence of different inhibitors of endocytosis; therefore, the authors
concluded that the entry mechanism of CADY/siRNA complexes is independent of
the endosomal pathway.
The same group (Crombez et al. 2009b) shortened the original MPG peptide by
six residues and mutated two residues to tryptophan, yielding a 21 amino acid
peptide called MPG-8. In addition to the cysteamide group at the C-terminus, a
b-alanine was added at the N-terminus to allow further functionalization of the
peptide. Concerning siRNA delivery, the optimal molar ratio was determined
to be 20:1 (peptide: siRNA), and under these conditions, MPG-8 exhibited a
Selected Strategies for the Delivery of siRNA In Vitro and In Vivo 45
significantly higher gene silencing activity than the parent peptide MPGDNLS. In
addition to target downregulation on the mRNA- as well as the protein-level, MPG-
8-mediated delivery of anticyclin B1 siRNA induced G2 arrest and blocked cell
proliferation specifically in cancer cells. Using a xenograft tumor mouse model,
local intratumoral administration but not intravenous injection of 0.25 mg/kg MPG-
8/siRNA particles prevented tumor growth completely. To improve systemic deliv-
ery, MPG-8/siRNA particles were functionalized with a cholesterol moiety through
activation of the N-terminal b-alanine group. This modification increased the
distribution level of anti-cyclin B1 siRNA and blocked tumor growth upon systemic
intravenous administration in a xenograft human prostate as well as human lung
cancer mouse model without activation of the innate immune system.
Similar synergistic effects had already been shown by Kim et al. (2006), who
combined oligoarginine with cholesterol (Chol-R9) for the non-covalent complex-
ation of an anti-VEGF siRNA. Chol-R9/siVEGF complexes suppressed VEGF
production in vitro in CT-26 cells as well as in an in vivo mouse model after
local administration to a subcutaneous tumor. Here, the lowered VEGF level was
accompanied by decreased tumor growth, which was probably due to the anti-
angiogenetic effect on tumor vascularization.
As briefly outlined above, the major rate-limiting step for most delivery
approaches is endosomal entrapment of the nucleic acids. Thus, many groups try
to improve their systems with the aim to increase endosomal escape of siRNA after
peptide-mediated delivery. Lundberg et al. (2007) rationally modified penetratin to
form a CPP (termed EB1) with improved endosomolytic properties. They achieved
a pH-dependent conformational change of the peptide to a higher degree of helicity
by the replacement of two basic amino acids with histidines and the N-terminal
addition of six amino acids. In this study, several CPPs were compared in a non-
covalent approach by measuring the overall cellular uptake via fluorescence and the
biological effect of siRNA targeted to luciferase mRNA. Penetratin- as well as
TP10-mediated transfection did not lead to any silencing of luciferase gene expres-
sion, despite high amounts of intracellular siRNA (Lundberg et al. 2007) in contrast
to previous reports using siRNA–penetratin-conjugates (Davidson et al. 2004) or
TP10/DNA-complexes (El-Andaloussi et al. 2005). EB1-mediated delivery of
100 nM siRNA led to approximately 50% reduction of luciferase activity. This
silencing effect was slightly better than for bPrPp and in the same range as for
MPGDNLS. As it was described earlier that addition of a pH-sensitive peptide
derived from hemagglutinin (HA2) can promote endosomal escape (Wadia et al.
2004), the authors linked HA2 to penetratin (Lundberg et al. 2007). It turned out
that although HA2-penetratin improved the silencing effect when coincubated with
penetratin, EB1 was more potent than this combination of peptides. Together with
confocal microscopy studies, the authors concluded that the lack of biological
effect after penetratin-mediated siRNA delivery is due to a lack of endosomal
escape and that EB1 has a superior endosomolytic activity in comparison to
HA2-penetratin.
Endoh et al. (2007, 2008) and Endoh and Ohtsuki (2009) recently presented
an innovative strategy, called CLIP-RNAi (i.e., CPP-linked RBP-mediated RNA
46 S.D. Laufer et al.
disruption (¼core) (Hu et al. 2009). Furthermore, they assemble into stable and
well-defined complexes with a high payload capacity and can be selectively sur-
face-functionalized to enable cell type-specific targeting. For example, Blackburn
et al. (2009) used the 12 amino acid peptide YSA for the delivery of anti-EGFR
siRNA to ovarian cancer cells via ligand-receptor binding mediated endocytosis.
Polycations, like PEI or PLL alone, can promote significant plasmid DNA
transfer efficiency but show only modest siRNA delivery activity. Therefore,
Meyer et al. (2008, 2009) functionalized these polycations with polyethylene glycol
(PEG) and a pH-responsive endosomolytic melittin peptide from bee venom (Ogris
et al. 2001). To minimize lytic activity in the extracellular environment, melittin
was further modified with dimethylmaleic anhydride (DMMAn), which is cleaved
in the endosome and therefore restores lytic activity in the intracellular compart-
ment. Modification of PEI or PLL with DMMAn-Mel greatly enhanced siRNA-
mediated luciferase gene knockdown (Meyer et al. 2009).
When it comes to in vivo delivery of siRNA, the situation gets much more
complicated than described above on a cellular level. In principle, the nucleic
acid molecules can be administered topically or locally to, for example, the eye,
skin, mucus membranes, and local tumors, or systemically through the blood
stream. Especially in the latter case, besides cellular uptake, there are many more
additional hurdles to consider like serum stability, aggregation with serum proteins,
uptake by phagocytes, and clearance by the kidneys (Alexis et al. 2008; Xie et al.
2006). Moreover, a significant challenge for siRNA delivery to many tissues
represents migration from the bloodstream across the vascular endothelium and
subsequently diffusion through the extracellular matrix, a dense network of poly-
saccharides and fibrous proteins.
There are a variety of techniques described to deliver siRNA in vivo. The
simplest option is the application of naked RNA into a target organ either non-
modified or chemically modified (e.g., 20 -O-methyl modifications). For systemic
delivery, siRNA can be conjugated for example with PEG, cholesterol, or small
peptides or alternatively complexed with peptides, lipids, polymers, polycations, or
even complex nanoparticles, in certain cases, in combination with antibodies or cell
surface-specific ligands for targeted delivery [for a review see, Jeong et al. (2009)].
To cover all of the different approaches described in the literature would be far
beyond the scope of this article. Therefore, we will focus on some of the most
promising examples described in recent years. A summary of these experiments can
be found in Table 3. For a more comprehensive coverage, the reader is referred to
these recent reviews (Aigner 2008; Castanotto and Rossi 2009; Whitehead et al.
2009) and references therein.
The first successful downregulation of a target mRNA by siRNA in mammals
was shown by McCaffrey et al. (2002). In this study, the authors showed that
48 S.D. Laufer et al.
Currently, the development of effective and safe delivery systems for therapeutic
oligonucleotides like siRNA is crucial to one day bring these molecules to the
clinic. Besides the development of viral vectors as delivery vehicles, there is a
highly diverse and constantly increasing number of nonviral systems evolving.
However, at present, even the most advanced systems either lack the efficiencies
required for downstream drug development or do show a substantial degree of
toxicity or both. Of the many factors that limit their use, cellular uptake of the
cargo/carrier complexes and particularly subsequent intracellular trafficking to
reach the target site are the most important. Moreover, for in vivo use, various
additional obstacles are to be taken into account like serum stability, pharmacoki-
netic considerations, and tissue barriers as well as target cell specificity. In spite of
these somewhat sobering insights, there is noticeable progress especially in recent
years. While most of the underlying problems are meanwhile identified, the
answers to these problems remain challenging. Most likely, there will be no
magic bullet but individual solutions for any given application.
Selected Strategies for the Delivery of siRNA In Vitro and In Vivo 51
Acknowledgments We apologize to those authors whose work was not cited directly owing to
space limitations. T.R. acknowledges funding by EC-grant LSHG-CT-2003-503480.
References
Aigner A (2008) Cellular delivery in vivo of siRNA-based therapeutics. Curr Pharm Des
14:3603–3619
Alexis F, Pridgen E, Molnar LK et al (2008) Factors affecting the clearance and biodistribution of
polymeric nanoparticles. Mol Pharm 5:505–515
Allinquant B, Hantraye P, Mailleux P et al (1995) Downregulation of amyloid precursor protein
inhibits neurite outgrowth in vitro. J Cell Biol 128:919–927
Anker P, Stroun M (2002) Progress in the knowledge of circulating nucleic acids: plasma RNA is
particle-associated. Can it become a general detection marker for a cancer blood test? Clin
Chem 48:1210–1211
Aouadi M, Tesz GJ, Nicoloro SM et al (2009) Orally delivered siRNA targeting macrophage
Map4k4 suppresses systemic inflammation. Nature 458:1180–1184
Bayele HK, Sakthivel T, O’Donell M et al (2005) Versatile peptide dendrimers for nucleic acid
delivery. J Pharm Sci 94:446–457
Bayele HK, Ramaswamy C, Wilderspin AF et al (2006) Protein transduction by lipidic peptide
dendrimers. J Pharm Sci 95:1227–1237
Berg K, Folini M, Prasmickaite L et al (2007) Photochemical internalization: a new tool for drug
delivery. Curr Pharm Biotechnol 8:362–372
Bernstein E, Caudy AA, Hammond SM et al (2001) Role for a bidentate ribonuclease in the
initiation step of RNA interference. Nature 409:363–366
Bitko V, Musiyenko A, Shulyayeva O et al (2005) Inhibition of respiratory viruses by nasally
administered siRNA. Nat Med 11:50–55
Blackburn WH, Dickerson EB, Smith MH et al (2009) Peptide-functionalized nanogels for
targeted siRNA delivery. Bioconjug Chem 20(5):960–968
Bøe S, Longva AS, Hovig E (2007) Photochemically induced gene silencing using small inter-
fering RNA molecules in combination with lipid carriers. Oligonucleotides 17:166–173
Bonsted A, Wagner E, Prasmickaite L et al (2008) Photochemical enhancement of DNA delivery
by EGF receptor targeted polyplexes. Methods Mol Biol 434:171–181
Bumcrot D, Manoharan M, Koteliansky V et al (2006) RNAi therapeutics: a potential new class of
pharmaceutical drugs. Nat Chem Biol 2:711–719
Castanotto D, Rossi JJ (2009) The promises and pitfalls of RNA-interference-based therapeutics.
Nature 457:426–433
Check E (2005) Gene therapy put on hold as third child develops cancer. Nature 433:561
Chen XQ, Bonnefoi H, Pelte MF et al (2000) Telomerase RNA as a detection marker in the serum
of breast cancer patients. Clin Cancer Res 6:3823–3826
Chen QR, Zhang L, Stass SA et al (2001) Branched co-polymers of histidine and lysine are
efficient carriers of plasmids. Nucleic Acids Res 29:1334–1340
Chiu YL, Ali A, Chu CY et al (2004) Visualizing a correlation between siRNA localization,
cellular uptake, and RNAi in living cells. Chem Biol 11:1165–1175
Console S, Marty C, Garcia-Echeverria C et al (2003) Antennapedia and HIV transactivator of
transcription (TAT) “protein transduction domains” promote endocytosis of high molecular
weight cargo upon binding to cell surface glycosaminoglycans. J Biol Chem 278:35109–35114
Crombez L, Morris MC, Deshayes S et al (2008) Peptide-based nanoparticle for ex vivo and
in vivo drug delivery. Curr Pharm Des 14:3656–3665
Crombez L, Aldrian-Herrada G, Konate K et al (2009a) A new potent secondary amphipathic cell-
penetrating peptide for siRNA delivery into mammalian cells. Mol Ther 17:95–103
52 S.D. Laufer et al.
Fattal E, Barratt G (2009) Nanotechnologies and controlled release systems for the delivery of
antisense oligonucleotides and small interfering RNA. Br J Pharmacol 157:179–194
Fawell S, Seery J, Daikh Y et al (1994) Tat-mediated delivery of heterologous proteins into cells.
Proc Natl Acad Sci USA 91:664–668
Filleur S, Courtin A, Ait-Si-Ali S et al (2003) SiRNA-mediated inhibition of vascular endothelial
growth factor severely limits tumor resistance to antiangiogenic thrombospondin-1 and slows
tumor vascularization and growth. Cancer Res 63:3919–3922
Fire A, Xu S, Montgomery MK et al (1998) Potent and specific genetic interference by double-
stranded RNA in Caenorhabditis elegans. Nature 391:806–811
Fittipaldi A, Ferrari A, Zoppe M et al (2003) Cell membrane lipid rafts mediate caveolar
endocytosis of HIV-1 Tat fusion proteins. J Biol Chem 278:34141–34149
Foged C, Nielsen HM (2008) Cell-penetrating peptides for drug delivery across membrane
barriers. Expert Opin Drug Deliv 5:105–117
Folini M, Bandiera R, Millo E et al (2007) Photochemically enhanced delivery of a cell-
penetrating peptide nucleic acid conjugate targeting human telomerase reverse transcriptase:
effects on telomere status and proliferative potential of human prostate cancer cells. Cell Prolif
40:905–920
Frankel AD, Pabo CO (1988) Cellular uptake of the tat protein from human immunodeficiency
virus. Cell 55:1189–1193
Futaki S, Suzuki T, Ohashi W et al (2001) Arginine-rich peptides. An abundant source of
membrane-permeable peptides having potential as carriers for intracellular protein delivery.
J Biol Chem 276:5836–5840
Futaki S, Masui Y, Nakase I et al (2005) Unique features of a pH-sensitive fusogenic peptide that
improves the transfection efficiency of cationic liposomes. J Gene Med 7:1450–1458
Futaki S (2006) Oligoarginine vectors for intracellular delivery: design and cellular-uptake
mechanisms. Biopolymers 84:241–249
Garcia-Olmo D, Garcia-Olmo DC, Ontanon J et al (2000) Horizontal transfer of DNA and the
“genometastasis hypothesis”. Blood 95:724–725
Green M, Loewenstein PM (1988) Autonomous functional domains of chemically synthesized
human immunodeficiency virus tat trans-activator protein. Cell 55:1179–1188
Grimm D, Streetz KL, Jopling CL et al (2006) Fatality in mice due to oversaturation of cellular
microRNA/short hairpin RNA pathways. Nature 441:537–541
Hacein-Bey-Abina S, Von Kalle C, Schmidt M et al (2003) LMO2-associated clonal T cell
proliferation in two patients after gene therapy for SCID-X1. Science 302:415–419
Haley B, Zamore PD (2004) Kinetic analysis of the RNAi enzyme complex. Nat Struct Mol Biol
11:599–606
Hammond SM, Boettcher S, Caudy AA et al (2001) Argonaute2, a link between genetic and
biochemical analyses of RNAi. Science 293:1146–1150
Haque ME, Koppaka V, Axelsen PH et al (2005) Properties and structures of the influenza and HIV
fusion peptides on lipid membranes: implications for a role in fusion. Biophys J 89:3183–3194
Heitz F, Morris MC, Divita G (2009) Twenty years of cell-penetrating peptides: from molecular
mechanisms to therapeutics. Br J Pharmacol 157:195–206
Hu Y, Atukorale PU, Lu JJ et al (2009) Cytosolic delivery mediated via electrostatic surface
binding of protein, virus, or siRNA cargos to pH-responsive core-shell gel particles. Bioma-
cromolecules 10:756–765
Hutvagner G, Simard MJ (2008) Argonaute proteins: key players in RNA silencing. Nat Rev Mol
Cell Biol 9:22–32
Hyndman L, Lemoine JL, Huang L et al (2004) HIV-1 Tat protein transduction domain peptide
facilitates gene transfer in combination with cationic liposomes. J Control Release 99:435–444
Jeong JH, Mok H, Oh YK et al (2009) siRNA conjugate delivery systems. Bioconjug Chem
20:5–14
Jinek M, Doudna JA (2009) A three-dimensional view of the molecular machinery of RNA
interference. Nature 457:405–412
54 S.D. Laufer et al.
Johnson LN, Cashman SM, Kumar-Singh R (2008) Cell-penetrating peptide for enhanced delivery
of nucleic acids and drugs to ocular tissues including retina and cornea. Mol Ther 16:107–114
Joliot AH, Triller A, Volovitch M et al (1991) Alpha-2, 8-Polysialic acid is the neuronal surface
receptor of antennapedia homeobox peptide. New Biol 3:1121–1134
Kale AA, Torchilin VP (2007a) “Smart” drug carriers: PEGylated TATp-modified pH-sensitive
liposomes. J Liposome Res 17:197–203
Kale AA, Torchilin VP (2007b) Design, synthesis, and characterization of pH-sensitive PEG-PE
conjugates for stimuli-sensitive pharmaceutical nanocarriers: the effect of substitutes at the
hydrazone linkage on the ph stability of PEG-PE conjugates. Bioconjug Chem 18:363–370
Kang H, Delong R, Fisher MH et al (2005) Tat-conjugated PAMAM dendrimers as delivery agents
for antisense and siRNA oligonucleotides. Pharm Res 22:2099–2106
Khalil IA, Kogure K, Futaki S et al (2007) Octaarginine-modified multifunctional envelope-type
nanoparticles for gene delivery. Gene Ther 14:682–689
Kilk K, El-Andaloussi S, J€arver P et al (2005) Evaluation of transportan 10 in PEI mediated
plasmid delivery assay. J Control Release 103:511–523
Kim WJ, Christensen LV, Jo S et al (2006) Cholesteryl oligoarginine delivering vascular endothe-
lial growth factor siRNA effectively inhibits tumor growth in colon adenocarcinoma. Mol Ther
14:343–350
Kleinman ME, Yamada K, Takeda A et al (2008) Sequence- and target-independent angiogenesis
suppression by siRNA via TLR3. Nature 452:591–597
Kogure K, Akita H, Yamada Y et al (2008) Multifunctional envelope-type nano device (MEND) as
a non-viral gene delivery system. Adv Drug Deliv Rev 60:559–571
Kootstra NA, Verma IM (2003) Gene therapy with viral vectors. Annu Rev Pharmacol Toxicol
43:413–439
Kopreski MS, Benko FA, Kwak LW et al (1999) Detection of tumor messenger RNA in the serum
of patients with malignant melanoma. Clin Cancer Res 5:1961–1965
Kumar P, Wu H, McBride JL et al (2007) Transvascular delivery of small interfering RNA to the
central nervous system. Nature 448:39–43
Kwon EJ, Bergen JM, Pun SH (2008) Application of an HIV gp41-derived peptide for enhanced
intracellular trafficking of synthetic gene and siRNA delivery vehicles. Bioconjug Chem
19:920–927
Laktionov PP, Dazard JE, Vives E et al (1999) Characterisation of membrane oligonucleotide-
binding proteins and oligonucleotide uptake in keratinocytes. Nucleic Acids Res 27:
2315–2324
Langel U € (ed) (2006) Handbook of cell-penetrating peptides. CRC Press, Boca Raton
Laufer SD, Restle T (2008) Peptide-mediated cellular delivery of oligonucleotide-based therapeu-
tics in vitro: quantitative evaluation of overall efficacy employing easy to handle reporter
systems. Curr Pharm Des 14:3637–3655
Lehmann MJ, Sczakiel G (2005) Spontaneous uptake of biologically active recombinant DNA by
mammalian cells via a selected DNA segment. Gene Ther 12:446–451
Leng Q, Scaria P, Zhu J et al (2005) Highly branched HK peptides are effective carriers of siRNA.
J Gene Med 7:977–986
Leung RK, Whittaker PA (2005) RNA interference: from gene silencing to gene-specific thera-
peutics. Pharmacol Ther 107:222–239
Lindgren M, H€allbrink M, Prochiantz A et al (2000) Cell-penetrating peptides. Trends Pharmacol
Sci 21:99–103
Liu Z, Li M, Cui D et al (2005) Macro-branched cell-penetrating peptide design for gene delivery.
J Control Release 102:699–710
Lundberg M, Johansson M (2001) Is VP22 nuclear homing an artifact? Nat Biotechnol 19:713–714
Lundberg M, Johansson M (2002) Positively charged DNA-binding proteins cause apparent cell
membrane translocation. Biochem Biophys Res Commun 291:367–371
Lundberg P, El-Andaloussi S, Sutlu T et al (2007) Delivery of short interfering RNA using
endosomolytic cell-penetrating peptides. FASEB J 21:2664–2671
Selected Strategies for the Delivery of siRNA In Vitro and In Vivo 55
Maiolo JR, Ferrer M, Ottinger EA (2005) Effects of cargo molecules on the cellular uptake of
arginine-rich cell-penetrating peptides. Biochim Biophys Acta 1712:161–172
Mano M, Teodosio C, Paiva A et al (2005) On the mechanisms of the internalization of S4(13)-PV
cell-penetrating peptide. Biochem J 390:603–612
McCaffrey AP, Meuse L, Pham TT et al (2002) RNA interference in adult mice. Nature 418:38–39
McNamara JO, Andrechek ER, Wang Y et al (2006) Cell type-specific delivery of siRNAs with
aptamer-siRNA chimeras. Nat Biotechnol 24:1005–1015
Meade BR, Dowdy SF (2008) Enhancing the cellular uptake of siRNA duplexes following
noncovalent packaging with protein transduction domain peptides. Adv Drug Deliv Rev
60:530–536
Mescalchin A, Detzer A, Wecke M et al (2007) Cellular uptake and intracellular release are major
obstacles to the therapeutic application of siRNA: novel options by phosphorothioate-stimu-
lated delivery. Expert Opin Biol Ther 7:1531–1538
Meyer M, Philipp A, Oskuee R et al (2008) Breathing life into polycations: functionalization with
pH-responsive endosomolytic peptides and polyethylene glycol enables siRNA delivery. J Am
Chem Soc 130:3272–3273
Meyer M, Dohmen C, Philipp A et al (2009) Synthesis and biological evaluation of a biorespon-
sive and endosomolytic siRNA-polymer conjugate. Mol Pharm 6:752–762
Michiue H, Tomizawa K, Wei FY et al (2005) The NH2 terminus of influenza virus hemaggluti-
nin-2 subunit peptides enhances the antitumor potency of polyarginine-mediated p53 protein
transduction. J Biol Chem 280:8285–8289
Midoux P, Monsigny M (1999) Efficient gene transfer by histidylated polylysine/pDNA com-
plexes. Bioconjug Chem 10:406–411
Morris MC, Vidal P, Chaloin L et al (1997) A new peptide vector for efficient delivery of
oligonucleotides into mammalian cells. Nucleic Acids Res 25:2730–2736
Morris MC, Deshayes S, Heitz F et al (2008) Cell-penetrating peptides: from molecular mechan-
isms to therapeutics. Biol Cell 100:201–217
Moschos SA, Jones SW, Perry MM et al (2007) Lung delivery studies using siRNA conjugated to
TAT(48–60) and penetratin reveal peptide induced reduction in gene expression and induction
of innate immunity. Bioconjug Chem 18:1450–1459
Muratovska A, Eccles MR (2004) Conjugate for efficient delivery of short interfering RNA
(siRNA) into mammalian cells. FEBS Lett 558:63–68
Nakamura Y, Kogure K, Futaki S et al (2007) Octaarginine-modified multifunctional envelope-
type nano device for siRNA. J Control Release 119:360–367
Ng EK, Tsui NB, Lam NY et al (2002) Presence of filterable and nonfilterable mRNA in the
plasma of cancer patients and healthy individuals. Clin Chem 48:1212–1217
Oehlke J, Scheller A, Wiesner B et al (1998) Cellular uptake of an alpha-helical amphipathic
model peptide with the potential to deliver polar compounds into the cell interior non-
endocytically. Biochim Biophys Acta 1414:127–139
Ogris M, Carlisle RC, Bettinger T et al (2001) Melittin enables efficient vesicular escape
and enhanced nuclear access of nonviral gene delivery vectors. J Biol Chem 276:
47550–47555
Oliveira S, Fretz MM, Hogset A et al (2007a) Photochemical internalization enhances silencing of
epidermal growth factor receptor through improved endosomal escape of siRNA. Biochim
Biophys Acta 1768:1211–1217
Oliveira S, van Rooy I, Kranenburg O et al (2007b) Fusogenic peptides enhance endosomal escape
improving siRNA-induced silencing of oncogenes. Int J Pharm 331:211–214
Oliveira S, Hogset A, Storm G et al (2008) Delivery of siRNA to the target cell cytoplasm:
photochemical internalization facilitates endosomal escape and improves silencing efficiency,
in vitro and in vivo. Curr Pharm Des 14:3686–3697
Overhoff M, W€unsche W, Sczakiel G (2004) Quantitative detection of siRNA and single-stranded
oligonucleotides: relationship between uptake and biological activity of siRNA. Nucleic Acids
Res 32:e170
56 S.D. Laufer et al.
Schubbert R, Renz D, Schmitz B et al (1997) Foreign (M13) DNA ingested by mice reaches
peripheral leukocytes, spleen, and liver via the intestinal wall mucosa and can be covalently
linked to mouse DNA. Proc Natl Acad Sci USA 94:961–966
Schwarze SR, Ho A, Vocero-Akbani A et al (1999) In vivo protein transduction: delivery of a
biologically active protein into the mouse. Science 285:1569–1572
Schwarze SR, Dowdy SF (2000) In vivo protein transduction: intracellular delivery of biologically
active proteins, compounds and DNA. Trends Pharmacol Sci 21:45–48
Simeoni F, Morris MC, Heitz F et al (2003) Insight into the mechanism of the peptide-based gene
delivery system MPG: implications for delivery of siRNA into mammalian cells. Nucleic
Acids Res 31:2717–2724
Snøve O Jr, Rossi JJ (2006) Toxicity in mice expressing short hairpin RNAs gives new insight into
RNAi. Genome Biol 7:231
Song E, Lee SK, Wang J et al (2003) RNA interference targeting Fas protects mice from fulminant
hepatitis. Nat Med 9:347–351
Song E, Zhu P, Lee SK et al (2005) Antibody mediated in vivo delivery of small interfering RNAs
via cell-surface receptors. Nat Biotechnol 23:709–717
Song JJ, Smith SK, Hannon GJ et al (2004) Crystal structure of Argonaute and its implications for
RISC slicer activity. Science 305:1434–1437
Soomets U, Lindgren M, Gallet X et al (2000) Deletion analogues of transportan. Biochim
Biophys Acta 1467:165–176
Sorensen DR, Leirdal M, Sioud M (2003) Gene silencing by systemic delivery of synthetic
siRNAs in adult mice. J Mol Biol 327:761–766
Soundara Manickam D, Oupický D (2006) Multiblock reducible copolypeptides containing
histidine-rich and nuclear localization sequences for gene delivery. Bioconjug Chem 17:
1395–1403
Soutschek J, Akinc A, Bramlage B et al (2004) Therapeutic silencing of an endogenous gene by
systemic administration of modified siRNAs. Nature 432:173–178
Takizawa PA, Yucel JK, Veit B et al (1993) Complete vesiculation of Golgi membranes and
inhibition of protein transport by a novel sea sponge metabolite, ilimaquinone. Cell 73:1079–1090
Tönges L, Lingor P, Egle R et al (2006) Stearylated octaarginine and artificial virus-like particles
for transfection of siRNA into primary rat neurons. RNA 12:1431–1438
Torchilin VP, Rammohan R, Weissig V et al (2001) TAT peptide on the surface of liposomes
affords their efficient intracellular delivery even at low temperature and in the presence of
metabolic inhibitors. Proc Natl Acad Sci USA 98:8786–8791
Torchilin VP, Levchenko TS, Rammohan R et al (2003) Cell transfection in vitro and in vivo with
nontoxic TAT peptide-liposome-DNA complexes. Proc Natl Acad Sci USA 100:1972–1977
Tu Y, Kim JS (2008) A fusogenic segment of glycoprotein H from herpes simplex virus enhances
transfection efficiency of cationic liposomes. J Gene Med 10:646–654
Tuerk C, Gold L (1990) Systematic evolution of ligands by exponential enrichment: RNA ligands
to bacteriophage T4 DNA polymerase. Science 249:505–510
Turner JJ, Arzumanov AA, Gait MJ (2005) Synthesis, cellular uptake and HIV-1 Tat-dependent
trans-activation inhibition activity of oligonucleotide analogues disulphide-conjugated to cell-
penetrating peptides. Nucleic Acids Res 33:27–42
Tyagi M, Rusnati M, Presta M et al (2001) Internalization of HIV-1 tat requires cell surface
heparan sulfate proteoglycans. J Biol Chem 276:3254–3261
Veldhoen S, Laufer SD, Trampe A et al (2006) Cellular delivery of small interfering RNA by a
non-covalently attached cell-penetrating peptide: quantitative analysis of uptake and biological
effect. Nucleic Acids Res 34:6561–6573
Veldhoen S, Laufer SD, Restle T (2008) Recent developments in peptide-based nucleic acid
delivery. Int J Mol Sci 9:1276–1320
Verma IM, Weitzman MD (2005) Gene therapy: twenty-first century medicine. Annu Rev
Biochem 74:711–738
58 S.D. Laufer et al.
Vives E, Brodin P, Lebleu B (1997) A truncated HIV-1 Tat protein basic domain rapidly
translocates through the plasma membrane and accumulates in the cell nucleus. J Biol Chem
272:16010–16017
Wadia JS, Stan RV, Dowdy SF (2004) Transducible TAT-HA fusogenic peptide enhances escape
of TAT-fusion proteins after lipid raft macropinocytosis. Nat Med 10:310–315
Wang HJ, Benlimame N, Nabi I (1997) The AMF-R tubule is a smooth ilimaquinone-sensitive
subdomain of the endoplasmic reticulum. J Cell Sci 110(Pt 24):3043–3053
Wang Y, Sheng G, Juranek S et al (2008a) Structure of the guide-strand-containing argonaute
silencing complex. Nature 456:209–213
Wang Y, Juranek S, Li H et al (2008b) Structure of an argonaute silencing complex with a seed-
containing guide DNA and target RNA duplex. Nature 456:921–926
Wesche-Soldato DE, Chung CS, Lomas-Neira J et al (2005) In vivo delivery of caspase-8 or Fas
siRNA improves the survival of septic mice. Blood 106:2295–2301
Whitehead KA, Langer R, Anderson DG (2009) Knocking down barriers: advances in siRNA
delivery. Nat Rev Drug Discov 8:129–138
Xie FY, Woodle MC, Lu PY (2006) Harnessing in vivo siRNA delivery for drug discovery and
therapeutic development. Drug Discov Today 11:67–73
Yu JY, DeRuiter SL, Turner DL (2002) RNA interference by expression of short-interfering RNAs
and hairpin RNAs in mammalian cells. Proc Natl Acad Sci USA 99:6047–6052
Zatsepin TS, Turner JJ, Oretskaya TS et al (2005) Conjugates of oligonucleotides and analogues
with cell penetrating peptides as gene silencing agents. Curr Pharm Des 11:3639–3654
Zimmermann TS, Lee AC, Akinc A et al (2006) RNAi-mediated gene silencing in non-human
primates. Nature 441:111–114
Zorko M, Langel U € (2005) Cell-penetrating peptides: mechanism and kinetics of cargo delivery.
Adv Drug Deliv Rev 57:529–545
RNAi Suppression and Its Application
Contents
1 RNA Interference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
2 RNAi-Directed Viral Immunity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3 Identification and Function Characterization of Viral RNAi Suppressors . . . . . . . . . . . . . . . . . . 68
3.1 Agrobacterium-Mediated Transient Suppression Assay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.2 Reversal of Transgene-Induced Gene Silencing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.3 Grafting Assay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.4 Replication Rescue of Mutant Viruses Defective in RNAi Suppression . . . . . . . . . . . . . . 70
4 Function Mechanism of Viral RNAi Suppressors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.1 Viral Suppressors that Bind Viral dsRNA to Interfere with viRNA Biogenesis . . . . . . 73
4.2 Viral Suppressors that Target Virus-Derived siRNA for RNAi Suppression . . . . . . . . . 75
4.3 Viral Suppressors that Target RNAi Effectors for Suppression . . . . . . . . . . . . . . . . . . . . . . . 77
4.4 Viral Suppressors that Suppress Systemic RNAi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5 RNAi Suppressors of Nonviral Origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
5.1 Suppression of RNAi by Bacterial Pathogens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
5.2 Cellular RNAi Suppressors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
6 Biotechnological Application of RNAi Suppressors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
6.1 Enhance Gene Expression for Rapid Function Analysis and Mass
Production of Valuable Protein . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
6.2 Serve as Molecular Tools to Probe Various RNAi-Directed Functions . . . . . . . . . . . . . . . 84
Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
X. Yi and R. Lu (*)
Department of Biological Sciences, Louisiana State University, Baton Rouge, LA 70803, USA
e-mail: ruilu@lsu.edu
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 59
RNA Technologies, DOI 10.1007/978-3-642-12168-5_3,
# Springer-Verlag Berlin Heidelberg 2010
60 X. Yi and R. Lu
Argonaut (AGO) proteins. These small RNAs mediate silencing of target genes
with complementary sequence at transcriptional or posttranscriptional level
thereby to control a wide variety of biological functions. In worms and plants,
RNA-dependent RNA polymerases (RdRPs) amplify RNAi by converting AGO
cleavage products into dsRNAs for the generation of secondary siRNAs. One of the
well characterized functions of RNAi is antiviral, which has been shown to serve as
major viral innate immunity in fungi, plants, and invertebrates. Typically, RNAi-
directed viral immunity (RDVI) is initiated with Dicer processing of viral dsRNAs,
usually the replication intermediates, into siRNAs. These virus-derived
siRNAs (viRNA) will then be used as sequence guide for target viral RNA
destruction. Host-encoded miRNAs also contribute to viral control in mammal or
bacterial control in plant. As a counterdefensive mechanism, many viruses and
some bacteria are found to encode RNAi suppressors, previously known as patho-
genicity factors. These RNAi antagonists target key components of RNAi for
suppression, which eventually leads to defects in viRNA biogenesis or function.
Since transgene expression in plants and invertebrates is often targeted by RNAi for
suppression but can be reversed by various RNAi suppressors, codelivery of a VSR
has been used to facilitate the isolation and biochemical characterization of a broad
range of proteins of interests. RNAi suppressors can also be used as genetic tool for
the study of biological functions controlled by certain class of endogenous sRNA
(siRNA or miRNA). This is because, when expressed as transgenes, some RNAi
suppressors can specifically target and interfere with the biogenesis or function of
certain class of endogenous sRNA but not the other.
Abbreviations
1 RNA Interference
Potent gene silencing triggered by double-stranded RNA (dsRNA) was first demon-
strated in an experimental setup in which delivery of artificial dsRNA into
nematode worm Caenorhabditis elegans (C. elegans) by microinjection initiated
62 X. Yi and R. Lu
Fig. 2 Schematic
representation of dsRNA-
triggered RNAi. Dicer, Type
III ribonuclease that
processes double-stranded
RNA (dsRNA) into small
interfering RNA (siRNA).
AGO, Argonaut protein that
cleaves target transcript
complementary to siRNA.
RdRP, RNA-dependent RNA
polymerase found in plants
and worms, which amplifies
RNAi by converting AGO
cleavage product into
secondary dsRNA
commonly called “RNase III domain”. Cleavage by RNase III enzymes produces a
characteristic dsRNA structure consisting of a 50 phosphate group and a two base
overhang at the 30 end. Dicer enzymes typically contain two ribonuclease domains,
a dsRNA binding domain, and an N-terminal DExD/H-box helicase domain fol-
lowed by a small domain of unknown function (DUF283) and a PAZ domain. Dicer
specifically recognizes and processes long dsRNA molecules into siRNA duplexes
of discrete size in a sequence-independent manner (Bernstein et al. 2001). Crystal-
lographic study suggests that Dicer functions as a molecular ruler that binds to the
ends of dsRNA through the PAZ domain and cleaves a set distance away (Song
et al. 2004; Macrae et al. 2006).
During RNAi-mediated gene silencing, one strand of the siRNA duplex, usually
the strand with a 50 end which is less stable in base pairing, is incorporated into an
RNA-induced silencing complex (RISC) and provides the sequence specificity
through Watson–Crick base pairing for target RNA destruction or translational
arrest. An Argonaut (AGO) protein can be found in all RISCs and its role is to
bind the small RNA and position it in a conformation that facilitates cognate target
recognition (Fig. 2). AGO proteins are ubiquitous in eukaryotes and some archaea.
The AGO protein superfamily falls into two major clades: the AGO clade found in
fungi, plants, and animals and the PIWI clade that are found only in animals so far.
The number of AGO genes found in different species varies from one [as in the
fission yeast Schizosaccharomyces pombe (S. pombe)] to over two dozens (27 in
C. elegans) (Volpe et al. 2002; Yigit et al. 2006). Although, in some cases, multiple
copies of an AGO gene function redundantly, it is common that AGO genes in an
organism have specialized and nonoverlapping functions. Typically, a PAZ domain
64 X. Yi and R. Lu
at the N-terminal and a PIWI domain at the C-terminal can be found for an AGO
protein. The PAZ domains function to bind the 30 end of the small guide RNA,
whereas some PIWI domain adopts a RNase H-like fold and can hydrolyze target
RNAs using an RNase H-like mechanism (Song et al. 2004). This activity is usually
termed as “slicing” of target RNAs. In plants and worms, RNA-dependent RNA
polymerase (RdRP) amplifies RNAi by converting the cleavage products of initial
RNAi targets into secondary dsRNAs, which are further processed into secondary
siRNAs (Fig. 2). RdRPs are responsible for systemic spreading of a silencing status
in both plants and worms and are believed to prime an antiviral status prior to viral
arrival (Sijen et al. 2001; Baulcombe and Molnar 2004).
Although the RNAi phenomenon was first observed in model plant, function
mechanism studies suggested that the genes involved in RNAi actually are con-
served in almost all eukaryotes. In addition to their response to artificial dsRNAs
delivered through transgene and microinjection, key RNAi genes also play essential
roles in the biogenesis or function of a wide variety of small RNA species derived
from endogenous transcripts and transposons under natural conditions. Based on
their origin and the effector proteins they associate with, these small RNAs can be
classified as microRNA (miRNA), endogenous siRNA (endo-siRNA), and PIWI-
interacting RNA (piRNA) (Lee and Ambros 2001; Girard et al. 2006; Yigit et al.
2006; Ambros and Chen 2007; Aravin et al. 2007; Brennecke et al. 2007).
miRNAs are a class of abundant 20–24 nt RNAs that are processed by Dicer
enzymes from stem–loop precursors found in intergenic regions, introns, and
coding regions. miRNA silences gene expression by guiding RISC to target
mRNA with fully or partly complementary sequences (Carrington and Ambros
2003; Bartel 2004; Voinnet 2009). In plant, miRNAs base pair with target mRNAs
through fully complementary sequences, and the common outcome of repression is
manifested as target mRNA cleavage. Animal miRNAs base pair with target
transcripts through a “seed” sequence, the 2–8 nucleotides at the 50 end. In contrast
to plant miRNA, base-pairing between animal miRNA and target in most cases
leads to translational arrest. It is of interest to note, since the “seed” sequences of
animal miRNAs are relatively short, 7 nt long, a single animal miRNA could, in
principle, potentially target and regulate a large number of protein-coding genes,
making miRNA target prediction and identification very difficult. miRNAs have
received the most extensive studies among these three known classes of endoge-
nous small RNAs and are shown to play essential roles in development, viral and
bacterial resistance, tumorigenesis, and stress response.
Endo-siRNAs are produced as populations through processing of long, perfectly
complementary dsRNA precursors by Dicer enzymes. The dsRNA precursors
might have derived from inverted repeats or convergent transcription of transpo-
sons or might have been produced by RdRPs using aberrant RNAs as templates
(Herr et al. 2005). Endo-siRNAs in many species are found to regulate gene
expression in trans or to suppress transposon mobility (Duchaine et al. 2006;
Ghildiyal et al. 2008; Okamura and Lai 2008; Tam et al. 2008).
piRNAs are a class of animal noncoding small RNAs whose major function is to
suppress the mobility of transposons. Interestingly, unlike miRNAs and endo-
RNAi Suppression and Its Application 65
plants ended up in a rapid recovery from severe virus disease symptoms and the
“recovered” plant became resistant to secondary infection by the same virus but
remained susceptible to secondary infection by unrelated viruses, such as Potato
virus X (PVX). Importantly, this nepovirus-induced viral resistance can be targeted
against PVX in secondary infection if the PVX genome is modified to carry a piece
of nepovirus sequence (Ratcliff et al. 1997). These observations together suggested
that plant viral infection triggers a natural antiviral response that operates in a
sequence-dependent manner, a characteristic feature of transgene-induced gene
silencing as reported in many studies (Ratcliff et al. 1997). In an independent
research, another group also reported gene silencing-mediated viral immunity
against DNA virus (Covey et al. 1997). These findings together suggested that
homology-dependent gene silencing, a term frequently used to describe RNAi at
that time, represents a novel viral immunity that occurs under natural condition.
RNAi as a natural antiviral defense mechanism was reconfirmed by the discovery
of viral suppressor of RNAi (VSR). Most of VSRs were previously known to be
dispensable for viral replication but responsible for enhanced viral infection and
disease symptom, and thus were often referred to as viral pathogenicity factors. The
first two VSRs identified are the helper component-proteinase (HC-Pro) encoded by
Tobacco etch virus and the 2b protein encoded by Cucumber mosaic virus (CMV)
(Anandalakshmi et al. 1998; Brigneti et al. 1998; Kasschau and Carrington 1998).
When expressed as transgene, HC-Pro reversed RNAi targeted against a transgene
in tobacco plants. HC-Pro is also capable of suppressing RNAi triggered by viral
replication when expressed from recombinant virus (Anandalakshmi et al. 1998;
Brigneti et al. 1998; Kasschau and Carrington 1998) (Fig. 3c, d). Later on, the potent
suppression activity of HC-Pro was ascribed to its ability to bind and interfere with
the stability of virus-derived siRNAs (viRNAs) (Lakatos et al. 2006; Lozsa et al.
2008). The 2b of CMV apparently adopts a different mechanism for RNAi suppres-
sion. When expressed from recombinant PVX, CMV 2b did not reverse the silencing
that has already been established but instead prevented the initiation of gene
silencing at the growing points of the plant (Brigneti et al. 1998) (Fig. 3e, f). Now
we know that this is because CMV 2b is able to suppress systemic silencing (Guo
and Ding 2002). As a result, target gene in the new emerging leaves will remain
nonsilenced because of the lack of systemic silencing signal.
RDVI was also found to be conserved in animal through studying the replication
of Flock house virus (FHV), a member of the nodavirus family, in Drosophila cell
culture (Li et al. 2002). Like what has been found in plant system, infection by FHV
virions results in the rapid accumulation of FHV-specific siRNAs of both plus and
minus polarities in infected cells. Moreover, increased accumulation of FHV RNAs
was observed in the fly cells depleted of AGO2, which is known to be a core
component of the effecter complex RISC. This finding strongly suggested that viral
replication is checked by an RNAi-directed antiviral defense. Most importantly, the
B2 protein of FHV, a suppressor of RNAi in diverse systems, is essential for
accumulation of FHV in wild-type fly cells but becomes dispensable in cells
depleted of AGO2. This perfect complementation unequivocally demonstrated
that the essential role of B2 in FHV infection is to suppress the AGO2-dependent
RNAi Suppression and Its Application 67
Fig. 3 Suppression of transgene-induced RNAi by PVY HC-Pro and CMV 2b. (a) Nicotiana
benthamiana plant (line 16c) showing high levels of GFP expression under UV illumination.
(b) 16c plant showing GFP silencing triggered by infiltration of an A. tumefaciens strain carrying
GFP T-DNA. The bright red color is from the chlorophyll fluorescence under UV illumination.
(c) GFP-silenced 16c plant infected with PVY under UV illumination (15 days post-inoculation).
The green fluorescence showing the reversal of GFP silencing by PVY. (d) GFP-silenced 16c plant
infected with recombinant PVX carrying PVY HC-Pro under UV illumination. The green fluore-
scence showing the reversal of GFP silencing by HC-Pro. (e) GFP-silenced 16c plant infected with
CMV (21 days post inoculation). GFP expression was restored in the newly emerging tissue after
systemic CMV infection had been established. (f) GFP-silenced 16c plant infected with recombi-
nant PVX carrying CMV 2b under UV illumination. All images in this figure are adapted, with
permission, from MacMillan Publisher Ltd: EMBO J (Brigneti et al. 1998), copyright 1998
antiviral RNAi against FHV (Li et al. 2002). RNAi suppressors were also isolated
from DNA viruses. These suppressors presumably prevent RNAi from targeting
highly structured viral transcripts produced from read-through transcription of
repeated DNA sequence or convergent transcription of the same DNA sequence
(Voinnet et al. 1999; Guillaume and Olivier 2004; Fukunaga and Doudna 2009).
Recently, induction and suppression of antiviral RNAi has also been documented in
fungi and other invertebrates such as shrimp, mosquito, silk moth, and tick cells.
These findings together establish RDVI as a viral innate immunity in diverse
organism species (Isobe et al. 2004; Li et al. 2004; Garcia et al. 2006; Segers
et al. 2007; Su et al. 2008).
68 X. Yi and R. Lu
Viruses are obligate intracellular parasites that must rely on host protein and nucleic
acid synthesis machinery to propagate. In order to survive in a largely hostile
environment, all viruses must have evolved a plethora of functions to suppress
host antiviral defense mechanisms such as RNAi. In agreement with this notion,
every single plant virus that has been closely examined is found to encode at least
one VSR. These include both single-stranded DNA viruses and RNA viruses with
positive-, negative-, or double-strand RNA genomes (see appendix, Table 1). Some
viruses, such as Citrus tristeza virus (CTV) and geminiviruses, encode multiple
VSRs with each of them appearing to have a distinct mode of action (Lu et al. 2004;
Vanitharani et al. 2004). So far, over thirty viral RNAi suppressors have been
identified, and this number is still growing. On prediction, there will be a very large
RNAi Suppression and Its Application 69
Agrobacterium-mediated transient assay has been frequently used for VSR isola-
tion mainly because of its simplicity and rapidity. In this assay, the candidate
suppressor protein is often delivered into plant with a transgene construct that
triggers silencing of a stable reporter transgene, usually a gene encoding green
fluorescent protein (GFP). The delivery is usually done using agroinfiltration.
Agroinfiltration is a technique that allows transient expression of target gene in
plant leaves. Typically, Agroinfiltration is carried out by vacuum-infiltrating plant
leaves with a recombinant agrobacterial culture containing target gene T-DNA.
When a GFP coding sequence is used as inducer in a transient assay, the GFP
mRNAs produced from the stable GFP transgene in the infiltrated zone will be
destroyed by RNAi in the absence of a suppressor but will accumulate in the
presence of an RNAi suppressor manifested as enhanced green fluorescence
Fig. 4 Agrobacterium-mediated transient suppression assay for the identification of RNAi sup-
pressors encoded by CTV. Leaves of the 16c plants were infiltrated with an A. tumefaciens strain
carrying 35S-GFP together with an A. tumefaciens strain carrying the empty binary plasmid (35S:–),
35S:TAV 2b, 35S:CMV 2b, 35S:CTV p23, 35S:CTV p20, or 35S:CTV CP. The green fluores-
cence images of the coinfiltrated leaves with the abaxial-side up were taken 3 days postinfiltration
under a long-wave UV lamp. Image in this figure is adapted, with permission, from National
Academy of Sciences, USA (Lu et al. 2004), copyright 2004
70 X. Yi and R. Lu
This strategy is based on the observation that many viral RNAi suppressors are able
to reverse transgene-induced silencing that occurs autonomously in some transgene
plants (Anandalakshmi et al. 1998; Brigneti et al. 1998; Kasschau and Carrington
1998). Thus, for identification purpose, candidate suppressors are assayed for their
ability to reverse an ongoing silencing targeting a transgene in a reporter plant.
Very often, the candidate suppressors are introduced into the reporter plants
through genetic crossing between reporter plants and transgenic plants expressing
candidate suppressors. The candidate suppressors can be also ectopically expressed
from a heterologous virus vector that is inoculated onto the reporter plants. How-
ever, if the latter strategy is used, an additive or synergistic effect resulted from
suppressors already carried by viral vector should be considered to access the
suppression activity of the candidate.
Grafting assay is so far the most reliable strategy used to assay VSR activity on
systemic silencing. This strategy is based on the observation that systemic silencing
signal generated from a silenced rootstock can spread into scion and cause
sequence-specific silencing targeting a scion transgene. Figure 5 illustrates a graft-
ing assay using a GUS-expressing tobacco line T19 as scion and a GUS-silenced
tobacco line 6b5 as rootstock. After introducing the candidate suppressor into line
6b5 through genetic crossing, whether or not the candidate suppressor is capable of
systemic silencing suppression is determined by assaying the GUS expression in the
T19 scion. Although a bit time-consuming, this strategy has not only allowed for
the identification of systemic suppression activity for VSRs that suppress local/
intracellular silencing but also allowed for the identification of VSR with specific
activity on systemic silencing (Guo and Ding 2002; Lu et al. 2004).
This strategy has been successfully used for identifying VSRs of animal viruses in
cell culture-based system (Li et al. 2004). The major component of this strategy is a
RNAi Suppression and Its Application 71
6b5
6b5/2b 6b5/2b
mutant virus whose genome is modified to carry a GFP reporter gene in the place of
an RNAi suppressor. Thus, replication of this virus is suppressed by RNAi in wild-
type cells but will be restored to produce GFP if an RNAi suppressor is provided in
trans. This strategy has also been adopted by plant researchers in VSR identifica-
tion. In this case, the rescue of virulence and systemic accumulation is often used as
an indication of activity in RNAi suppression (Yelina et al. 2002).
It is important to note that, so far, a large number of VSRs have been identified
and characterized through transient expression together with a second reporter
transgene. This simple and fast approach may have limited application in charac-
terizing VSRs from viruses like CTV and geminiviruses, which are shown to
produce multiple VSRs with distinct modes of action (Lu et al. 2004; Vanitharani
et al. 2004). In the case of CTV, because the CP does not suppress local/intracellu-
lar silencing (see Sect. 4 for more details), its function in systemic silencing
suppression, as demonstrated in grafting assay, would have been overlooked in
transient expression assay. This transgenic approach may have its limitation in
identifying VSRs that exhibit temporal or spatial expression pattern during the
course of viral infection. For example, VSRs produced early during viral infection
and targeting viRNA biogenesis for suppression will fail to suppress siRNA-
induced RNAi, whereas those capable of inhibiting secondary viRNA synthesis
will not be identified as RNAi suppressors in the context of RNAi triggered by
inverted repeat transgenes. As we know, production of foldback RNAs from
inverted repeat transgenes does not require RdRP activities. Additionally, because
the level and timing of VSR and RNAi trigger expression in most transgenic or
transient approaches are usually set arbitrarily such that the results from indepen-
dent tests might not be comparable or inaccurate (Ding and Voinnet 2007).
Recently, a genetic rescue strategy has been demonstrated in a few reports, which
72 X. Yi and R. Lu
may serve as supplementary strategy for identification of novel VSRs (Deleris et al.
2006; Diaz-Pendon et al. 2007; Wang et al. 2006). In these reports, rescue of viral
replication was assayed in RNAi-defective hosts using mutant viruses with disabled or
modified VSRs. These assays not only reconfirmed the suppression activity of
known VSRs but also unequivocally implicate key RNAi genes in antiviral RNAi.
Thus, to fully appreciate the functionality of VSRs during authentic viral infections, a
combination of complementary strategies will be needed for function characteriza-
tion.
The fact that viruses are intracellular parasites that entirely rely on host macro-
molecule synthesis and metabolism machineries entails that viruses must overcome
various host antiviral mechanisms to survive and evolve. Since RNAi-based viral
immunity is not conserved in prokaryotes, it can be inferred that viral RNAi
suppressors must have emerged as RNAi antagonists after viruses expanded their
host range from prokaryotes to eukaryotes. The lack of any obvious sequence or
structure similarity in between known VSRs further suggests that they must have
emerged independently as a result of viral adaptation to RDVI. In agreement with
this hypothesis, many RNAi suppressors are produced from out-of-frame over-
lapped region of viral genome, and their function is dispensable for some basic viral
function such as viral replication and packaging. For example, both CMV 2b and
FHV B2 are produced from out-of-frame overlapped region of respective viral
genomes. Both will become dispensable for viral replication when host antiviral
RNAi is made defective (Wang et al. 2006; Diaz-Pendon et al. 2007). The over-
lapped suppressor genes might have been created through overprinting, a phenom-
enon in which a single coding sequence is translated in different reading frames
(Keese and Gibbs 1992). Very often, for each pair of genes created through over-
printing, one is more ancient and widespread whereas the other is novel and has a
confined lineage in the phylogeny of viruses (Li and Ding 2006).
In plants and worms, cellular proteins can also serve as negative regulators of
RNAi (Anandalakshmi et al. 2000; Kennedy et al. 2004; Ramachandran and Chen
2008). Some of these proteins target and destabilize siRNA or miRNA for RNAi
suppression (Kennedy et al. 2004; Ramachandran and Chen 2008). Presumably,
under natural conditions, these cellular proteins contribute to a fine-tuning mecha-
nism that ensures various RNAi directed functions to be carried out in a controlled
manner.
complex nature of viral counterdefensive mechanisms, so far, only very few VSRs
are better characterized for their modes of action in RNAi suppression. Based on
their biological targets of RNAi machinery, these VSRs can be loosely classified
into four independent groups as discussed below.
Despite the extreme sequence diversity, many VSRs are in fact dsRNA-binding
proteins. This probably reflects the fact that all RNAi mediated antiviral responses
invariably begin with dicer processing of virus-derived dsRNAs. Thus, targeting
dicer substrate dsRNA for protection would serve as common strategy for many
VSRs. The B2 protein encoded by FHV is a VSR of this category that has been
subject to extensive functional and structural analyses. The full length FHV B2
protein is only 106 amino acids (aa) long. A dsRNA-binding domain can be found
at the N-terminal region. B2 is synthesized at high levels in the early stage of viral
replication and, like many other VSRs, suppresses RNAi across kingdoms (Li et al.
2002). B2 binds both siRNA and long dsRNA in a sequence-independent manner
(Chao et al. 2005; Lu et al. 2005). This finding suggests dual modes of action for B2
in RNAi suppression, in that long dsRNA binding would interfere with dicer
function in dsRNA processing whereas siRNA binding would inhibit active RISC
formation, leading to defect in target RNA cleavage. In agreement with this
hypothesis, B2 was found to inhibit dicer processing of long dsRNA in vitro
(Chao et al. 2005; Lu et al. 2005). Both NMR and crystallization structural analyses
have been performed for B2. These studies revealed an all-helix structure for the
N-terminal 72 aa (Chao et al. 2005; Lingel et al. 2005). A cocrystal structure of B2
and an 18-bp dsRNA shows that B2 recognizes dsRNA as a homodimer that forms a
four-helix bundle (Fig. 6a). B2 interacts exclusively with the ribose-phosphate
backbone of two successive minor grooves and the intervening major groove
such that when multiple B2 proteins bind to the same dsRNA molecule, they
would form a “coat” for the bound dsRNA, making the target dsRNA inaccessible
to Dicer (Fig. 6b, c). In consistence with structural data, replacement of Arg at
position 54, which is in the center of the dsRNA-binding surface, by Gln abolished
B2 activity in both dsRNA binding and dicing inhibition in vitro (Lu et al. 2005).
Despite sharing very limited sequence identity with FHV B2, a recent report shows
that the B2 protein from Nodamura virus (NoV), another member of the nodavirus
family, also forms dimmers with helical bundle structure. The crystal packing
places the RNA-binding residues along one face of symmetry-related molecules,
suggesting that NoV B2 suppresses RNAi using a similar mechanism (Shaik Syed
Ali and Chen 2009).
Compared to FHV B2, the V2 protein from Tomato yellow leaf curl geminivirus
(TYLCV) seems to have a unique taste in dsRNA binding in that it specifically
binds dsRNA with 50 end overhangs. Previously, it has been shown that V2 inhibits
74 X. Yi and R. Lu
Fig. 6 Structure basis of B2 and p19 function in RNAi suppression. Upper panel: Blue and green,
B2 dimer; Magenta, dsRNA. (a) B2 binds dsRNA as a dimmer and recognizes two successive
minor grooves and a major groove that separates them. (b) the dimmer is rotated by approximately
10 relative to the RNA to maximize contacts with both minor grooves. (c) the region of RNA that
is bound by B2 dimers is localized to one face of the duplex [adapted, with permission, from
MacMillan Publisher Ltd: Nat. Struct. Mol. Biol. (Chao et al. 2005), copyright 2005]. Lower
panel: Blue and magenta, individual monomers of the p19 dimer; Orange and pink, the two siRNA
strands. Two tryptophans from each monomer of the p19 dimer, which bracket the terminal base
pairs at either end of the siRNA duplex, are shown in stick representation. (d) A stereo view
perpendicular to the twofold axis. (e) and (f), Alternative mono views of the complex rotated by
90 along different axes (modified and adapted, with permission, from MacMillan Publisher Ltd:
Nature (Ye et al. 2003), copyright 2003)
silencing targeting a GFP transgene but does not affect the biogenesis of GFP-
specific siRNAs, suggesting that V2 targets a step in the RNAi pathway that is
downstream of Dicer processing of dsRNA precursor (Zrachya et al. 2007). RNA-
binding assay revealed that V2 binds to dsRNA of various lengths with 50 end
overhangs. Interestingly, dsRNA binding of V2 is inhibited by the presence of
phosphate group at the 30 ends, but the 30 end methylation modification does not
seem to affect V2 activity in dsRNA binding, indicating that V2 may recognize and
bind to the junction of a dsRNA region with a 50 end overhanging strand (Fukunaga
and Doudna 2009). dsRNA-binding competition assay suggests that V2 may
RNAi Suppression and Its Application 75
The first viral protein characterized in this category is the p19 protein from Tomato
bushy stunt virus (TBSV), a member of the tombusvirus family that also include
Cucumber necrosis virus (CNV) and Cymbidium ringspot virus (CRV). Initially, it
was found that the p19 proteins from these viruses are not essential for virus cell-
to-cell movement but, instead, are required for virus systemic spread and symptom
development. p19 was identified as an RNAi suppressor based on its ability to
reverse the silencing targeting GFP transgene in the systemic leaves of plants
infected with either TBSV or PVX carrying a p19 insert (Voinnet et al. 1999).
Subsequently, several groups independently demonstrated the suppression activity
of p19 proteins from a number of different tombusviruses using the agroinfiltration
assay (Qiu et al. 2002; Qu and Morris 2002). Through the study of p19-mediated
RNAi suppression in Drosophila embryo extracts, Burgyan and colleagues found
out that p19 inhibited siRNA-directed slicing of the target mRNA only when p19
were added to the embryo extracts at the same time as the siRNA duplex. The fact
that p19 failed to inhibit slicing when added 20 minutes later than the siRNA duplex
suggested that p19 may bind siRNA duplexes and prevent them from being
incorporated into RISC (Lakatos et al. 2004). In supporting this hypothesis, P19
binds 21-nt siRNA duplex with high affinity independent of the 2-nt overhangs at
the 30 end of siRNA, and its affinity is much weaker for dsRNAs of 22 nt or longer.
Interestingly, P14, a P19 homologue encode by aureusviruses of the same family,
binds dsRNA for RNAi suppression without a size preference (Merai et al. 2005).
p19 structure studies through crystallography reconfirmed the siRNA binding
specificity and established a structural explanation for how dimerization of p19
was essential for binding siRNA (Vargason et al. 2003; Ye et al. 2003). From these
studies, it is clear that p19 binds to a siRNA as a homodimer with two molecules of
p19 per siRNA duplex. In a crystal structure of p19 bound to a 21-nucleotide
76 X. Yi and R. Lu
siRNA, the 19-basepair RNA duplex is “cradled” within the concave face of a
continuous eight-stranded b-sheet, formed across the p19 homodimer interface
(Fig. 6d). In consistence with the sequence-independent siRNA recognition of
p19, the direct and water-mediated intermolecular contacts are restricted to the
backbone phosphates and sugar 20 -OH groups. Two a-helical “reading heads”
project from opposite ends of the p19 homodimer and position pairs of tryptophans
for stacking over the terminal base pairs, thereby measuring and bracketing both
ends of the siRNA duplex (Fig. 6e, f). These studies provide a perfect structural
explanation of siRNA sequestering by p19, making it one of the best characterized
VSRs.
Recently, siRNA binding capacity was also demonstrated for p21 of clostero-
viruses and AC4 protein from geminivirus, suggesting a common strategy used by
viral suppressors in counteracting on RDVI (Chellappan et al. 2005; Lakatos et al.
2006). Interestingly, p21 binds siRNA duplex thereby to inhibit the formation of
mature RISC, whereas AC4 preferentially binds the single-stranded siRNA or
miRNA in in vitro RNA binding assay presumably to inhibit the activity of mature
RISC.
The RNase3 protein of Sweet potato chlorotic stunt virus (SPCSV) suppresses
RNAi in a unique way, in that it targets viRNA for destruction. SPCSV is a single-
stranded RNA (ssRNA) crinivirus whose infection is phloem limited. SPCSV
synergizes Seet potato Fathery mottle virus (SPFMV), an HC-Pro producing
virus, in coinfected sweet potato plants as manifested in enhanced disease symp-
toms and elevated titer of SPFMV. SPCSV RNase3 is a class I endoribonuclease III
that specifically binds and cleaves dsRNA but not ssRNA of various length. When
expressed from a sweet potato transgene, SPCSV RNase3 alone is sufficient to
break down resistance to SPFMV and other unrelated viruses, leading to higher
accumulation of beneficiary viruses and severe disease symptoms (Cuellar et al.
2009). Notably, SPCSV RNase3 suppresses RNAi in an endonuclease activity-
dependent manner. In vitro, it cleaves synthetic siRNAs duplexes of 21, 22, and
24 bp in length, and the cleavage product is approximately 14 bp long. Thus, it
appears that SPCSV RNase3 suppresses antiviral defense by destroying infecting
virus-derived siRNAs. In agreement with this notion, SPCSV RNase3 treatment
caused a clear reduction of total siRNA isolated from SPFMV-infected sweet potato
plants (Cuellar et al. 2009). Interestingly, deep sequencing analysis suggested that
only a small margin of SPFMV-derived siRNAs, 3.95%, are siRNA duplexes. It is
thus possible that clearance of this small portion of siRNAs may allow beneficiary
viruses to get an upper hand on silencing. It is also possible that the siRNA duplexes
targeted by RNase3 may play an essential role in establishing systemic antiviral
silencing, which can attenuate/slow down viral infection in the presence of RNAi
suppressor encoded by beneficiary viruses (Voinnet 2005).
In Arabidopsis, miRNA and endo-siRNA are methylated at the 30 end by HEN1
for enhanced stability (Yu et al. 2005). Virus-derived siRNAs are also the target of
HEN1, suggesting a role of HEN1 in antiviral defense. As a counterdefensive
mechanism, some viral RNAi suppressors target HEN1-directed modification for
RNAi suppression. The HC-Pro from potyvirus (Ebhardt et al. 2005) is one of
RNAi Suppression and Its Application 77
such suppressors whose expression from transgene was found to be associated with
marked decrease of the 30 end modification of viral siRNAs. Interestingly, HC-Pro
does not seem to significantly affect the modification of endogenous miRNAs and
24-nt siRNAs. The silencing suppressor from tobamovirus, the 126-kDa protein,
appears to use similar strategy in RNAi suppression (Ding et al. 2004; Vogler et al.
2007). Previously, it was shown that the 126-kDa protein of TMV is dispensable for
viral replication in protoplast. In N. benthamiana plants, TMV infection leads to the
production of virus-specific siRNAs in the presence of 126-kDa protein, indicating
that this TMV suppressor targets a step downstream of viRNA biogenesis. b-
elimination assay confirmed that TMV infection indeed interferes with HEN1-
mediated methylation of viRNAs and that this interference and the formation of
virus-induced disease symptoms are experimentally linked to the silencing suppres-
sor activity of the 126-kDa protein. This RNAi suppression strategy seems to be
shared within the tobamovirus family as the infection by Olseed rape mosaic
tobamovirus (ORMV) also leads to interference with HEN1-mediated methylation
of siRNA and miRNA. Previously, it was shown that, for unknown reason, RNAi
suppression by the coat protein of Turnip crinkle virus (TCV) is often associated
with diminished level of target-derived siRNAs (Qu et al. 2003; Dunoyer et al.
2004). Now, it turns out that this is because TCV CP interferes with the methylation
of siRNA, leading to its instability. Interestingly, in contrast to suppressors of
tobamovirus origin, TCV CP does not interfere with the methylation of host
miRNA.
The first viral protein identified in this category is the P0 protein from poleroviruses
(Pfeffer et al. 2002). P0 protein was first identified as RNAi suppressor based on an
agroinfiltration-based transient assay in which overexpression of P0 protein from
Beet western yellows virus (BWYV) suppressed the silencing targeting GFP
reporter in transgenic N. benthamiana plants. Later on, it was found that P0
physically interacts with Arabidopsis orthologs of S-phase kinase-related protein
1 (SKP1) through a conserved F-box-like motif (Pazhouhandeh et al. 2006). It
appears that P0 functions in an SKP1-like protein-dependent manner, in that down-
regulation of a SKP1 ortholog in N. benthamiana rendered the plants resistant to
polerovirus infection. In agreement with this observation, point mutations in the
F-box-like motif of P0 abolished not only the P0-SKP1 ortholog interaction but also
diminished virus pathogenicity and the RNAi suppression activity of P0. SKP1 is a
component of the SCF (Skp1, Cul1/Cdc53, F-box proteins) family of ubiquitin E3
ligases, which are known to add polyubiquitin tracts on selected lysine residues,
thereby marking a protein for proteasome-mediated degradation. These findings
therefore strongly suggested that P0 may target key RNAi components for degra-
dation. In supporting this hypothesis, later on, it was shown that P0 targets the PAZ
motif and its adjacent upstream sequence in Arabidopsis AGO1 and mediates its
78 X. Yi and R. Lu
a
GU
5' CU UGA
UAGCUUAUCAGACUGAUGUUGA pre-miR-21
A
GUCGGGUAG CUGACCACAACG U
CU GU UC
3' AC
b
GAUA
G A
5' U A
A G G U UUCGA
GGGC CUCUUCCGU GUCUG UOGCAAGGGUAUCAUGGOGGAOGAOOGGGG A
CUCG GAGGGGGCA CAGACUGCAGCGU OOCA AGUGOOGOCUGCCGGOOUA C
C C A G A GA U GGCCC
U UGG C C
U UG C
U VA1
UA
3' GU
CG
GC
CG
c CG
C U 800
G U 1
CCA
In plants and worms, a silencing status can spread out of the initial silencing site
(Voinnet and Baulcombe 1997; Tijsterman et al. 2004; Winston et al. 2002). The
target sequence-specific feature of systemic silencing suggests the involvement of a
class of RNA molecules. Systemic silencing in plants is believed to be responsible
for initiating an antiviral condition prior to virus arrival (Baulcombe 2002). Accord-
ingly, as a counter-defense mechanism, some viruses encode suppressors that are
capable of specific targeting of systemic silencing signal. This was first demon-
strated for the P25 protein of PVX. An earlier study suggested a role for P25 in PVX
systemic spread in that deletion of P25 coding sequence does not affect PVX
accumulation in inoculated protoplasts but abolishes spread of PVX out of the
initially infected cells (Angell et al. 1996). A functional role of P25 in systemic
80 X. Yi and R. Lu
Transient expression of target gene for rapid function assay has been frequently used
in diverse systems. However, horizontal gene delivery through transfection, electro-
poration, bombardment, microinjection, and agrobacterium-mediated transformation
often results in poor expression of target gene. Although other possibilities exist, a
major host factor that is responsible for suboptimal expression of target gene is an
RNAi-based gene surveillance system that guards against nucleic acid intruders.
Thus, as a circumventing strategy, codelivery of an RNAi suppressor has been used
to enhance target gene expression in diverse systems, and this strategy has been
shown to be particularly successful in plants (Johansen and Carrington 2001;
Voinnet et al. 2003).
In plants, transient expression of target gene can be easily achieved using
recombinant A. tumefaciens strain that contains the target gene in the T-DNA
region of the binary-Ti plasmid. When the bacterial culture is vacuum-infiltrated
into leaves, and, upon T-DNA transfer, the target gene will be ectopically
expressed in cells that received the T-DNA. However, very often, the ectopic
gene expression ceases after 2–3 days mainly due to RNAi triggered by transgene
corresponding to the target (see Sect. 5.1 for detailed discussion). By coinfiltration
of a viral RNAi suppressor, Johansen and Voinnet have shown that the target gene
expression can be significantly enhanced (Johansen and Carrington 2001; Voinnet
et al. 2003). In the case of p19 suppressor from TBSV, expression of a range of
proteins was enhanced 50-fold or more in the presence of p19. Quantitative
analysis using GFP as target protein suggested that in coinfiltrated tissues, GFP
proteins accumulated up to 7% of total soluble protein (Voinnet et al. 2003). Based
on these observations, many research labs have developed various transient gene
expression systems for fast, flexible, and reproducible function assay. In fact,
because of its simplicity and rapidity, the p19-enhanced expression system is
also used in industrial production for isolation of a broad range of proteins without
the need for the time-consuming regeneration of stably transformed plants.
Viruses multiply very efficiently and thereby drive their gene expression to
extremely high level. This has inspired researchers to utilize replicating virus for
target gene overexpression. To facilitate large-scale virus inoculation, the recom-
bined virus is often delivered into host as stable transgene. In principle, this
technology, often referred to “amplicon,” will allow for much higher yield of the
target gene products compared to conventional transgene approaches that utilize
strong promoters. However, in plants, the first attempt of this strategy failed, in that
all of the transformants consistently exhibited RNA silencing of the amplicon
transgene (Angell and Baulcombe 1997). Now we know that the dsRNA form of
viral replication intermediates produced in every single cell of the transgenic plants
84 X. Yi and R. Lu
would have served as potent triggers of the RNAi-based defense mechanism that
normally suppresses viral replication during natural infections. Conceivably, coex-
pression of viral suppressors would diminish the suppression effect of antiviral
RNAi, thereby allowing high-level target gene expression initially envisaged for
amplicons. This idea was once tested using transgenic tobacco plants expressing
TEV HC-Pro. When the HC-Pro expressing line is crossed with amplicon line that
is designed to express a GUS reporter gene from the PVX genome (Mallory et al.
2002), a dramatic increase in virus accumulation and GUS expression was observed
in progeny plants that contain both the suppressor transgene and the amplicon
locus. The GUS expression was so much enhanced that leaves of mature plants
accumulated the GUS protein up to 3% of total soluble proteins.
The target sequence-independent mode of action of all VSRs has limited their
application in function analysis of specific genes. However, the fact that some
of VSRs function through specific interaction with key components of RNAi
machinery suggests that we may be able to use VSRs as molecular tools to for
function and mechanism study of RNAi. For example, the CP of CTV has been
shown to specifically suppress systemic silencing, probably by interfering with the
long distance spread of systemic silencing signal (Lu et al. 2004). Current studies
support a role for 21 nt siRNAs in short distance spread of gene silencing but give
no clue about the identity of the signal molecules responsible for long distance
spread of silencing (Dunoyer et al. 2005). The unique feature of CTV CP makes it
an ideal molecular tool for identifying such a signal molecule. It can be anticipated
that detection and biochemical identity characterization of protein and/or RNA
molecules that are physically associated with CTV CP might allow us to gain
insight into the identity of this mysterious class of signal molecules. Similarly,
VSRs that function through specific interaction with key RNAi factors may allow
us to study RNAi-mediated function in organisms for which a genetic approach is
not available. For example, since VA1 is known to specifically bind and inhibit the
function of mammalian Dicer, overexpression of VA1 in a tissue-specific manner
may allow for the characterization of miRNA-mediated function in the target tissue.
In plants, some viral suppressors, when expressed as transgene, are able to
suppress miRNA function, leading to the accumulation of miRNA targets (Chapman
et al. 2004; Dunoyer et al. 2004). Although these observations were made on model
plant Arabidopsis, it can be inferred that transgenic expression of these VSRs would
also lead to accumulation of miRNA targets in plant species for which a genetic
approach for miRNA function study is still lacking or impossible. Thus, these VSRs
can be used as molecular probes to study miRNA-mediated function in economically
important crop plants. It is also possible that overexpression of these VSRs in some
animal species will compromise miRNA function, thereby leading to enrichment of
RNAi Suppression and Its Application 85
miRNA targets. Once proved true, this strategy may greatly facilitate miRNA target
identification in animal. Animal miRNA target identification has been difficult
mainly because only as few as 7 nt complementary sequence is needed for animal
miRNA to confer suppression effect.
endo-siRNAs of animal are a class of relatively new small RNAs isolated from
both germline and soma. In addition to transposon control, animal endo-siRNAs
may also modulate the gene expression in trans like that found for plant endo-
siRNAs (Czech et al. 2008; Okamura and Lai 2008; Watanabe et al. 2008).
Recently, it was shown that expression of various plant and insect VSRs in
transgenic flies led to no perturbation of miRNA pathway but instead inhibited
harpin RNA-triggered RNAi as well as transposon silencing conferred by endo-
siRNAs (Berry et al. 2009). These findings thus identify VSRs as genetic tools for
the study of endo-siRNA-mediated functions in animals.
Appendix
Table 1 (continued)
Virus genus Virus name VSR Motif
implicated in
VSR activity
Cucurbit aphid-born yellows virus
Potexvirus Potato virus X P25 Helicase
Potyvirus Tobacco etch virus Hc-Pro
Potato virus Y
Turnip mosaic virus
Sobemovirus Rice yellow mottle virus P1
Tobamovirus Tobacco mosaic viruses P130
Tomato mosaic viruses
Tobravirus Tobacco rattle virus 16K Cysteine-rich
protein
Tombusvirus Tomato bushy stunt virus P19 dsRNA
bindinga
Cymbidium ringspot virus
Tymovirus Turnip yellow mosaic virus P69
Vitiviruses Grapevine virus A P10
Negative-strand RNA viruses of plant origin
Tenuivirus Rice hoja blanca virus NS3
Tospovirus Tomato spotted wilt virus NSs
Double-stranded RNA viruses of plant origin
Phytoreovirus Rice dwarf virus Pns10
DNA viruses of plant origin
Begomovirus Tomato leaf curl virus C2 DNA binding,
NLS
Tomato Yellow Leaf Curl Virus V2 dsRNA
bindingd
TYLCCNV-Y10 Y10b bC1 DNA binding,
NLS
African cassava mosaic virus (KE) AC2 DNA binding,
NLS, AD
EACMCV, ICMV, TGMV
Mungbean yellow mosaic virus
African cassava mosaic virus (CM) AC4 miRNA
bindingb
Curtovirus Beet curly top virus L2 Protein binding
Positive-strand RNA viruses of animal origin
Nodavirus Flock house virus, nodamura virus, B2 dsRNA binding
Striped jack nervous necrosis virus,
Greasy grouper nervous necrosis virus
Negative-strand RNA viruses of animal origin
Orthomyxovirus Influenza virus A NS1 dsRNA binding
Orthobunyavirus La Crosse virus NSs
Filovirus Ebola virus VP35
Double-stranded RNA viruses of animal origin
Orthoreovrivus s3 dsRNA
bindingc
(continued)
RNAi Suppression and Its Application 87
Table 1 (continued)
Virus genus Virus name VSR Motif
implicated in
VSR activity
Retroviruses of animal origin
Lentivirus HIV-1 Tat
Spumavirus PFV-1 Tas
DNA viruses of animal origin
Adenovirus Adenovirus VA1 RNA Dicer binding
Poxvirus Vaccinia virus E3L dsRNA binding
This table is adapted and modified, with permission, from Annual Reviews: Annu. Rev. Microbiol.
(Li and Ding 2006), copyright 2006
a
Prefer 19-nt RNA duplex
b
Single-strand mature miRNA
c
Prefer dsRNA longer than 30 nt
d
Prefer dsRNA with 50 end overhangs
References
Ambros V, Chen X (2007) The regulation of genes and genomes by small RNAs. Development
134:1635–1641
Anandalakshmi R, Marathe R, Ge X et al (2000) A calmodulin-related protein that suppresses
posttranscriptional gene silencing in plants. Science 290:142–144
Anandalakshmi R, Pruss GJ, Ge X et al (1998) A viral suppressor of gene silencing in plants. Proc
Natl Acad Sci USA 95:13079–13084
Andersson MG, Haasnoot PC, Xu N et al (2005) Suppression of RNA interference by adenovirus
virus-associated RNA. J Virol 79:9556–9565
Angell SM, Baulcombe DC (1997) Consistent gene silencing in transgenic plants expressing a
replicating potato virus X RNA. EMBO J 16:3675–3684
Angell SM, Davies C, Baulcombe DC (1996) Cell-to-cell movement of potato virus X is asso-
ciated with a change in the size-exclusion limit of plasmodesmata in trichome cells of
Nicotiana clevelandii. Virology 216:197–201
Aravin AA, Sachidanandam R, Girard A et al (2007) Developmentally regulated piRNA clusters
implicate MILI in transposon control. Science 316:744–747
Bartel DP (2004) MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 116:281–297
Baulcombe D (2002) Viral suppression of systemic silencing. Trends Microbiol 10:306–308
Baulcombe DC (1999) Fast forward genetics based on virus-induced gene silencing. Curr Opin
Plant Biol 2:109–113
Baulcombe DC, Molnar A (2004) Crystal structure of p19 – a universal suppressor of RNA
silencing. Trends Biochem Sci 29:279–281
Baumberger N, Tsai CH, Lie M et al (2007) The Polerovirus silencing suppressor P0 targets
ARGONAUTE proteins for degradation. Curr Biol 17:1609–1614
Bayne EH, Rakitina DV, Morozov SY et al (2005) Cell-to-cell movement of potato potexvirus X is
dependent on suppression of RNA silencing. Plant J 44:471–482
Bennasser Y, Le SY, Benkirane M et al (2005) Evidence that HIV-1 encodes an siRNA and a
suppressor of RNA silencing. Immunity 22:607–619
Bernstein E, Caudy AA, Hammond SM et al (2001) Role for a bidentate ribonuclease in the
initiation step of RNA interference. Nature 409:363–366
Berry B, Deddouche S, Kirschner D et al (2009) Viral Suppressors of RNA silencing hinder
exogenous and endogenous small RNA pathways in Drosophila. PLoS ONE 4:e5866
88 X. Yi and R. Lu
Brennecke J, Aravin AA, Stark A et al (2007) Discrete small RNA-generating loci as master
regulators of transposon activity in Drosophila. Cell 128:1089–1103
Brigneti G, Voinnet O, Li WX et al (1998) Viral pathogenicity determinants are suppressors of
transgene silencing in Nicotiana benthamiana. EMBO J 17:6739–6746
Buck AH, Santoyo-Lopez J, Robertson KA et al (2007) Discrete clusters of virus-encoded
microRNAs are associated with complementary strands of the genome and the 7.2-kilobase
stable intron in murine cytomegalovirus. J Virol 81:13761–13770
Cai X, Schafer A, Lu S et al (2006) Epstein-Barr virus microRNAs are evolutionarily conserved
and differentially expressed. PLoS Pathog 2:e23
Carrington JC, Ambros V (2003) Role of microRNAs in plant and animal development. Science
301:336–338
Chao JA, Lee JH, Chapados BR et al (2005) Dual modes of RNA-silencing suppression by Flock
House virus protein B2. Nat Struct Mol Biol 12:952–957
Chapman EJ, Prokhnevsky AI, Gopinath K et al (2004) Viral RNA silencing suppressors inhibit
the microRNA pathway at an intermediate step. Genes Dev 18:1179–1186, Erratum in: Genes
Dev. 2004, 18:1510
Chellappan P, Vanitharani R, Fauquet CM (2005) MicroRNA-binding viral protein interferes with
Arabidopsis development. Proc Natl Acad Sci USA 102:10381–10386
Cogoni C, Irelan JT, Schumacher M et al (1996) Transgene silencing of the Al-1 gene in vegetative
cells of Neurospora is mediated by a cytoplasmic effector and does not depend on DNA-DNA
interactions or DNA methylation. EMBO J 15:3153–3163
Covey SN, Al-Kaff NS, Langara A et al (1997) Plants combat infection by gene silencing. Nature
385:781–782
Cuellar WJ, Kreuze JF, Rajam€aki ML et al (2009) Elimination of antiviral defense by viral RNase
III. Proc Natl Acad Sci USA 106:10354–10358
Cullen BR (2006) Is RNA interference involved in intrinsic antiviral immunity in mammals? Nat
Immunol 7:563–567
Czech B, Malone CD, Zhou R et al (2008) An endogenous small interfering RNA pathway in
Drosophila. Nature 453:798–802
Deleris A, Gallego-Bartolome J, Bao J et al (2006) Hierarchical action and inhibition of plant
Dicer-like proteins in antiviral defense. Science 313:68–71
Diaz-Pendon JA, Li F, Li WX et al (2007) Suppression of antiviral silencing by cucumber mosaic
virus 2b protein in Arabidopsis is associated with drastically reduced accumulation of three
classes of viral small interfering RNAs. Plant Cell 19:2053–2063
Ding SW, Voinnet O (2007) Antiviral immunity directed by small RNAs. Cell 130:413–426
Ding XS, Liu J, Cheng NH et al (2004) The Tobacco mosaic virus 126-kDa protein associated with
virus replication and movement suppresses RNA silencing. Mol Plant Microbe Interact
17:583–592
Duchaine TF, Wohlschlegel JA, Kennedy S et al (2006) Functional proteomics reveals the
biochemical niche of C. elegans DCR-1 in multiple small-RNA-mediated pathways. Cell
124:343–354
Dunoyer P, Himber C, Voinnet O (2005) DICER-LIKE 4 is required for RNA interference and
produces the 21-nucleotide small interfering RNA component of the plant cell-to-cell silencing
signal. Nat Genet 37:1356–1360
Dunoyer P, Himber C, Voinnet O (2006) Induction, suppression and requirement of RNA
silencing pathways in virulent Agrobacterium tumefaciens infections. Nat Genet 38:258–263
Dunoyer P, Lecellier CH, Parizotto EA et al (2004) Probing the microRNA and small interfering
RNA pathways with virus-encoded suppressors of RNA silencing. Plant Cell 16:1235–1250
Ebhardt HA, Thi EP, Wang MB et al (2005) Extensive 30 modification of plant small RNAs is
modulated by helper component-proteinase expression. Proc Natl Acad Sci USA
102:13398–13403
Elbashir SM, Martinez J, Patkaniowska A et al (2001) Functional anatomy of siRNAs for
mediating efficient RNAi in Drosophila melanogaster embryo lysate. EMBO J 20:6877–6888
RNAi Suppression and Its Application 89
Li F, Ding SW (2006) Virus counterdefense: Diverse strategies for evading the RNA-silencing
immunity. Annu Rev Microbiol 60:503–531
Li H, Li WX, Ding SW (2002) Induction and suppression of RNA silencing by an animal virus.
Science 296:1319–1321
Li WX, Li H, Lu R et al (2004) Interferon antagonist proteins of influenza and vaccinia viruses are
suppressors of RNA silencing. Proc Natl Acad Sci USA 101:1350–1355
Lichner Z, Silhavy D, Burgyan J (2003) Double-stranded RNA-binding proteins could suppress
RNA interference-mediated antiviral defences. J Gen Virol 84:975–980
Lindbo JA, Dougherty WG (1992) Pathogen-derived resistance to a potyvirus: immune and
resistant phenotypes in transgenic tobacco expressing altered forms of a potyvirus coat protein
nucleotide sequence. Mol Plant Microbe Interact 5:144–153
Lingel A, Simon B, Izaurralde E et al (2005) The structure of the flock house virus B2 protein, a
viral suppressor of RNA interference, shows a novel mode of double-stranded RNA recogni-
tion. EMBO Rep 6:1149–1155
Lozsa R, Csorba T, Lakatos L et al (2008) Inhibition of 30 modification of small RNAs in virus-
infected plants require spatial and temporal co-expression of small RNAs and viral silencing-
suppressor proteins. Nucleic Acids Res 36:4099–4107
Lu R, Folimonov A, Shintaku M et al (2004) Three distinct suppressors of RNA silencing encoded
by a 20-kb viral RNA genome. Proc Natl Acad Sci USA 101:15742–15747
Lu R, Maduro M, Li F et al (2005) Animal virus replication and RNAi-mediated antiviral silencing
in Caenorhabditis elegans. Nature 436:1040–1043
Lu S, Cullen BR (2004) Adenovirus VA1 noncoding RNA can inhibit small interfering RNA and
MicroRNA biogenesis. J Virol 78:12868–12876
Macrae IJ, Zhou K, Li F et al (2006) Structural basis for double-stranded RNA processing by
Dicer. Science 311:195–198
Mallory AC, Parks G, Endres MW et al (2002) The amplicon-plus system for high-level expres-
sion of transgenes in plants. Nat Biotechnol 20:622–625
Mangwende T, Wang M-L, Borth W et al (2009) The P0 gene of sugarcane yellow leaf virus
encodes an RNA silencing suppressor with unique activities. Virology 384:38–50
Matzke MA, Mette MF, Matzke AJ (2000) Transgene silencing by the host genome defense:
implications for the evolution of epigenetic control mechanisms in plants and vertebrates. Plant
Mol Biol 43:401–415
Merai Z, Kerenyi Z, Molnar A et al (2005) Aureusvirus P14 is an efficient RNA silencing
suppressor that binds double-stranded RNAs without size specificity. J Virol 79:7217–7226
Morel JB, Godon C, Mourrain P et al (2002) Fertile hypomorphic ARGONAUTE (ago1)
mutants impaired in post-transcriptional gene silencing and virus resistance. Plant Cell
14:629–639
Mourrain P, Beclin C, Elmayan T et al (2000) Arabidopsis SGS2 and SGS3 genes are required for
posttranscriptional gene silencing and natural virus resistance. Cell 101:533–542
Napoli C, Lemieux C, Jorgensen RA (1990) Introduction of a chimeric chalcone synthase gene
into Petunia results in reversible co-suppression of homologous genes in trans. Plant Cell
2:279–289
Navarro L, Dunoyer P, Jay F et al (2006) A plant miRNA contributes to antibacterial resistance by
repressing auxin signaling. Science 312:436–439
Navarro L, Jay F, Nomura K et al (2008) Suppression of the microRNA pathway by bacterial
effector proteins. Science 321:964–967
Okamura K, Lai EC (2008) Endogenous small interfering RNAs in animals. Nat Rev Mol Cell Biol
9:673–678
Pazhouhandeh M, Dieterle M, Marrocco K et al (2006) F-box-like domain in the polerovirus
protein P0 is required for silencing suppressor function. Proc Natl Acad Sci USA
103:1994–1999
Pfeffer S, Dunoyer P, Heim F et al (2002) P0 of beet Western yellows virus is a suppressor of
posttranscriptional gene silencing. J Virol 76:6815–6824
RNAi Suppression and Its Application 91
Matthias Bauer
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
2 Cellular Sensors of siRNA-Triggered Innate Immune Response . . . . . . . . . . . . . . . . . . . . . . . . . . 96
2.1 TLR-Mediated Innate Immunity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
2.2 Non-TLR-Mediated Innate Immunity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
3 Cellular Sequels After siRNA-Triggered Innate Immune System Activation . . . . . . . . . . . . . . 98
4 Overcoming Synthetic siRNA-Triggered Innate Immune Response . . . . . . . . . . . . . . . . . . . . . . . 99
5 Overcoming shRNA-Triggered Innate Immune Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
6 shRNA-Mediated Disruption of the Endogenous miRNA Machinery . . . . . . . . . . . . . . . . . . . . 102
7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Abstract RNA interference (RNAi) allows selective gene silencing, is widely used
for functional analysis of individual genes in mammalian cells, and represents an
attractive therapeutic option for treating various diseases. However, growing evi-
dence exists that chemically synthesized small interfering RNAs (siRNAs) as well
as promoter-expressed short-hairpin RNAs (shRNAs) may cause cellular toxicity
resulting in unspecific cellular phenotypes and, in case of therapeutic interventions,
severe side effects. Various mechanisms have been identified including the induc-
tion of interferon-stimulated gene (ISG) expression as well as the disruption of
natural microRNA biogenesis and function by competition of exogenous shRNAs
for the endogenous RNAi machinery. This review highlights recent progress in
the understanding of siRNA-triggered toxicity and outlines strategies to prevent
undesirable side effects.
M. Bauer
Department of Neurology, Hertie Institute for Clinical Brain Research, University of T€
ubingen,
T€ubingen, Germany
Department of Protein Sciences, Helmholtz Center M€ unchen, German Research Center for
Environmental Health, M€unchen-Neuherberg, Germany
Institute for Human Genetics, Klinikum rechts der Isar, TU M€unchen, M€unchen, Germany
e-mail: matthias.bauer@helmholtz-muenchen.de
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 93
RNA Technologies, DOI 10.1007/978-3-642-12168-5_4,
# Springer-Verlag Berlin Heidelberg 2010
94 M. Bauer
Abbreviations
1 Introduction
a
Conventional siRNA
N(21)
5′- NNNNNNNNNNNNNNNNNNNNN U U -3′
3′- U U NNNNNNNNNNNNNNNNNNNNN -5′
Antisense strand
b
“1stGeneration” shRNA
N(21) C
U A
5′- NNNNNNNNNNNNNNNNNNNNNU
A
3′- U U NNNNNNNNNNNN NNNNNNNNN A
G G
Antisense strand A
c
miR-30
A
5′- G C G C U G U A A A C A U C C G U C A C U GG A A G C U GU G A A
3′- C GU C G A C G U U U G U A G G C U G A C U U U C GG CAC C G
G UA GA
Drosha Dicer
d
“2ndGeneration” miRNA-designed shRNA
N(13) N(8)
A UC A
5′- G C G N N N N N N N N N N N N N NNNNNNN N C U GUG A
3′- C GU C NNNNNNNNNNNN N NNNNNNN N GG CA C C G
GU AGA
Antisense strand
Fig. 1 Schematic structure of a synthetic siRNA, first and second generation miRNA-designed
shRNA. (a) Structure of a conventional siRNA with a length of 21 nucleotides and 30 dinucleotide
overhangs. (b) first generation shRNA construct with a stem–loop–stem structure (Brummelkamp
et al. 2002). The sequence of the target site (sense orientation) is shown in blue (passenger strand)
and the antisense strand (guide strand) is shown in red. (c) Endogenous miR-30 primary transcript;
putative cleavage sites for Dicer and Drosha are indicated with arrows. (d) second generation
miR-30-designed shRNA construct
the cytosol [for comprehensive overview see Juliano et al. (2008)]. In the case of
widely used cationic lipid reagents, siRNA-containing lipoplexes are internalized
by endocytosis followed by entering the endosomal/lysosomal pathway after cellu-
lar uptake (Doherty and McMahon 2009; Zuhorn et al. 2002). After destabilization
of the endosomal membrane by carrier cationic lipids and subsequent dissociation
from the carrier, dsRNA is released into the cytosol and enters the RNAi pathway
by its incorporation into the RNA-induced silencing complex (RISC) (Carthew and
Sontheimer 2009).
96 M. Bauer
Two years after the first demonstration of RNAi in mammalian cells, two groups
have independently shown that chemically synthesized 21 nucleotide siRNAs
(Sledz et al. 2003) as well as lentivirally expressed shRNAs (Bridge et al. 2003)
resulted in a global upregulation of interferon-stimulated genes (ISG) in targeted
cells. In mammalian cells, several membrane-bound as well as cytosolic receptors
exist for the detection of dsRNAs and ssRNAs, which belong to the cellular
antiviral defence machinery (de Veer et al. 2005; Garcı́a-Sastre and Biron 2006).
Innate immunity to siRNA can be classified as (1) Toll-like receptor (TLR)-
mediated and (2) non-TLR-mediated (Robbins et al. 2009). As a general rule, it is
believed that endogenously expressed shRNAs predominantly activate non-TLR-
mediated dsRNA sensors, whereas exogenously delivered siRNAs, which have
entered the endosomal/lysosomal pathway, may activate both the TLR-mediated
as well as the non-TLR-mediated innate immune system.
Strategies to Prevent siRNA-Triggered Cellular Toxicity 97
Thirteen TLR receptors have been identified in human and mouse cells and three –
TLR3, 7, and 8 – have been linked to siRNA-mediated activation of the innate
immune system so far. In humans, TLR3 is expressed in myeloid dendritic cells
(Muzio et al. 2000) and primary endothelial cells (Kleinman et al. 2008), whereas
its expression in mice is found in a wider range of cell types (Applequist et al.
2002). At the subcellular level, TLR3 is expressed at the cell surface as well as in
the endosome and might therefore detect siRNAs in the blood stream and cationic
lipid-complexed siRNAs after endocytosis by leukocytes. After in vivo delivery of
21 nucleotide siRNAs, TLR3 on endothelial cells has been activated followed by
the expression of several cytokines including IFNg and IL-12 (Kleinman et al.
2008), which then can eventually trigger a more widespread immune response.
TLR7 and TLR8 are expressed in dendritic cells and other immune cell types in
humans and are located in the endosomal membrane (Zarember and Godowski
2002). siRNA-triggered TLR7 activation in primary immune cells resulted in the
induction of type 1 interferon expression and expression of other proinflammatory
cytokines (Karikó et al. 2004; Hornung et al. 2005; Gorden et al. 2005). Unlike
TLR3, TLR7 activation seems to be dependent on the presence of immunostimu-
latory motifs including GU-rich sequences (Judge et al. 2005). Interestingly, TLR8
does not respond to conventional TLR7/8 ligands in murine immune cells (Gorden
et al. 2006). In fact, TLR8 expression has been detected in the embryonic mouse
brain localizing in axonal structures, which points towards a function of TLRs in
murine neuronal development (Ma et al. 2007) and reminds us of the importance to
consider interspecies differences in innate immune activation. Even more complex,
different culture conditions as well as cellular stress may by itself modulate or
induce components of the TLR pathway and may prime cells to respond to siRNAs
in an unintended way (Robbins et al. 2009).
the cytoplasm and is associated with ribosomes (Thomis et al. 1992). Although it
has been previously thought that activation of PKR requires >30 base pair dsRNA,
(Manche et al. 1992), kinase assays have disclosed that synthetic 21 nucleotide
siRNAs can induce PKR to some extent (Zhang et al. 2006; Lemaire et al. 2008).
Activated PKR leads to the phosphorylation of elongation/initiation factor eIF-
2alpha and blocks overall cellular translation (de Veer et al. 2005). siRNA-acti-
vated PKR also activates transcription factors including NFkB, IRFs, and ATF2,
which then trigger interferon (IFN) expression within an affected cell (Garcı́a-
Sastre and Biron 2006). IFN then binds to cell surface receptors in an auto- and/
or paracrine fashion and confers a more global antiviral state by inducing a complex
array of additional IFN-stimulated genes (Garcı́a-Sastre and Biron 2006).
Various pathogen-associated molecular patterns (PAMPs) helicases including
RIG-I have been identified as cytoplasmic sensors of foreign RNAs in human cells.
RIG-I acts as a sensor for either single-stranded or double-stranded blunt-ended
RNAs containing uncapped 50 -triphosphates, which is a characteristic of some viral
RNAs (Hornung et al. 2006) and is present in phage polymerase-transcribed
siRNAs (Kim et al. 2004). When activated, RIG-I interacts with IFNb promoter
stimulator 1 (IPS-1) followed by signal transmission to IRF transcription factors
and NFkB, which then leads to the synthesis of proinflammatory cytokines and type
1 IFNs (Yoneyama et al. 2005).
So far, three members of the OAS family, OAS1 to OAS3, have been identified
(Hovnanian et al. 1998). dsRNA-activated OAS enzymes convert ATP to 20 –50 -
oligoadenylates, which then activate the cellular endoribonuclease RNase L. Acti-
vated RNase L unselectively cleaves mRNAs leading to a general inhibition of
protein synthesis (de Veer et al. 2005) and apoptosis (Li et al. 2004). Synthetical 21
nucleotide dsRNAs (Sledz et al. 2003) as well as shRNA-derived siRNAs (Bridge
et al. 2003) are capable of activating OAS proteins, and it has been shown
previously that even dsRNAs below a length of 20 nucleotides can activate OAS-
mediated immune response (Sarkar et al. 1999).
Depending on the origin and developmental stage, cells are differentially sensitive
to siRNA-triggered IFN response. We have tested over 30 different shRNA-expressing
(Fig. 1b) lentiviral vectors in various cell lines as well as in primary embryonic
mouse neuronal cultures. Whereas none of the shRNA constructs induced an IFN
response in the tested cell lines (i.e., NIH3t3, 293T, HELA, and COS cells), shRNA
expression in primary cultures almost regularly resulted in elevated Oas1 expres-
sion accompanied by significant adverse cellular phenotypes (own unpublished
observations). Moreover, different neuronal cell populations from different parts
of the embryonic brain seem differentially susceptible and even cells derived from
Strategies to Prevent siRNA-Triggered Cellular Toxicity 99
the same anatomical brain region have shown different degrees of IFN response
depending on the embryonic age of the donor (own unpublished observations).
The induction of interferon-stimulated genes (ISGs) by synthetic siRNAs as well
as shRNA-derived siRNAs can result in a multitude of morphological changes and
cellular phenotypes in vitro and in vivo. For example, we have shown previously
that shRNA-triggered activation of innate antiviral response in primary cortical
neurons resulted in neurite retraction and apoptotic cell death (Bauer et al. 2009).
In another study with primary hippocampal neurons, shRNA-induced IFN-response
resulted in more subtle morphological changes with rarefication of dendritic spines
and electrophysiological perturbations (Alvarez et al. 2006). In vivo experiments
with transgenic mice expressing shRNAs directed against arylamine N-acetyltransferase
showed increased early lethality, which was – at least in part – associated with
elevated Oas1 mRNA levels (Cao et al. 2005) and adenovirus-mediated shRNA
expression in the striatum of adult mice resulted in loss of striatal neuron identity
and microglial activation (McBride et al. 2008). However, in the latter study, ISG
expression has not been monitored and therefore other shRNA or vector-related
immunological sequels than IFN response may be involved.
Due to the aforementioned findings and in consideration of the fact that siRNA-
based technologies have been and will be applied extensively in basic research as
well as for therapeutic approaches, there is a strong need for the identification of
determinants triggering IFN response and to invent strategies to prevent cellular
immune response-mediated toxicity.
Over the past years, extensive research has been conducted to identify IFN inducing
determinants of siRNAs, and it became clear that RNAi mediated by different RNA
sources – i.e., synthetic siRNA and shRNA-derived siRNAs – are associated with
different modes of innate immune system activation. One of the obvious advantages
in using synthetic siRNAs is that they are clearly defined in length and structure and
that chemical modifications can be introduced to abrogate or at least reduce
unintended immunostimulatory effects.
Undoubtedly, the size of applied dsRNAs is one of the major factors for the
induction of innate immune response, whereby duplexes shorter than 20 base pairs –
at least in HeLa S3 cells – do not activate PKR and TLR3 even at high RNA
concentrations (100 nM). With increased siRNA length, the RNA concentration
threshold for IFN response induction continuously declined, showing toxic effects
with 23 base pair siRNA at concentrations as low as 10 nM (Reynolds et al. 2006).
In another study, OAS protein family sensors have been activated by 21 base
pair siRNAs at even lower concentrations in RCC1 cells (Sledz et al. 2003),
which strongly suggests the use of 19–21 base pair siRNAs at the lowest effective
concentration.
100 M. Bauer
shRNAs, which are endogenously expressed from plasmidal or viral vectors differ
from exogenously administered siRNAs regarding the involvement of cellular
mechanisms responsible for the activation of innate immune response, and there-
fore different strategies must be taken into consideration. In general, it is believed
that endogenously expressed shRNAs are more toxic to cells than synthetic siRNA
delivery and that some IFN inducing determinants cannot be readily controlled due
to prior endogenous processing of shRNAs before they become functional siRNAs.
For example, due to variable Dicer cleavage and heterogenous 30 -ends caused by
the polyT transcription termination signal, siRNAs derived from polymerase III
promoter-transcribed shRNAs have shown a marked heterogeneity in length
(21–25 nucleotides) (Olejniczak et al. 2009). However, polymerase II and III
driven expression of shRNA is still the method of choice when long-lasting
silencing effects are required and considerable progress has been made to make
shRNA-mediated RNAi a safer and more predictable gene silencing tool.
After the first report in 2003 showing IFN induction after lentivirus-mediated
overexpression of shRNAs in human lung fibroblasts (Bridge et al. 2003), the same
group published a paper in which they systematically modified the U6 promoter
transcription start site and 50 shRNA sequences showing that an AA-dinucleotide
near the transcription start site has been the prime determinant for ISG induction
(Pebernard and Iggo 2004). In contrast, removal of the AA-dinucleotide at the
transcription initiation site in a H1-driven shRNA construct capable to induce ISG
expression in primary cortical neurons did not abolish activation of innate immune
response (Bauer et al. 2009). This finding indicates that U6 promoter-driven shRNA
expression differs from H1-driven constructs with respect to stimulate dsRNA-
triggered immune response and general rules regarding polymerase III promoter
sequences may not have been deduced so far. However, it seems more favorable to
express shRNAs under the control of the H1 polymerase III promoter, since it has
Strategies to Prevent siRNA-Triggered Cellular Toxicity 101
been repeatedly reported that U6 vectors give a higher frequency of ISG induction
(Bridge et al. 2003; Pebernard and Iggo 2004).
In contrast to exogenously administered siRNAs, activation of innate immune
response by specific sequences of the guide and/or passenger strand of shRNA
constructs is less clear. Although chemically synthesized shRNAs may theoreti-
cally activate TLRs within the endosome in the presence of GU-rich immunosti-
mulatory sequence motifs (Karikó et al. 2004), endogenously vector-mediated
expression of shRNAs is rather unlikely to do so, since the resulting dsRNAs
normally do not enter the endosomes, therefore bypassing TLR recognition.
Thus, the presence of immunostimulatory sequence motifs in shRNAs expressed
from either plasmidal or viral vectors may not be the cause and its avoidance
might not be sufficient to circumvent interferon response, when present (Robbins
et al. 2006).
Due to the fact that a multitude of endogenously expressed, cell-derived short
dsRNA species, including miRNAs, are present in the cytosol (H€uttenhofer et al.
2005; Carthew and Sontheimer 2009), it is crucial that cellular sensors, known as
pattern-recognition receptors, can distinguish between self and nonself dsRNAs. It
is therefore reasonable to think that the implementation of design features of
naturally occurring short dsRNAs in shRNA constructs lowers the risk of induction
of innate immune response. A proof of principal study conducted by the Cullen
laboratory in 2002 has shown that the introduction of target gene-derived sequences
within a miRNA-30 backbone was capable of decreasing the complementary
mRNA target in human cells, providing first experimental evidence that artificial
miRNAs can be used as siRNA shuttles (Zeng et al. 2002). Following studies
demonstrated that miRNA-designed or “second generation” shRNAs are (1) more
efficient in gene silencing than “first generation” shRNA constructs (Boden et al.
2004; Bauer et al. 2009) and (2) seem to circumvent induction of ISG expression at
least in primary neurons in vitro (Bauer et al. 2009 and unpublished data). Interest-
ingly, only an exact implementation of all design features of naturally occurring
miRNA-30 precursors, including the introduction of a dinucleotide bulge structure
within the stem duplex was capable of avoiding induction of ISGs (Bauer et al.
2009) (Fig. 1c, d). In addition, switching from first generation to second generation
miRNA-designed shRNA constructs in a study targeting huntingtin expression in
the CNS resulted in reduced microglia activation and inflammation in vivo
(McBride et al. 2008). These studies give a strong indication that the use of
miRNA-based shRNA constructs is favorable with respect to the avoidance of
toxic, immune response-related side-effects in target cells.
In general, it seems advantageous when sh/siRNAs comprise natural features of
endogenous small RNA species (i.e., O-methylation of the ribose sugar backbone of
synthetic siRNAs, inclusion of bulges in the passenger strand of shRNAs, and usage
of miRNA-designed shRNAs) to avoid activation of innate viral defense machin-
ery. Therefore, ongoing research to understand the components and mechanisms of
the endogenous si/miRNA machinery as well as innate viral defense mechanisms
will further help to improve and rationalize the design of synthetic siRNA and
shRNA constructs.
102 M. Bauer
7 Conclusions
References
Alvarez VA, Ridenour DA, Sabatini BL (2006) Retraction of synapses and dendritic spines
induced by off-target effects of RNA interference. J Neurosci 26:7820–7825
Applequist SE, Wallin RP, Ljunggren HG (2002) Variable expression of Toll-like receptor in
murine innate and adaptive immune cell lines. Int Immunol 14:1065–1074
Bartlett DW, Su H, Hildebrandt IJ et al (2007) Impact of tumor-specific targeting on the
biodistribution and efficacy of siRNA nanoparticles measured by multimodality in vivo imag-
ing. Proc Natl Acad Sci USA 104:15549–15554
Bauer M, Kinkl N, Meixner A et al (2009) Prevention of interferon-stimulated gene expression
using microRNA-designed hairpins. Gene Ther 16:142–147
Birmingham A, Anderson EM, Reynolds A et al (2006) 30 UTR seed matches, but not overall
identity, are associated with RNAi off-targets. Nat Methods 3:199–204
Boden D, Pusch O, Silbermann R et al (2004) Enhanced gene silencing of HIV-1 specific siRNA
using microRNA designed hairpins. Nucleic Acids Res 32:1154–1158
Boudreau RL, Monteys AM, Davidson BL (2008) Minimizing variables among hairpin-based
RNAi vectors reveals the potency of shRNAs. RNA 14:1834–1844
Boudreau RL, Martins I, Davidson BL (2009) Artificial microRNAs as siRNA shuttles: improved
safety as compared to shRNAs in vitro and in vivo. Mol Ther 17:169–175
Bridge AJ, Pebernard S, Ducraux A et al (2003) Induction of an interferon response by RNAi
vectors in mammalian cells. Nat Genet 34:263–264
Brummelkamp TR, Bernards R, Agami R (2002) A system for stable expression of short interfer-
ing RNAs in mammalian cells. Science 296:550–553
104 M. Bauer
Cao W, Hunter R, Strnatka D et al (2005) DNA constructs designed to produce short hairpin,
interfering RNAs in transgenic mice sometimes show early lethality and an interferon
response. J Appl Genet 46:217–225
Carthew RW, Sontheimer EJ (2009) Origins and mechanisms of miRNAs and siRNAs. Cell
136:642–655
Castanotto D, Sakurai K, Lingeman R et al (2007) Combinatorial delivery of small interfering
RNAs reduces RNAi efficacy by selective incorporation into RISC. Nucleic Acids Res
35:5154–5164
de Veer MJ, Sledz CA, Williams BR (2005) Detection of foreign RNA: implications for RNAi.
Immunol Cell Biol 83:224–228
Diebold SS, Massacrier C, Akira S et al (2006) Nucleic acid agonists for Toll-like receptor 7 are
defined by the presence of uridine ribonucleotides. Eur J Immunol 36:3256–3267
Doherty GJ, McMahon HT (2009) Mechanisms of endocytosis. Annu Rev Biochem 78:
857–902
Elbashir SM, Harborth J, Lendeckel W et al (2001a) Duplexes of 21-nucleotide RNAs mediate
RNA interference in cultured mammalian cells. Nature 411:494–498
Elbashir SM, Lendeckel W, Tuschl T (2001b) RNA interference is mediated by 21- and 22-
nucleotide RNAs. Genes Dev 15:188–200
Garcı́a-Sastre A, Biron CA (2006) Type 1 interferons and the virus-host relationship: a lesson in
détente. Science 312:879–882
Gorden KB, Gorski KS, Gibson SJ et al (2005) Synthetic TLR agonists reveal functional
differences between human TLR7 and TLR8. J Immunol 174:1259–1268
Gorden KK, Qiu XX, Binsfeld CC et al (2006) Cutting edge: activation of murine TLR8 by a
combination of imidazoquinoline immune response modifiers and polyT oligodeoxynucleo-
tides. J Immunol 177:6584–6587
Grimm D, Streetz KL, Jopling CL et al (2006) Fatality in mice due to oversaturation of cellular
microRNA/short hairpin RNA pathways. Nature 441:537–541
Hébert SS, De Strooper B (2009) Alterations of the microRNA network cause neurodegenerative
disease. Trends Neurosci 32:199–206
Hornung V, Guenthner-Biller M, Bourquin C et al (2005) Sequence-specific potent induction of
IFN-alpha by short interfering RNA in plasmacytoid dendritic cells through TLR7. Nat Med
11:263–270
Hornung V, Ellegast J, Kim S et al (2006) 50 -Triphosphate RNA is the ligand for RIG-I. Science
314:994–997
Hovnanian A, Rebouillat D, Mattei MG et al (1998) The human 20 ,50 -oligoadenylate synthetase
locus is composed of three distinct genes clustered on chromosome 12q24.2 encoding the 100-,
69-, and 40-kDa forms. Genomics 52:267–277
H€uttenhofer A, Schattner P, Polacek N (2005) Non-coding RNAs: hope or hype? Trends Genet
21:289–297
Jackson AL, Linsley PS (2004) Noise amidst the silence: off-target effects of siRNAs? Trends
Genet 20:521–524
Jackson AL, Bartz SR, Schelter J et al (2003) Expression profiling reveals off-target gene
regulation by RNAi. Nat Biotechnol 21:635–637
Judge AD, Sood V, Shaw JR et al (2005) Sequence-dependent stimulation of the mammalian
innate immune response by synthetic siRNA. Nat Biotechnol 23:457–462
Judge AD, Bola G, Lee AC et al (2006) Design of noninflammatory synthetic siRNA mediating
potent gene silencing in vivo. Mol Ther 13:494–505
Juliano R, Alam MR, Dixit V et al (2008) Mechanisms and strategies for effective delivery of
antisense and siRNA oligonucleotides. Nucleic Acids Res 36:4158–4171
Karikó K, Bhuyan P, Capodici J et al (2004) Small interfering RNAs mediate sequence-indepen-
dent gene suppression and induce immune activation by signaling through toll-like receptor 3.
J Immunol 172:6545–6549
Strategies to Prevent siRNA-Triggered Cellular Toxicity 105
Kim DH, Longo M, Han Y et al (2004) Interferon induction by siRNAs and ssRNAs synthesized
by phage polymerase. Nat Biotechnol 22:321–325
Kleinman ME, Yamada K, Takeda A et al (2008) Sequence- and target-independent angiogenesis
suppression by siRNA via TLR3. Nature 452:591–597
Lecellier CH, Dunoyer P, Arar K et al (2005) A cellular microRNA mediates antiviral defense in
human cells. Science 308:557–560
Lemaire PA, Anderson E, Lary J et al (2008) Mechanism of PKR Activation by dsRNA. J Mol Biol
381:351–360
Li G, Xiang Y, Sabapathy K et al (2004) An apoptotic signaling pathway in the interferon antiviral
response mediated by RNase L and c-Jun NH2-terminal kinase. J Biol Chem 279:1123–1131
Ma Y, Haynes RL, Sidman RL et al (2007) TLR8: an innate immune receptor in brain, neurons and
axons. Cell Cycle 6:2859–2868
Manche L, Green SR, Schmedt C et al (1992) Interactions between double-stranded RNA
regulators and the protein kinase DAI. Mol Cell Biol 12:5238–5248
McBride JL, Boudreau RL, Harper SQ et al (2008) Artificial miRNAs mitigate shRNA-mediated
toxicity in the brain: implications for the therapeutic development of RNAi. Proc Natl Acad Sci
USA 105:5868–5873
Merlin G, Chebath J, Benech P et al (1983) Molecular cloning and sequence of partial cDNA for
interferon-induced (20 -50 )oligo(A) synthetase mRNA from human cells. Proc Natl Acad Sci
USA 80:4904–4908
Meurs E, Chong K, Galabru J (1990) Molecular cloning and characterization of the human double-
stranded RNA-activated protein kinase induced by interferon. Cell 62:379–390
Muzio M, Bosisio D, Polentarutti N et al (2000) Differential expression and regulation of toll-like
receptors (TLR) in human leukocytes: selective expression of TLR3 in dendritic cells.
J Immunol 164:5998–6004
Olejniczak M, Galka P, Krzyzosiak WJ (2010) Sequence-non-specific effects of RNA interference
triggers and microRNA regulators. Nucleic Acids Res. 38:1–16
Paddison PJ, Caudy AA, Bernstein E et al (2002) Short hairpin RNAs (shRNAs) induce sequence-
specific silencing in mammalian cells. Genes Dev 16:948–958
Pebernard S, Iggo RD (2004) Determinants of interferon-stimulated gene induction by RNAi
vectors. Differentiation 72:103–111
Provost P, Dishart D, Doucet J (2002) Ribonuclease activity and RNA binding of recombinant
human Dicer. EMBO J 21:5864–5874
Reynolds A, Anderson EM, Vermeulen A et al (2006) Induction of the interferon response by
siRNA is cell type- and duplex length-dependent. RNA 12:988–993
Robbins MA, Li M, Leung I, Li H et al (2006) Stable expression of shRNAs in human CD34þ
progenitor cells can avoid induction of interferon responses to siRNAs in vitro. Nat Biotechnol
24:566–571
Robbins M, Judge A, Liang L et al (2007) 20 -O-methyl-modified RNAs act as TLR7 antagonists.
Mol Ther 15:1663–1669
Robbins M, Judge A, MacLachlan I (2009) siRNA and innate immunity. Oligonucleotides
19:89–102
Sarkar SN, Bandyopadhyay S, Ghosh A et al (1999) Enzymatic characteristics of recombinant
medium isozyme of 20 -50 oligoadenylate synthetase. J Biol Chem 274:1848–1855
Sledz CA, Holko M, de Veer MJ et al (2003) Activation of the interferon system by short-
interfering RNAs. Nat Cell Biol 5:834–839
Thomis DC, Doohan JP, Samuel CE (1992) Mechanism of interferon action: cDNA structure,
expression, and regulation of the interferon-induced, RNA-dependent P1/eIF-2 alpha protein
kinase from human cells. Virology 188:33–46
Tseng YC, Mozumdar S, Huang L (2009) Lipid-based systemic delivery of siRNA. Adv Drug
Deliv Rev 61:721–731
Yi R, Qin Y, Macara IG, Cullen BR (2003) Exportin-5 mediates the nuclear export of pre-
microRNAs and short hairpin RNAs. Genes Dev 17:3011–3026
106 M. Bauer
Yoneyama M, Kikuchi M, Natsukawa T et al (2004) The RNA helicase RIG-I has an essential
function in double-stranded RNA-induced innate antiviral responses. Nat Immunol 5:730–737
Yoneyama M, Kikuchi M, Matsumoto K et al (2005) Shared and unique functions of the DExD/
H-box helicases RIG-I, MDA5, and LGP2 in antiviral innate immunity. J Immunol 175:
2851–2858
Zarember KA, Godowski PJ (2002) Tissue expression of human Toll-like receptors and differen-
tial regulation of Toll-like receptor mRNAs in leukocytes in response to microbes, their
products, and cytokines. J Immunol 168:554–561
Zeng Y, Wagner EJ, Cullen BR (2002) Both natural and designed micro RNAs can inhibit the
expression of cognate mRNAs when expressed in human cells. Mol Cell 9:1327–1333
Zhang Z, Weinschenk T, Guo K et al (2006) siRNA binding proteins of microglial cells: PKR is an
unanticipated ligand. J Cell Biochem 97:1217–1229
Zuhorn IS, Kalicharan R, Hoekstra D (2002) Lipoplex-mediated transfection of mammalian cells
occurs through the cholesterol-dependent clathrin-mediated pathway of endocytosis. J Biol
Chem 277:18021–18028
RNAi in Malignant Brain Tumors: Relevance
to Molecular and Translational Research
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
1.1 Diagnostic Characteristics of Diffuse Astrocytic Tumor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
1.2 Clinical Course . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
1.3 Obstacles in Treatment of Diffuse Astrocytic Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
2 RNAi for Glioma: from Bench to Clinic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
2.1 Preclinical Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
2.2 Target Genes for Silencing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
2.3 Problems in Clinical Translation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
3 miRNAs in Glioma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
3.1 Molecular Pathology of Aberrant miRNAs in Glioblastoma . . . . . . . . . . . . . . . . . . . . . . . 120
4 Conclusions and Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
Abstract Gliomas are the most frequent malignant intracranial tumors arising from
the brain or spinal cord tissue. The most malignant among them is glioblastoma
multiforme (GBM), which constitutes approximately 20–25% of all primary intra-
cranial neoplasms with an incidence of 3–4/100,000. This type of tumor is character-
ized by progressive overgrowth of neoplastic glial cells with widespread and
relentless invasion, resulting in acquisition of resistance to treatment and a poor
prognosis due to recurrence. These features hamper efficient surgical intervention of
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 107
RNA Technologies, DOI 10.1007/978-3-642-12168-5_5,
# Springer-Verlag Berlin Heidelberg 2010
108 M. Nakada et al.
the disease. Current chemo/radiotherapy conditions act sublethally but cannot effec-
tively suppress the proliferation of glioma cells. Despite some progress, the thera-
peutic options are yet limited, and novel therapeutic strategies are clearly needed.
Targeting of disease-specific molecules involved in the proliferation, apoptosis, and
invasion of the tumor cells as well as in tumor angiogenesis may offer a high potential
for the development of more effective therapies for GBM. RNA interference (RNAi)
has been emerging not only for in vitro target validation, but also for a novel
therapeutic strategy based on the highly specific and efficient silencing of a target
gene. Indeed, RNAi has shown to act against GBM efficiently in numerous preclini-
cal studies. Many efforts have been devoted to overcome the three major obstacles
in use of RNAi in vivo; their specificity, instability, and poor cellular delivery
of bioactive small interfering RNA (siRNA) across the blood–brain barrier. The
identification of effective target and the establishment of novel siRNA delivery
systems are needed for the clinical applications of RNAi for treatment of GBM.
1 Introduction
Diffuse astrocytic tumors are the most frequently observed intracranial neoplasms;
they account for more than 60% of all primary brain tumors. A signature character-
istic of astrocytic tumor cells is their histological resemblance to astrocytes.
Therefore, the presence of cells showing fine fibrillary processes and expression
of glial fibrillary acidic protein (GFAP) is a major diagnostic feature for these
tumors. These findings in a certain proportion of tumor cells are important for
astrocytic tumor cell identification, even in the more undifferentiated and anaplastic
form of the tumor. Since the tumor cells typically exhibit diffuse infiltration of the
adjacent brain parenchyma, they are termed diffuse astrocytic tumors.
Although many classifications and grading systems for diffuse astrocytic tumors
have been introduced, the World Health Organization (WHO) classification is
adopted in this chapter because of its widespread recognition (Louis et al. 2007).
According to the WHO classification and grading system, diffuse astrocytic tumors
are divided into three categories: (1) diffuse astrocytoma (WHO grade II); (2) ana-
plastic astrocytoma (WHO grade III); and (3) the “glioblastoma” (GBM; WHO
grade IV), which is the most undifferentiated. GBM is the most malignant among
the astrocytic tumors and is composed of poorly differentiated neoplastic astro-
cytes. The presence of microvascular proliferation and/or necrosis is essential for
histopathological diagnosis of GBM (Fig. 1).
RNAi in Malignant Brain Tumors: Relevance to Molecular and Translational Research 109
1.2.1 Symptoms
Clinical signs and symptoms induced by diffuse astrocytic tumors are dependent on
the site where the tumors locate within the central nervous system (Buckner et al.
2007; Chandana et al. 2008). Diffuse astrocytomas usually develop slowly; how-
ever, some of them may present suddenly with an onset of epilepsy. Frontal lobe
lesions are associated with intellectual deterioration. When the tumor locates more
posteriorly in the hemisphere (into the motor cortex or pyramidal tract), progressive
hemiparesis tends to occur. A more centrally located tumor involving the thalamus
frequently presents with raised intracranial pressure due to hydrocephalus caused
by occlusion of the cerebrospinal-fluid pathway. If located within the brain stem,
the tumor may induce cranial nerve palsies resulting from invasion of the tumor into
the cranial nerve nuclei.
110 M. Nakada et al.
After a varying time course after the diagnosis of diffuse astrocytoma, anaplastic
change occurs in the major proportion of anaplastic astrocytomas. Some cases of
anaplastic astrocytoma are diagnosed as tumors arising de novo after the onset of
symptoms due to increased intracranial pressure, such as headache (typically in the
morning), drowsiness, and vomiting.
As is the case for other grades of diffuse astrocytic tumors, the pattern of GBM
clinical presentation depends on the site of the tumor. In addition, GBM’s rapid
growth pattern is an important sign for preoperative diagnosis. Complications such
as hemorrhage and infarction of the tumor, which are caused by its rapid growth,
may result in sudden and rapid enlargement of the lesion and an increase in
intracranial pressure.
1.2.2 Prognosis
Due to the diffuse infiltrating nature of diffuse astrocytic tumors, complete removal
of the tumor is quite difficult (Nakada et al. 2007). In addition, there are many
regions, such as eloquent cortices (motor cortex, sensory cortex, language cortex,
etc.), thalamus, internal capsule, brain stem, and spinal cord, where tumor excision
RNAi in Malignant Brain Tumors: Relevance to Molecular and Translational Research 111
damages critical neural structures and functions. Damages to these neural structures
are associated with irreversible neurological deficits and impairment of the patient’s
QOL. Due to the infiltrating nature of these tumors, the risk of recurrence or
regrowth from a residual tumor cannot be eliminated. Thus, one of the most
important strategies for adjuvant treatment is the control of residual tumor cells
to minimize invasion and regrowth. In terms of such a tumor control strategy,
conventional treatment with radiation and/or chemotherapies has so far failed to
suppress tumor cell activity completely.
Recent advances in molecular biology and genetic engineering have provided
new insights into tumor cell biology and into the distinct genetic alterations
underlying formation and progression of diffuse astrocytic tumors. Several genetic
alterations have been revealed in diffuse astrocytic tumors (Hayashi et al. 2004;
Ohgaki 2005; Ohgaki and Kleihues 2007; Watanabe et al. 2009; Yan et al. 2009).
The TP53, PDGFRA, and IDH1 alterations and/or their overexpression have been
observed in early stages of diffuse astrocytic tumor development. In addition to the
alteration of these genes, GBM frequently carries the loss of heterozygosity (LOH)
in chromosome 10 and exhibits EGFR, PTEN, and CDKN2A alterations. Moreover,
many studies have demonstrated several other genetic alterations occurring in
diffuse astrocytic tumors. Along with a deeper understanding of the genetic features
of tumor cells, newly developed treatments such as immunotherapy, gene therapy,
and molecular-targeted therapy may challenge the inherent difficulties of diffuse
astrocytic tumor treatment. However, until now, most of these challenges have
failed to achieve their clinical goal of controlling these tumors. In the future, new
treatment directions for diffuse astrocytic tumors will include unique and specific
therapies targeting molecules responsible for tumor cell proliferation, invasion, and
angiogenesis with the assistance of newly developed techniques, including RNA
interference (RNAi) and RNAi-related mechanisms.
performed in vitro, evolving strategies for in vivo application of RNAi will lead to
the efficient use of this targeted strategy for future therapeutic paradigms.
RNAi represents a useful experimental approach for use in future clinical trials,
including RNAi-mediated targeting in vitro and in vivo for functional studies of
various genes whose expressions are shown to be upregulated in tumor tissues.
Preclinical studies, mostly in in vitro models, confirm that RNAi techniques can be
used to silence glioma-related target transcripts (Miyashita et al. 2009). In vivo
studies have also shown favorable outcomes using RNAi targeting of gene products
critical for glioma cell growth (Purow et al. 2005; Saydam et al. 2005; Grzelinski
et al. 2006; Pallini et al. 2006; Gillespie et al. 2007), invasion (Gondi et al. 2004a;
Gondi et al. 2004b; Lakka et al. 2004; Nakada et al. 2006; Gondi et al. 2007),
and angiogenesis (Zhen et al. 2007). These studies show the promising possibility of
developing novel therapeutic approaches against GBM based on RNAi gene targeting.
The RNAi phenomenon consists of a multistep intracellular process that can be
divided into two phases. In the first phase, endogenous or exogenous dsRNA
molecules that are present in the cell are processed through the cleavage activity
of RNase III-type (Dicer) into short 20–30-nucleotide fragments called small
interfering RNAs (siRNAs). In the second step, siRNAs as well as many proteins,
including nucleases and helicase, form the RNA-induced silencing complex
(RISC). Through unwinding of the double-stranded siRNA, this complex becomes
activated with single-stranded, noncoding siRNA, which guides the RISC to its
complementary target mRNA, resulting in its endonucleolytic cleavage (Fig. 2)
(Mathupala et al. 2006).
The applications of RNAi can be mediated by two types of molecules – the
chemically synthesized double-stranded siRNA and the vector-based short hairpin
RNA (shRNA). Effective RNAi was initially demonstrated by the application of
synthetic siRNA (Fire et al. 1998). Although siRNA and shRNA are used to achieve
similar functional outcomes, they are intrinsically different molecules. siRNAs
comprise 21–23-nucleotide dsRNA molecules. Once incorporated into the RISC,
they facilitate the cleavage and degradation of its recognized mRNA. shRNA is an
RNA molecule that contains a sense strand, an antisense strand, and a short-loop
sequence that intervenes between both strands. The complementarity of the sense
and antisense fragments in their sequence allows these RNA molecules to form
hairpin-shaped dsRNA structures. The cloning of shRNA into a vector allows for its
expression by an RNA polymerase III promoter. The expressed shRNA is then
exported into the cytoplasm, where it is processed by Dicer into siRNA and
incorporated into the RISC. The use of shRNA allows for more efficient gene
silencing using endogenous processing machinery as compared with siRNA. The
feasibility of siRNA manufacturing and the transient nature of the effect per dose
optimally meet the treatment requirement of certain diseases in which high vector
doses are required, e.g., viral injections.
RNAi in Malignant Brain Tumors: Relevance to Molecular and Translational Research 113
Dicer
1st step
Processing
cleavage
siRNA
2nd step
Unwinding
Target mRNA
RISC
Fig. 2 Molecular process of RNAi machinery. In the first step, long dsRNA molecules of
exogenous or endogenous origin are cleaved to produce siRNA by the enzyme, Dicer. In the
second step, siRNA molecules are incorporated into a RISC. The duplex RNA is unwound leaving
the antisense strand to guide RISC to complementary mRNA for subsequent endonucleolytic
cleavage. dsRNA: double-stranded RNA, siRNA: small interfering RNA, RISC: RNA-inducing
silencing complex
The target molecules for RNAi therapeutics in glioma should meet the following
characteristics.
1. Target genes are overexpressed in glioma cells.
2. Target genes are not expressed in normal cells in the brain or behave as
bystanders in physiological cells even if the genes are expressed.
3. Target genes are associated with the malignant potential of glioma.
114 M. Nakada et al.
Several issues must first be overcome in order to achieve successful clinical trials;
these issues are discussed herein. The specific knockdown of the target genes is
certainly important. Related undesirable consequences caused by off-target effects
due to lack of specificity and by “interferon response” should be prevented. The
stability of RNAi chemicals in brain is also critical for the continuous silencing of
target genes. The delivery method is the most crucial issue for glioma therapy,
because entry into glioma cells is essential if RNAi chemicals are to exert their
effects.
RNAi in Malignant Brain Tumors: Relevance to Molecular and Translational Research 115
2.3.1 Specificity
RNAi is a natural cellular process through which expression of a targeted gene can
be knocked down with high specificity and selectivity. In order to knockdown target
genes, the single-stranded antisense oligonucleotides were first developed for anti-
mRNA strategies. RNAi has since been proven to target mRNA more efficiently
than antisense techniques. It works more specifically than the antisense oligonucle-
otide to decrease expression of a gene or to eliminate it entirely (Bertrand et al.
2002; Chi et al. 2003; Semizarov et al. 2003). Thus, siRNAs are effective at
concentrations that are several orders of magnitude below the concentrations
typically used in antisense experiments (Meister et al. 2004). Since RNAi is
considered an effective strategy to knockdown gene expression in mammalian
cells, several technical limitations relating to the specificity of the RNAi response
are highlighted below.
Off-Target Effect
Interferon Response
Another obstacle against specific RNAi found during in vitro studies is interferon
response in the targeted cells. This undesirable effect was discovered in studies
116 M. Nakada et al.
2.3.2 Instability
siRNA is fairly unstable in vivo and its half-life in peripheral blood is only a few
minutes after intravenous administration. The short half-lives of siRNA in vivo,
even with chemical modification, prohibit long-term application of RNAi therapeu-
tics (Braasch et al. 2003). In order to obtain a continuous RNAi effect, technical
improvements are necessary to stabilize synthetic siRNA in vivo and to maintain
long-term in vivo expression of shRNA in the targeted cell. One of these improve-
ments is the development of an expression vector-based system, which is the most
effective method available at present for prolonged application of RNAi in vivo.
Many groups have examined the use of both plasmid- and virus-based delivery of
shRNA-generating gene therapy constructs in the mammalian brain and have found
varying shRNA delivery efficiencies. Obviously, once applied in situ, regulation of
the expression of the RNAi within the targeted glioma will be important. Moreover,
the influence of RNAi on the adjacent normal brain tissue should be minimized in
long-term survivors with glioma to prevent any side effects. Tumor cell-specific
control of RNAi is exemplified by a tetracycline-regulatable system of gene
expression (Gossen et al. 1995) that is routinely used in molecular biology research
and has been tested as a potential strategy for controlled expression of siRNA in the
lesion (Szulc et al. 2006).
2.3.3 Delivery
The delivery of si/shRNA to target cells in vivo is a limiting factor that determines
the therapeutic effects of RNAi (Zeng and Cullen 2002). The main obstacle to
delivery is the presence of the blood–brain barrier (BBB) in the cerebral vascula-
ture. The BBB is a physiological mechanism that controls the permeability of brain
capillaries so that some substances, such as certain drugs, are prevented from
entering the brain tissue while other substances are allowed to enter freely. This
barrier functions in the following two ways. First, the adjoining capillary endothe-
lial cells are sealed together at their edges by tight junctions that form a mechanical
RNAi in Malignant Brain Tumors: Relevance to Molecular and Translational Research 117
Vessels
BBB
RNAi
Viral
vehicle
Non- viral
Glioma cells
endocytosis
Fig. 3 Schematic illustration of the routs for delivery of RNAi chemicals to glioma cells in vivo
barrier. These barriers prevent water-soluble substances in the blood from extra-
vasating and entering the extra- and intracellular fluid in the brain. Such excluded
substances include siRNA (where the average molecular weight is around 14 kDa)
and large plasmid vector-based shRNA-generating constructs. Second, these capil-
laries are enclosed by the flattened “end-feet” of astrocytes, which also act as a
functional barrier. These types of barriers need to be bypassed by any therapeutic
strategy, including siRNA therapeutics. Numerous efforts have been made to
bypass the BBB for efficient delivery of si/shRNA to the brain (Pardridge 2004;
Mathupala et al. 2006; Pardridge 2007). These delivery systems are divided mainly
into systemic and local routes (Fig. 3).
Systemic Route
The best method for selective delivery of RNAi-based agents to glioma tissue is
intravenous injection of them by using a virus or nonvirus system as a vehicle. Viral
vectors are often used in a laboratory setting for delivery of shRNA to cells and
rodents because of their high transfection efficiency and effective integration of
exogenous DNA into the host genome (Saydam et al. 2005). However, concerns
over safety issues and recent reports about immunogenicity (Check 2002; Nguyen
et al. 2008) have slowed the clinical application of viral vectors. Nonviral poly-
meric delivery, in particular delivery using biodegradable vehicles, is much safer
118 M. Nakada et al.
than a viral delivery system although its transfection efficiency is generally lower
(Aigner 2007; Luten et al. 2008). There are four major classes of vehicles for
nonviral delivery systems: liposomes (Zhang et al. 2003; Pardridge 2004; Zhang
et al. 2004), biopolymers (Lesniak et al. 2005), nanoparticles (Lesniak 2005;
Moreira et al. 2008), and avidin–biotin complexes (Xia et al. 2007). Many of the
vehicles that have recently shown promise are actually a combination of these
classes.
Liposomes are commonly used vehicles in which an aqueous substance is
entirely enclosed by a membrane composed of phospholipid molecules. They can
entrap materials for delivery in their aqueous compartment (water-soluble materials
such as RNAi chemicals) as well as within the membrane (organic solvent-soluble
materials). The presence of hydrophilic polymers on the surface of the liposomes
gives rise to a steric barrier that inhibits the adsorption of blood components.
Studies using immune liposomes in vivo have demonstrated the feasibility of this
approach for glioma therapy (Zhang et al. 2003; Pardridge 2004; Zhang et al. 2004).
Lipid-based nanoparticles are also potential vehicles for the delivery of shRNA and
siRNA. These nanoparticles mimic low-density lipoproteins (LDLs) and thus
interact with the LDL receptors on endothelial cells, resulting in their uptake across
the BBB (Lesniak 2005).
One strategy that is applicable to delivery of vehicles containing RNAi chemi-
cals to the brain via a systemic route is the hydrodynamic approach, which enables
distribution of the RNAi chemicals in various organs by rapid high-dose injections
of these chemicals (McCaffrey et al. 2002; Campbell et al. 2008). Since the new
vessels formed via angiogenesis in GBM are usually immature and have a less
functional BBB, “naked” RNAi chemicals may pass through the pathological
vascular endothelial layer.
Local Route
The best way to allow water-soluble substances to cross the BBB is to bypass the
walls of cerebral capillaries. Local route application obviates any concerns about
the BBB and makes it possible to achieve very high concentrations of chemothera-
peutic agents at the tumor site and to avoid the side effects associated with systemic
high-dose chemotherapy. Direct delivery of RNAi chemicals to the residual tumor
bed and/or dead space formed by tumor resection may be an effective option. The
two approaches that have yielded promising results are polymerically controlled
release and convection-enhanced delivery (CED). Controlled-release polymers are
implanted directly at the resection site to allow for restricted slow release of
chemotherapeutic agents into the residual tumor bed (Sampath and Brem 1998).
Similar to the Gliadel wafers routinely used for localized delivery of the chemo-
therapeutic agent carmustine (BCNU), this strategy involves the entrapment of
siRNA in biodegradable polymers for effective delivery (Valtonen et al. 1997).
CED is an alternative method of controlled local drug release that has been
RNAi in Malignant Brain Tumors: Relevance to Molecular and Translational Research 119
developed to deliver compounds throughout the brain. This method overcomes the
diffusion barrier seen with polymerically controlled release systems (Bobo et al.
1994) by using an applied external positive pressure infusion to generate fluid
convection in the brain, thus distributing chemotherapeutic drugs throughout the
parenchyma via interstitial spaces. Currently, the CED system is actively employed
in clinical trials (Stukel and Caplan 2009).
Even if RNAi chemicals are successfully distributed to the glioma, these che-
micals need to penetrate glioma cell membranes selectively in order to knockdown
the target genes. Efficiency of tumor targeting and cell entry are enhanced by
modification of the vehicle with targeting moieties, such as monoclonal antibodies,
peptides, small molecule ligands, and aptamers, which recognize tumor cell-specific
surface markers (Hughes and Rao 2005; Vorhies and Nemunaitis 2007). The
positive charge of nonviral delivery vehicles facilitates complex formation with
negatively charged nucleic acids and binding to the negatively charged glycocalyx
on external cell membranes, thereby promoting endocytosis. Within cells, the
vehicle’s positive charge facilitates immediate escape from the endosome (Godbey
et al. 2000; Thomas and Klibanov 2002). However, although the positive charge of
these vehicles improves their transfection efficiency, it is also associated with
increased toxicity (Zhang et al. 2007; Kim et al. 2009).
3 miRNAs in Glioma
miRNAs (also called miRs) are a recently discovered class of small (18–25
nucleotides in length), noncoding RNAs that modulate gene expression posttran-
scriptionally. Recently, alterations in miRNAs and their relevant targets have been
identified in various types of cancers, including GBM (Nicoloso and Calin 2008;
Lawler and Chiocca 2009). The global expression profile of GBM-related miRNAs
was first determined by Ciafre et al. (Ciafre et al. 2005) using a microarray
technique. They found upregulation of miR-221 and downregulation of miR-128,
miR-181a, miR-181b, and miR-181c. Since then, aberrant expression of miRNAs in
GBM has been reported, including upregulation of miR-21 (Chan et al. 2005), miR-26a
(Huse et al. 2009), miR-125b (Xia et al. 2009c), miR-221-222 cluster (le Sage et al.
2007), and miR-296 (Wurdinger et al. 2008), and downregulation of miR-7 (Kefas
et al. 2008), miR-124/137 (Silber et al. 2008), and miR-128 (Godlewski et al. 2008).
In tumors, downregulated miRNAs target oncogenes while upregulated miRNAs
target tumor-suppressor genes.
Each miRNA is identified and annotated according to the guidelines proposed by
Ambros et al. (2003). The Rfam database (http://rfam.janelia.org/) is provided as an
online clearinghouse for annotation of RNA families, including miRNAs. In the
miRNA database and registry miRBase (http://www.mirbase.org/), 10,883 entries
have been registered as of September, 2009.
120 M. Nakada et al.
miR-21
One of the most frequently and highly expressed miRNAs in various types of
malignancies, miR-21 plays important roles in tumor cell proliferation, invasion,
angiogenesis, and metastasis (Calin et al. 2005; Chan et al. 2005; Landgraf et al.
2007; Si et al. 2007; Zhu et al. 2008). Chan et al. (Chan et al. 2005) found marked
elevation of miR-21 levels in GBM tissues and cell lines as compared with non-
neoplastic human brain tissues and nonneoplastic cultured glial cells, respectively.
This miRNA acts as an antiapoptotic factor; pathway analyses of computer-
generated lists of miR-21 target genes have shown that miR-21 targets a network
of p53, TGF-b, and mitochondrial apoptosis factors in GBM (Papagiannakopoulos
et al. 2008). miR-21 promotes glioma invasion by targeting matrix metalloprotei-
nase (MMP) regulators and by suppressing RECK (a reversion-inducing cysteine-
rich protein with Kazal motifs, a tumor suppressor gene) and TIMP3 (a tissue
inhibitor of metalloproteinase 3) (Gabriely et al. 2008). Although their functions are
not sufficiently known in GBM, current studies have identified PTEN (phosphatase
and tensin homologue) (Meng et al. 2007), TPM1 (tropomyosin 1) (Zhu et al. 2007),
PDCD4 (programmed cell death 4) (Frankel et al. 2008), and Bcl-2 (Si et al. 2007)
as direct targets of miR-21. Thus, overexpression of miR-21 is pathological in
GBM, which provides a significant survival advantage to tumor cells.
miR-26a
Huse et al. (Huse et al. 2009) showed that miR-26a was overexpressed in a subset of
high-grade gliomas and that it directly targets the PTEN transcript. They also
demonstrated that overexpression of miR-26a in glioma primarily resulted from
amplification at the DNA level, a genomic event strongly associated with mono-
allelic PTEN loss. This association suggests that amplification of miR-26a may
contribute to silencing residual PTEN transcript in PTEN+/ tumors, which is
analogous to a loss-of-heterozygosity event (Huse et al. 2009).
miR-125b
has also been observed in oligodendroglial tumors (Nelson et al. 2006). The
miRNAs – miR-125b1 and miR-125b2 – are precursors of a mature miRNA
sequence designated as miR125b, but the sequences flanking the mature miRNA
are different. The expression of miR-125b1 was significantly upregulated in GBMs
as compared with normal brain tissues (Ciafre et al. 2005); however, Lee et al. (Lee
et al. 2005) revealed that miR-125b2 was the source of most of the miR-125b they
detected in human cell lines. In glioma U373 cells treated by all-trans-retinoic acid
(ATRA), an inverse correlation between the expression of miR-125b and the cell
apoptosis-related protein Bcl-2 modifying factor (Bmf) was shown, and both miR-
125b1 and miR-125b2 restored cell viability and inhibited cell apoptosis (Xia et al.
2009c).
miR-221-222 Cluster
miR-221 and miR-222, both of which are encoded by their respective genes on
chromosome Xp11.3, were found to be upregulated in GBM (Ciafre et al. 2005) as
well as in papillary thyroid carcinomas (He et al. 2005), hepatocellular carcinomas
(Gramantieri et al. 2007), breast cancers (Miller et al. 2008; Zhao et al. 2008a), and
prostate cancers (Galardi et al. 2007). The miR-221-222 cluster downregulates
p27/Kip1 (Zhang et al. 2009) and p57/Kip2 – well-known cyclin-dependent kinase
inhibitors (Medina et al. 2008). In a large proportion of GBM, elevated levels of
miR-221 and miR-222 are correlated with low levels of its target p27/Kip1 tran-
script (le Sage et al. 2007).
miR-296
miR-7
miR-15
Deregulation of miR-15 in GBM tissue samples obtained from Chinese patients was
investigated by Xia et al. (Xia et al. 2009a). They observed that overexpression of
miR-15b in glioma cell lines U87 and U118 resulted in cell cycle arrest at the
G0/G1 phase, while suppression of miR-15b expression resulted in a decrease of
cell populations in the G0/G1 phase and a proportional increase of cell populations
in the S phase. They also showed that CCNE1 (encoding cyclin E1) is one of the
downstream targets of miR-15b. These findings indicate that miR-15b regulates
cell cycle progression in glioma cells by targeting cell cycle-related molecules
(Xia et al. 2009a).
miR-124 and miR-137 are known to belong to the family of upregulated miRNAs in
developing neuronal cells (Sempere et al. 2004; Visvanathan et al. 2007). They are
markedly downregulated in high-grade glioma and are associated with upregulation
of CDK6 (cyclin-dependent kinase 6) (Silber et al. 2008). miR-124 and miR-137
collectively inhibit proliferation of GBM cells and induce differentiation of brain
tumor stem cells by inhibiting G1 cell cycle arrest (Silber et al. 2008).
miR-128
miR-146b
miR-181 Family
Using microarray and northern blot analyses, Ciafre et al. found that expression of
miR-181a, b, and c were significantly downregulated in primary GBMs and in
human GBM cell lines as compared with normal brain tissue (Ciafre et al. 2005).
Shi et al. showed that miR-181a and miR-181b function as tumor suppressors that
trigger growth inhibition, induce apoptosis, and inhibit invasion of glioma cells (Shi
et al. 2008). They also revealed that the tumor-suppressive effect of miR-181b was
stronger than miR-181a. While their target genes in GBM are not yet sufficiently
understood, the miR-181 family has been shown to be one of the regulators of the
Tcl-1 oncogene in B-cell chronic lymphocytic leukemia (Pekarsky et al. 2006).
Distinct genes, miR-181a1 and miR-181a2, and miR-181b1 and miR-181b2 encode
mature miR-181a and miR-181b miRNAs, respectively (Table 1).
Our understanding and application of RNAi has dramatically advanced since the
RNAi phenomenon was discovered in 1998. The promise of RNAi as a future
clinical strategy in targeting glioma has been sufficiently demonstrated at bench side.
As with any novel therapeutic tool, it is evident that several issues, such as target
selection, effector potency, off-target effects, and delivery vehicle design, must be
addressed and resolved before future application of these RNA technologies for
treatment of refractory diseases, as represented by GBM, can be implemented.
References
Aigner A (2007) Nonviral in vivo delivery of therapeutic small interfering RNAs. Curr Opin Mol
Ther 9:345–352
Ambros V, Bartel B, Bartel DP et al (2003) A uniform system for microRNA annotation. RNA
9:277–279
Bates DC, Sin WC, Aftab Q et al (2007) Connexin43 enhances glioma invasion by a mechanism
involving the carboxy terminus. Glia 55:1554–1564
Bertrand JR, Pottier M, Vekris A et al (2002) Comparison of antisense oligonucleotides and
siRNAs in cell culture and in vivo. Biochem Biophys Res Commun 296:1000–1004
Bigner SH, Humphrey PA, Wong AJ et al (1990) Characterization of the epidermal growth factor
receptor in human glioma cell lines and xenografts. Cancer Res 50:8017–8022
Boado RJ (2005) RNA interference and nonviral targeted gene therapy of experimental brain
cancer. NeuroRx 2:139–150
Bobo RH, Laske DW, Akbasak A et al (1994) Convection-enhanced delivery of macromolecules
in the brain. Proc Natl Acad Sci USA 91:2076–2080
Braasch DA, Jensen S, Liu Y et al (2003) RNA interference in mammalian cells by chemically-
modified RNA. Biochemistry 42:7967–7975
Bridge AJ, Pebernard S, Ducraux A et al (2003) Induction of an interferon response by RNAi
vectors in mammalian cells. Nat Genet 34:263–264
RNAi in Malignant Brain Tumors: Relevance to Molecular and Translational Research 125
Gondi CS, Lakka SS, Dinh DH et al (2007) Intraperitoneal injection of a hairpin RNA-expressing
plasmid targeting urokinase-type plasminogen activator (uPA) receptor and uPA retards
angiogenesis and inhibits intracranial tumor growth in nude mice. Clin Cancer Res
13:4051–4060
Gossen M, Freundlieb S, Bender G et al (1995) Transcriptional activation by tetracyclines in
mammalian cells. Science 268:1766–1769
Gramantieri L, Ferracin M, Fornari F et al (2007) Cyclin G1 is a target of miR-122a, a microRNA
frequently down-regulated in human hepatocellular carcinoma. Cancer Res 67:6092–6099
Grzelinski M, Urban-Klein B, Martens T et al (2006) RNA interference-mediated gene silencing
of pleiotrophin through polyethylenimine-complexed small interfering RNAs in vivo exerts
antitumoral effects in glioblastoma xenografts. Hum Gene Ther 17:751–766
Hayashi Y, Yamashita J, Watanabe T (2004) Molecular genetic analysis of deep-seated glioblastomas.
Cancer Genet Cytogenet 153:64–68
He H, Jazdzewski K, Li W et al (2005) The role of microRNA genes in papillary thyroid
carcinoma. Proc Natl Acad Sci U S A 102:19075–19080
Hoelzinger DB, Nakada M, Demuth T et al (2008) Autotaxin: a secreted autocrine/paracrine factor
that promotes glioma invasion. J Neurooncol 86:297–309
Hughes JA, Rao GA (2005) Targeted polymers for gene delivery. Expert Opin Drug Deliv
2:145–157
Huse JT, Brennan C, Hambardzumyan D et al (2009) The PTEN-regulating microRNA miR-26a is
amplified in high-grade glioma and facilitates gliomagenesis in vivo. Genes Dev 23:1327–1337
Jackson AL, Bartz SR, Schelter J et al (2003) Expression profiling reveals off-target gene
regulation by RNAi. Nat Biotechnol 21:635–637
Jiang W, Xiang C, Cazacu S et al (2008) Insulin-like growth factor binding protein 7 mediates
glioma cell growth and migration. Neoplasia 10:1335–1342
Karpel-Massler G, Schmidt U, Unterberg A et al (2009) Therapeutic inhibition of the epidermal
growth factor receptor in high-grade gliomas: where do we stand? Mol Cancer Res
7:1000–1012
Kefas B, Godlewski J, Comeau L et al (2008) MicroRNA-7 inhibits the epidermal growth factor
receptor and the Akt pathway and is down-regulated in glioblastoma. Cancer Res
68:3566–3572
Kim Y, Tewari M, Pajerowski JD et al (2009) Polymersome delivery of siRNA and antisense
oligonucleotides. J Control Release 134:132–140
Kita D, Ciernik IF, Vaccarella S et al (2009) Age as a predictive factor in glioblastomas:
population-based study. Neuroepidemiology 33:17–22
Kuan CT, Wikstrand CJ, Bigner DD (2001) EGF mutant receptor vIII as a molecular target in
cancer therapy. Endocr Relat Cancer 8:83–96
Lakka SS, Gondi CS, Yanamandra N et al (2004) Inhibition of cathepsin B and MMP-9 gene
expression in glioblastoma cell line via RNA interference reduces tumor cell invasion, tumor
growth and angiogenesis. Oncogene 23:4681–4689
Landgraf P, Rusu M, Sheridan R et al (2007) A mammalian microRNA expression atlas based on
small RNA library sequencing. Cell 129:1401–1414
Lawler S, Chiocca EA (2009) Emerging functions of microRNAs in glioblastoma. J Neurooncol
92:297–306
le Sage C, Nagel R, Egan DA et al (2007) Regulation of the p27(Kip1) tumor suppressor by
miR-221 and miR-222 promotes cancer cell proliferation. EMBO J 26:3699–3708
Lee YS, Kim HK, Chung S et al (2005) Depletion of human micro-RNA miR-125b reveals that it
is critical for the proliferation of differentiated cells but not for the down-regulation of putative
targets during differentiation. J Biol Chem 280:16635–16641
Lesniak MS (2005) Novel advances in drug delivery to brain cancer. Technol Cancer Res Treat
4:417–428
Lesniak MS, Upadhyay U, Goodwin R et al (2005) Local delivery of doxorubicin for the treatment
of malignant brain tumors in rats. Anticancer Res 25:3825–3831
RNAi in Malignant Brain Tumors: Relevance to Molecular and Translational Research 127
Lin W, Zhang J, Zhang J et al (2008) RNAi-mediated inhibition of MSP58 decreases tumor growth,
migration and invasion in a human glioma cell line. J Cell Mol Med 13(11–12):4608–4622
Louis DN, Ohgaki H, Wiestler OD et al (2007) The 2007 WHO classification of tumours of the
central nervous system. Acta Neuropathol 114:97–109
Luten J, van Nostrum CF, De Smedt SC et al (2008) Biodegradable polymers as non-viral carriers
for plasmid DNA delivery. J Control Release 126:97–110
Mathupala SP, Guthikonda M, Sloan AE (2006) RNAi based approaches to the treatment of
malignant glioma. Technol Cancer Res Treat 5:261–269
McCaffrey AP, Meuse L, Pham TT et al (2002) RNA interference in adult mice. Nature 418:38–39
Medina R, Zaidi SK, Liu CG et al (2008) MicroRNAs 221 and 222 bypass quiescence and
compromise cell survival. Cancer Res 68:2773–2780
Meister G, Landthaler M, Dorsett Y et al (2004) Sequence-specific inhibition of microRNA- and
siRNA-induced RNA silencing. RNA 10:544–550
Meng F, Henson R, Wehbe-Janek H et al (2007) MicroRNA-21 regulates expression of the PTEN
tumor suppressor gene in human hepatocellular cancer. Gastroenterology 133:647–658
Miller TE, Ghoshal K, Ramaswamy B et al (2008) MicroRNA-221/222 confers tamoxifen
resistance in breast cancer by targeting p27Kip1. J Biol Chem 283:29897–29903
Miyashita K, Kawakami K, Nakada M et al (2009) Potential therapeutic effect of glycogen
synthase kinase 3b inhibition against human glioblastoma. Clin Cancer Res 15:887–897
Moreira JN, Santos A, Moura V et al (2008) Non-viral lipid-based nanoparticles for targeted
cancer systemic gene silencing. J Nanosci Nanotechnol 8:2187–2204
Nakada M, Drake KL, Nakada S et al (2006) Ephrin-B3 ligand promotes glioma invasion through
activation of Rac1. Cancer Res 66:8492–8500
Nakada M, Nakada S, Demuth T et al (2007) Molecular targets of glioma invasion. Cell Mol Life
Sci 64:458–478
Nelson PT, Baldwin DA, Kloosterman WP et al (2006) RAKE and LNA-ISH reveal microRNA
expression and localization in archival human brain. RNA 12:187–191
Nguyen T, Menocal EM, Harborth J et al (2008) RNAi therapeutics: an update on delivery. Curr
Opin Mol Ther 10:158–167
Nicoloso MS, Calin GA (2008) MicroRNA involvement in brain tumors: from bench to bedside.
Brain Pathol 18:122–129
Ohgaki H (2005) Genetic pathways to glioblastomas. Neuropathology 25:1–7
Ohgaki H (2009) Epidemiology of brain tumors. Methods Mol Biol 472:323–342
Ohgaki H, Kleihues P (2007) Genetic pathways to primary and secondary glioblastoma. Am J
Pathol 170:1445–1453
Pallini R, Sorrentino A, Pierconti F et al (2006) Telomerase inhibition by stable RNA interference
impairs tumor growth and angiogenesis in glioblastoma xenografts. Int J Cancer
118:2158–2167
Papagiannakopoulos T, Shapiro A, Kosik KS (2008) MicroRNA-21 targets a network of key
tumor-suppressive pathways in glioblastoma cells. Cancer Res 68:8164–8172
Pardridge WM (2004) Intravenous, non-viral RNAi gene therapy of brain cancer. Expert Opin Biol
Ther 4:1103–1113
Pardridge WM (2007) shRNA and siRNA delivery to the brain. Adv Drug Deliv Rev 59:141–152
Pekarsky Y, Santanam U, Cimmino A et al (2006) Tcl1 expression in chronic lymphocytic
leukemia is regulated by miR-29 and miR-181. Cancer Res 66:11590–11593
Purow BW, Haque RM, Noel MW et al (2005) Expression of notch-1 and its ligands, delta-like-1
and jagged-1, is critical for glioma cell survival and proliferation. Cancer Res 65:2353–2363
Rao DD, Vorhies JS, Senzer N et al (2009) siRNA vs. shRNA: similarities and differences. Adv
Drug Deliv Rev 61:746–759
Salhia B, Rutten F, Nakada M et al (2005) Inhibition of Rho-kinase affects astrocytoma morphol-
ogy, motility, and invasion through activation of Rac1. Cancer Res 65:8792–8800
Salhia B, Tran NL, Chan A et al (2008) The guanine nucleotide exchange factors trio, Ect2, and
Vav3 mediate the invasive behavior of glioblastoma. Am J Pathol 173:1828–1838
128 M. Nakada et al.
Sampath P, Brem H (1998) Implantable slow-release chemotherapeutic polymers for the treatment
of malignant brain tumors. Cancer Control 5:130–137
Saxena S, Jonsson ZO, Dutta A (2003) Small RNAs with imperfect match to endogenous mRNA
repress translation. Implications for off-target activity of small inhibitory RNA in mammalian
cells. J Biol Chem 278:44312–44319
Saydam O, Glauser DL, Heid I et al (2005) Herpes simplex virus 1 amplicon vector-mediated
siRNA targeting epidermal growth factor receptor inhibits growth of human glioma cells
in vivo. Mol Ther 12:803–812
Scacheri PC, Rozenblatt-Rosen O, Caplen NJ et al (2004) Short interfering RNAs can induce
unexpected and divergent changes in the levels of untargeted proteins in mammalian cells.
Proc Natl Acad Sci USA 101:1892–1897
Semizarov D, Frost L, Sarthy A et al (2003) Specificity of short interfering RNA determined
through gene expression signatures. Proc Natl Acad Sci USA 100:6347–6352
Sempere LF, Freemantle S, Pitha-Rowe I et al (2004) Expression profiling of mammalian micro-
RNAs uncovers a subset of brain-expressed microRNAs with possible roles in murine and
human neuronal differentiation. Genome Biol 5:R13
Shi L, Cheng Z, Zhang J et al (2008) hsa-mir-181a and hsa-mir-181b function as tumor suppres-
sors in human glioma cells. Brain Res 1236:185–193
Si ML, Zhu S, Wu H et al (2007) miR-21-mediated tumor growth. Oncogene 26:2799–2803
Silber J, Lim DA, Petritsch C et al (2008) miR-124 and miR-137 inhibit proliferation of glioblas-
toma multiforme cells and induce differentiation of brain tumor stem cells. BMC Med 6:14
Sledz CA, Holko M, de Veer MJ et al (2003) Activation of the interferon system by short-
interfering RNAs. Nat Cell Biol 5:834–839
Smirnova L, Grafe A, Seiler A et al (2005) Regulation of miRNA expression during neural cell
specification. Eur J Neurosci 21:1469–1477
Stegh AH, Kim H, Bachoo RM et al (2007) Bcl2L12 inhibits post-mitochondrial apoptosis
signaling in glioblastoma. Genes Dev 21:98–111
Stukel JM, Caplan MR (2009) Targeted drug delivery for treatment and imaging of glioblastoma
multiforme. Expert Opin Drug Deliv 6:705–718
Szulc J, Wiznerowicz M, Sauvain MO et al (2006) A versatile tool for conditional gene expression
and knockdown. Nat Methods 3:109–116
Thomas M, Klibanov AM (2002) Enhancing polyethylenimine’s delivery of plasmid DNA into
mammalian cells. Proc Natl Acad Sci USA 99:14640–14645
Tran NL, McDonough WS, Savitch BA et al (2006) Increased fibroblast growth factor-inducible
14 expression levels promote glioma cell invasion via Rac1 and nuclear factor-kappaB and
correlate with poor patient outcome. Cancer Res 66:9535–9542
Valtonen S, Timonen U, Toivanen P et al (1997) Interstitial chemotherapy with carmustine-loaded
polymers for high-grade gliomas: a randomized double-blind study. Neurosurgery 41:44–48,
discussion 48–49
Visvanathan J, Lee S, Lee B et al (2007) The microRNA miR-124 antagonizes the anti-neural
REST/SCP1 pathway during embryonic CNS development. Genes Dev 21:744–749
Vollmann A, Vornlocher HP, Stempfl T et al (2006) Effective silencing of EGFR with RNAi
demonstrates non-EGFR dependent proliferation of glioma cells. Int J Oncol 28:1531–1542
Vorhies JS, Nemunaitis J (2007) Nonviral delivery vehicles for use in short hairpin RNA-based
cancer therapies. Expert Rev Anticancer Ther 7:373–382
Wang HB, Zhang SY, Wang S et al (2009) REV3L confers chemoresistance to cisplatin in human
gliomas: the potential of its RNAi for synergistic therapy. Neuro Oncol 11(6):790–802
Watanabe T, Nobusawa S, Kleihues P et al (2009) IDH1 mutations are early events in the
development of astrocytomas and oligodendrogliomas. Am J Pathol 174:1149–1153
Webster RJ, Giles KM, Price KJ et al (2009) Regulation of epidermal growth factor receptor
signaling in human cancer cells by microRNA-7. J Biol Chem 284:5731–5741
RNAi in Malignant Brain Tumors: Relevance to Molecular and Translational Research 129
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
2 A New Approach to Understanding Normal Function of Wild-Type Huntingtin . . . . . . . . . 133
3 Nonallele-Specific Silencing of Huntingtin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
4 Allele-Specific Silencing of Mutant Huntingtin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
5 Current Challenges and Future Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 131
RNA Technologies, DOI 10.1007/978-3-642-12168-5_6,
# Springer-Verlag Berlin Heidelberg 2010
132 Y. Zhang and R.M. Friedlander
1 Introduction
in striatal neurons (Inagaki et al. 2008). Granule cells lost their resistance to the
toxicity of mutant HD when Hsp70 was inhibited by siRNA (Tagawa et al. 2007).
SiRNA-mediated suppression of Omi-HtrA accelerated mutant htt-induced cell
death of primary neurons (Inagaki et al. 2008). Hsp70 is a chaperone protein that
protects neurons from the toxicity of mutant HD (Chai et al. 1999; Warrick et al.
1999; Zhou et al. 2001; Wacker et al. 2004). Omi/HtrA2, a mitochondrial protein
essential to the maintenance of mitochondrial homeostasis, suppresses apoptosis of
neurons by preventing the accumulation of activated Bax in the mitochondrial outer
membrane (Chao et al. 2008). Neuron type – selective dysregulation of Hsp70 and
Omi-HtrA2 might be related to striatal neuron-specific pathology in HD (Tagawa
et al. 2007; Inagaki et al. 2008).
Besides mitochondria, the functions of huntingtin have also been related to other
organelles such as the endoplasmic reticulum (ER) and Golgi apparatus and
they are cell-type specific (Omi et al. 2005; del Toro et al. 2009). Depletion of
endogenous wild-type huntingtin by siRNA in N2a and T98G cells resulted in a
congregated ER network as shown by staining with anticalnexin, while other
organelles showed no significant changes when immunostained with their specific
markers (Omi et al. 2005). However, depletion of huntingtin by RNAi in HeLa cells
results in partial Golgi disruption (Caviston et al. 2007). Some huntingtin might
associate with dynein and facilitate dynein-based vesicle motility (Caviston et al.
2007). Using full-length unspliced genomic HD it was found that huntingtin coloca-
lized in the cytoplasm with trans-Golgi and clathrin-coated vesicles (Strehlow et al.
2007). Another important observation regarding Hdh-null mouse embryonic stem
cells and neurons is that they have fewer transcripts destined for the extracellular
space and show less lysosomal activity and apoptosis than their normal counterparts.
It has been suggested that huntingtin plays a role in trafficking a discrete set of
proteins between Golgi and the extracellular space (Strehlow et al. 2007). Optineurin
is a huntingtin-interacting protein colocalized with huntingtin in the Golgi apparatus.
Knocking down wild-type huntingtin by siRNA resulted in a more diffuse distribu-
tion of optineurin and GT-YFP (a special label for the Golgi apparatus) and decreased
mannose-6-phosphate receptor postGolgi transport (del Toro et al. 2009). Optineurin
and Rab8 form a complex that regulates postGolgi transport. While huntingtin is
crucial for localization of optineurin, disruption of huntingtin by RNAi results in
partial Golgi disruption and impaired optineurin/Rab8-dependent postGolgi traffick-
ing to lysosomes (del Toro et al. 2009). These reports indicate an important role for
huntingtin in the regulation of postGolgi trafficking.
RNAi was also used in a high-throughput screen for modifiers of aggregate
formation in Drosophila larval CNS-derived cells expressing mutant htt exon 1
with 62 CAG repeats. The 68 candidates from this screen were further analyzed in
degenerative Drosophila eyes in in vivo models of HD. The degeneration of fly eyes
is characterized by loss of pigment cells and increased formation of dark necrotic
spots on the external eye. In a second functional screen, 21 genes were found to
have effects on mutant huntingtin-induced toxicity in fly eyes. These newly identi-
fied modifiers are related to nuclear transport, nucleotide processing, and the
ubiquitin-proteasome system (Doumanis et al. 2009).
Silencing Huntington’s Disease Gene with RNAi 135
deficits in motor behavior. Later studies also found that AAV-shRNA causes
sequence-specific neurotoxicity to the mouse striatum (McBride et al. 2008;
Boudreau et al. 2009). A particularly successful strategy was to infect HD transgenic
mice with AAV-microRNA instead of AAV-shRNA (McBride et al. 2008; Boudreau
et al. 2009). AAV-mi2.4 is an artificial microRNA-based cassette driven by RNA
polymerase II promoter that can generate siRNA that targets the same sequence
within exon 2 of both introduced and endogenous HD transcripts (McBride et al.
2008; Boudreau et al. 2009). AAV-mir2.4 produced less antisense RNA and less
toxicity but achieved similar gene silencing effects as AAV-sh2.4 (McBride et al.
2008; Boudreau et al. 2009). In both CAG140 heterozygous knock-in mice and
HD-171-82Q mice, AAV-mi2.4 injection resulted in more cells exhibiting immuno-
reactivity toward DARPP-32 compared with AAV-sh2.4 injection. Moreover,
AAV-mi2.4 did not increase Iba1 staining as much as AAV-sh2.4 and is unlikely to
have elicited the same extent of microglial activation. Although there is controversy
regarding the immunoreaction as judged by real-time analysis of CD11b mRNA in
these two transgenic mice (McBride et al. 2008; Boudreau et al. 2009), AAV-mi2.4
clearly exhibited less neuronal toxicity than AAV-sh2.4. In addition, RNA polymer-
ase II promoter has the potential to be used in cell-specific expression. This artificial
AAV-microRNA-based RNAi delivery system might provide a new tool to efficiently
silence the mutant HD gene and improve the safety of AAV-shRNA.
The R6/2 transgenic mouse is the most thoroughly studied animal model for HD.
The insertion in its germ line of exon 1 of the human HD gene with 105 CAG
repeats causes rapid neurologic decline with death after about 3 months (Mangiarini
et al. 1996). To decrease expression of the mutant human HD-derived transgene,
postnatal day 2 mice were injected with synthetic siRNA targeting human HD exon
1 in the brain’s lateral ventricle (siRNA-HDexon1) (Wang et al. 2005). This
treatment caused a drop in expression of mutant HD mRNA in the animals’ brain
4–7 days later. Staining with antibodies against huntingtin protein revealed less
protein aggregation in the striatum of 8-week-old R6/2 mice treated with siRNA-
HDexon1 than in mock-treated or untreated R6/2 mice brains of the same age.
Moreover, siRNA-HDexon1-treated R6/2 mice retained greater motor control than
untreated animals (Wang et al. 2005). Up to the age of 8 or 9 weeks, the animals’
body weight remained relatively stable compared with untreated R6/2s. In addition,
the onset of the feet-clasping response was delayed by the synthetic siRNA, and
it was repeated at a lower frequency once it became manifest. At 12 weeks, an
age when symptoms were already advanced, test animals exhibited more activity in
the open field test than untreated controls. Finally, intraventricular injection with
synthetic siRNA extended the transgenic animals’ lifespan by 15 days (Wang et al.
2005). Although siRNA has been reported to degrade rapidly in vivo, intraven-
tricular injection on the day after birth had unexpectedly long-lived effects (Wang
et al. 2005). Apparently, the molecule’s ability to cause physiological changes goes
beyond simple silencing of gene expression. It seems that the loss of mutant HD
expression early on has long-lasting effects that persist after protein synthesis
resumes. Besides the experiments using synthetic siRNA, R6/2 mice were treated
with a high-capacity adenoviral vector expressing shRNA against HD exon1
Silencing Huntington’s Disease Gene with RNAi 141
Most HD patients are heterozygotic at the HD locus, carrying one allele for the wild-
type protein and one for its mutant counterpart. The former protein is essential to cell
survival; the latter is harmful, with greatest toxicity to striatal neurons. Developing
RNAi therapy for HD faces a dilemma: how to silence expression of one protein
while maintaining expression of the other? The technology of RNA interference
offers the requisite efficiency and specificity to satisfy both imperatives. The critical
difference between wild-type and mutant HD genes is the number of tandem CAG
triplets at the 50 -end of the sequence – over 36 in deleterious alleles but generally
below 30 in healthy ones. Though this variable sequence is an obvious target for
RNAi, it is of minimal practical use. RNAi molecules must have 19–23 bases to act
efficiently yet not elicit a strong immune response. Such molecules are too short to
span the CAG repeat, making them incapable of distinguishing between mutant and
wild-type alleles. An alternative approach is to target a polymorphic site that is
genetically linked to the CAG expansion. There are many candidate sequences.
Silencing Huntington’s Disease Gene with RNAi 145
A total of 190 SNPs (single-nucleotide polymorphisms) are known in the introns and
exons at the HD locus at chromosome 4p16.3. Analysis of HD alleles from 65 patients
of European origin revealed that disease-associated SNPs form a cluster of similar
haplotypes (haplogroup A) found on 95% of CAG-expanded chromosomes. Two
variants of haplogroup A (A1 and A2) are dramatically and specifically enriched on
HD chromosomes and are therefore markers for persons at risk of the disease (Warby
et al. 2009). In a study of 327 European Caucasian HD patients, 86% were heterozy-
gous at one or more of 26 SNPs analyzed (Lombardi et al. 2009). In these cases,
allele-specific RNAi is potentially applicable should the shRNA target polymorphic
sites linked to CAG expansion in a downstream part of the gene. Many questions
remain about methods to approach allele-specific RNAi therapy for HD and how
successful it will be. The following issues are important: (1) Is it possible to find one
or a few SNPs that are common and have a sufficiently strong linkage to the CAG
expansion to be generally useful for a relative larger HD population and can be tested
in large-scale clinical trials to obtain permits from drug regulatory agencies? Alter-
natively, can one use allele-specific RNAi therapy only for personal medicine? (2) On
an individual basis, can one rapidly identify SNPs with sufficiently tight linkage to
CAG expansion? (3) When wild type and mutant alleles only differ by one to three
nucleic acids, can one design a RNAi that silences the mutant allele enough to render
it innocuous but preserves sufficient expression of the normal species to execute
essential functions? While there is a long road ahead, researchers have already made
a significant start. Searches of the human-genome database have revealed many
polymorphic sites within the HD gene. Some of these SNPs occur at a high frequency
among test groups of substantial size. The most promising results come from several
studies on fibroblasts from HD patients. In cultures of these cells, RNAis have been
used to silence the mutant HD allele while preserving expression of its wild-type
counterpart (Table 2).
A noteworthy polymorphism that is linked to CAG expansion is the D2642
triplet deletion in exon 58 of the HD gene (Ambrose et al. 1994; Novelletto et al.
1994; Almqvist et al. 1995; Rubinsztein et al. 1995; Vuillaume et al. 1998). The
typical sequence beginning at this position is GAG.GAG.GAG.GAG; that for
D2642 is GAG.GAG.GAG. The deletion of a codon causes the loss of one of four
tandem glutamate residues in the huntingtin protein. The three-glutamate species
occur substantially more frequently among HD alleles than in those without CAG
expansion. According to previous study, the D2642 deletion occurred in 38% of HD
alleles but in only 7% of healthy ones (Ambrose et al. 1994). Efficient and specific
silencing of a D2642-tagged HD allele has been achieved in an in vitro system.
Immortalized cells were transfected with the two HD alleles, each of which had
been fused to the 30 end of a luciferase reporter gene. Monitoring the intensity of
fluorescence from these cells showed that the appropriate siRNA specifically
silenced expression of the mutant copy. One skin fibroblast from an HD patient
was identified that carried both an HD allele marked by the D2642 deletion and its
wild-type counterpart. These cells were transfected with each of four siRNAs
designed to target the polymorphic site. One of these molecules efficiently and
specifically silenced the D2642-marked, CAG-expanded allele in this HD fibroblast
146 Y. Zhang and R.M. Friedlander
Table 2 List of experimentally tested polymorphic sites linked to HD mutant allele with CAG
expansion
Reference SNP sites HD Location Identified HD Maxim References
number population fibroblast to reduction of
validate mutant
isoform-specific allele (%)
siRNA
rs363125 A/C 11% Exon 39 HD fibroblast 80% (van Bilsen
GM04022 et al.
2008)
D2642 Deletion 38% Exon 58 HD fibroblast 43% (Zhang
deletion of one GM09197 et al.
GAG 2009)
Rs362307 C/T 48% 30 UTR Luciferase reporter (Palfi et al.
for each isoform 2007)
Rs362273 G/A 35% Exon 57 Luciferase reporter (Palfi et al.
for each isoform 2007)
Rs362331 C/T 39% Exon 57 HD fibroblast 58% (Lombardi
GM09197 et al.
2009)
Rs2276881 A/G 8.6% Exon 60 Luciferase reporter (Lombardi
for each isoform et al.
2009)
CAG Expanded 100% Exon 1 HD fibroblasts: 100% (with (Hu et al.
repeats CAG GM09197, 30% 2009)
GM04281, reduction
GM04869, in wt-
GM04719, huntingtin)
GM04717
(Zhang et al. 2009). Similar allele-specific silencing results were obtained from a
later study of this HD fibroblast (Hu et al. 2009).
Only a minority of mutant HD alleles are marked by the D2642 deletion
mutation. Ongoing research aims to identify additional SNP sites in the HD gene
that could be used for RNAi silencing directed specifically to alleles with the CAG
expansion (Liu et al. 2008; van Bilsen et al. 2008). SiRNAs can be designed to
discriminate between two alleles that differ at a single nucleotide, i.e., a SNP
(Schwarz et al. 2006). It was found that the difference between half maximal
silencing (IC50) of mismatch and match reached the maximum discrimination is
from siRNA-target at position 16 (start from 50 guide strand of siRNA) of purine:
purine mismatch (Schwarz et al. 2006). Because the HD locus is large and both
isoforms are over 10 kb, SNPs in that gene are usually distant from the CAG repeat
at the 50 terminal. For this reason, it is difficult to identify SNPs that are linked to the
harmful triplet expansion in mutated HD alleles. Two methods have been used to
identify allele-specific SNPs (Fig. 1). One is to use SNP-specific RT primers to
selectively generate cDNA from a single allele of HD. A second round of PCR is
conducted on the resulting cDNA using primers spanning the CAG repeat sequence
(van Bilsen et al. 2008). Using this method, 11 known SNP heterozygote sites in
fibroblast cells from 21 different HD patients were allele-specifically determined
Silencing Huntington’s Disease Gene with RNAi 147
(van Bilsen et al. 2008). For example, fibroblasts from one patient had the rs363125
SNP with cytosine at that position in the mutant HD allele but adenine in its wild-
type counterpart. Using the appropriate RNAi molecule, it was possible to silence
the mutant allele while preserving expression of the wild-type one (van Bilsen et al.
2008). Another method used the circularization method to link the SNP with CAG
repeats (Liu et al. 2008). It was reverse-transcribed as usual. A primer flanking the
SNP with a KasI site at each end was used to amplify DNA ranging from SNP to the
CAG expansion. The key is to circularize this PCR product by intramolecular KasI
ligation. A PCR spanning SNP and CAG repeats was then conducted on two wild-
type and mutant individual ligation products. The length of CAG repeats was
evaluated on the PCR products by sequencing. Because the allele-specific SNP
site and CAG repeats are adjacent, it was easy for direct sequencing to identify the
SNP–(CAG)n linkage (Liu et al. 2008). This method was used in later studies with
large sample sizes (Palfi et al. 2007). Evaluation of 225 human HD and control
samples from American and European carriers revealed that over 48% of patients
were heterozygous at one of 24 identified SNP sites. Incidence of the U isoform of
rs362307 SNP at exon 67 on HD alleles much exceeded that on control samples
(Palfi et al. 2007). Seven out of 16 HD blood samples evaluated have expanded
CAG-linked heterozygous U isoform at the rs362307 site, an approximate 50%
linkage. Moreover, disease-associated SNPs at sites rs363125, rs362273, and
rs362307 are so frequent within this sample that five siRNAs that specifically
silence the harmful allele have the potential to treat three-quarters of these HD
148 Y. Zhang and R.M. Friedlander
are particularly stringent for allele-specific silencing, where on- and off-target
alleles differing by only 1–3 nucleotides. The technique of nonallele-specific
silencing is beset with a different problem, i.e., preserving sufficient expression
of endogenous huntingtin to prevent molecular changes that cause cell death,
especially over the long term.
The foremost difficulty when designing a safe and effective RNAi-based
therapy for HD comes from off-target effects of the RNAi. The first step toward
understanding (and thus solving) this problem is to identify the off-target mRNAs
that are partially or fully silenced. The current approach uses microarrays to
determine the global pattern of mRNA expression with and without changes
effected by RNAi. As is to be expected, the mRNAs responsible for off-target
effects have sequences resembling those of the target species. A “BLAST search”
of the gene database enables the investigator to identify additional mRNAs with
close sequence homology to the target mRNA. Even mRNAs that are perfectly
complementary to the siRNA’s seed region are not always silenced by that
interfering molecule (Birmingham et al. 2006). Moreover, it is difficult to deter-
mine whether a gene is silenced because of an siRNA:mRNA interaction or
simply from nonspecific changes due to systemic perturbations. Another con-
founding phenomenon arises when an siRNA mimics the function of a microRNA
or simply causes its cleavage. MicroRNAs are themselves regulatory elements
that inhibit translation of target messages but do not necessarily degrade them.
In that situation, the cell’s mRNA profile remains the same, whereas the spectrum
of expressed proteins changes. Analysis by proteomics reveals effects due to altered
translation, but the subsequent procedures necessary to evaluate sequence-specific
and concentration-related off-target effects in individual mRNA and translated
proteins are labor-intensive. In addition, the nature of off-target effects varies
among both cell lines and mouse strains (McBride et al. 2008; Boudreau et al.
2009). An example of this confounding variable is the different changes in CD11b
regulation when AAV1-mi2.4 is injected into B6C3F1/J and C57BL/6 mice. Upon
treatment with this RNAi, the former mouse exhibits increased levels of CD11, a
change indicative of microglial activation. By contrast, the latter animals are not
affected by introducing that very RNAi in a microRNA cassette (McBride et al. 2008;
Boudreau et al. 2009).
The above discussion underlines the complexity expected when developing
RNAi therapies for HD. As noted earlier, the clinical responses to RNAi-mediated
gene silencing are likely to vary among patients. The extent and nature of off-target
effects will largely determine who tolerates the therapy and who does not. In
anticipation of such variability, preclinical studies should be conducted using a
number of mouse models. As described above, one variable in the biology of HD
mice is the nature of the transgene. Mice that express a fragment of the HD gene
(i.e., R6/2) suffer rapidly progressing neurodegeneration. By contrast, animals that
carry a full-length HD gene (i.e., YAC128) exhibit more gradual disease progres-
sion that is reminiscent of the chronic symptoms of human HD patients. Conse-
quently, while mice of the former sort are easier to study, the latter are more reliable
models of the human disease.
150 Y. Zhang and R.M. Friedlander
even as its dosage was reduced. Unlike the situation for liver tissue, unprocessed
shRNA molecules did not accumulate to high levels in the striatum of these test
animals. The character and vulnerability of brain cells in the one study contrasts
with those qualities of liver cells in the other (Grimm et al. 2006; McBride et al.
2008; Boudreau et al. 2009). A reasonable explanation is that the affected micro-
RNA-processing machinery differs between liver cells and striatal cells (Grimm
et al. 2006; McBride et al. 2008; Boudreau et al. 2009).
In summary, HD therapy must use AAV-shRNA or AAV-miRNA at a dosage
that silences target genes but does not overwhelm the cell’s ability to process
endogenous small RNAs. Complicating this task is the difficulty of controlling
expression levels of virus-borne shRNA in vivo. A compliant system must allow for
fine adjustment of shRNA expression. The effectiveness of an inducible shRNA
vector for HD has been demonstrated in vivo using a lentivirus (Drouet et al. 2009).
Transcription from a doxycycline/tetracycline-responsive promoter may lend itself
to reversible repression, though such sensitivity and reversibility in processed
RNAi has yet to undergo sophisticated evaluation in vivo (Drouet et al. 2009).
The other important strategy for introducing RNAi into the host cell or animal is
to use naked siRNA. This technique has a major disadvantage when compared with
virus-based methods: unlike virus constructs, synthetic siRNA is easily degraded
in vivo. Nevertheless, naked siRNA has several properties that recommend it.
Simple nucleic acids are less immunogenic than viral complexes. While this
property is not relevant to in vitro systems, it removes a major barrier to application
in vivo (i.e., the ultimate goal – a clinical therapy). At the molecular level, naked
siRNA interferes less with endogenous pathways of microRNA biogenesis. The
central challenge when developing therapies built on this strategy is to identify
nucleic acids sequences with high specificity for the harmful HD. Any attempt to
solve this puzzle must be based on the molecular biology of RNAi. The first step
is to optimize sequence design. Within the molecule’s total length of 19–23
nucleotide pairs, positions 2 through 8 at the 50 -end of the guide strands of RNAi
are of exceptional importance to the recognition of and binding to target mRNAs
(Doench and Sharp 2004; Haley and Zamore 2004). In this way, it is similar to
microRNA, in which the corresponding portion also determines target mRNA
selection. Nucleotides at the 30 -end of the guide strands dictate cleavage efficiency
(Doench and Sharp 2004; Haley and Zamore 2004). In general, the affinity of
a siRNA for an mRNA increases as the number of mismatches decreases. Never-
theless, there can be partial gene silencing of mRNAs that match the introduced
siRNA at as few as 11 nucleotides (Jackson et al. 2003). Bioinformatic analysis
reveals that off-target effects most often arise from binding of the siRNA to a
sequence in an mRNA’s 30 UTR (Jackson et al. 2003; Lin et al. 2005; Qiu et al.
2005). Moreover, it is usually the siRNA’s seed region that matches the sequence
within the off-targeted mRNA (Lin et al. 2005; Birmingham et al. 2006; Jackson
et al. 2006b). The rules of mRNA-target selection are too complex to be fully
determined by the 16,348 possible combinations for nucleotides in positions 2–8.
Evidently, nucleotides beyond the seed region – some of them part of the siRNA’s
guest strand – contribute to target selection by siRNAs. Consequently, the next
152 Y. Zhang and R.M. Friedlander
years, the latter branch of experimental medicine is strictly regulated. For this reason,
siRNA-based therapies are more likely to obtain FDA approval. BevasiranibTM, the
first such drug, has already passed phase II and entered phase III clinical trials for
treating age-related macular degeneration (AMD).
Besides off-target effects, another big challenge in the development of RNAi-
based therapy for HD concerns delivery. In previous in vivo studies, viral and
nonviral application and local intrastriatal or intraventricular administration were
successfully used to silence HD in vivo. While virus-carried shRNA showed high
efficiency in transducing cells in the striatum, coupling siRNA with lipids or
cholesterol are other strategies for introducing HD RNAi into a living organism.
In one study, siRNA was coupled to liposomes (lipofectamine2000) to form amor-
phous lipoplex particles. Intraventricular injection with lipofectamine2000-siRNA
into day-2 (P2) postnatal mice countered both the generation of striatal nuclear
inclusion and brain atrophy. The molecular and anatomical changes were correlated
with a reduction in abnormal behavioral and increased lifespan (Wang et al. 2005).
Because this cationic lipid-based transfection reagent is toxic to primary neurons,
novel delivery strategies have been used. Lipophilic polypeptides such as stearylated
octaarginine (steary1-R8), MPG-based particles, and lipid-based artificial virus-like
particles (AVPs) are all peptide-based particles with potential applications in siRNA
therapy. They achieve transfection efficiency comparable to liposomal reagents, but
are less detrimental to primary neurons and embryonic stem cells (Futaki et al. 2001;
Simeoni et al. 2005; Tonges et al. 2006; Crombez et al. 2007). Another experimental
system for siRNA-based therapy has been explored in mice transduced with the
AAV-htt 100Q viral construct. Intrastriatal injection with a cholesterol-coupled HD
siRNA silenced the mutant transgene for 3 days and relieved neuropathological
symptoms for 1 week (DiFiglia et al. 2007). In addition, siRNA coupled to cell-
penetrating peptides provides a potential tool for cell type-specific siRNA targeting.
SiRNA coupled with the penetrating peptide via a disulfide bond is far more
effective than lipofectamine2000 in entering and silencing genes in primary neurons
(Davidson et al. 2004; Muratovska and Eccles 2004; Jankowski et al. 2009). siRNA
can also be conjugated to aptamers for cell-specific delivery (Chu et al. 2006).
Moreover, electroporation by a nucleofector enables siRNA to be introduced into
over 50% of cells in a population of mouse striatal primary neurons (Jin et al. 2005).
This technology also enables siRNA to reach 70% and shRNA to reach 59% of cells
within a culture of primary neurons (Dail et al. 2006; Hood et al. 2006; Seng et al.
2006; Yang et al. 2006). These new delivery methods will increase the silencing
efficiency of RNAi against HD and its related proteins to further elucidate its
physiological and pathological roles in HD and guide us in developing new drug
targets (Schwartz 2009). It would be interesting to compare the effects of siRNA
with virus-carried shRNA targeting the same HD sequence in the same transgenic
mouse model.
For local delivery, in collaboration with Dr. Don M. Gash of the University of
Kentucky, Alnylam Pharmaceuticals and Medtronic are developing an implantable
pump to infuse ALN-HTT (an RNAi-based drug against HD) into the brain. Using
an experimental catheter and the commercially available Medtronic SynchroMed II
154 Y. Zhang and R.M. Friedlander
pump, this technology permits the local delivery of siRNA to striatum and
subsequent transport to distant regions of the brain. Early studies have shown that
continuous infusion of an appropriate siRNA over 7 days caused an approximately
45% drop in the level of HD mRNA in the putamen. One big advantage of
implanted pumps is their ability to deliver drugs with much greater efficiency
than other techniques. However, the long-term efficacy of gene silencing is not
yet known. Another critical issue is the safety of the procedure. While no clinical
abnormalities were observed following 28 consecutive days of siRNA infusion into
a nonhuman primate, a comprehensive study must still be conducted. Ongoing
research on the promising technology of implanted pumps is likely to reach the
level of clinical trials in the not-so-distant future.
Whereas the most desirable way to get a drug into the CNS is by injection into
the bloodstream (or by ingestion), such systemic administration works only for
compounds that cross the blood–brain barrier. Unfortunately, this protective barrier
prevents RNAi molecules from entering the brain and spinal cord. However, the
situation is different in the liver. ApoB mRNA in liver has been silenced by 90% for
at least 11 days after systemic administration of one dose of siRNA encapsulated in
stable nucleic acid-lipid particles (SNALPS) (a bilayer consisting of cationic and
neutral lipids and coated with hydrophilic polyethylene) (Morrissey et al. 2005;
Zimmermann et al. 2006). Recently, investigators have developed a method to
circumvent this physiological barrier to the siRNA’s entrance into the CNS (Kumar
et al. 2007). In this technique, the siRNA is conjugated to a nona-D-arginine
polypeptide, which associates with a portion of a rabies virus glycoprotein
(RVG). Due to its affinity for the acetylcholine receptor (AChR), the molecular
complex binds to and is endocytosed by cells expressing that surface protein. The
internalized siRNA has been shown to silence the target gene (Kumar et al. 2007).
Moreover, the AChR is expressed in medium spiny projection (MSP) neurons of the
striatum and cerebral cortex (Weiner et al. 1990; Bernard et al. 1992; Ince et al.
1997). Theoretically, this technology should enable investigators to silence HD
mRNAs by targeting siRNA to the very cells with selective sensitivity to HD.
Practical considerations, i.e., the efficiency with which the molecular complex can
be delivered into the CNS and its toxicity once there, must still be determined.
Ultimately, it may be possible to target siRNAs to specific classes of cells by
coupling them to fragments of the appropriate viral glycoproteins. This technology
could be used to target other cell types besides MSNs involved in the pathological
mechanism of HD. For example, besides damage from mutant huntingtin’s toxicity
to MSNs, it may diminish the glia’s protective ability against excitotoxicity caused
by glutamate (Shin et al. 2005). Should the above technology be adapted to target
many cell types, it will prove a more effective tool with which to counter HD.
The rapidly developing technology of RNA interference has already proven
useful in experimental therapies for HD. Studies using RNAi in cellular and animal
systems have increased our understanding of the function of wild-type huntingtin.
In addition, they have already been used in preclinical trials in animals. Findings
from these investigations suggest that targeted gene silencing may be used in therapy
for what is now an untreatable disease. As a rule, difficult and time-consuming
Silencing Huntington’s Disease Gene with RNAi 155
References
Allerson CR, Sioufi N, Jarres R et al (2005) Fully 20 -modified oligonucleotide duplexes with
improved in vitro potency and stability compared to unmodified small interfering RNA. J Med
Chem 48:901–904
Almqvist E, Spence N, Nichol K et al (1995) Ancestral differences in the distribution of the delta
2642 glutamic acid polymorphism is associated with varying CAG repeat lengths on normal
chromosomes: insights into the genetic evolution of Huntington disease. Hum Mol Genet
4:207–214
Amarzguioui M, Holen T, Babaie E et al (2003) Tolerance for mutations and chemical modifica-
tions in a siRNA. Nucleic Acids Res 31:589–595
Ambrose CM, Duyao MP, Barnes G et al (1994) Structure and expression of the Huntington’s
disease gene: evidence against simple inactivation due to an expanded CAG repeat. Somat Cell
Mol Genet 20:27–38
Bernard V, Normand E, Bloch B (1992) Phenotypical characterization of the rat striatal neurons
expressing muscarinic receptor genes. J Neurosci 12:3591–3600
Birmingham A, Anderson EM, Reynolds A et al (2006) 30 UTR seed matches, but not overall
identity, are associated with RNAi off-targets. Nat Methods 3:199–204
Boudreau RL, Davidson BL (2006) RNAi therapy for neurodegenerative diseases. Curr Top Dev
Biol 75:73–92
Boudreau RL, McBride JL, Martins I et al (2009) Nonallele-specific silencing of mutant and wild-
type huntingtin demonstrates therapeutic efficacy in Huntington’s disease mice. Mol Ther
17:1053–1063
Bumcrot D, Manoharan M, Koteliansky V et al (2006) RNAi therapeutics: a potential new class of
pharmaceutical drugs. Nat Chem Biol 2:711–719
Burger C, Gorbatyuk OS, Velardo MJ et al (2004) Recombinant AAV viral vectors pseudotyped
with viral capsids from serotypes 1, 2, and 5 display differential efficiency and cell tropism
after delivery to different regions of the central nervous system. Mol Ther 10:302–317
156 Y. Zhang and R.M. Friedlander
Schwartz EI (2009) Potential application of RNAi for understanding and therapy of neurodegenerative
diseases. Front Biosci 14:297–320
Schwarz DS, Ding H, Kennington L et al (2006) Designing siRNA that distinguish between genes
that differ by a single nucleotide. PLoS Genet 2:e140
Schwarz DS, Hutvagner G, Du T et al (2003) Asymmetry in the assembly of the RNAi enzyme
complex. Cell 115:199–208
Seng S, Avraham HK, Jiang S et al (2006) KLHL1/MRP2 mediates neurite outgrowth in a
glycogen synthase kinase 3beta-dependent manner. Mol Cell Biol 26:8371–8384
Senut MC, Suhr ST, Kaspar B et al (2000) Intraneuronal aggregate formation and cell death after
viral expression of expanded polyglutamine tracts in the adult rat brain. J Neurosci 20:219–229
Shin JY, Fang ZH, Yu ZX et al (2005) Expression of mutant huntingtin in glial cells contributes to
neuronal excitotoxicity. J Cell Biol 171:1001–1012
Simeoni F, Morris MC, Heitz F et al (2005) Peptide-based strategy for siRNA delivery into
mammalian cells. Methods Mol Biol 309:251–260
Slow EJ, van Raamsdonk J, Rogers D et al (2003) Selective striatal neuronal loss in a YAC128
mouse model of Huntington disease. Hum Mol Genet 12:1555–1567
Sobczak K, de Mezer M, Michlewski G et al (2003) RNA structure of trinucleotide repeats
associated with human neurological diseases. Nucleic Acids Res 31:5469–5482
Steffan JS, Kazantsev A, Spasic-Boskovic O et al (2000) The Huntington’s disease protein
interacts with p53 and CREB-binding protein and represses transcription. Proc Natl Acad Sci
USA 97:6763–6768
Strehlow AN, Li JZ, Myers RM (2007) Wild-type huntingtin participates in protein trafficking
between the Golgi and the extracellular space. Hum Mol Genet 16:391–409
Tagawa K, Marubuchi S, Qi ML et al (2007) The induction levels of heat shock protein 70
differentiate the vulnerabilities to mutant huntingtin among neuronal subtypes. J Neurosci
27:868–880
Taymans JM, Vandenberghe LH, Haute CV et al (2007) Comparative analysis of adeno-associated
viral vector serotypes 1, 2, 5, 7, and 8 in mouse brain. Hum Gene Ther 18:195–206
The Huntington’s Disease Collaborative Research Group (1993) A novel gene containing a
trinucleotide repeat that is expanded and unstable on Huntington’s disease chromosomes.
The Huntington’s Disease Collaborative Research Group. Cell 72:971–983
Tonges L, Lingor P, Egle R et al (2006) Stearylated octaarginine and artificial virus-like particles
for transfection of siRNA into primary rat neurons. RNA 12:1431–1438
van Bilsen PH, Jaspers L, Lombardi MS et al (2008) Identification and allele-specific silencing of
the mutant huntingtin allele in Huntington’s disease patient-derived fibroblasts. Hum Gene
Ther 19:710–719
Velier J, Kim M, Schwarz C et al (1998) Wild-type and mutant huntingtins function in vesicle
trafficking in the secretory and endocytic pathways. Exp Neurol 152:34–40
Vuillaume I, Vermersch P, Destee A et al (1998) Genetic polymorphisms adjacent to the CAG
repeat influence clinical features at onset in Huntington’s disease. J Neurol Neurosurg
Psychiatry 64:758–762
Wacker JL, Zareie MH, Fong H et al (2004) Hsp70 and Hsp40 attenuate formation of spherical and
annular polyglutamine oligomers by partitioning monomer. Nat Struct Mol Biol 11:1215–1222
Waelter S, Scherzinger E, Hasenbank R et al (2001) The huntingtin interacting protein HIP1 is a
clathrin and alpha-adaptin-binding protein involved in receptor-mediated endocytosis. Hum
Mol Genet 10:1807–1817
Wang YL, Liu W, Wada E et al (2005) Clinico-pathological rescue of a model mouse of
Huntington’s disease by siRNA. Neurosci Res 53:241–249
Warby SC, Montpetit A, Hayden AR et al (2009) CAG expansion in the Huntington disease gene is
associated with a specific and targetable predisposing haplogroup. Am J Hum Genet
84:351–366
Warrick JM, Chan HY, Gray-Board GL et al (1999) Suppression of polyglutamine-mediated
neurodegeneration in Drosophila by the molecular chaperone HSP70. Nat Genet 23:425–428
160 Y. Zhang and R.M. Friedlander
Weiner DM, Levey AI, Brann MR (1990) Expression of muscarinic acetylcholine and dopamine
receptor mRNAs in rat basal ganglia. Proc Natl Acad Sci USA 87:7050–7054
Woda JM, Calzonetti T, Hilditch-Maquire P et al (2005) Inactivation of the Huntington’s disease
gene (Hdh) impairs anterior streak formation and early patterning of the mouse embryo. BMC
Dev Biol 5:17
Yamamoto A, Lucas JJ, Hen R (2000) Reversal of neuropathology and motor dysfunction in a
conditional model of Huntington’s disease. Cell 101:57–66
Yang YC, Lin CH, Lee EH (2006) Serum- and glucocorticoid-inducible kinase 1 (SGK1) increases
neurite formation through microtubule depolymerization by SGK1 and by SGK1 phosphory-
lation of tau. Mol Cell Biol 26:8357–8370
Zeitlin S, Liu JP, Chapman DL et al (1995) Increased apoptosis and early embryonic lethality in
mice nullizygous for the Huntington’s disease gene homologue. Nat Genet 11:155–163
Zhang H, Das S, Li QZ et al (2008) Elucidating a normal function of huntingtin by functional and
microarray analysis of huntingtin-null mouse embryonic fibroblasts. BMC Neurosci 9:38
Zhang Y, Engelman J, Friedlander RM (2009) Allele-specific silencing of mutant Huntington’s
disease gene. J Neurochem 108:82–90
Zhang Y, Leavitt BR, van Raamsdonk JM et al (2006) Huntingtin inhibits caspase-3 activation.
Embo J 25:5896–5906
Zhou H, Li SH, Li XJ (2001) Chaperone suppression of cellular toxicity of huntingtin is independent
of polyglutamine aggregation. J Biol Chem 276:48417–48424
Zimmermann TS, Lee AC, Akinc A et al (2006) RNAi-mediated gene silencing in non-human
primates. Nature 441:111–114
Zuccato C, Ciammola A, Rigamonti D et al (2001) Loss of huntingtin-mediated BDNF gene
transcription in Huntington’s disease. Science 293:493–498
Application of Dicer-Substrate siRNA
in Pain Research
Contents
1 RNAi in Pain Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
2 Potential Applications of RNA Interference for Pain Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . 163
2.1 Ion Channels as Therapeutic Targets for Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
2.2 G-Protein-Coupled Receptors as Pain Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
3 Advantages of DsiRNA Over Conventional siRNA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
3.1 Inherent Character of DsiRNA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
3.2 Molecular Mechanism of Action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
4 DsiRNA for Efficient Silencing In Vitro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
4.1 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
4.2 Validation of Knockdown Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
4.3 Targets In Vitro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
5 DsiRNA In Vivo: One Step Forward . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
5.1 Working Evidence of DsiRNA In Vivo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
5.2 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
6 Conclusion and Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 161
RNA Technologies, DOI 10.1007/978-3-642-12168-5_7,
# Springer-Verlag Berlin Heidelberg 2010
162 P. Sarret et al.
medication to maintain the same level of pain relief. Although a large number of
pharmacological treatments are currently available, there is still a pressing need for
the development of new alternative therapeutic approaches to alleviate pain and to
increase patient comfort. With this perspective, RNA interference (RNAi) repre-
sents a major tool for discovery and validation of targets implicated in pain
disorders and may be useful therapeutically to treat some intractable conditions
where conventional therapy is ineffective or inexistent. In this chapter, we first
provide a brief overview of the use of RNAi in pain research, with emphasis on ion
channels and G-protein-coupled receptors (GPCRs) as potential targets for thera-
peutic intervention. We then discuss the advantages of designing longer exogenous
siRNA to serve as Dicer substrates (DsiRNAs). Finally, we focus on the potency of
these new synthetic 27-mer RNA duplexes to efficiently knockdown pain targets in
both in vitro and in vivo studies.
In the last 5 years, siRNA strategies have been widely used to study pain mechani-
sms in vivo, providing a straightforward approach to validate new pain targets. Due
to space limitations, however, only a limited number of reports can be described
here. For further details, the reader can refer to recent reviews (Ganju and Hall
2004; Kurreck 2004; Sah 2006; Jain 2008).
Dorsal root ganglia (DRG) of the spinal cord represent an attractive site for gene
therapy designed to treat chronic pain. The first successful in vivo proof of concept
of the use of siRNA in a pain model was conducted through the intrathecal infusion
of siRNA targeting the cation-channel subunit P2X3 (Dorn et al. 2004). This P2X
purinergic receptor is highly expressed in nociceptive sensory neurons and plays a
crucial role in ATP-induced inflammatory and neuropathic pain. It has been
demonstrated that 21-nt siRNA sequences can specifically knockdown the rat
P2X3 receptor in vitro and in vivo (Hemmings-Mieszczak et al. 2003; Dorn et al.
2004). Continuously administered via osmotic mini-pump into neuropathic rats at a
high dose of 0.4 mg/day for up to 6 days, intrathecal naked P2X3 siRNA reduced
both mechanical allodynia and thermal hyperalgesia induced by partial sciatic
nerve ligation. In line with these results, siRNA-treated rats showed diminished
pain responses following injection of the P2X3 receptor agonist, ab-methylene ATP
(stable analog of ATP). At a molecular level, siRNA administration resulted in
down-regulation of P2X3 mRNA and protein pools in DRG and spinal cord.
Members of the transient receptor potential (TRP) family of ion channels also
represent attractive targets for pain therapy. TRPV1 is a cation channel that is
164 P. Sarret et al.
From then on, the development of small molecules to specifically block each sodium
channel subtype has presented significant challenges for conventional medicinal
chemistry. As a therapeutic class, sodium channel blockers such as lidocaine,
amitriptyline, and lamotrigine have been widely used to treat disorders involving
neuronal hyperexcitability, including management of pain. However, none of the
currently used sodium channel blockers is capable to distinguish between the various
subtypes that have been identified to date. This lack of selectivity is thought to
contribute to side effects of these drugs such as motor dysfunction, therefore
compromising seriously the use of these medications in a clinical pain setting. A
critical role of NaV1.8 in mediating pathologic pain has been suggested by the
observation that tetrodotoxin-resistant sodium channel a-subunit knockout mice
showed strong analgesia to noxious mechanical stimuli and delayed development
of inflammatory hyperalgesia (Akopian et al. 1999). The restricted expression of
NaV1.8 to the peripheral sensory neurons suggests that the blockade of this channel
has therapeutic potential in various pain paradigms and may improve the compliance
in patients actually treated with existing sodium channel blockers. In 2006, Alnylam
Pharmaceutical Inc. and collaborators were the first to release preclinical data
demonstrating that intrathecal injection of siRNA targeting NaV1.8 provided nearly
complete pain relief in an animal model of chronic neuropathic pain (Sah et al. 2006).
Since then, Dong et al. (2007) demonstrated that NaV1.8 siRNA delivered to the
lumbar DRG via an indwelling epidural cannula caused a significant reduction of
NaV1.8 mRNA expression in L4–L5 DRG neurons and consequently reversed
mechanical allodynia in the rat chronic constriction nerve-injury model.
Inward rectifying K+ channels and voltage-gated Ca2+ channels, which regulate
either cellular excitability or synaptic transmission, also play an important role in
the detection and transmission of nociceptive stimuli in primary sensory neurons.
The Kir4.1 potassium channel subunit and the N-type calcium channel Cav2.2 were
validated as pain targets in vivo by RNAi. Kir4.1 expressed in the trigeminal
ganglion is reduced following chronic constriction injury of the infraorbital nerve
and specific silencing of Kir4.1 by RNAi produces a facial pain-like behavior in
freely moving rats (Vit et al. 2008). Two Cav2.2 N-type channel isoforms (e37a and
e37b) can be generated by alternative splicing. Altier et al. (2007) demonstrated by
selective down-regulation of each Cav2.2 isoform by RNAi that channels contain-
ing exon 37a are specifically required for developing thermal and mechanical
hyperalgesia during inflammatory and neuropathic pain states.
GPCRs represent the largest and most diverse family of cell surface proteins that act
as sensors for extracellular stimuli, including hormones, neuropeptides, neurotrans-
mitters, or sensory stimuli such as light and smell. Despite their molecular and
function diversity, all GPCRs share a similar structure, which consists of seven
transmembrane domains linked by alternating intracellular and extracellular loops.
Extracellular domains, which vary among the different classes of GPCR, contribute
Application of Dicer-Substrate siRNA in Pain Research 167
In the last decade, RNA species of varying lengths have been tested for their
potency to silence target mRNA. Double-stranded siRNA (i.e., duplexes) was
shown more stable than single-stranded RNA, but additionally, those shorter than
30 base pairs were reported to be devoid of severe immune stimulation. Among
those tested, Kim and colleagues demonstrated that 25–30 bp duplexes were more
potent than corresponding 21-mers; 27-mers dsRNA presenting a 100-fold amelio-
ration (Kim et al. 2005). The hypothesis proposed implies a role for Dicer in
promoting the processing of the double stranded RNA (dsRNA) in an appropriate
size and its loading in the RISC complex. Classic 21-mers RNAs bypass this step
explaining the poor odds for their RISC loading. Although DsiRNA are intrinsically
more potent, optimal in vivo RNAi can only be achieved if the duplexes reach the
cells of interest, are properly up-taken into the cytoplasm and loaded in the
silencing machinery. To this day, no universal delivery system is yet available to
assist siRNA to cross all those barriers while remaining an inert carrier. Thus,
modifications made to the duplex structure are the best solution to circumvent
stability issues and specific/nonspecific off-target effects (OTEs). Evading from
nucleases action, decreasing the activation of the innate immune system, and
limiting the interference with the endogenous micro RNA pathway while optimiz-
ing the cellular uptake are challenges to reach a proper level of RNAi.
Inherently, 27-nt dsRNA such as DsiRNAs are somewhat more stable in 10% FBS
serum than siRNA (t1/2 ¼ 83 h vs. 1.6 h) or in active serum like in mammals
(30 min vs. 5 min) (Kubo et al. 2007). Though, their stability can still be enhanced
to optimize RNAi. The most straightforward approach to increase siRNAs stability
is to modify their internucleotide phosphate linkage. Importantly, modifications
added to protect from nuclease degradation should not interfere with the gene-
silencing capacity of the duplex. In other words, the cellular proteins of the RNAi
machinery should maintain an adequate access to the guide strand. Indeed, for
DsiRNAs, attention is required when adding modifications to protect from serum
nucleases, since some degree of nuclease activity is still desired for Dicer endoribo-
nuclease to perform its cleavage of the substrate (Behlke 2008). Chemical
substitutions of a nonbridging oxygen are multiple to confer stability (nitrogen,
sulfur, boron, methyl). The most secure, easy to perform and studied substitution
remains the phosphorothioate. However, in the case of DsiRNA, the retained
Application of Dicer-Substrate siRNA in Pain Research 169
strategy involves the introduction of 20 -O-Methyl RNA (20 OMe) in the strands. This
variant is naturally occurring in mammalian rRNA and tRNA (Behlke 2008) and it
was demonstrated that its inclusion in DsiRNA had minimal impact on potency at
multiples sites (Collingwood et al. 2008). Duplexes presenting 20 OMe RNA on both
sense (S) and antisense (AS) strands were however found inactive, just as too heavy
modification protocols did. The “evader” pattern was retained, introducing ten
20 OMe RNA bases in alternation in the AS strand. Modified DsiRNA were shown
to remain intact after 24 h of incubation in serum (Collingwood et al. 2008). This
pattern was exploited for research purposes until recently when experimenters
observed an in vitro loss of potency for certain sequences. Again, no rule could
help to define which targeted sequence would be affected or not. Behlke’s team thus
switched to a “m7 pattern”, with fewer 20 OMe bases in the middle of the AS strand
(seven in total), a hotspot recognized as sensitive to sequence substitutions. This
M7 modification pattern confers adequate stability while maintaining potency in
every sequence tested. Moreover, 20 -F modifications could be introduced in addi-
tion to 20 OMe, since this type of add-on is tolerated at the Ago2 site of cleavage.
However, 20 -F sugars are artificial and once available in the cells, they can incor-
porate in the endogenous transcriptome or genome (Behlke 2008). Toxicity sur-
veillance should be performed if a modified DsiRNA presenting both substitutions
is used in in vivo sustained release. Since, 20 OMe bases are present in the nature, no
apparent toxicity is anticipated if used as the sole modification.
Cell membrane
25/27mer DsiRNA
transfection Sites of
Dicer processing Degraded target mRNA
5’ pCCUACUGACCCAGUGCAUACUGC a g
3’ CUGGAUGACUGGGUCACGUAUGACGUC
5
1 Ago 2 mRNA
NNNNNNNNNNNGACCUACUGACCCAGUGCAUACUANNNNNAAAA
3’ CUGGAUGACUGGGUCACGUAU
Dicer
5’ pCCUACUGACCCAGUGCAUACUGCa g
3’ CUGGAUGACUGGGUCACGUAUGACGUC
Target recognition
2 4
CU
A
Ago 2 GCAU
5’ CCUACUGACCCAGUGC AUAtt 5’ pCCUACUGACCC AGUGCAUACU U
ACCCAG
3’ ttGGAUGACUGGGUCACGUAU 3’ CUGGAUGACUGGGUC ACGUAUp 5’ pCCUACUG
3’ CUGGAUGACUGGGUCACGUAUp
5’ pCCUACUGACCC AGUGCAUACUG
Dicer product 3’ CUGGAUGACUGGGUC ACGUAUGp
(21-mer and 22-mer)
3
Standard synthetic
Transfection Ago 2
21-mer siRNA
5’ pCCUACUGACCCAGUGC AUACU
3’ CUGGAUGACUGGGUCACGUAUp
Cytoplasm RISC
two siRNA strand referred to as the guide strand (or AS) is loaded onto Ago
(step 2). Afterwards, RISC is activated by the cleavage of the other strand called
the passenger strand (or S) and the guide strand directs the sequence specificity of
RISC and therefore the RNAi silencing (step 3). The incorporated strand that serves
as the guide strand is generally the one which 50 terminus is at the thermodynami-
cally less stable end of the duplex. For instance, in the case of asymmetric siRNAs,
it is easier to unwind the duplex from the less stable end, thus preferentially
facilitating the incorporation into the RISC complex. In contrast, a symmetric
siRNA possesses two equally stable ends, thus both strands of the siRNA are
assembled into the RISC complex with an equal efficiency (Schwarz et al. 2003).
Ago is a multidomain protein of 100 kDa comprising the PAZ domain and the
C-terminal PIWI domain. Cleavage catalysis is mediated by the PIWI domain
showing extensive homology to RNase H, an endonuclease that nicks RNA strands
in 10 nucleotides segments from the 50 end siRNA, leaving the siRNA intact for
another round of cleavage (Ameres et al. 2007; Hock and Meister 2008). A third
functionally important domain that resides between the PAZ and the PIWI domains
is termed the middle domain (MID), containing a high basic pocket that binds the
characteristic 50 phosphate of small RNAs (Hock and Meister 2008). Thus, evidence
supports that the 20 -nucleotide-long 30 overhangs and the 50 phosphate of siRNA are
anchored onto Ago proteins by the PAZ and MID domains, respectively. The
number of Ago genes is highly variable between species. Eight Ago proteins are
encoded by the human genome and of these, only hAgo2 shows an endonucleolytic
activity (Ameres et al. 2007). Therefore, several forms of RISC have been reported.
Indeed in human cells, a third player called TRBP, a RNA-binding protein, is
necessary for efficient transfer of siRNA from Dicer to Ago2, and thus essential
for the RISC loading complex formation (Wang et al. 2009).
Depending on both nature of Ago protein and the degree of complementary between
the DsiRNA and the mRNA-target sequence, gene silencing results as the cleavage
or translational repression of mRNA. If the base-pairing complementary between
siRNA and its target mRNA is not optimal, the target may be physically unreachable
by the active center of the endonuclease domain of RISC complex. Consequently,
this will result in translational repression instead of efficient cleavage of target
mRNA. By contrast, a noncleaving RISC complex resulting in a lack of slicing
activity of Ago can direct the target mRNA only for translational repression regard-
less of high complementarity between siRNA and target (van den Berg et al. 2008).
The way that RISC complex efficiently recognizes its target remains unclear.
However, evidences suggest that a RISC-mediated siRNA-target interaction is
more complex than simple nucleic acid hybridization and some factors within
RISC may facilitate target recognition (Hutvagner et al. 2004).
Overall, DsiRNAs demonstrate a higher efficiency in RISC loading. This is
imputable to the fact that DsiRNAs are a substrate of Dicer enzyme instead of a
174 P. Sarret et al.
product (i.e., siRNAs) and that the complex formed binds directly to Ago in the core
of the RISC complex. The advanced design elaborated from experimental observa-
tions done on silencing efficiency allowed to further modify the duplex to increase
its potency. DsiRNAs are far more stable and their uptake and processing surpasses
the standards in in vitro knockdown research.
In order to obtain a potent and efficient knockdown when using RNAi technologies,
the first step consists in a sturdy design. Thereafter, rigorous controls and the choice
of a representative cell assay are the basis to optimize the validation toward a
further use in in vivo research. In the next section, we will describe the preparation
process of DsiRNA candidates suitable for in vivo pain research.
4.1 Methodology
Various design methods exist, but for all cases, there are basic rules to follow.
Because it has become more convincing in the literature that chemically modified
asymmetric 27-mers siRNA are more potent than unmodified blunt 27-mers, we
will place more emphasis on general guidelines for this particular kind of mole-
cules. Indeed, as reported in the previous section, processing of unmodified 27-mers
duplexes by Dicer is unpredictable and can result in the generation of siRNAs of
poor activity, sometimes below that of an optimal 21-mer siRNA (Amarzguioui and
Rossi 2008). Indeed, the chemically modified asymmetric 27-mers DsiRNA pro-
posed by Rose and colleagues or similar species with different modifications
suggested by Kubo et al. are more stable than classical 21-nt dsRNAs and possess
higher long-term RNAi activity (Rose et al. 2005; Kubo et al. 2007). However,
because a range of potencies is frequently seen among different target sites within a
same gene, it is highly recommended to design multiple siRNA candidates for the
same gene of interest.
By analyzing the relevant structural features of endogenous pre-miRNAs, the
problem of making DsiRNA processing predictable has been solved (Amarzguioui
et al. 2006). There are now many siRNA design algorithms that are freely accessible
on the web. The aim here is to help the reader to access a list of algorithms to choose
suitable criteria for optimal designing. Charité’s University, in Germany, has listed
many of the currently available Web-based siRNA design tools (http://itb.biologie.
hu-berlin.de/nebulus/sirna/siDesign.htm) separated in commercial- or academic-
supported Web site. Fully automated DsiRNA design is also available through
Integrated DNA Technology (IDT) (http://www.idtdna.com/Scitools/Applications/
RNAi/RNAi.aspx). But whatever design algorithm is used, to increase the chance of
Application of Dicer-Substrate siRNA in Pain Research 175
success, it is suggested to blast more than one and choose candidates with high
scores in multiple or all of the algorithms (Amarzguioui et al. 2006).
Once the DsiRNA candidates are designed and manufactured, in vitro testing is
prime. The choice of the cell lineage is important in regard to the targeted gene. The
choice should be made toward cells natively expressing the gene of interest and
ideally in a cell line specific to the pathology or tissue aimed for future in vivo
studies or clinical therapy (e.g., neuronal cell line in the event of in vivo pain
research). However, expression levels of a gene of interest can be very low or
absent in commercially available cell lines. In such a case, the use of a transgenic
expression cell line is suggested. Nevertheless, overexpression of a gene can bias
the expected silencing outcomes of a DsiRNA. Indeed, silencing efficiency can be
high due to an imposing “artificial” pool of target mRNA available in the cytoplasm
of transgenic cells, but results could be rather disappointing in ex vivo or in vivo
systems attributable to more realistic stable levels of various mRNAs. Therefore,
consideration of more than one potent in vitro candidate is essential to act as a
backup in case of a weak hit in vivo.
Once the cell line is chosen, the selection of an optimal formulation agent is
required to facilitate cell uptake. Indeed, naked RNA usually gives poor silencing
rates at low doses and is toxic at high doses at which cellular uptake is though
greater. Formulation is a complex research domain and there is no universal recipe
for an effective delivery. Yet, in vitro research allows performing extensive evalu-
ation assays in parallel conversely to in vivo research for which costly and time-
consuming experiments limit the margin for error. Thus, because there is no
universal reagent, every cell line, tested for DsiRNA uptake, should be screened
with a battery of transfection reagents to identify the most potent. The final goal
remains to find the best balance between an efficient knockdown and the least
toxicity and variability between assays. The best protocol consists in using a
“control” DsiRNA (e.g., HPRT is a constitutively expressed housekeeping gene)
to screen the optimal reagent formulation for a specific cell line. Once the reagent is
chosen, then the assays using the target DsiRNA at study can be performed.
The most frequently employed class or reagents for classical 21-mers siRNA,
i.e., cationic lipids, can work as well for DsiRNA. Note that an adjustment of the
N:P ratio should be considered to optimize the formulation (i.e., the equilibrium
cationic lipid nitrogen/anionic RNA phosphate charge ratio to insure optimal
formulation but also appropriate release once in the cell). Nevertheless, polymer
or peptide-based reagents are also conceivable alternatives for delivery where
cationic lipids failed to yield high transfection levels. Hard-to-transfect cells such
176 P. Sarret et al.
as neurons are ideal candidates for new formulations such as the protein-based
transfection reagent Transductin™. Indeed, this novel transfection reagent, com-
posed of a peptide transduction domain (PTD) and a double-stranded RNA-binding
domain (DRBD), has recently shown to be able to efficiently transfect siRNA into
primary cultures of hard-to-transfect cells (Eguchi and Dowdy 2009; Eguchi et al.
2009). New avenues are frequently exposed in specialized journals such as
Advanced Drug Delivery Reviews to which the reader can refer for updates.
0.1 nM. However, other carrier reagents such as peptides or liposomes often require
greater concentrations of DsiRNA. Positive and negative controls should be set at
the maximal concentration used in the dose-response curve. Keep in mind that
when looking at the knockdown efficiency, the results greatly depend upon the cell
type used and the targeted gene. For instance, easy to transfect HeLa cells are in
general expected to provide a knockdown of about 95% with the maximal chosen
dose for positive control DsiRNA. However, weaker silencing might be achieved if
targeting a gene with a strong compensation mechanism. On the other hand, with
hard-to-transfect cells (e.g., RAW), reaching 70–80% of knockdown with a high
dose of DsiRNA is considered satisfactory.
Whichever cells or transfection agent are at use, the key step for obtaining
optimal DsiRNA silencing remains to extensively optimize each cell line with a
series of reagents using a housekeeping positive control DsiRNA. The number of
DsiRNA candidates to be tested is user-dependent. An experimenter that requests
only one or two good DsiRNAs will approximately analyze a dozen issued
from appropriate algorithms. However, for a researcher who wants several potent
DsiRNAs, tiling the entire gene with DsiRNA may be the only way to find the
optimal one.
cell line and siLentFect reagent from Bio-Rad, they managed to obtain successful
knockdown for all genes with a range of DsiRNA concentration from 100 pM to
50 nM. IDT R&D team further demonstrated the potency of different chemical
modifications on DsiRNA, including 20 -O-methyl (20 OMe) RNA, 20 -fluoro (20 -F)
RNA and DNA bases (Collingwood et al. 2008). Based on their study, a modifica-
tion pattern that includes a 20 OMe bases was identified as the one having minimal
impact on potency, did not trigger immune responses in human immune cells
in vitro and showed improved serum stability. This study was performed either in
HeLa cells with TriFECT reagent (IDT), HCT116 cells with Lipofectamine 2000,
human peripheral blood mononuclear cells (PBMCs), again with Lipofectamine
2000, or T98 cells with siLentFect reagent. In an interesting study recently pub-
lished by Takanori Kubo and colleagues, this group demonstrated the efficacy of
amino-modified DsiRNA by using a luciferase gene reporter assay in HeLa cell line
using the transfection reagent Lipofectamine 2000 (Kubo et al. 2008). The authors
further proposed two different routes (Fig. 2) by which DsiRNA can be processed.
Depending on the modification made to the DsiRNA structure, the second route of
activation, the only one leading to induction of gene silencing, would be favored
thus increasing the knockdown potency. This group also used an unconventional
transfection method represented by the conjugation of 27-mers DsiRNA with
cholesterol bound at the 50 -sense end. This construct possesses high membrane
permeability in the absence of delivery reagent. These results were in agreement
with the data previously demonstrating this phenomenon with classical 21-nt
siRNAs (Soutschek et al. 2004). Another recent study evoked the use of DsiRNA
without a conventional transfection method (Zhou et al. 2009). They created a
chimera composed of DsiRNA linked to an anti-gp120 aptamer, which specifically
binds to and internalizes into cells expressing HIV gp160. Once inside the cells,
DsiRNAs can be processed by dicer and further inhibit HIV-1 replication and
infectivity in CEM T-cells or PBMCs. This study demonstrated the increasing
possibilities of DsiRNA for future therapies. DsiRNAs complexed to chitosan
nanoparticles is also a method proposed to transfect primary peritoneal macro-
phages (Howard et al. 2009). With this technique, they managed to obtain a
knockdown of approximately 66% of TNF-a mRNA pools. This study supports
the one of Amarzguioui in 2006, and both demonstrate an effective knockdown of
TNF-a and a potential for those DsiRNA to be tested in a pain paradigm.
The first studies directly targeting pain mechanisms with DsiRNAs were
published by our group in 2008 (Dore-Savard et al. 2008) and more recently by
others (LaCroix-Fralish et al. 2009). In the first study, in vitro validation of DsiRNAs
targeting the NTS2 receptor, a GPCR known to be involved in nociceptive modula-
tion, was performed (Roussy et al. 2009). Indeed, six DsiRNA and an appropriate
mismatch control were evaluated in vitro for their ability to specifically reduce
NTS2 mRNA levels in an NTS2-stable Chinese Hamster Ovary (CHO) cell line. The
cells were treated with a dose-response curve (0.1–10 nM) of DsiRNA formulated to
RNAiMAX reagent. The capacity for DsiRNA to silence NTS2 expression was
further analyzed using quantitative reverse-transcription PCR. On the six DsiRNA
tested, three resulted in nearly complete inhibition of the targeted gene at 10 nM and
Application of Dicer-Substrate siRNA in Pain Research 179
mRNA
mRNA
Fig. 2 Relationship between dicing and RNAi activity of asymmetric duplex RNA and 50 -sense
amino-modification. (a) Asymmetric duplex increases the interaction of Dicer with the 30 -sense
dangling end, thus promoting route B, the only one leading to efficient gene silencing, over
route A. (b) Amino-modified 50 -sense strand impedes the access of Dicer to the 30 -antisense
end, thus blocking route A and leaving route B as the only one active. This figure represents an
hypothetical mechanism of the higher efficacy of modified asymmetric DsiRNA (Kubo et al. 2008)
180 P. Sarret et al.
remained effective at a very low dose (0.1 nM). In vitro validation was assessed for
functional potency in NIH3T3 cell line (LaCroix-Fralish et al. 2009). In this case, 0.1
or 10 nM DsiRNAs were transfected with siLentFect reagent and incubated for 24 h
on cells before RNA isolation and subsequent quantitative real-time PCR analysis.
Atp1b3, a gene coding for the b3 subunit of the Na+, K+-ATPase pump that
contributes to interstrain differences in the early phase of the formalin test, was
the targeted gene in this study. When testing five different Atp1b3 DsiRNAs, they
managed to obtain more than 90% knockdown at the highest dose, and two of the
candidates remained potent even at 0.1 nM. These data are expected to stimulate
in vivo pain research in the next few years.
RNAi studies rapidly reached the step of applicability to animal models and clinical
trials. This transfer from the cell to the organism is impressively fast considering
that this mechanism was discovered in the late 1990s. The first experiments using
siRNA have been performed in peripheral organs so was the first ex vivo assay
with DsiRNA (Amarzguioui et al. 2006). Only when the application was proven
functional, the technology moved onward the central nervous system. The CNS
environment being hermetic, difficult to access without invasive procedures and
having low tolerance for exogenous contaminants contributes to making this system
less targeted. The situation is nowadays different as technologies evolve. As dis-
cussed earlier in this chapter, DsiRNAs are particularly suitable for such studies
since the effective dose used is very low (ng to low mg), thus lowering side effects.
The formulations used are also more potent in delivering their content in hard-
to-transfect neuronal cells. In the following section, we will describe the application
of DsiRNA in vivo at large, with a final focus on the CNS and pain research.
5.2 Methodology
5.2.1 Formulation
The first choice the experimenter has to make in order to use RNAi in vivo is the
administration technique in combination with the best carrier to favor an optimal
penetration in the targeted tissue and cells. Recently, high efficiency has been
182 P. Sarret et al.
reached with cationic lipids. Their relative toxicity limited the amount used for
systemic studies but the lipid iFect proved to be very efficient in the central nervous
system when locally injected (Luo et al. 2005). DsiRNAs are first diluted in sterile
physiological saline and then formulated with the iFect at a ratio of 1:4–1:5 to avoid
precipitation impairments. A 10 ml volume of the formulation is injected in the
intrathecal space of the rat (5 ml for mice) between L5 and L6 vertebrae under brief
(but deep) anesthesia. It is important to mention that, despite its obvious efficiency,
some concern has been raised with the repeated administration of iFect. Indeed,
transient ataxia on the rotarod was observed (LaCroix-Fralish et al. 2009) while
Sarret’s group noticed a certain level of interference with the pain response at the
formalin test (Dore-Savard et al. 2008). Sarret’s group managed this issue by testing
different cationic lipids, polymers and finally obtained a satisfactory result with the
peptide agent Transducin. As opposed to the other reagents tested, no adverse
effects were identified in tonic pain tests.
When it comes to the knockdown of a targeted protein, many strategies could be
used in terms of dose and timing of injection. However, it is now accepted that
repeated daily injections (generally 2 or 3) in the low mg range are enough to induce a
significant decrease in the expression of the selected molecule. For instance, 1 mg
of DsiRNA daily injected intrathecally for two consecutive days resulted in a
knockdown for three subsequent days (Dore-Savard et al. 2008). In the mouse, a
dose of 0.5 mg has been administered for three consecutive days. The effectiveness of
the knockdown was verified 24 h after the last injection (LaCroix-Fralish et al. 2009).
The next step before running any knockdown experiment is to ensure that delivery
of a fluorophore-coupled DsiRNA is observable in extracted tissue. This strategy was
used by Dore-Savard et al. (2008) with a Texas Red-coupled scramble duplex injected
i.t. and then observed in DRG neurons and superficial layers of the spinal cord
24 h after the last injection (Fig. 3a, b). LaCroix-Fralish and colleagues performed a
similar validation using a Cy3-labeled scrambled DsiRNA to visualize its entry
in dorsal root ganglia (DRG) neurons after an intrathecal injection (Fig. 3c, d)
(LaCroix-Fralish et al. 2009).
The level of remaining expression of the target mRNA after the knockdown repre-
sents crucial information. Real-time quantitative PCR is the approach of choice to
confirm the knockdown at the molecular level. However, to avoid misinterpretation
of the phenotypic behavioral observations, protein expression levels should not be
neglected. If the target mRNA content is decreasing but the protein of interest still
present (e.g., secretion vesicles still loaded), the conclusions observed at the behav-
ioral level will not be the same. Immunohistochemical studies using an antibody for
Application of Dicer-Substrate siRNA in Pain Research 183
a b
c d
Fig. 3 Distribution of fluorescence in dorsal root ganglia and spinal cord following intrathecal
injection of a fluorophore-tagged dicer-substrate siRNA. (a and b) Texas Red-labeled control
DsiRNA distribution is shown in rat DRG and lumbar spinal cord after two spinal injections (1 mg).
The fluorescence is heterogeneous and expressed in every neuronal subpopulation. (c and d) The
same observations are made with a Cy3-labeled DsiRNA (red) injected in the mouse intrathecal
space (0.5 mg). Nuclei were counterstained with DAPI for visualization purposes (Blue). Scale bar:
(100 mm)
the targeted protein or a western blot can be performed to ensure that the mRNA
silencing resulted in an effective knockdown of the protein as well.
activity of a close gene from the same family can be assessed. It was verified that
the activity of the neurotensin receptor NTS1 was not modified by NTS2 knock-
down (Dore-Savard et al. 2008). Finally, the most frequent sequence-independent
side effect of RNAi being the induction of an immune response, the evaluation of
the expression levels of pro-inflammatory genes like IFN-g, TNF-a, Il-6, or COX
can be determined. The latter can be detected in tissue lysates using qPCR or in
cerebrospinal fluid using large-scale immunoassays (e.g., Bioplex: Bio-Rad). Once
all the important controls are performed, the functional observations regarding
the effects of a knockdown can be analyzed without any bias, even if we start
with the concept that low doses of DsiRNA make these events presumably rare.
In order to study the role of a gene in pain perception, the effect of the deletion
needs to be observed at the behavioral or electrophysiological level. Acute pain
tests such as the tail flick, hot plate, von Frey hair, or Hargreaves radiant plantar test
are adequate to detect any modification of the pain threshold at a basal state. A step
further, the formalin tonic pain test is a well-characterized biphasic inflammatory
test to evaluate persistent pain in rodents. The acute phase (0–9 min), induced by
direct stimulation of the nociceptors by the chemical substance, is followed by an
interphase of active inhibition (10–20 min) and ended by a longer tonic inflamma-
tory phase (20–60 min) consisting in secondary hyperalgesia, inflammation, and
central sensitization. The above tests are ultimately performed to predict the
behavior of a new treatment in chronic pain.
In order to provide a reliable “proof of concept” for DsiRNA in the central
nervous system, Sarret’s team chose to aim at a GPCR involved in pain modulation
mechanisms. In 2008, they proved that the intrathecal injection of DsiRNAs could
result in an efficient knockdown of the neurotensin NTS2 receptor known for its
central analgesic effects (Dore-Savard et al. 2008). First, they exposed that the
central delivery of a texas-red DsiRNAs formulated to iFect efficiently penetrated
DRG and spinal cord neurons. Then, they started by determining the efficacy of this
technology in acute pain. Two different duplexes targeting the mRNA coding for
the third intracellullar loop of the receptor were formulated to the cationic lipid
iFect. The complex induced a decrease in the amount of NTS2 mRNA up to 70.4%
in the spinal cord and 85.5% in the DRG following daily intrathecal injections
of only 1 mg for two consecutive days. The behavioral validation of the down-
regulation consisted in the observation of the analgesic effect of a selective NTS2
pharmacological agonist at the tail-flick test, in control and treated rats. The
scrambled-treated group displayed increased latency to tail withdrawal as expected,
while DsiRNA-treated animals showed no analgesia following the intrathecal
injection of the pharmacological agonist, demonstrating the functional deletion of
the receptor. The knockdown lasted for 3 days following the last injection of
DsiRNA (Fig. 4a, b).
Application of Dicer-Substrate siRNA in Pain Research 185
a ** * * iFect + mismatch
14
iFect + DsiRNAv1-1
12 iFect + JMV-431
10 iFect + DsiRNAv1-1
Latency (sec)
+ JMV-431
8
b *** iFect
14
iFect + mismatch
12 **
iFect + DsiRNAv2-5
Latency (sec)
10 iFect + DsiRNAv1-1
8 iFect + JMV-431
iFect + mismatch
6 + JMV-431
c Ctl saline n = 6
3
DsiRNAv2-5 +
JMV-431 n = 6
DsiRNA_NC1 + JMV-431 n=6
2
Pain Score
0
0 10 20 30 40 50 60
Time (min)
Fig. 4 Effects of NTS2 receptor knockdown using dicer-substrate siRNA on pain perception in
rats. (a) Two daily injections of NTS2-targeting DsiRNA induced a marked decrease in the
186 P. Sarret et al.
In the process to add data to this “proof of concept”, knockdown of NTS2 was
confirmed using the formalin test (Tetreault et al. 2009). This step toward chronic
pain assessment was capital to validate sustained pain in a context of RNAi
technology. The exercise was appropriate since Sarret’s team uncovered that the
initial transfection reagent induced undesired side effects in the formalin test. As
previously described, Transducin was selected to perform the formalin test ade-
quately without adverse effects. Using a similar injection protocol and a low dose of
5 mg, the NTS2 receptor was appropriately silenced and the selective NTS2
pharmacological agonist lost its efficiency to inhibit pain behavior during the late
phase of the formalin test (Fig. 4c).
This proof of concept was also reported later by another team looking at the
functional role of the b3 subunit of the Na+, K+-ATPase in the modulation of the
response to formalin in different mice strains (LaCroix-Fralish et al. 2009). In this
study, the hypothesis was that this pump was responsible for the differential pain
responses of two mouse strains to the formalin test. After the uptake of the
DsiRNAs was validated in DRG neurons, it was determined that the intrathecal
infusion of two different DsiRNAs targeting the b3 subunit abolished the difference
between the strains (Fig. 5a). The electrophysiological activity of extracted DRG
neurons in response to an inhibitor of this pump (ouabain) also confirmed the loss of
phenotype in the knockdown mice (Fig. 5b).
<
Fig. 4 (continued) analgesic effect of the selective agonist JMV-431. Indeed, the latency to
withdrawal at the tail flick test was not increased in the treated group, as opposed to the scramble
DsiRNA group, where the analgesic effect of the agonist is conserved. (b) Both DsiRNAs tested did
not modify the tail withdrawal latency compared with the vehicle i-Fect or scramble (mismatch)
DsiRNA treatments. Additionally, mismatch duplexes did not change the antinociceptive effects of
the JMV-431. (c) The same observations were made with the tonic pain formalin test. NTS2-
targeting DsiRNA treatment abolished the analgesic effect (decrease of the pain score in the late
phase of the test) of JMV-431, this time with the Transducin as a carrier (From: Dore-Savard 2008)
Application of Dicer-Substrate siRNA in Pain Research 187
2
ec
lin
A-
A-
N
and scramble DsiRNA
i-F
R
Sa
N
si
R
(scrm siRNA) group.
si
si
rm
sc
(b) Consequently, Na+, K+
pump density of current is
modified in A/J mice
b
4.0
following b3 knockdown
(From: Lacroix-Fralish 2009) 3.5
Na+,K+ Pump Current
3.0
2.5
(pA/pF)
2.0
**
1.5
1.0
0.5 7 8
0.0
scrm siRNA siRNA-2
levels, confirmation of target gene knockdown, safe and efficient delivery as well as
elimination of toxicity and immunogenicity. DsiRNAs represent a technological
improvement in regards to their ability to circumvent the drawbacks while yielding
a high silencing potency for pain research. DsiRNA bioavailability to supraspinal
sites remains the most significant barrier to their use in a clinical pain setting. At
present, considerable effort is being made to develop appropriate vehicle to opti-
mize the delivery and stability for siRNA (Whitehead et al. 2009). Alternatively,
recent advances in blood–brain barrier disruption for delivery of chemotherapy in
the treatment of brain tumor patient may also be applied as a brain delivery strategy
to facilitate siRNA penetration (Bellavance et al. 2008). In summary, carefully
planned preclinical analyses of therapeutic RNAi are required to prevent premature
failures in human trials that could delay or dampen maturation of the field.
188 P. Sarret et al.
References
Akopian AN, Souslova V, England S et al (1999) The tetrodotoxin-resistant sodium channel SNS
has a specialized function in pain pathways. Nat Neurosci 2:541–548
Altier C, Dale CS, Kisilevsky AE et al (2007) Differential role of N-type calcium channel splice
isoforms in pain. J Neurosci 27:6363–6373
Amarzguioui M, Rossi J (2008) Principles of Dicer substrate (D-siRNA) design and function.
Methods Mol Biol 442:3–10
Amarzguioui M, Lundberg P, Cantin E et al (2006) Rational design and in vitro and in vivo
delivery of Dicer substrate siRNA. Nat Protoc 1:508–517
Ameres SL, Martinez J, Schroeder R (2007) Molecular basis for target RNA recognition and
cleavage by human RISC. Cell 130:101–112
Behlke MA (2008) Chemical modification of siRNAs for in vivo use. Oligonucleotides
18:305–319
Bellavance MA, Blanchette M, Fortin D (2008) Recent advances in blood-brain barrier disruption
as a CNS delivery strategy. AAPS J 10:166–177
Cenac N, Altier C, Chapman K et al (2008) Transient receptor potential vanilloid-4 has a major
role in visceral hypersensitivity symptoms. Gastroenterology 135:937–946, e931–932
Christoph T, Grunweller A, Mika J et al (2006) Silencing of vanilloid receptor TRPV1 by RNAi
reduces neuropathic and visceral pain in vivo. Biochem Biophys Res Commun 350:238–243
Christoph T, Bahrenberg G, De Vry J et al (2008) Investigation of TRPV1 loss-of-function
phenotypes in transgenic shRNA expressing and knockout mice. Mol Cell Neurosci
37:579–589
Collingwood MA, Rose SD, Huang L et al (2008) Chemical modification patterns compatible with
high potency dicer-substrate small interfering RNAs. Oligonucleotides 18:187–200
Deval E, Noel J, Lay N et al (2008) ASIC3, a sensor of acidic and primary inflammatory pain.
EMBO J 27:3047–3055
Dobner PR (2006) Neurotensin and pain modulation. Peptides 27:2405–2414
Dong X-W, Goregoaker S, Engler H et al (2007) Small interfering RNA-mediated selective
knockdown of Na(V)1.8 tetrodotoxin-resistant sodium channel reverses mechanical allodynia
in neuropathic rats. Neuroscience 146:812–821
Dore-Savard L, Roussy G, Dansereau MA et al (2008) Central delivery of Dicer-substrate siRNA:
a direct application for pain research. Mol Ther 16:1331–1339
Dorn G, Patel S, Wotherspoon G et al (2004) siRNA relieves chronic neuropathic pain. Nucleic
Acids Res 32:e49
Eguchi A, Dowdy SF (2009) siRNA delivery using peptide transduction domains. Trends Phar-
macol Sci 30:341–345
Eguchi A, Meade BR, Chang Y-C et al (2009) Efficient siRNA delivery into primary cells by a
peptide transduction domain-dsRNA binding domain fusion protein. Nat Biotechnol
27:567–571
Fan J, Wu X, Cao Z et al (2009) Up-regulation of anterior cingulate cortex NR2B receptors
contributes to visceral pain responses in rats. Gastroenterology 136(1732–1740):e1733
Gabra BH, Kessler FK, Ritter JK et al (2007) Decrease in N-methyl-D-aspartic acid receptor-NR2B
subunit levels by intrathecal short-hairpin RNA blocks group I metabotropic glutamate
receptor-mediated hyperalgesia. J Pharmacol Exp Ther 322:186–194
Ganju P, Hall J (2004) Potential applications of siRNA for pain therapy. Expert Opin Biol Ther
4:531–542
Garraway SM, Xu Q, Inturrisi CE (2007) Design and evaluation of small interfering RNAs that
target expression of the N-methyl-D-aspartate receptor NR1 subunit gene in the spinal cord
dorsal horn. J Pharmacol Exp Ther 322:982–988
Garraway SM, Xu Q, Inturrisi CE (2009) siRNA-mediated knockdown of the NR1 subunit gene of
the NMDA receptor attenuates formalin-induced pain behaviors in adult rats. J Pain
10:380–390
Application of Dicer-Substrate siRNA in Pain Research 189
Grimm D, Streetz KL, Jopling CL et al (2006) Fatality in mice due to oversaturation of cellular
microRNA/short hairpin RNA pathways. Nature 441:537–541
Hefner E, Clark K, Whitman C et al (2008) Increased potency and longevity of gene silencing
using validated Dicer substrates. J Biomol Tech 19:231–237
Hemmings-Mieszczak M, Dorn G, Natt FJ et al (2003) Independent combinatorial effect of
antisense oligonucleotides and RNAi-mediated specific inhibition of the recombinant rat
P2X3 receptor. Nucleic Acids Res 31:2117–2126
Hock J, Meister G (2008) The Argonaute protein family. Genome Biol 9:210
Howard KA, Paludan SR, Behlke MA et al (2009) Chitosan/siRNA nanoparticle-mediated TNF-
alpha knockdown in peritoneal macrophages for anti-inflammatory treatment in a murine
arthritis model. Mol Ther 17:162–168
Hutvagner G, Simard MJ, Mello CC et al (2004) Sequence-specific inhibition of small RNA
function. PLoS Biol 2:E98
Jackson AL, Bartz SR, Schelter J et al (2003) Expression profiling reveals off-target gene
regulation by RNAi. Nat Biotechnol 21:635–637
Jain KK (2008) Gene therapy for pain. Expert Opin Biol Ther 8:1855–1866
Ji X (2008) The mechanism of RNase III action: how dicer dices. Curr Top Microbiol Immunol
320:99–116
Jinek M, Doudna JA (2009) A three-dimensional view of the molecular machinery of RNA
interference. Nature 457:405–412
John M, Constien R, Akinc A et al (2007) Effective RNAi-mediated gene silencing without
interruption of the endogenous microRNA pathway. Nature 449:745–747
Kariko K, Bhuyan P, Capodici J et al (2004) Small interfering RNAs mediate sequence-
independent gene suppression and induce immune activation by signaling through toll-like
receptor 3. J Immunol 172:6545–6549
Kasama S, Kawakubo M, Suzuki T et al (2007) RNA interference-mediated knock-down of
transient receptor potential vanilloid 1 prevents forepaw inflammatory hyperalgesia in rat.
Eur J Neurosci 25:2956–2963
Kim D-H, Behlke MA, Rose SD et al (2005) Synthetic dsRNA Dicer substrates enhance RNAi
potency and efficacy. Nat Biotechnol 23:222–226
Kortylewski M, Swiderski P, Herrmann A et al (2009) In vivo delivery of siRNA to immune cells
by conjugation to a TLR9 agonist enhances antitumor immune responses. Nat Biotechnol
27:925–932
Kubo T, Zhelev Z, Ohba H et al (2007) Modified 27-nt dsRNAs with dramatically enhanced
stability in serum and long-term RNAi activity. Oligonucleotides 17:445–464
Kubo T, Zhelev Z, Ohba H et al (2008) Chemically modified symmetric and asymmetric duplex
RNAs: an enhanced stability to nuclease degradation and gene silencing effect. Biochem
Biophys Res Commun 365:54–61
Kurreck J (2004) Antisense and RNA interference approaches to target validation in pain research.
Curr Opin Drug Discov Devel 7:179–187
Kurreck J (2009) RNA interference: from basic research to therapeutic applications. Angew Chem
Int Ed Engl 48:1378–1398
LaCroix-Fralish ML, Mo G, Smith SB et al (2009) The beta3 subunit of the Na+, K+-ATPase
mediates variable nociceptive sensitivity in the formalin test. Pain 144:294–302
Lima WF, Murray H, Nichols JG et al (2009) Human Dicer binds short single-strand and double-
strand RNA with high affinity and interacts with different regions of the nucleic acids. J Biol
Chem 284:2535–2548
Lin X, Ruan X, Anderson MG et al (2005) siRNA-mediated off-target gene silencing triggered
by a 7 nt complementation. Nucleic Acids Res 33:4527–4535
Lin CR, Amaya F, Barrett L et al (2006) Prostaglandin E2 receptor EP4 contributes to inflammatory
pain hypersensitivity. J Pharmacol Exp Ther 319:1096–1103
Liu TT, Bi HS, Lv SY, Wang XR, Yue SW (2010) Inhibition of the expression and function of
TRPV4 by RNA interference in dorsal root ganglion. Neurol Res 32(5):466–471
190 P. Sarret et al.
Luo MC, Zhang DQ, Ma SW et al (2005) An efficient intrathecal delivery of small interfering
RNA to the spinal cord and peripheral neurons. Mol Pain 1:29
Macrae IJ, Zhou K, Li F et al (2006) Structural basis for double-stranded RNA processing by
Dicer. Science 311:195–198
Merkel OM, Beyerle A, Librizzi D et al (2009) Nonviral siRNA delivery to the lung: investigation
of PEG-PEI polyplexes and their in vivo performance. Mol Pharm 6:1246–1260
Ndong C, Pradhan A, Puma C et al (2009) Role of rat sensory neuron-specific receptor (rSNSR1)
in inflammatory pain: contribution of TRPV1 to SNSR signaling in the pain pathway. Pain
143:130–137
Rose SD, Kim D-H, Amarzguioui M et al (2005) Functional polarity is introduced by Dicer
processing of short substrate RNAs. Nucleic Acids Res 33:4140–4156
Roussy G, Dansereau MA, Dore-Savard L et al (2008) Spinal NTS1 receptors regulate nociceptive
signaling in a rat formalin tonic pain model. J Neurochem 105:1100–1114
Roussy G, Dansereau MA, Baudisson S et al (2009) Evidence for a role of NTS2 receptors in the
modulation of tonic pain sensitivity. Mol Pain 5:38
Sah DW (2006) Therapeutic potential of RNA interference for neurological disorders. Life Sci
79:1773–1780
Sah D, Guo W, Luo M et al (2006) Small interfering RNA targeting NaV1.8 alleviates experimen-
tally-induced chronic pain. Abstr Soc Neurosci 245.11/N10
Sarret P, Esdaile MJ, Perron A et al (2005) Potent spinal analgesia elicited through stimulation of
NTS2 neurotensin receptors. J Neurosci 25:8188
Saxena S, Jonsson ZO, Dutta A (2003) Small RNAs with imperfect match to endogenous mRNA
repress translation. Implications for off-target activity of small inhibitory RNA in mammalian
cells. J Biol Chem 278:44312–44319
Schwarz DS, Hutvagner G, Du T et al (2003) Asymmetry in the assembly of the RNAi enzyme
complex. Cell 115:199–208
Sledz CA, Holko M, de Veer MJ et al (2003) Activation of the interferon system by short-
interfering RNAs. Nat Cell Biol 5:834–839
Soutschek J, Akinc A, Bramlage B et al (2004) Therapeutic silencing of an endogenous gene by
systemic administration of modified siRNAs. Nature 432:173–178
Tan PH, Yang LC, Shih HC et al (2005) Gene knockdown with intrathecal siRNA of NMDA
receptor NR2B subunit reduces formalin-induced nociception in the rat. Gene Ther 12
(1):59–66
Tang G (2005) siRNA and miRNA: an insight into RISCs. Trends Biochem Sci 30:106–114
Tetreault PG, Beaudet N, Dore-Savard L et al (2009). Functional inhibition of NTS2 receptors by
Dicer-substrate siRNA modulates tonic pain perception. 39th annual meeting of the Society for
Neuroscience, Chicago
van den Berg A, Mols J, Han J (2008) RISC-target interaction: cleavage and translational
suppression. Biochim Biophys Acta 1779:668–677
Vit JP, Ohara PT, Bhargava A et al (2008) Silencing the Kir4.1 potassium channel subunit in
satellite glial cells of the rat trigeminal ganglion results in pain-like behavior in the absence of
nerve injury. J Neurosci 28:4161–4171
Wang HW, Noland C, Siridechadilok B et al (2009) Structural insights into RNA processing by the
human RISC-loading complex. Nat Struct Mol Biol 16:1148–1153
Whitehead KA, Langer R, Anderson DG (2009) Knocking down barriers: advances in siRNA
delivery. Nat Rev Drug Discov 8:129–138
Zhang HM, Chen SR, Cai YQ et al (2009) Signaling mechanisms mediating muscarinic
enhancement of GABAergic synaptic transmission in the spinal cord. Neuroscience
158:1577–1588
Zhou J, Swiderski P, Li H et al (2009) Selection, characterization and application of new RNA
HIV gp 120 aptamers for facile delivery of Dicer substrate siRNAs into HIV infected cells.
Nucleic Acids Res 37:3094–3109
RNAi Treatment of HIV-1 Infection
Contents
1 The RNAi Pathway for Expression of Therapeutic siRNAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
2 RNAi Gene Therapy Against HIV-1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
3 Finding the Optimal Viral Target Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
4 Combinatorial RNAi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
5 Targeting Cellular Cofactors of HIV-1 Replication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
6 Alternative Inhibitory RNA Molecules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
7 Preclinical Test Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
8 Sequence-Specificity of RNAi Action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
9 Safety Issues Raised in Clinical Trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
10 Clinical Trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
11 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 191
RNA Technologies, DOI 10.1007/978-3-642-12168-5_8,
# Springer-Verlag Berlin Heidelberg 2010
192 K.J. von Eije and B. Berkhout
in vitro and in vivo test systems to evaluate the efficacy and safety of an RNAi
gene therapy.
One RNA strand gets preferentially incorporated into the complex (Tomari et al.
2004; Tomari and Zamore 2005).
In mammals, RNAi-mediated gene silencing is mainly elicited by translational
repression of the targeted mRNA (Ambros 2004). An important determinant is the
level of base pairing complementarity between the miRNA and the mRNA target,
leading to mRNA cleavage (perfect complementarity) or translational repression
(near-perfect complementarity) (Brennecke et al. 2003; Doench and Sharp 2004;
Lewis et al. 2003; Kiriakidou et al. 2004; Lai 2002). RISC typically forms com-
plexes when the “seed” region of the miRNA (50 end) that finds multiple target
sequences in the 30 untranslated region (30 UTR) of the mRNA. The number
of 30 UTR targets and their distance determines the silencing efficiency (Saetrom
et al. 2007). Most mammalian miRNAs anneal through imperfect base pairing
complementarity with the mRNA to cause translational repression, but at least
one case of perfect complementarity and mRNA cleavage is known in humans
(Yekta et al. 2004). Endonucleolytic cleavage of the targeted mRNA occurs
opposite of nucleotide position 10–11 in the miRNA, and the cleaved mRNA is
subsequently degraded.
In contrast to natural miRNAs, small interfering RNA (siRNA) with full base
pairing complementarity can direct mRNA cleavage with only a single target site
that can be located anywhere within the mRNA. Such artificial double-stranded (ds)
RNA can be produced by several methods. Synthetic mature siRNAs can be
transfected into cells (Elbashir et al. 2001), but short hairpin RNAs (shRNAs)
(Brummelkamp et al. 2002; Paddison et al. 2002) and artificial miRNAs are
expressed intracellularly from a transgene construct (Zeng et al. 2002). The natural
miRNA pathway can be instructed with the man-made inhibitors for therapeutic
downregulation of a specific mRNA. This therapeutic approach is relevant for
diseases caused by overexpression of a specific mRNA or to specifically target
the RNA genomes of invading microbes such as HIV-1.
Instruction of the cellular miRNA pathway with new siRNA specificity is
associated with certain risks. One potential problem is direct competition of the
artificial siRNAs with the endogenous siRNAs and/or saturation of the miRNA
pathway. Since the miRNA pathway is important in the control of cellular gene
expression, this can have unwanted side effects such as cell death, disturbances
in cell differentiation programs, or even cancer. Saturation of the miRNA pathway
can occur and lead to death when high doses of shRNAs were delivered by an
adeno-associated virus (AAV) vector in mice (Grimm et al. 2006; McBride et al.
2008; Boudreau et al. 2008; Vickers et al. 2007; Castanotto et al. 2007; Ter Brake
et al. 2009). Another potential problem is targeting of other mRNAs as the miRNAs
require only a seed sequence complementarity of 7–8 base pairs within the 30 UTR
of a given mRNA for function (Brennecke et al. 2005). Such “off-target” effects can
be elicited by the passenger or guide strand (Jackson et al. 2003; Fedorov et al.
2006; Jackson et al. 2006). Yet another problem relates to the induction of an
immune response by siRNAs and shRNAs (Bridge et al. 2003; Sledz et al. 2003),
which can be avoided by optimal design of the si/shRNA molecule (Marques and
Williams 2005).
194 K.J. von Eije and B. Berkhout
The goal of an RNAi-based gene therapy approach against HIV-1 is to protect the
cells of the immune system that are susceptible to HIV-1 infection. This includes
the CD4+ T cells, monocytes, macrophages, and dendritic cells. Such “intracellular
immunization” will prevent the depletion of these immune cells during disease
progression. Maintenance of the immune function should prevent opportunistic
infections and progression towards AIDS.
HIV-1 causes a chronic infection without the possibility of viral clearance. Thus,
a continuously active treatment regime is required. Repeated delivery of exogenous
siRNAs as anti-HIV therapy has been described (Kumar et al. 2008) in a humanized
immune system (HIS) mouse model. Effective virus inhibition was obtained in this
animal model, with a concomitant prevention of the loss of CD4+ T cells. However,
it is doubtful whether such an approach would be suitable in a patient setting, where
the prevention of viral escape requires the continuous presence of an effective
siRNA dose in all infected human cells that are located in many different body
compartments. Instead, we advocate the concept of continuous expression of anti-
HIV molecules obtained by a single transduction of susceptible cells with a
lentiviral vector. A lentiviral vector is based on the genome of HIV-1 itself. The
pathogenic genes are replaced by novel control and therapeutic sequences. The
lentiviral vector does infect the target cell and deposits the transgene, but cannot
replicate. A benefit of the lentiviral vector compared to other viral delivery methods
is that dividing and nondividing cell types can be transduced efficiently. Further-
more, the lentiviral vector is stably integrated into the host cell genome, thus
yielding permanent transduction (Naldini et al. 1996).
Some problems can be encountered when using lentiviruses to target HIV-1 with
RNAi, such as self-targeting of the transgene RNA by the shRNA expressed in the
producer cell, and targeting of HIV-derived sequences in the vector genome. These
problems and solutions have previously been discussed in detail by Ter Brake et al.
(Ter Brake and Berkhout 2007). There are other viral vector systems available for
delivering a therapeutic RNAi transgene, and these have been extensively discussed
by others (Nguyen et al. 2008; de Fougerolles 2008).
We depicted a possible gene therapy procedure for HIV-infected individuals
using the lentiviral vector in Fig. 1. Hematopoietic stem cells seed the different
lineages of immune cells in the blood and are therefore interesting candidates for an
ex vivo gene therapy followed by autolog transplantation. The lentiviral vector will
durably equip all hematopoietic stem cell-derived immune cells with the antiviral
arsenal. In the presence of HIV-1, one expects the preferential survival of the
shRNA-expressing immune cells over untreated cells, which will result in a gradual
increase in the percentage of protected cells. The treatment should result in partial
or complete reconstitution of the immune system, thus preventing HIV-1 infection
to progress towards AIDS. Ideally, a single gene therapy treatment should achieve a
durable effect because the transduced stem cells continue to generate immune cells
of the different lineages. Hematopoietic stem cells transduced with a retroviral
vector encoding an anti-HIV-1 ribozyme have already been used in clinical trials
RNAi Treatment of HIV-1 Infection 195
Fig. 1 RNAi gene therapy for HIV-1. The HIV-1 infected patient that fails on regular antiretro-
viral therapy (1) could be offered the RNAi-based gene therapy with a lentiviral vector. The
lentiviral vector is produced in 293T cells (2) transfected with the lentiviral vector (e.g., JS1) and a
standard set of packaging plasmids (pRSV-Rev, pVSV-g and pSYNGP). The lentiviral vector will
produce viral genomes and the packaging plasmids will produce the proteins required to assemble
new viral particles. pVSV-g produces the Vesicular Stomatitis Virus glycoprotein that is used for
virus pseudotyping. Virus particles are collected after 2 or 3 days. The patient will undergo an
apheresis for the collection of hematopoietic stem cells after pre-treatment with granulocyte
colony stimulatory factor (G-CSF) that mobilizes these cells from the bone marrow into the
periphery (3). The hematopoietic stem cells will be purified and transduced with the therapeutic
lentiviral construct (4). This “intracellular immunization” with the antiviral shRNA will protect
these cells against HIV-1. Transduced cells will be infused back into the patient (5) and the
HIV-resistant immune cells will hopefully prevent disease progression towards AIDS (6)
(Amado et al. 2004; Mitsuyasu et al. 2009). These trials demonstrate the feasibility
and safety of the proposed stem cell approach. Another option is the treatment of
the mature CD4+ T-cell population, in which case repetitive gene therapy should be
applied because T cells have only a limited life span (Dropulic 2001).
Potent and sequence-specific HIV-1 inhibition has been reported with RNAi-
inducing reagents in cell culture infections. It soon became apparent that HIV-1 is
196 K.J. von Eije and B. Berkhout
prone to viral escape in a mono-shRNA therapy (Boden et al. 2003; Das et al. 2004;
Nishitsuji et al. 2006; Sabariegos et al. 2006; Ter Brake et al. 2006; Unwalla et al.
2006; Westerhout et al. 2005), similar to single antiretroviral drug regimens. The
therapeutic vector used in a clinical trial should therefore always tackle the virus with
multiple inhibitors at the same time. Such a combinatorial RNAi attack can target the
virus (Ter Brake et al. 2008), host-encoded cofactors, or both. One could also
combine RNAi molecules with other RNA effector molecules such as decoys and
ribozymes (Li et al. 2005). Another elegant solution to avoid viral escape is the use of
the second generation shRNAs that specifically target viral escape variants (Brake
and Berkhout 2005). However, the relatively high number of viral escape routes
available to HIV-1 reduces the feasibility of this approach (von Eije et al. 2008).
How to identify the optimal target sites for RNAi attack on the 9 kb HIV-1 RNA
genome? One option is the selection of targets in the early spliced mRNAs that
encode the early viral proteins Tat, Rev, and Nef. An early block in viral gene
expression will severely impact the expression of structural viral proteins and
consequently virion assembly. Alternatively, one could target HIV-1 genome
regions that are represented in all spliced viral mRNAs in the untranslated 50 -leader
and 30 -trailer domains (Muesing et al. 1985). Target RNA structure can block an
RNAi attack (Westerhout et al. 2005; Westerhout and Berkhout 2007). Thus,
targeting of “open” RNA domains is beneficial. In addition, the selection of targets
with highly conserved sequences seems important to allow inhibition of as many
virus strains as possible. Targeting of highly conserved genome regions may also
restrict the evolution of viral escape mutants because sequence variation may have
an impact on the viral fitness. An extensive shRNA screen against highly conserved
sequences of the HIV-1 genome has been performed, yielding approximately
20 candidate shRNAs (Ter Brake et al. 2006). Stable shRNA-expressing T cell
lines were infected with HIV-1, which yielded four durable shRNA inhibitors that
restricted virus replication for more than 100 days (von Eije et al. 2009). We and
others have identified effective shRNAs and siRNAs targeting regulatory HIV-1
sequences, e.g., in the long terminal repeat (LTR) and untranslated leader
RNA (Jacque et al. 2002; Ter Brake et al. 2006) and most viral genes: gag
(Chang et al. 2005; Novina et al. 2002; Park et al. 2002; Ter Brake et al. 2006),
pol (Chang et al. 2005; Berkhout and Brake 2008; Surabhi and Gaynor 2002), vif
(Jacque et al. 2002), tat (Ter Brake et al. 2006; Coburn and Cullen 2002; Lee et al.
2002a; Surabhi and Gaynor 2002), rev (Ter Brake et al. 2006; Coburn and Cullen
2002; Lee et al. 2002a), vpu (Chang et al. 2005), env (Park et al. 2002), and nef
(Jacque et al. 2002). Follow-up analyses should include prolonged culturing of
stably transduced T cells to score the impact on cell viability. Prolonged culturing
in the presence of HIV-1 can result in viral escape. The appearance of mutations in
the viral target demonstrates the exquisite sequence-specificity of RNAi action.
RNAi Treatment of HIV-1 Infection 197
4 Combinatorial RNAi
An advantage of targeting host cell cofactors that are important for HIV-1
replication is the reduced chance of viral escape. Silencing of several cofactors
resulted in HIV-1 inhibition: nuclear factor kappa B (Surabhi and Gaynor 2002),
198 K.J. von Eije and B. Berkhout
CD4 (Novina et al. 2002; Anderson et al. 2003a), CXCR4 (Martinez et al. 2002;
Anderson et al. 2003a; Anderson and Akkina 2005; Anderson et al. 2003b),
DDX-3 (Ishaq et al. 2008), LEDGF/p75 (Vandekerckhove et al. 2006), and
CCR5 (An et al. 2007a; Anderson et al. 2003a; Anderson and Akkina 2005).
CCR5 is a critical receptor for HIV-1 entry and a promising and well-studied
target. Individuals with the D-32 mutation in CCR5 are resistant to HIV-1 infec-
tion, yet are healthy. Only an increased risk for infection with the West Nile virus
has been reported (Lim et al. 2006). A potent shRNA targeting this host cell
factor has been developed (Anderson et al. 2003a; An et al. 2007a). CCR5-tropic
viruses are generally responsible for HIV-1 transmission to other individuals, but
the virus can also use the alternative CXCR4 receptor. When the CCR5 receptor
is downregulated, this will potentially set the stage for selection of CXCR4-tropic
HIV-1 variants, but this remains to be tested. Many cellular targets will obviously
not be good candidates for a gene therapy because they are essential for the cell
and the host. For example, CXCR4 is required for homing of hematopoietic stem
cells to the bone marrow and subsequent T-cell differentiation (Lapidot 2001).
High-throughput RNAi gene knockdown screens have recently identified many
candidate host cofactors of HIV-1 replication (Brass et al. 2008; Konig et al.
2008; Zhou et al. 2008). Identified cofactors should first be validated using
alternative assay systems to exclude false positives. Such studies should increase
the arsenal of cellular targets available for a combinatorial RNAi approach.
Other types of inhibitory RNA molecules can be used to target the virus. The
currently ongoing phase I clinical trial at the City of Hope uses a lentiviral vector
that encodes a TAR-decoy, CCR5-ribozyme, and a shRNA targeting the HIV-1
genome in the tat–rev region (Li et al. 2005). The TAR-decoy is a small nucleolar
RNA molecule that binds Tat, which will prevent the Tat–TAR interaction that is
essential for enhanced viral promoter activity (Lisziewicz et al. 1993). The CCR5
ribozyme cleaves the CCR5 mRNA to cause reduced CCR5 expression on the
cell surface (Sarver et al. 1990). Other anti-HIV-1 RNA molecules include
antisense transcripts (Chatterjee et al. 1992; Levine et al. 2006), decoys (Kohn
et al. 1999), ribozymes (Sarver et al. 1990), and aptamers (Symensma et al.
1996). A new addition to this arsenal is an antisense molecule that can elicit
transcriptional gene silencing of the viral LTR promoter (Weinberg et al. 2006).
The novel RNAu method is based on the expression of a modified U1 small
nuclear RNA that blocks polyadenylation of the targeted mRNA, which is
subsequently degraded (Abad et al. 2008). Comprehensive reviews on combina-
torial RNA approaches are available (Grimm and Kay 2007; Liu and Berkhout
2008).
RNAi Treatment of HIV-1 Infection 199
When potent antiviral shRNAs are successfully expressed from a single vector, one
can move to relevant preclinical models to critically assess the safety and efficacy.
A simple and efficient in vitro test system to measure the impact on cell viability is
to perform a co-culture of GFP+ transduced cells and non-treated cells (unpublished
results). A reduction over time in the percentage of GFP+ cells is an indication of
delayed cell growth and RNAi toxicity. Outgrowth of the transduced cells should
occur in the presence of HIV-1, again using simple FACS analysis to screen the
mixed cell culture.
The SIV-macaque model (Lackner and Veazey 2007), which has been used
extensively for vaccination studies, can be considered for testing of an anti-HIV-1
RNAi gene therapy, although it has several limitations. First, anti-HIV shRNAs
cannot easily be tested against SIV because of sequence dissimilarity, and the same
likely holds for the genes of cofactors in man versus macaque. Thus, the anti-HIV
shRNAs should either be converted into anti-SIV shRNAs, which may affect their
inhibitory power, or HIV-1 target sequences can be incorporated into the SIV
genome. Second, transduction of the HIV-based lentiviral vector is restricted by
TRIM5a in macaque cells (Stremlau et al. 2004). Third, macaque experiments are
rather expensive, and the number of animals that can be used is restricted. A
minimally modified simian-tropic (st) HIV-1 strain has recently been developed
that produces an acute viraemie and persistent infection in pig-tailed macaques
(Hatziioannou et al. 2009). In contrast to most infected humans, stHIV-1 infection
is controlled in the macaque model after several months.
Most of these limitations do not apply to HIS mouse models (Traggiai et al.
2004; Gimeno et al. 2004). All major human myeloid and lymphoid cellular
compartments develop and mature from input human stem cells in the most recent
HIS mouse (Legrand et al. 2006b; Shultz et al. 2007; Manz 2007). This model
provides access to in vivo and ex vivo experimentation on human T cells (Legrand
et al. 2006a). HIS mice can be infected by injection of the virus but also via rectal
and vaginal transmission routes. Infection results in viremia and the depletion of
human CD4+ cells as seen in the disease course of infected patients (Baenziger
et al. 2006; Zhang et al. 2007; Watanabe et al. 2007; Berges et al. 2006; Berges et al.
2008; An et al. 2007b). We used this model to test safety and efficacy of a lentiviral-
based gene therapy of human hematopoietic stem cells (Ter Brake et al. 2009).
These and other animal models have recently been reviewed (Goldstein 2008).
An important lesson to be learned from various siRNA tests concerns the inclusion
of appropriate control experiments. Several studies on the inhibition of infections
and inflammation used a control siRNA that targets GFP. Results were in favor of a
200 K.J. von Eije and B. Berkhout
therapeutic effect, but it turned out that the GFP control is a particular siRNA of low
immunogenicity compared to other shRNAs, including the therapeutic ones. Most
siRNAs trigger the TLR-7/8 interferon pathway, but the GFP siRNA control does
not (Robbins et al. 2008). Anyhow, the therapeutic effect was not elicited by
specific downregulation of the targeted mRNA.
Another lesson comes from a study on an siRNA therapeutic designed for the
eye against age-related macular degeneration. The siRNA exhibited a therapeutic
effect, but this was not elicited by the RNAi mechanism since the effector molecule
cannot penetrate cells. Instead, the clinical effect was obtained through TLR-3
signaling (Kleinman et al. 2008). Both examples illustrate the importance of
selecting the correct controls to ensure that one is looking at RNAi-specific effects.
For HIV-1 therapies that target the viral genome, exclusive specificity can be
demonstrated with escape variants with one or more point mutations in the target
sequence. The sequence-specificity of a particular RNAi effector molecule must be
demonstrated in vitro and in the animal model before moving to the clinical phase
(von Eije et al. 2008; Ter Brake et al. 2008; Ter Brake et al. 2009).
The first patient treated with a gene therapy in 1990 suffered from adenosine
deaminase deficiency, a form of severe combined immunodeficiency (SCID)
(Culliton 1990). In 1999, a patient died due to the administration of a gene therapy.
This patient was treated for a genetic liver disease – ornithine transcarbamylase
deficiency – and received an adenovirus treatment with the wild-type gene. He died
4 days later of a massive immune response, most likely triggered by the use of
the viral vector (Marshall 1999). A trial where patients with SCID received a
g-retroviral gene transfer with the wild-type interleukin-2 gene started in 2000.
Although this procedure improved the condition of all patients – a true success
(Hacein-Bey-Abina et al. 2002) – two patients developed a leukemia-like condition
of clonal lymphocyte proliferation. Both cases were caused by integration of the
retroviral vector near the promoter of the LMO2 proto-oncogene, leading to
enhanced expression of the LMO2 protein, which has a crucial role in hematopoie-
tic development (Hacein-Bey-Abina et al. 2003). More patients in this and a similar
trial subsequently developed leukemia-like conditions due to insertional oncogene-
sis. By now, more than 1,300 clinical trials involving a gene therapy have been
performed (Edelstein et al. 2007). From these clinical trials, lessons can be learned
for future improvement of gene therapies. For instance, retroviral vectors have been
replaced by lentiviral vectors, which are much more safe because all transcriptional
enhancer motifs have been removed (“self-inactivating” design (Bushman et al.
2005)) and because these vectors tend to integrate in genes, and not near the
promoter region. In addition, experiments with a lentiviral vector and hematopoie-
tic stem cells in tumor-prone mice did not, in contrast to the retroviral vector, show
signs of insertional oncogenesis (Montini et al. 2006). Other safety and regulatory
RNAi Treatment of HIV-1 Infection 201
10 Clinical Trials
An overview of gene therapy trials for HIV-1 is available (Rossi et al. 2007). While
positive in vitro results were obtained for these antiviral gene therapies, the clinical
trials failed to demonstrate a therapeutic benefit. In studies where T cells or
hematopoietic stem cells were treated with the original retroviral vectors, one of
the bottlenecks was effective gene delivery to a clinically relevant number of cells
(Mitsuyasu et al. 2009). This problem has disappeared with the use of lentiviral
vectors that have a much higher transduction efficiency in a variety of cell types. In
addition, many of the previously used inhibitory RNA molecules seem suboptimal
when compared to antiviral shRNAs. RNAi is therefore a promising candidate for
development of a future anti-HIV-1 gene therapy.
The first clinical trial with a lentiviral vector was in fact directed against HIV-1
by expression of an extended antisense transcript against the viral RNA genome.
Persistent in vivo expression of the therapeutic antisense molecule was documented
by the VirXsys company (Levine et al. 2006). In addition, vector integration sites in
blood cells revealed a preference for gene-rich regions, which is typical for a
lentivirus, and no signs of insertional oncogenesis were observed. Another lenti-
viral vector trial was initiated recently in children with X-linked adrenoleukody-
strophy. Promising results were presented by Nathalie Cartier-Lacave (Necker
Hospital, Paris) at the 2008 meeting of the European Society for Gene and Cell
Therapy (ESGCT). Hematopoietic stem cells were effectively transduced ex vivo,
persistence of the modified cells was documented upon reinjection, and clinical
improvement was reported. Another anti-HIV gene therapy trial with a triple RNA
payload (ribozyme, decoy, shRNA) was initiated at the City of Hope by the team of
John Rossi. Persistent gene marking and consistent shRNA expression was docu-
mented in blood cells (John Rossi, personal communication).
11 Conclusion
vector that encodes four antiviral shRNAs to evaluate its safety and efficacy.
This vector yielded very potent antiviral effects in prolonged in vitro cell cultures
(Ter Brake et al. 2008). Safety and efficacy are currently addressed in a humanized
mouse model (Ter Brake et al. 2009), and we expect to initiate a clinical trial within
2 years.
References
Aagaard LA, Zhang J, von Eije KJ et al (2008) Engineering and optimization of the miR-106b
cluster for ectopic expression of multiplexed anti-HIV RNAs. Gene Ther 15:1536–1549
Abad X, Vera M, Jung SP et al (2008) Requirements for gene silencing mediated by U1 snRNA
binding to a target sequence. Nucleic Acids Res 36:2338–2352
Amado RG, Mitsuyasu RT, Rosenblatt JD et al (2004) Anti-human immunodeficiency virus
hematopoietic progenitor cell-delivered ribozyme in a phase I study: myeloid and lymphoid
reconstitution in human immunodeficiency virus type-1-infected patients. Hum Gene Ther
15:251–262
Ambros V (2001) microRNAs: tiny regulators with great potential. Cell 107:823–826
Ambros V (2004) The functions of animal microRNAs. Nature 431:350–355
An DS, Donahue RE, Kamata M et al (2007a) Stable reduction of CCR5 by RNAi through
hematopoietic stem cell transplant in non-human primates. Proc Natl Acad Sci USA
104:13110–13115
An DS, Poon B, Ho Tsong FR et al (2007b) Use of a novel chimeric mouse model with a
functionally active human immune system to study human immunodeficiency virus type 1
infection. Clin Vaccine Immunol 14:391–396
Anderson J, Akkina R (2005) CXCR4 and CCR5 shRNA transgenic CD34+ cell derived macro-
phages are functionally normal and resist HIV-1 infection. Retrovirology 2:53
Anderson J, Banerjea A, Akkina R (2003a) Bispecific short hairpin siRNA constructs targeted to
CD4, CXCR4, and CCR5 confer HIV-1 resistance. Oligonucleotides 13:303–312
Anderson J, Banerjea A, Planelles V, Akkina R (2003b) Potent suppression of HIV type 1 infection
by a short hairpin anti-CXCR4 siRNA. AIDS Res Hum Retrovir 19:699–706
Baehrecke EH (2003) miRNAs: micro managers of programmed cell death. Curr Biol
13:R473–R475
Baenziger S, Tussiwand R, Schlaepfer E et al (2006) Disseminated and sustained HIV infection in
CD34+ cord blood cell-transplanted Rag2-/-gamma c-/-mice. Proc Natl Acad Sci USA
103:15951–15956
Banerjea A, Li MJ, Bauer G et al (2003) Inhibition of HIV-1 by lentiviral vector-transduced
siRNAs in T lymphocytes differentiated in SCID-hu mice and CD34+ progenitor cell-derived
macrophages. Mol Ther 8:62–71
Barichievy S, Saayman S, von Eije KJ et al (2007) The inhibitory efficacy of RNA POL
III-expressed long hairpin RNAs targeted to untranslated regions of the HIV-1 50 long terminal
repeat. Oligonucleotides 17:419–431
Bartel DP (2004) MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 116:281–297
Berges BK, Akkina SR, Folkvord JM et al (2008) Mucosal transmission of R5 and X4 tropic
HIV-1 via vaginal and rectal routes in humanized Rag2(-/-)gammac(-/-) (RAG-hu) mice.
Virology 373:342–351
RNAi Treatment of HIV-1 Infection 203
Berges BK, Wheat WH, Palmer BE et al (2006) HIV-1 infection and CD4 T cell depletion in the
humanized Rag2-/-gamma c-/-(RAG-hu) mouse model. Retrovirology 3:76
Berkhout B, Brake Ot (2008) Towards an RNAi-based gene therapy. BIOforum Europe 4:35–37
Bernstein E, Caudy AA, Hammond SM et al (2001) Role for a bidentate ribonuclease in the
initiation step of RNA interference. Nature 409:363–366
Boden D, Pusch O, Lee F et al (2003) Human immunodeficiency virus type 1 escape from RNA
interference. J Virol 77:11531–11535
Bohnsack MT, Czaplinski K, Gorlich D (2004) Exportin 5 is a RanGTP-dependent dsRNA-
binding protein that mediates nuclear export of pre-miRNAs. RNA 10:185–191
Boudreau RL, Monteys AM, Davidson BL (2008) Minimizing variables among hairpin-based
RNAi vectors reveals the potency of shRNAs. RNA 14:1834–1844
Brake Ot, Berkhout B (2005) A novel approach for inhibition of HIV-1 by RNA interference:
counteracting viral escape with a second generation of siRNAs. J RNAi Gene Silencing
1:56–65
Brass AL, Dykxhoorn DM, Benita Y et al (2008) Identification of host proteins required for HIV
infection through a functional genomic screen. Science 319:921–926
Brennecke J, Hipfner DR, Stark A et al (2003) Bantam encodes a developmentally regulated
microRNA that controls cell proliferation and regulates the proapoptotic gene hid in
Drosophila. Cell 113:25–36
Brennecke J, Stark A, Russell RB, et al (2005) Principles of microRNA-target recognition. PLoS
Biol 3:e85
Bridge AJ, Pebernard S, Ducraux A et al (2003) Induction of an interferon response by RNAi
vectors in mammalian cells. Nat Genet 34:263–264
Brummelkamp TR, Bernards R, Agami R (2002) A system for stable expression of short
interfering RNAs in mammalian cells. Science 296:550–553
Bushman F, Lewinski M, Ciuffi A et al (2005) Genome-wide analysis of retroviral DNA
integration. Nat Rev Microbiol 3:848–858
Carrington JC, Ambros V (2003) Role of microRNAs in plant and animal development. Science
301:336–338
Castanotto D, Sakurai K, Lingeman R et al (2007) Combinatorial delivery of small interfering
RNAs reduces RNAi efficacy by selective incorporation into RISC. Nucleic Acids Res
35:5154–5164
Chan SP, Slack FJ (2007) And now introducing mammalian mirtrons. Dev Cell 13:605–607
Chang LJ, Liu X, He J (2005) Lentiviral siRNAs targeting multiple highly conserved RNA
sequences of human immunodeficiency virus type 1. Gene Ther 12:1133–1144
Chatterjee S, Johnson PR, Wong KK Jr (1992) Dual-target inhibition of HIV-1 in vitro by means of
an adeno- associated virus antisense vector. Science 258:1485–1488
Chendrimada TP, Gregory RI, Kumaraswamy E et al (2005) TRBP recruits the dicer complex to
Ago2 for microRNA processing and gene silencing. Nature 436:740–744
Coburn GA, Cullen BR (2002) Potent and specific inhibition of human immunodeficiency virus
type 1 replication by RNA interference. J Virol 76:9225–9231
Culliton BJ (1990) Gene therapy begins. Science 249:1372
Das AT, Brummelkamp TR, Westerhout EM et al (2004) Human immunodeficiency virus type 1
escapes from RNA interference-mediated inhibition. J Virol 78:2601–2605
Davidson BL, Paulson HL (2004) Molecular medicine for the brain: silencing of disease genes
with RNA interference. Lancet Neurol 3:145–149
de Fougerolles AR (2008) Delivery vehicles for small interfering RNA in vivo. Hum Gene Ther
19:125–132
Denli AM, Tops BB, Plasterk RH et al (2004) Processing of primary microRNAs by the
microprocessor complex. Nature 432:231–235
Ding H, Schwarz DS, Keene A et al (2003) Selective silencing by RNAi of a dominant allele that
causes amyotrophic lateral sclerosis. Aging Cell 2:209–217
204 K.J. von Eije and B. Berkhout
Doench JG, Sharp PA (2004) Specificity of microRNA target selection in translational repression.
Genes Dev 18:504–511
Dropulic B (2001) Lentivirus in the clinic. Mol Ther 4:511–512
Edelstein ML, Abedi MR, Wixon J (2007) Gene therapy clinical trials worldwide to 2007 – an
update. J Gene Med 9:833–842
Elbashir SM, Harborth J, Lendeckel W et al (2001) Duplexes of 21-nucleotide RNAs mediate
RNA interference in cultured mammalian cells. Nature 411:494–498
Fedorov Y, Anderson EM, Birmingham A et al (2006) Off-target effects by siRNA can induce
toxic phenotype. RNA 12:1188–1196
Gimeno R, Weijer K, Voordouw A et al (2004) Monitoring the effect of gene silencing by
RNA interference in human CD34+ cells injected into newborn RAG2-/- gammac-/- mice:
functional inactivation of p53 in developing T cells. Blood 104:3886–3893
Goldstein H (2008) Summary of presentations at the NIH/NIAID new humanized rodent models
2007 workshop. AIDS Res Ther 5:3
Gregory RI, Chendrimada TP, Cooch N et al (2005) Human RISC couples microRNA biogenesis
and posttranscriptional gene silencing. Cell 123:631–640
Gregory RI, Yan KP, Amuthan G et al (2004) The Microprocessor complex mediates the genesis
of microRNAs. Nature 432:235–240
Grimm D, Kay MA (2007) Combinatorial RNAi: a winning strategy for the race against evolving
targets? Mol Ther 15:878–888
Grimm D, Streetz KL, Jopling CL et al (2006) Fatality in mice due to oversaturation of cellular
microRNA/short hairpin RNA pathways. Nature 441:537–541
Hacein-Bey-Abina S, Le Deist F, Carlier F et al (2002) Sustained correction of X-linked severe
combined immunodeficiency by ex vivo gene therapy. N Engl J Med 346:1185–1193
Hacein-Bey-Abina S, von Kalle C, Schmidt M et al (2003) LMO2-associated clonal T cell
proliferation in two patients after gene therapy for SCID-X1. Science 302:415–419
Hammond SM, Boettcher S, Caudy AA et al (2001) Argonaute2, a link between genetic and
biochemical analyses of RNAi. Science 293:1146–1150
Han J, Lee Y, Yeom KH et al (2004) The Drosha–DGCR8 complex in primary microRNA
processing. Genes Dev 18:3016–3027
Hatziioannou T, Ambrose Z, Chung NP et al (2009) A macaque model of HIV-1 infection. Proc
Natl Acad Sci USA 106:4425–4429
Ishaq M, Hu J, Wu X et al (2008) Knockdown of cellular RNA helicase DDX3 by short hairpin
RNAs suppresses HIV-1 viral replication without inducing apoptosis. Mol Biotechnol
39:231–238
Jackson AL, Bartz SR, Schelter J et al (2003) Expression profiling reveals off-target gene
regulation by RNAi. Nat Biotechnol 21:635–637
Jackson AL, Burchard J, Schelter J et al (2006) Widespread siRNA “off-target” transcript
silencing mediated by seed region sequence complementarity. RNA 12:1179–1187
Jacque JM, Triques K, Stevenson M (2002) Modulation of HIV-1 replication by RNA interference.
Nature 418:435–438
Kapadia SB, Brideau-Andersen A, Chisari FV (2003) Interference of hepatitis C virus RNA
replication by short interfering RNAs. Proc Natl Acad Sci USA 100:2014–2018
Kiriakidou M, Nelson PT, Kouranov A et al (2004) A combined computational-experimental
approach predicts human microRNA targets. Genes Dev 18:1165–1178
Kleinman ME, Yamada K, Takeda A et al (2008) Sequence- and target-independent angiogenesis
suppression by siRNA via TLR3. Nature 452:591–597
Kohn DB, Bauer G, Rice CR et al (1999) A clinical trial of retroviral-mediated transfer of a
rev-responsive element decoy gene into CD34(+) cells from the bone marrow of human
immunodeficiency virus-1-infected children. Blood 94:368–371
Konig R, Zhou Y, Elleder D et al (2008) Global analysis of host-pathogen interactions that regulate
early-stage HIV-1 replication. Cell 135:49–60
RNAi Treatment of HIV-1 Infection 205
Marshall E (1999) Gene therapy death prompts review of adenovirus vector. Science
286:2244–2245
Martinez J, Patkaniowska A, Urlaub H et al (2002) Single-stranded antisense siRNAs guide target
RNA cleavage in RNAi. Cell 110:563–574
McBride JL, Boudreau RL, Harper SQ et al (2008) Artificial miRNAs mitigate shRNA-mediated
toxicity in the brain: implications for the therapeutic development of RNAi. Proc Natl Acad Sci
USA 105:5868–5873
McCaffrey AP, Nakai H, Pandey K et al (2003) Inhibition of hepatitis B virus in mice by RNA
interference. Nat Biotechnol 21:639–644
McManus MT (2004) Small RNAs and immunity. Immunity 21:747–756
Mitsuyasu RT, Merigan TC, Carr A et al (2009) Phase 2 gene therapy trial of an anti-HIV
ribozyme in autologous CD34(+) cells. Nat Med 15:285–292
Montini E, Cesana D, Schmidt M et al (2006) Hematopoietic stem cell gene transfer in a tumor-
prone mouse model uncovers low genotoxicity of lentiviral vector integration. Nat Biotechnol
24:687–696
Muesing MA, Smith DH, Cabradilla CD et al (1985) Nucleic acid structure and expression of the
human AIDS/lymphadenopathy retrovirus. Nature 313:450–458
Naldini L, Blomer U, Gallay P et al (1996) In vivo gene delivery and stable transduction of
nondividing cells by a lentiviral vector. Science 272:263–267
Nguyen T, Menocal EM, Harborth J et al (2008) RNAi therapeutics: an update on delivery. Curr
Opin Mol Ther 10:158–167
Nishitsuji H, Kohara M, Kannagi M et al (2006) Effective suppression of human immunodefi-
ciency virus type 1 through a combination of short- or long-hairpin RNAs targeting essential
sequences for retroviral integration. J Virol 80:7658–7666
Novina CD, Murray MF, Dykxhoorn DM et al (2002) siRNA-directed inhibition of HIV-1
infection. Nat Med 8:681–686
Paddison PJ, Caudy AA, Bernstein E et al (2002) Short hairpin RNAs (shRNAs) induce sequence-
specific silencing in mammalian cells. Genes Dev 16:948–958
Park WS, Miyano-Kurosaki N, Hayafune M et al (2002) Prevention of HIV-1 infection in human
peripheral blood mononuclear cells by specific RNA interference. Nucleic Acids Res
30:4830–4835
Robbins M, Judge A, Ambegia E et al (2008) Misinterpreting the therapeutic effects of siRNA
caused by immune stimulation. Hum Gene Ther 19(10):991–999
Rossi JJ, June CH, Kohn DB (2007) Genetic therapies against HIV. Nat Biotechnol 25:1444–1454
Ruby JG, Jan CH, Bartel DP (2007) Intronic microRNA precursors that bypass Drosha processing.
Nature 448:83–86
Sabariegos R, Gimenez-Barcons M, Tapia N et al (2006) Sequence homology required by human
immunodeficiency virus type 1 to escape from short interfering RNAs. J Virol 80:571–577
Saetrom P, Heale BS, Snove O Jr et al (2007) Distance constraints between microRNA target sites
dictate efficacy and cooperativity. Nucleic Acids Res 35:2333–2342
Sano M, Li H, Nakanishi M et al (2008) Expression of long anti-HIV-1 hairpin RNAs for the
generation of multiple siRNAs: advantages and limitations. Mol Ther 16:170–177
Sarver N, Cantin EM, Chang PS et al (1990) Ribozymes as potential anti-HIV-1 therapeutic
agents. Science 247:1222–1225
Shultz LD, Ishikawa F, Greiner DL (2007) Humanized mice in translational biomedical research.
Nat Rev Immunol 7:118–130
Sledz CA, Holko M, de Veer MJ et al (2003) Activation of the interferon system by short-
interfering RNAs. Nat Cell Biol 5:834–839
Snyder LL, Esser JM, Pachuk CJ et al (2008) Vector design for liver-specific expression of
multiple interfering RNAs that target hepatitis B virus transcripts. Antiviral Res 80:36–44
Stremlau M, Owens CM, Perron MJ et al (2004) The cytoplasmic body component TRIM5alpha
restricts HIV-1 infection in old world monkeys. Nature 427:848–853
RNAi Treatment of HIV-1 Infection 207
Surabhi RM, Gaynor RB (2002) RNA interference directed against viral and cellular targets
inhibits human immunodeficiency virus type 1 replication. J Virol 76:12963–12973
Symensma TL, Giver L, Zapp M et al (1996) RNA aptamers selected to bind human immunodefi-
ciency virus type 1 Rev in vitro are Rev responsive in vivo. J Virol 70:179–187
Takeshita F, Ochiya T (2006) Therapeutic potential of RNA interference against cancer. Cancer
Sci 97:689–696
Ter Brake O, ’t Hooft K, Liu YP et al (2008) Lentiviral vector design for multiple shRNA
expression and durable HIV-1 Inhibition. Mol Ther 16:557–564
Ter Brake O, Berkhout B (2007) Lentiviral vectors that carry anti-HIV shRNAs: problems and
solutions. J Gene Med 9:743–750
Ter Brake O, Konstantinova P, Ceylan M et al (2006) Silencing of HIV-1 with RNA interference: a
multiple shRNA approach. Mol Ther 14:883–892
Ter Brake O, Legrand N, von Eije KJ et al (2009) Evaluation of safety and efficacy of RNAi
against HIV-1 in the human immune system (Rag-2(-/-)(c)(-/-)) mouse model. Gene Ther
16:148–153
Tomari Y, Matranga C, Haley B et al (2004) A protein sensor for siRNA asymmetry. Science
306:1377–1380
Tomari Y, Zamore PD (2005) Perspective: machines for RNAi. Genes Dev 19:517–529
Traggiai E, Chicha L, Mazzucchelli L et al (2004) Development of a human adaptive immune
system in cord blood cell-transplanted mice. Science 304:104–107
Unwalla HJ, Li HT, Bahner I et al (2006) Novel Pol II fusion promoter directs human immunode-
ficiency virus type 1-inducible coexpression of a short hairpin RNA and protein. J Virol
80:1863–1873
Vandekerckhove L, Christ F, Van Maele B et al (2006) Transient and stable knockdown of
the integrase cofactor LEDGF/p75 reveals its role in the replication cycle of human immuno-
deficiency virus. J Virol 80:1886–1896
Vickers TA, Lima WF, Nichols JG et al (2007) Reduced levels of Ago2 expression result in
increased siRNA competition in mammalian cells. Nucleic Acids Res 35:6598–6610
von Eije KJ, Ter Brake O, Berkhout B (2008) Human immunodeficiency virus type 1 escape is
restricted when conserved genome sequences are targeted by RNA interference. J Virol
82:2895–2903
von Eije KJ, Ter BO, Berkhout B (2009) Stringent testing identifies highly potent and escape-proof
anti-HIV short hairpin RNAs. J Gene Med 11:459–467
Watanabe S, Terashima K, Ohta S et al (2007) Hematopoietic stem cell-engrafted NOD/SCID/
IL2Rgamma null mice develop human lymphoid systems and induce long-lasting HIV-1
infection with specific humoral immune responses. Blood 109:212–218
Weinberg MS, Villeneuve LM, Ehsani A et al (2006) The antisense strand of small interfering
RNAs directs histone methylation and transcriptional gene silencing in human cells. RNA
12:256–262
Westerhout EM, Berkhout B (2007) A systematic analysis of the effect of target RNA structure on
RNA interference. Nucleic Acids Res 35:4322–4330
Westerhout EM, Ooms M, Vink M et al (2005) HIV-1 can escape from RNA interference by
evolving an alternative structure in its RNA genome. Nucleic Acids Res 33:796–804
Wiznerowicz M, Trono D (2003) Conditional suppression of cellular genes: lentivirus vector-
mediated drug-inducible RNA interference. J Virol 77:8957–8961
Yekta S, Shih IH, Bartel DP (2004) MicroRNA-directed cleavage of HOXB8 mRNA. Science
304:594–596
Yi R, Qin Y, Macara IG et al (2003) Exportin-5 mediates the nuclear export of pre-microRNAs
and short hairpin RNAs. Genes Dev 17:3011–3016
Zeng Y, Wagner EJ, Cullen BR (2002) Both natural and designed micro RNAs can inhibit the
expression of cognate mRNAs when expressed in human cells. Mol Cell 9:1327–1333
Zhang H, Kolb FA, Jaskiewicz L et al (2004) Single processing center models for human Dicer and
bacterial RNase III. Cell 118:57–68
208 K.J. von Eije and B. Berkhout
Zhang L, Kovalev GI, Su L (2007) HIV-1 infection and pathogenesis in a novel humanized mouse
model. Blood 109:2978–2981
Zhou H, Xu M, Huang Q et al (2008) Genome-scale RNAi screen for host factors required for HIV
replication. Cell Host Microbe 4:495–504
Zhou X, Symons J, Hoppes R et al (2007) Improved single-chain transactivators of the Tet-On
gene expression system. BMC Biotechnol 7:6
Application of RNA Interference to Treat
Conditions Associated with Dysregulation
of Transient Receptor Potential Vanilloid 1
Channel
Contents
1 TRPV1 Ion Channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
2 Neuropathic Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
2.1 Current Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
2.2 TRPV1 and Neuropathic Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
2.3 Diabetic Peripheral Neuropathy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
3 Drug-Induced Hearing Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
3.1 TRPV1 and Cisplatin Ototoxicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
3.2 Utility of RNAi in Treating Cisplatin Ototoxicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
4 Other Potential Uses of RNAi Targeting TRPV1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
4.1 Inflammation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
4.2 Arthritis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
4.3 Cystitis and Bladder Hyperactivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
4.4 Cancer Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
4.5 Obesity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 209
RNA Technologies, DOI 10.1007/978-3-642-12168-5_9,
# Springer-Verlag Berlin Heidelberg 2010
210 V. Ramkumar et al.
channel is expressed primarily in small diameter neurons (Ad and C fibers) within
sensory ganglia comprising the pain pathway but expression is observed in larger
diameter neurons under conditions of inflammation. TRPV1 is a nonselective cation
channel, which responds to heat (activation threshold 43 C) and transmits pain
sensations in response to noxious heat. It is also expressed in nonneuronal tissue such
as keratinocytes, bladder uroepithelium, knee joints, the gastrointestinal tract and the
cochlea, suggesting additional physiological roles of channel activation. Several
disease states associated with dysregulation of TRPV1 include visceral and peripheral
inflammatory pain observed in irritable bowel disease, bone cancers, arthritis and
bladder inflammation. Other conditions include diabetic peripheral neuropathy, obe-
sity, and drug-induced hearing loss. Animal studies have shown beneficial effects of
targeting TRPV1, using both agonist and antagonist drugs, in the treatment of some of
these conditions. However, the propensity of TRPV1 antagonists to produce hyper-
thermia could limit their future application. Such problems might be resolved by the
focal administration of short interfering (si) RNA to the affected organ in order to
reduce TRPV1 expression. While demonstration of TRPV1 knockdown by RNA
interference (RNAi) is widely used in vitro, results obtained from a limited number
of in vivo studies suggest that RNAi could be applied to the treatment of diseases
associated with TRPV1 hyperactivity. To date, RNAi has proven beneficial in reduc-
ing inflammatory pain and in treating hearing loss associated with cisplatin chemo-
therapy. This review will focus on the current progress of RNAi in the treatment of
diseases associated with TRPV1 dysfunction, discuss potential future applications of
this technology, and highlight factors that could affect its use clinically.
Abbreviations
Transient receptor potential (TRP) channels are sensors of temperature and pressure
stimuli and are therefore the frontline indicators of excessive stimulation from these
modalities. These receptors are localized primarily to peripheral sensory and spinal
cord neurons where they can participate fully in pain perception. A subset of these
channels detect a wide range of temperatures, spanning the spectrum from hot to
cold, with each channel subtype exhibiting a discrete temperature range for activa-
tion. TRPV1 (also known as VR1), which is activated by capsaicin, the active
ingredient of “hot” chili pepper (Szallasi and Blumberg 1999) shows activation
threshold of 43 C. TRPV2 (also known as VRL-1) responds to extremely hot
temperatures, and TRPV3 and TRPV4 (also known as VROAC or OTRPC4) are
activated by warm temperatures (Tominaga and Caterina 2004). Two additional
members of this group, which are activated by cold stimuli, include TRPV8 (also
known as CMR1) and TRPA1 (also known as ANKTM1) (Patapoutian et al. 2003;
Jordt et al. 2004).
In addition to its function as a thermal sensor, TRPV1 is activated by vanilloids
(capsaicin, resiniferatoxin) and nonvanilloids (protons, anandamide) and serves as a
primary mediator of thermal and inflammatory pain. This receptor is highly
expressed on sensory afferents (C and Ad fibers), which terminate in laminae
I and II of the spinal cord. These fibers might also serve an efferent role since
they release substance P and calcitonin gene-related peptide (CGRP) upon activa-
tion. TRPV1 is also highly expressed in the nucleus of the solitary tract, an area that
receives vagal projections from the nodose ganglion. Other central locations of
TRPV1 include the hypothalamus (Hori et al. 1988), substantia nigra (Marinelli
et al. 2003), locus coeruleus (Marinelli et al. 2002), and hippocampus (Al-Hayani
and Davies 2002), where its functions are still undefined. Nonneuronal localization
of TRPV1 include epithelial cells of the skin and bladder (Birder et al. 2001),
kidney, GI mucosa, and the inner ear (Mukherjea et al. 2008).
Studies using TRPV1-deficient mice have been critical to demonstrating the
involvement of this receptor in thermal and inflammatory pain. This mouse model
exhibited diminished responses to noxious stimuli, such as capsaicin, proton, or
heat, in addition to decreased thermal hyperalgesia induced by inflammation
(Caterina et al. 2000; Davis et al. 2000). Surprisingly, these mice also showed
deficits in neuropathic pain, mechanical allodynia, and mechanical hyperalgesia,
suggesting a role of TRPV1 in integrating multiple pain stimuli.
Treatment of conditions associated with increased expression or dysregulation
of TRPV1 have focused primarily on the use of agonist and antagonist drugs. The
usefulness of agonists derives from the observation that they desensitize TRPV1 or
destroy TRPV1 containing nerve fibers and thereby provide different degrees of
pain relief. Antagonists have shown efficacy in the treatment of inflammation,
arthritis, and cancer pain in animal models. The ability of most TRPV1 antagonists
to produce hyperthermia in animal models (Gavva et al. 2007) and humans (Gavva
et al. 2008) has led to concerns as to the future clinical utility of these compounds.
212 V. Ramkumar et al.
However, since the hyperthermia is transient and disappears with repeated dosing,
these agents might still play a significant role in pain management. The utility of
RNAi technology for effecting TRPV1 knockdown has been shown in a limited
number of disease models in animals. These include animal models of neuropathic
pain and drug-induced hearing loss. Information detailing the involvement of
TRPV1 in these and other conditions and the potential utility of RNAi as a
treatment option are provided below.
2 Neuropathic Pain
TRPV1 in this chronic pain model. In a rat model of neuropathic pain, capsazepine
blocked A-fibers evoked responses in the dorsal horn, and selective TRPV1 antago-
nists attenuated mechanical allodynia and hyperalgesia (Christoph et al. 2006;
Honore et al. 2005; Kanai et al. 2005; Pomonis et al. 2003). These studies implicate
TRPV1 in mediating mechanical allodynia, at least in these in vivo models. In a
recent study, Christoph et al. (2006) demonstrated the utility of intrathecal adminis-
tration of siRNA against TRPV1. These investigators showed that knockdown of
TRPV1 by RNAi reduced cold allodynia of mononeuropathic rats and spontaneous
visceral pain. The effectiveness of RNAi was comparable to the TRPV1 antagonist,
N-(4-tertiarybutylphenyl)-4-(3-chloropyridin-2-yl)tetrahydropyrazine-1(2H)-carbox
amide (BCTC), and persisted over a period of 5 days. In a more recent study, Christoph
et al. (2008) showed that a transgenic mouse expressing short hairpin RNA (shRNA)
for TRPV1 exhibited decreased response to capsaicin-induced hypothermia,
decreased heat sensitivity on hot plates, and decreased mechanical allodynia. These
data will add further support to the contention that TRPV1 is a mediator of neuropathic
pain and are in stark contrast to data showing development of mechanical allodynia
and hypersensitivity to spinal nerve injury in TRPV1 knockout mice (Caterina et al.
2000). While the reason for this difference is unclear, it might reflect differences in the
animal models used and/or differences in compensatory measures produced by
knockout versus knockdown of TRPV1 by RNAi. One unexpected finding between
these two models is an increase in TRPV3 expression in dorsal root ganglions (DRGs)
of transgenic mice expressing TRPV1 shRNA but a decrease in TRPV3 expression in
TRPV1 knockout mice (Christoph et al. 2008). Overall, these data would support the
utility of TRPV1 RNAi in the treatment of neuropathic pain but might be limited by
the inability to effectively target the “drug” to the affected area.
Platinum containing drugs have been successfully used in the treatment of various
solid tumors of the head and neck. One such drug, cisplatin, is an important
component of chemotherapeutic regimen for treating solid tumors. This drug
produces ototoxicity, in part, through the generation of reactive oxygen species
(ROS) (Rybak and Ramkumar 2007). One target of ROS include the organ of Corti
(Rybak 1999), where it destroys outer hair cells (Kopke et al. 1997). Current
treatment strategy involves the use of antioxidant (Rybak 1999). However, con-
comitant antioxidant use could interfere with the anticancer efficacy of cisplatin and
limit its usefulness in chemotherapy. As such, other treatment targets have been
sought after. One such target that we have identified is TRPV1, expressed in the
organ of Corti and spiral ganglion cells. In a recent study (Mukherjea et al. 2008),
we showed that TRPV1 is a target of ROS generated by cisplatin. ROS promote
activation and induction of TRPV1 and the NOX3 isoform of NADPH oxidase
(a major source of ROS generation in the cochlea) in the rat organ of Corti and
spiral ganglion cells. Generation of ROS via NOX3 was shown to be crucial to the
activation and induction of TRPV1.
Application of RNA Interference to Treat Conditions Associated with Dysregulation 215
a b
Untreated Scrambled siRNA siTRPV1
HOOK
100
90
80 Scrambled siRNA
70 siTRPV1
BASE
60
50
40
30 *
20 *
MIDDLE
10
TURN
0 *
Tu ddle
se
ok
Ho
Ba
rn
i
M
Fig. 1 TRPV1 siRNA protects against cisplatin-induced outer hair cell damage. (a), Rats were
pretreated with either a scrambled siRNA sequence or with siRNA against TRPV1 by round
window application for 48 h. This was followed by cisplatin administration (13 mg/kg, i.p.), the
cochleae was collected 72 h later and were processed from SEM. TRPV1 siRNA protects against
cisplatin-mediated damage and loss of cochlear outer hair cells. SEM of samples collected 72 h
following cisplatin administration indicate damage to or loss of outer hair cells in cochleae
pretreated with scrambled siRNA sequence, with greatest effects obtained in the hook, followed
by the base and the middle turn (see arrows). Cochleae obtained from rats pretreated with siRNA
against TRPV1 showed statistically significant reductions in hair cell loss in all three regions
examined. (b) The percentage of outer hair cell damage in (a) is presented in graphical format.
Asterisk (*) indicate statistically significant reductions in the cochleae treated with TRPV1 siRNA
versus those treated with a scrambled siRNA (n ¼ 5; p < 0.05). (Reprinted with the permission
from Society of Neuroscience)
216 V. Ramkumar et al.
a b
expression
150
β-Actin 100
Cisplain – + – + 50
TRPV1 siRNA – – + + *
Scrambled + + – – 0
l
siRNA
ntro P V1 la
tin
Co TR isp
si C
c
50
siTRPV1 (0.9μg /3μl)
45
Scrambled siRNA
Threshold change (dB)
40
35
30
25
20
15
* *
10
5
0
8k 16k 32k clix
Fig. 2 siRNAs against TRPV1 reduced cisplatin-induced ototoxicity in rats. (a) siRNA against
TRPV1 suppressed the basal and cisplatin-stimulated TRPV1 protein levels in the cochlea
assessed 24 h following cisplatin administration. This is a representative of three independent
experiments showing similar responses. (b) Cochlear mRNA obtained from rats administered
TRPV1 siRNA and assessed 48 h later indicated a 85% reduction in basal TRPV1 mRNA.
Asterisk (*) indicate statistically significant reductions in the cochleae treated with TRPV1 siRNA
versus those treated with a scrambled siRNA (n ¼ 5; p < 0.05). (c) Pretreatment ABRs were
determined in the rats, which were then administered either a scrambled siRNA sequence in one
ear or 0.9 mg siRNA against TRPV1 by round window application in the other ear. Cisplatin
(13 mg/kg i.p) was administered 48 h later and posttreatment ABRs were determined after an
additional 72 h period. Cisplatin produced a shift in ABR thresholds (determined by comparing
pre- and posttreatment ABR thresholds) of 25–40 dB over an 8–32 kHz frequency range in the
ears of animals pretreated with the scrambled siRNA. However, statistically significant reductions
in ABR thresholds (p < 0.05; n ¼ 5) were obtained at 8 and 16 kHz frequency range and a trend
for protection observed at the 32 kHz range and for clicks (clix) in rats pretreated with siRNA
against TRPV1. (Reprinted with the permission from Society of Neuroscience)
were measured in rats pretreated with scrambled or TRPV1 siRNA and treated
with cisplatin 48 h later. ABRs performed 72 h following administration of cisplatin
(in rats pretreated with scrambled siRNA) showed increased ABR thresholds
by 38 6, 30 9, 24 4 and 30 7 dB at testing frequencies of 8, 16,
32 kHz or clicks, respectively. However, rats pretreated with TRPV1 siRNA
Application of RNA Interference to Treat Conditions Associated with Dysregulation 217
4.1 Inflammation
Inflammation is initiated by tissue injury and/or infection that sensitizes the tissue to
stimuli, which normally do not produce pain (allodynia) or increase pain to normally
noxious stimuli (hyperalgesia). This hypersensitivity is believed to result from
increased release of inflammatory mediators, such as prostaglandins, adenosine,
serotonin, bradykinin, and ATP, which act on Gq coupled receptors to sensitize
TRPV1 to endogenous or exogenous activators. These mediators also lower the
temperature threshold for activation of TRPV1 to below normal body temperature,
leading to chronic pain at the site of inflammation. A role of TRPV1 in mediating this
thermal hyperalgesia is provided by the observation that TRPV1 knockout mice
failed to show thermal hypersensitivity from inflammation induced by carrageenan
(Davis et al. 2000) or complete Freund’s adjuvant (Caterina et al. 2000).
One form of inflammation, termed neurogenic inflammation, is produced by
TRPV1-mediated release of mediators such as CGRP and substance P from nerve
terminals. The close proximity of these terminals to mast cells can trigger mast cell
degranulation by the mediators, leading to the release of additional inflammatory
mediators such as histamine, proteoglycans, serotonin, interleukin (IL), and tumor
necrosis-a (TNF-a). Mast cell mediators can positively regulate sensory neuron to
release CGRP and substance P and thereby serve as a positive feedback loop to
enhance neurogenic inflammation (Bı́ró et al. 1998). TRPV1 activation on mast
cells promotes release of inflammatory cytokines, such as IL-4 (Bı́ró et al. 1998).
As such, activation of TRPV1 can contribute indirectly (via stimulation of nerve
terminal) or directly (via mast cells) to neurogenic inflammation. TRPV1 also
appears to mediate a portion of the inflammatory response observed in caerulein-
induced acute pancreatitis. Activation of sensory fibers innervating the pancreas
leads to increased release of inflammatory neuropeptides (such as substance P),
which promote edema and the inflammatory response via the neurokinin 1 (NK1)
receptor. Blockade of NK1 receptor or TRPV1 reduces experimental pancreatitis
(Liddle 2007). These findings would support the clinical utility of TRPV1 blockade
or knockdown by RNAi in suppressing pancreatitis.
4.2 Arthritis
One aspect of inflammatory pain where TRPV1 plays a role is arthritis. TRPV1 has
been studied for its potential role in arthritic conditions such as osteoarthritis and
rheumatoid arthritis. TRPV1 levels were higher in the iodoacetate osteoarthritis
model (Fernihough et al. 2005) and in the complete Freund’s adjuvant-model of
inflammation (Amaya et al. 2003; Carlton and Coggeshall 2001), suggesting that it
mediates inflammation or is induced by inflammation. In a model of knee joint
inflammation, TRPV1 knockout mice showed reduced knee swelling in comparison
to the wild-type TRPV1 mice (Keeble et al. 2005), implicating TRPV1 in the
Application of RNA Interference to Treat Conditions Associated with Dysregulation 219
inflammatory process. These animals show similar levels of TNF-a, but TNF-a
induced a greater degree of knee swelling in the wild-type mice as compared to the
TRPV1-knockout mice (Keeble et al. 2005). These findings suggest an indirect
rather than a direct role of TRPV1 in arthritis. In a separate study by Szabó et al.
(2005), TRPV1 activation was shown to mediate adjuvant-induced chronic arthritis
in mice. The proinflammatory effect of adjuvant was further enhanced by bradyki-
nin and lipoxygenase products (Szabó et al. 2005), which are able to sensitize
TRPV1 through protein kinase C activation (Premkumar and Ahern 2000). Activa-
tion of TRPV1 under these conditions promote the release of substance P and
CGRP from peripheral nerve terminals (Szolcsanyi 1996), which mediate local
arteriolar vasodilation and increased plasma extravasation and inflammation in the
synovium (Lam and Ferrell 1991). As the inflammation progresses, these neuropep-
tides also increase the proliferation of synoviocytes (Lambert et al. 1998) and
increase secretion of inflammatory cytokines and hyperplasia of synovial tissues.
TRPV1 was also detected in the human synoviocytes, where it increased intracel-
lular calcium release (Kochukov et al. 2006). The increased intracellular calcium
release promoted ROS generation and death of synovial fibroblasts (Hu et al. 2008).
Recent studies have shown that the expression of TRPV1 is higher in the synovial
fibroblasts from patients with osteoarthritis and rheumatoid arthritis (Engler et al.
2007), implicating these receptors in the pathophysiology of osteoarthritis and
rheumatoid arthritis. In addition, studies have implicated TRPV1 not only in the
propagation of the inflammation but also in the generation of the pain associated
with the disease process. Hence, targeting the local TRPV1 receptors for inhibition
or knockdown by RNAi strategies could provide relief from arthritic pain and
inflammation. Direct injections of TRPV1 siRNA into the joint cavity might enable
localized knockdown of TRPV1 without significant systemic side effects. In this
respect, capsaicin creams (which are expected to destroy sensory nerve terminals in
the area) have been used topically on joints to relieve joint swelling and pain of
arthritis (Hautkappe et al. 1998).
In contrast to a purported role of TRPV1 in mediating the pain and inflammation
of arthritis, antiinflammatory role of capsaicin has also been demonstrated in the
treatment of arthritis (Brand et al. 1990; Jarreau et al. 1994; Joe and Lokesh 1994),
mediated presumably via inhibition of nuclear factor (NF)-kB and activator protein-1
(AP-1) (Singh et al. 1996, Surh et al. 2000). However, this response is observed at
higher doses of capsaicin and appears to be independent of TRPV1 activation.
Cancer pain can significantly reduce the quality of life in patients who might also
have reduced life expectancy. This type of pain has historically been poorly
treated by analgesics. The pain likely results from increased pressure of the
tumor on nerves in the vicinity, increased inflammation and nerve damage or
altered expression of receptors involved in pain transmission. For example, the
levels of TRPV1 in DRGs were increased in an animal model of bone cancer pain
(Shinoda et al. 2008) and in squamous cell carcinoma of the human tongue
Application of RNA Interference to Treat Conditions Associated with Dysregulation 221
(Marincsák et al. 2009). In an effort to better treat cancer pain, the World Health
Organization has developed a stepwise ladder, which includes a stepwise use of
nonopioid drugs followed by opioid drugs (Zech et al. 1995). Effective control of
pain was reported by 70–90% of patients on this regimen. However, this treatment
protocol does not produce adequate pain control in 10–30% of individuals,
especially those suffering from bone cancer pain. While morphine is generally
used for these patients, higher doses are generally prescribed and are associated
with significant side effects such as sedation, constipation, and drug abuse liabil-
ity. Several reports have indicated that TRPV1 antagonists could be useful as
adjuncts to morphine in the treatment of animal models of bone cancer pain
(Shinoda et al. 2008; Niiyama et al. 2009). Administration of the antagonist
N-(3-methoxyphenyl)-4-chlorocinnamide (SB366791) with morphine decreased
pain induced by pressure on the affected limb or by movement (Niiyama et al.
2009). Interestingly, no significant pain relief was observed in these animals when
the drugs were administered individually, reinforcing the need for combined drug
administration. These findings, if confirmed, would support the clinical utility of
targeting TRPV1 in the treatment of cancer pain. Such treatment could include
TRPV1 antagonists or localized delivery of siRNA directed against TRPV1.
4.5 Obesity
fat loss (Snitker et al. 2009). Overall, these studies support a role of TRPV1
activation in controlling obesity and would obviate the need to the use of TRPV1
antagonist or knockdown by RNAi to treat this condition.
In contrast to the aforementioned study by Zhang et al. (2007), it was reported
that TRPV1 knockout mice on a low fat diet (4.5%) gained similar amount of
weight as their wild-type counterparts (Motter and Ahern 2008). However, on a
high fat diet (11% fat), the wild type mice showed increased weight gain after
17 weeks, compared to the TRPV1 knockout mice. When these mice were tested
over a 44 week period, wild type mice showed a significant increase in the mean
body mass compared to the knockouts. Thus, a deficiency in TRPV1 rendered these
mice more resistant to diet-induced weight gain. It was also reported that TRPV1-
null mice were able to withstand 1 h of cold exposure (0–2 C) without significant
change in their core body temperature, suggesting a greater capacity for thermo-
genesis. The discrepancy in results between Zhang et al. (2007) and Motter and
Ahern (2008) may be due to the difference in the amount of fat in the diets as well as
time period for which the animals were observed. However, these findings shed
some doubt as to whether activation of TRPV1 is indeed important for controlling
obesity and would suggest that knockdown of TRPV1 by RNAi and capsaicin in a
high fat diet could control obesity.
5 Conclusions
Over the past few years, TRPV1 has expanded its reach into a number of physio-
logical and pathological processes beyond its initially characterized role as a
mediator of thermal and inflammatory pain and a target of vanilloids. This receptor
is implicated in the mediation of pain in various disease processes including
neuropathic pain, arthritis, cystitis, and diabetic neuropathy and is implicated in
conditions such as hearing loss and possibly obesity. As such, there is increasing
interest at research institutions and pharmaceutical companies to produce drugs that
would eliminate or block TRPV1 activation. Since the clinical utility of TRPV1
antagonist has been shadowed by their induction of hyperthermia, it is expected that
RNAi would contribute to the overall available treatment options for these and
other conditions associated with TRPV1 dysregulation.
Conditions such as ototoxicity, arthritis, and cystitis, which would be alleviated
by localized administration of drugs could be specifically targeted for TRPV1
knockdown by RNAi. This technology is expected to provide target selectivity
and longer duration of action than existing therapeutics (especially if DNA-based
RNAi drugs and viral vectors are utilized) and provide limited side effects espe-
cially if delivered locally. As such, RNAi would be ideal for the long-term
management of chronic pain conditions. Unlike TRPV1 agonists, which could
induce pain relief through destroying TRPV1 containing nerve terminals, RNAi
is expected to selectivly target TRPV1 for knockdown without affecting sensory
inputs by different receptors present on those terminals. One concern is that the
Application of RNA Interference to Treat Conditions Associated with Dysregulation 223
Acknowledgments This work was supported by a grants from the National Institute of Health
(DC02396) to LPR and (CA135494) to VR, a grant from the National Organization for Hearing
Research and SIU School of Medicine Excellence in Academic Medicine Award to VR.
References
Ahern GP (2003) Activation of TRPV1 by the satiety factor oleoylethanolamide. J Biol Chem
278:30429–30434
Al-Hayani A, Davies SN (2002) Effect of cannabinoids on synaptic transmission in the rat
hippocampal slice is temperature-dependent. Eur J Pharmacol 442:47–54
Amaya F, Oh-hashi K, Naruse Y et al (2003) Local inflammation increases vanilloid receptor 1
expression within distinct subgroups of DRG neurons. Brain Res 963:190–196
Apostolidis A, Brady CM, Yiangou Y et al (2005) Capsaicin receptor TRPV1 in urothelium of
neurogenic human bladders and effect of intravesical resiniferatoxin. Urology 65:400–405
Birder LA, Kanai AJ, de Groat WC et al (2001) Vanilloid receptor expression suggests a sensory
role for urinary bladder epithelial cells. Proc Natl Acad Sci 98:13396–13401
Bı́ró T, Maurer M, Modarres S et al (1998) Characterization of functional vanilloid receptors
expressed by mast cells. Blood 9:1332–1340
Bölcskei K, Helyes Z, Szabó A et al (2005) Investigation of the role of TRPV1 receptors in acute
and chronic nociceptive processes using gene-deficient mice. Pain 117:368–376
Brady CM, Apostolidis AN, Harper M et al (2004) Parallel changes in bladder suburothelial
vanilloid receptor TRPV1 and pan-neuronal marker PGP9.5 immunoreactivity in patients with
neurogenic detrusor overactivity after intravesical resiniferatoxin treatment. BJU Int
93:770–776
Brand LM, Skare KL, Loomans MF et al (1990) Anti-inflammatory pharmacology and mechanism
of the orally active capsaicin analogs, NE-19550 and NE-28345. Agents Actions 31:329–340
Campbell KC, Meech RP, Klemens JJ et al (2007) Prevention of noise- and drug-induced hearing
loss with D-methionine. Hear Res 226:92–103
Carlton SM, Coggeshall RE (2001) Peripheral capsaicin receptors increase in the inflamed rat
hindpaw: a possible mechanism for peripheral sensitization. Neurosci Lett 310:53–56
Caterina MJ, Leffler A, Malmberg AB et al (2000) Impaired nociception and pain sensation in
mice lacking the capsaicin receptor. Science 288:306–313
Chancellor MB, de Groat WC (1999) Intravesical capsaicin and resiniferatoxin therapy: spicing up
the ways to treat the overactive bladder. J Urol 162:3–11
Charrua A, Cruz CD, Narayanan S et al (2009) GRC-6211, a new oral specific TRPV1 antagonist,
decreases bladder overactivity and noxious bladder input in cystitis animal models. J Urol
181:379–386
Christoph T, Gr€unweller A, Mika J et al (2006) Silencing of vanilloid receptor TRPV1 by RNAi
reduces neuropathic and visceral pain in vivo. Biochem Biophys Res Commun 350:238–243
Christoph T, Bahrenberg G, De Vry J et al (2008) Investigation of TRPV1 loss-of-function
phenotypes in transgenic shRNA expressing and knockout mice. Mol Cell Neurosci
37:579–589
224 V. Ramkumar et al.
Cruz CD, Charrua A, Vieira E et al (2008) Intrathecal delivery of resiniferatoxin (RTX) reduces
detrusor overactivity and spinal expression of TRPV1 in spinal cord injured animals. Exp
Neurol 214:301–308
Davis JB, Gray J, Gunthorpe MJ et al (2000) Vanilloid receptor-1 is essential for inflammatory
thermal hyperalgesia. Nature 405:183–187
Dinis P, Charrua A, Avelino A et al (2004a) Intravesical resiniferatoxin decreases spinal c-foc
expression and increases bladder volume to reflex micturition in rats with chronic inflamed
urinary bladder. BJU Int 94:153–157
Dinis P, Charrua A, Avelino A et al (2004b) Anandamide-evoked activation of vanilloid receptor 1
contributes to the development of bladder hyperreflexia and nociceptive transmission to spinal
dorsal horn neurons in cystitis. J Neurosci 24:11253–11263
Engler A, Aeschlimann A, Simmen BR et al (2007) Expression of transient receptor potential
vanilloid 1 (TRPV1) in synovial fibroblasts from patients with osteoarthritis and rheumatoid
arthritis. Biochem Biophys Res Commun 359:884–888
Facer P, Casula MA, Smith GD et al (2007) Differential expression of the capsaicin receptor
TRPV1 and related novel receptors TRPV3, TRPV4 and TRPM8 in normal human tissues and
changes in traumatic and diabetic neuropathy. BMC Neurol 7:11
Fernihough J, Gentry C, Bevan S et al (2005) Regulation of calcitonin gene-related peptide and
TRPV1 in a rat model of osteoarthritis. Neurosci Lett 388:75–80
Gavva NR, Bannon AW, Hovland DN Jr et al (2007) Repeated administration of vanilloid receptor
TRPV1 antagonists attenuates hyperthermia elicited by TRPV1 blockade. J Pharmacol Exp
Ther 323:128–137
Gavva NR, Treanor JJ, Garami A et al (2008) Pharmacological blockade of the vanilloid receptor
TRPV1 elicits marked hyperthermia in humans. Pain 136:202–210
Hautkappe M, Roizen MF, Toledano A et al (1998) Review of the effectiveness of capsaicin for
painful cutaneous disorders and neural dysfunction. Clin J Pain 14:97–106
Hong S, Wiley JW (2005) Early painful diabetic neuropathy is associated with differential changes
in the expression and function of vanilloid receptor 1. J Biol Chem 280:618–627
Honore P, Wismer CT, Mikusa J et al (2005) A-425619 [1-isoquinolin-5-yl-3-(4-trifluoromethyl-
benzyl)-urea], a novel transient receptor potential type V1 receptor antagonist, relieves patho-
physiological pain associated with inflammation and tissue injury in rats. J Pharmacol Exp
Ther 314:410–421
Hori T, Shibata M, Kiyohara T et al (1988) Responses of anterior hypothalamic-preoptic thermo-
sensitive neurons to locally applied capsaicin. Neuropharmacology 27:135–142
Hu F, Sun WW, Zhao XT et al (2008) TRPV1 mediates cell death in rat synovial fibroblasts
through calcium entry-dependent ROS production and mitochondrial depolarization. Biochem
Biophys Res Commun 369:989–993
Hunt SP, Pini A, Evan G (1987) Induction of c-fos like protein in spinal cord neurons following
sensory stimulation. Nature 328:632–634
Ishizuka O, Igawa Y, Mattiasson A et al (1994) Capsaicin-induced bladder hyperactivity in normal
conscious rats. J Urol 152:525–530
Jarreau PH, D’Ortho MP, Boyer V et al (1994) Effects of capsaicin on the airway responses to
inhaled endotoxin in the guinea pig. Am J Crit Care Med 149:128–133
Joe B, Lokesh BR (1994) Role of capsaicin, curcumin and dietary n-3 fatty acids in lowering the
generation of reactive oxygen species in rat peritoneal macrophages. Biochem Biophys Acta
1224:255–263
Jordt SE, Bautista DM, Chuang HH et al (2004) Mustard oils and cannabinoids excite sensory
nerve fibres through the TRP channel ANKTM1. Nature 427:260–265
Kanai Y, Nakazato E, Fujiuchi A (2005) Involvement of an increased spinal TRPV1 sensitization
through its up-regulation in mechanical allodynia of CCI rats. Neuropharmacology
49:977–984
Keeble J, Russell F, Curtis B et al (2005) Involvement of transient receptor potential vanilloid 1
in the vascular and hyperalgesic components of joint inflammation. Arthritis Rheum
52:3248–3256
Application of RNA Interference to Treat Conditions Associated with Dysregulation 225
Kochukov MY, McNearney TA, Fu Y et al (2006) Thermosensitive TRP ion channels mediate
cytosolic calcium response in human synoviocytes. Am J Physiol Cell Physiol 291:C424–C432
Kopke RD, Liu W, Gabaizadeh R et al (1997) Use of organotypic cultures of Corti’s organ to study
the protective effects of antioxidant molecules on cisplatin-induced damage of auditory hair
cells. Am J Otol 18:559–571
Lam FY, Ferrell WR (1991) Specific neurokinin receptors mediate plasma extravasation in the rat
knee joint. Br J Pharmacol 103:1263–1267
Lambert N, Lescoulie PL, Yassine-Diab B et al (1998) Substance P enhances cytokine-induced
vascular cell adhesion molecule-1 (VCAM-1) expression on cultured rheumatoid fibroblast-
like synoviocytes. Clin Exp Immunol 113:269–275
Lecci A, Giuliani S, Santicioli P et al (1994) Involvement of spinal tachykinin NK1 and NK2
receptors in detrusor hyperreflexia during chemical cystitis in anaesthetized rats. Eur J Pharmacol
259:129–135
Liddle RA (2007) The role of transient receptor potential vanilloid 1 (TRPV1) channels in
pancreatitis. Biochim Biophys Acta 1772:869–878
MacLean DB (1985) Abrogation of peripheral cholecystokinin-satiety in the capsaicin treated rat.
Regul Pept 11:321–333
Marincsák R, Tóth BI, Czifra G et al (2009) Increased expression of TRPV1 in squamous cell
carcinoma of the human tongue. Oral Dis 15:328–335
Marinelli S, Vaughan CW, Christie MJ et al (2002) Capsaicin activation of glutamatergic synaptic
transmission in the rat locus coeruleus in vitro. J Physiol 543:531–540
Marinelli S, Di Marzo V, Berretta N et al (2003) Presynaptic facilitation of glutamatergic synapses
to dopaminergic neurons of the rat substantia nigra by endogenous stimulation of vanilloid
receptors. J Neurosci 23:3136–3144
Motter AL, Ahern GP (2008) TRPV1-null mice are protected from diet-induced obesity. FEBS
Lett 582:2257–2262
Mukherjea D, Jajoo S, Whitworth C et al (2008) Short interfering RNA against transient receptor
potential vanilloid 1 attenuates cisplatin-induced hearing loss in the rat. J Neurosci
28:13056–13065
Mukherjea D, Jajoo S, Kaur T et al (2010) Transtympanic administration of short interfering (si)
RNA for NOX3 isoform of NADPH oxidase protects against cisplatin-induced hearing loss in
the rat. Antiox. Redox Signal. [Epub June 2010]
Niiyama Y, Kawamata T, Yamamoto J et al (2009) SB366791, a TRPV1 antagonist, potentiates
analgesic effects of systemic morphine in a murine model of bone cancer pain. Br J Anaesth
102:251–258
O’Connor AB, Dworkin RH (2009) Treatment of neuropathic pain: an overview of recent guide-
lines. Am J Med 122:S22–32
Pabbidi RM, Yu SQ, Peng S et al (2008) Influence of TRPV1 on diabetes-induced alterations in
thermal pain sensitivity. Mol Pain 14:9
Patapoutian A, Peier AM, Story GM et al (2003) ThermoTRP channels and beyond: mechanisms
of temperature sensation. Nat Rev Neurosci 4:529–539
Pomonis JD, Harrison JE, Mark L (2003) N-(4-Tertiarybutylphenyl)-4-(3-cholorphyridin-2-yl)
tetrahydropyrazine -1(2H)-carbox-amide (BCTC), a novel, orally effective vanilloid receptor 1
antagonist with analgesic properties: II. in vivo characterization in rat models of inflammatory
and neuropathic pain. J Pharmacol Exp Ther 306:387–393
Premkumar LS, Ahern GP (2000) Induction of vanilloid receptor channel activity by protein
kinase C. Nature 408:985–990
Ramkumar V, Whitworth CA, Pingle SC et al (2004) Noise induces A1 adenosine receptor
expression in the chinchilla cochlea. Hear Res 188:47–56
Ritter RC, Ladenheim EE (1985) Capsaicin pretreatment attenuates suppression of food intake by
cholecystokinin. Am J Physiol 248:R501–504
Rybak LP (1999) Ototoxicity: bioprotective mechanisms. Curr Opin Otolaryngol Head Neck Surg
11:328–333
226 V. Ramkumar et al.
Contents
1 Melanin: A Ubiquitous Pigment with a Multitude of Functions . . . . . . . . . . . . . . . . . . . . . . . . . . 228
2 Melanogenesis: Insights Uncovered by the Detailed Analysis of Genetic Disorders of
Pigmentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
3 An RNAi-Based Functional Genomics Approach to Identify
Novel Regulators of Melanogenesis in Human Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
4 Integration of Multiple Systems-Level Approaches to Uncover
Additional Regulators of Melanogenesis in Our Functional Genomics Dataset . . . . . . . . . . 237
5 Identification of Novel Pathways that Regulate the Transcription
of Melanogenic Enzymes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
6 Identification of Novel Pathways that Regulate Melanosome Biogenesis . . . . . . . . . . . . . . . . 244
7 Identification of Regulators of Human Pigment Variation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
8 Identification of Pharmacologic Agents that Impact
Melanin Accumulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
9 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
Abstract Melanin, the primary chromophore in human skin, protects the skin and
eyes from the harmful effects of UV irradiation, protects neural cells from toxic
insults, and is required for sound conduction in the inner ear. Aberrant regulation of
melanogenesis underlies skin disorders (melasma and vitiligo), neurologic disor-
ders (Parkinson’s disease), auditory disorders (Waardenburg’s syndrome), and
ophthalmologic disorders (age-related macular degeneration). Extensive studies
have identified over 150 genes that regulate the production of melanin pigment in
human cells. Despite this extensive investigation, the phenotypic variation in
human skin color cannot simply be explained by the aberrant regulation of these
150 genes. We recently utilized RNAi-based functional genomics to unravel the
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 227
RNA Technologies, DOI 10.1007/978-3-642-12168-5_10,
# Springer-Verlag Berlin Heidelberg 2010
228 H. Ho et al.
Melanin plays a critical role in protecting the skin, eyes, and brain from toxic
insults. Melanin is produced from tyrosine via a highly regulated enzymatic path-
way (Slominski et al. 2004). It absorbs light over a broad spectrum, encompassing
both the UV and visible spectrum (Stein 1955). This property of melanin serves to
protect the skin from the harmful effects of UV irradiation (Slominski et al. 2004).
There is an inverse correlation between epidermal melanin content and skin cancer
incidence (Kadekaro et al. 2003), demonstrating that melanin is the key pigment
that protects the epidermis from UV damage. In addition to its protective effects
against UV irradiation, melanin, particularly neuromelanin, is a chelator of heavy
metals (Double 2006). The protective role of melanin in the brain has been linked
with its ability to block the toxic effects of environmental metals (Zecca et al.
2008). Loss of striatal melanin in the inner ear correlates with age-associated
hearing loss, suggesting that melanin may have protective effects in this organ
Harnessing RNAi-Based Functional Genomics to Unravel the Molecular Complexity 229
(Ohlemiller et al. 2009). The protective effect of melanin in the ear is thought to be
secondary to its ability to bind both trace metals and ototoxic drugs (Breathnach
1988). Melanin also protects against high-intensity noise damage by an unknown
mechanism (Breathnach 1988). In the eye, melanin protects choroidal blood vessels
from the toxic effects of UV radiation (Ohlemiller et al. 2009). The absorptive
properties of retinal pigment epithelium melanin prevents reflection and protects
the photoreceptors from harmful effects of UV (Breathnach 1988). The critical
protective role of melanin in both epithelial and mesenchymal tissues has spawned
interest in further understanding the normal cellular mechanisms that control the
production and metabolism of melanin.
In addition to its protective effects, melanin plays a key role in the etiology of
human disease. Several studies have suggested that aberrant production of melanin
may be an initial event in melanoma formation (Meyskens et al. 2007). Melanoma
cells accumulate abnormal melanosome structures both in human and rodent
melanoma models (Bomirski et al. 1987). Aberrant regulation of melanin in the
skin underlies the pigmentary disorders vitiligo and melasma (Grimes et al. 2006).
Vitiligo melanocytes are also characterized by abnormal accumulation of melano-
somal structures (Boissy et al. 1983, 1991a, b). Several Waardenburg’s syndrome
variants have both skin pigment defects and hearing loss, indicating that defects in
melanin production directly impact sound conduction (Price et al. 1998). Interest-
ingly, these variants have different patterns of epidermal pigment loss, demonstrat-
ing that there may be differences in the regulation of melanogenesis in different
body segments (Delezoide and Vekemans 1994). Age-related macular degeneration
is characterized by irregular deposition of melanin in the macula, suggesting a link
between melanin dysfunction and the most common cause of blindness in the
United States (Sarangarajan and Apte 2005). Additional studies have indicated
that melanin plays a role in the etiology of Parkinson’s disease (Zecca et al.
2006), a disorder characterized by the loss of both dopaminergic neurons and
melanin in the substantia nigra and locus ceruleus, a part of the brain that produces
the majority of neural melanin (Double and Halliday 2006). While the current
treatment of Parkinson’s disease involves the administration of agents that stimu-
late dopaminergic neurons, these agents are also known precursors of melanin
(L-dopa in particular) (Lewitt 2008). The impact of L-dopa treatment on melanin
production in the brain is currently unclear.
Melanin also plays a critical role in an organism’s adaptation to its environment
(Hoekstra 2006). Color adaptation can provide camouflage for certain animal
species, protecting them from predators or toxic insult. Variations in melanin
pigment define adaptive pigment variation in wolves (Anderson et al. 2009),
adaptive pigment variation in mice (Nachman 2005), and coat color variation in
cattle (Seo et al. 2007). Variation in plumage can provide a protective advantage for
bird species (Palleroni et al. 2005). Similarly, several species of fish can selectively
change color to adapt to their environment (Sugimoto 2002) to help them avoid
predators. In humans, skin color varies with latitude, suggesting an adaptive
relationship between variation in skin pigmentation and relative UV exposure
(Parra 2007). The mechanisms that underlie this differential regulation are
230 H. Ho et al.
Extensive studies have identified over 150 genes that regulate melanin production
and have centered on characterizing inherited pigmentary disorders (Bennett and
Lamoreux 2003). Melanin production is precisely regulated within the cell, involv-
ing specific mechanisms of transcriptional, enzymatic, and spatial regulation
(Yamaguchi et al. 2007). Recent studies have identified several different pathways
that upregulate melanin production in the melanocyte. UV irradiation is the primary
carcinogen in the epidermis, and can induce DNA crosslinks (Gorner 1994).
A central role of the melanocyte in the skin is to protect adjacent epidermal cells
from the harmful effects of UV irradiation (Costin and Hearing 2007). To do this,
the melanocyte must be able to sense UV damage and stimulate the production of
melanin in response to ultraviolet light (Costin and Hearing 2007). Recent studies
have demonstrated that the keratinocyte play a vital role in inducing melanogenesis.
UV irradiation of keratinocytes in the epidermis leads DNA crosslinks, which
eventually result in p53 activation (Cui et al. 2007). Recent studies have demon-
strated that p53 activation in keratinocytes is critical for inducing melanogenesis in
adjacent melanocytes (Cui et al. 2007). This phenomenon can occur by one of the
two mechanisms. Some studies have determined that p53 activation in keratino-
cytes leads to increased MSH production, which can then bind to the melanocortin
receptor on melanocytes and stimulate pigment production (Cui et al. 2007). Other
studies have determined that p53 activation leads to increased Kit ligand expression
in keratinocytes, which can then bind to the Kit receptor on adjacent melanocytes
and induce pigment production (McGowan et al. 2008). The precise role of each of
these pathways in UV-induced melanogenesis is currently unclear.
While UV irradiation can stimulate melanin production, other mechanisms exist
to regulate the baseline levels of intracellular melanin. One such mechanism
involves the EDNRB pathway. This pathway is initiated when endothelin binds
to the endothelin receptor, EDNRB. EDNRB can then stimulate an intracellular
signaling cascade involving the G-protein GNA11, ultimately leading to an increase
in MITF transcriptional activity (Hou et al. 2004). Other ligands can regulate MC1R
activity in a UV-independent manner. In wolves, coat color variation is partially
regulated by b-defensin, a molecule that also regulates innate immunity (Anderson
et al. 2009). b-defensin binds to the melanocortin 1 receptor, stimulating melanin
production via a mechanism similar to MSH (Anderson et al. 2009). The precise
regulation of b-defensin and endothelin in human skin and the roles of these
molecules in pigment regulation is poorly understood.
Harnessing RNAi-Based Functional Genomics to Unravel the Molecular Complexity 231
trafficking (Setty et al. 2007). This pathway is specifically required for the transport
of ATP7A, a copper transporter that loads copper onto tyrosinase, to the melano-
some (Setty et al. 2008). BLOC-2 mutations have more subtle pigment dilution
phenotypes, suggesting that these two complexes are components of the same
trafficking pathway.
Despite the significant advances gained from studying genetic disorders of
pigmentation (Bennett and Lamoreux 2003), the complex regulation of melanogen-
esis in human cells still remains a mystery. Two aspects of melanin regulation have
been particularly difficult to decipher. The regulation of the transcription of mela-
nogenic enzymes is extremely complex, involving the interaction of multiple
different signaling pathways and signaling intermediates. Similarly, the transport
of melanosomal enzymes is a complex process involving the interaction of several
different transport pathways. RNAi-based functional genomics is a systems-level
approach to identify key regulators of biological phenotypes. We sought to use this
technology to uncover the underappreciated molecular complexity that governs
melanin production. Specifically, we hoped to use this approach to identify cellular
components that regulate the transcription of melanogenic enzymes and the trans-
port of melanosomal enzymes.
to four unique siRNA duplexes, targeting each of the 21,127 unique human genes
arrayed in a one-gene/one-well format on 96 well microtiter plates was used to
identify siRNAs that inhibit pigment production. A spectrophotometric melanin
quantitation assay was coupled with a luminescence cell viability assay (CellTiter-
Glo) to identify siRNAs that impact melanin production but do not have impacts on
cell survival (Ganesan et al. 2008). To identify genes that impact both pheomelanin
and eumelanin production, we measured melanin content at 405 nm (Ozeki et al.
1996), a wavelength at which both pheomelanin and eumelanin absorb light.
Previous studies have demonstrated that MNT-1 cells readily secrete melanin into
the media, facilitating the identification of genes that regulate all stages of melanin
secretion. Optimization studies using tyrosinase siRNAs determined the optimal
time to measure the impact of siRNA treatment on pigment production in MNT-1
cells (Ganesan et al. 2008). Other preliminary studies developed protocols to
normalize our dataset to control for variations in pigment due to plate or position
effects. The entire screen was completed in duplicate (Ganesan et al. 2008). The
mean and standard deviation for each siRNA pool was calculated, and the
corresponding log 2 value or Z-score (Fig. 1) was used to plot the data distribution.
2
Z-score
0
0 5000 10000 15000 20000
–1
–2
–3
–4
siRNAs
Table 1 (continued)
Category Symbol Comments Motifs
Mutation causes
porphyria cutanea
tarda
HPD Mutation causes Glyoxalase
tyrosinemia type III
ALDH9A1 Aldedh
PLTP BPI1, BPI2
MSRA Downregulated in vitiligo PMSR
(hypopigmentation)
SMOX Amino_oxidase, DAO
UEVLD UBCc, Ldh_1_N, Ldh_1_C
GMPPB NTP_transferase, Hexapep
ALDH1A1 Expression lost in Aldedh
Parkinson’s disease
MGC4172 adh_short, Epimerase, KR
ENO2 Enolase_N, Enolase_C
Protein phosphorylation NLK S_TKc
PKN2 Hr1, C2, S_TKc, S_TK_X
RIOK1 RIO
PPP1R15A Expression lost in
melanoma
transformation
PPP2CB PP2Ac
Helicase RTEL1 DEXDc, HELICc
LOC389901 Ku, SAP DNA bd
Peptidase ARTS-1 Peptidase_M1
KLK13 Tyrp_SPc
LYZ Amyloidosis LYZ1
ADAM19 Pep_M12B_propep,
Reprolysin, DISIN,
ACR, EGF_2
CPZ FRI, Zn_pept
TRY1 Tryp_SPc
SENP1 DSS1_SEM1
SHFM1 Split hand/foot Peptdiase_C48
malformation
Translation EEF1A1 GTP_EFTU,
GTP_EFTU_D2,
GTP_EFTU_D3
VARS2 tRNA-synt_1, Anticodon_1
Other unknown NPM3 Nucleoplasmin
STX18 Syntaxin
KRTAP4- Keratin_B2
11
FGF23 Overexpressed in FGF
hyperpigmentation
syndrome
SFRS2 RRM
SLC17A5 Mutation causes Salla MFS_1
disease
USHBP1
(continued)
Harnessing RNAi-Based Functional Genomics to Unravel the Molecular Complexity 237
Table 1 (continued)
Category Symbol Comments Motifs
UBE2V1 UBCc
TEX11 TPR_2
TANC2 ANK, TPR
FATE1
LRRC1 LRR
RTN3 Reticulon
SPATA22
ETAA1 Tumor antigen,
melanoma of soft
parts
c12orf49
FAM125B
HSPC049 WD40
AFAP1L2 PH
FLJ41423
MAGEA6 Melanoma antigen MAGE
MUC3b EGF, SEA
C1orf194 NuA4
FAM89B
Our siRNA-based screening approach identified 98 siRNAs that significantly inhibited pigment
production. Four of these 98 genes did not retest, while two of the 98 genes were removed from the
Refseq database. Gene Ontology databases were used to segregate the 92 remaining genes into
function classes and identify conserved domains within the corresponding proteins. Gene ontology
databases, OMIM, and Pubmed searching was utilized to identify associations between our genes
and human diseases. This table was taken with permission from Ganesan et al. (2008)
Initial studies utilized tyrosinase siRNA to set a threshold for hit determination.
Using this approach, 96 genes were identified that significantly impacted pigment
production in MNT-1 cells (Table 1).
Upon completion of the screen, our first task was to identify those genes that
significantly impacted melanin production. Determination of hit threshold in
RNAi screens is a somewhat arbitrary process relying on a combination of statisti-
cal parameters and information about known genes that regulate the pathway under
study. Examination of our RNAi dataset suggested that our approach had a high
false negative rate, as many known regulators of melanogenesis did not score as
“hits” using our scoring threshold although they did have some impacts on pigment
production (Z-score less than 1) (Ganesan et al. 2008). While our screen had an
apparent high false negative rate, it also likely had a high false positive rate similar
238 H. Ho et al.
to other RNAi screens. Our first challenge was to identify those targets that most
likely impacted melanogenesis.
To adjust our hit threshold, we initially examined our target list to see if our hits
were components of known melanogenesis pathways or were components of other
novel pathways that regulate melanogenesis. While annotation analysis of our hits
did define a novel melanogenesis regulatory pathway (autophagy) that impacts
melanogenesis, the majority of our hits had no apparent connection with known
melanogenesis regulatory pathways, nor were they components of discrete biological
machines that could be linked to melanogenesis. Therefore, we sought to develop
better methods to identify novel melanogenesis regulators and novel melanogenesis
regulatory pathways hidden in our screen dataset.
To uncover novel melanogenesis regulatory pathways hidden within our dataset,
we integrated multiple systems-level approaches to identify additional genes in our
screen that may have impacts on melanogenesis. For this purpose, we utilized two
different PPI network algorithms to better identify “hits” in our RNAi screen.
Recent studies have developed algorithms (RNAicut) (Kaplow et al. 2009) to adjust
the score threshold in RNAi screens by integrating protein–protein interaction
network connectivity algorithms with functional genomics datasets. These
approaches rely on the contention that true hits in RNAi screens are more densely
connected within the PPI network. RNAiCut computes the edge count of the
induced subgraph for a given list of genes and estimates the p-value for obtaining
a PPI subgraph of at least that size within the PPI network. The program then
computes a cutoff based on the distribution of p-values. To better define the
spectrum of biological processes that regulate melanogenesis identified in our
screen, we utilized the RNAicut algorithm to identify GO biological process that
were enriched in our RNAi screen (Fig. 2). For this analysis, we entered all genes
that had a Z score less than 1 into the RNAicut algorithm (2,583 genes) and
identified 2,497 genes that met the cutoff. This cutoff (Z-score < 1 or >1) was
selected because this list contained other known regulators of melanogenesis
(Ganesan et al. 2008). We examined the GO biological process data for each of
these genes and sorted the genes into respective categories. While the majority of
genes did not sort into discrete functional categories, we did identify a relative
enrichment of genes that are components of intracellular signaling pathways (7.7%)
(Fig. 2). The finding that a large proportion of these genes are G-protein-coupled
receptors (4.3%) is consistent with other studies showing that G-protein-coupled
receptors and their signaling intermediates play a critical role in regulating mela-
nogenesis (Van Raamsdonk et al. 2004). Our hits were also enriched in transcrip-
tional regulators, further illustrating the complexity of transcriptional networks
that impact melanogenesis (Steingrimsson et al. 2004). Additionally, we identified
components of protein transport machinery and metabolic machinery that do
impact melanogenesis, consistent with the established role of metabolic pathways
(Scislowski et al. 1985) and protein transport (Raposo and Marks 2007) in melanin
synthesis. While the RNAicut algorithm did identify some biological pathways that
were enriched in our screen, it was not able to significantly pare down our list of
targets. Future validation studies will determine if the biological pathways
Harnessing RNAi-Based Functional Genomics to Unravel the Molecular Complexity 239
Angiogenesis
Apoptosis
Ion transport
Cell adhesion
Cell cycle
Cell differentiation
Cell proliferation
Cell-Cell signaling
Chromatin metabolism
Development/ Morphogenesis
DNA Metabolism
Electron transport
Epidermis development
Intracellular metabolism
Microtubule-based movement
Neurogenesis
Protein folding
Protein modification
Protein transport
Proteolysis and peptidolysis
Transcription regulation
RNA processing
RNA splicing
Signaling pathways
Ubiquitin
Other
Fig. 2 Identification of novel regulators of melanogenesis by coupling our RNAi dataset with PPI
network connectivity algorithms. To further identify novel regulators of melanogenesis, siRNAs
that negatively impacted melanogenesis (Z-score less than 1, 2,583 genes) were inputted into the
RNAicut Web-based program. Using the established p-value cutoff, 2,497 genes were identified
that were highly connected in the PPI network. GO biological process annotation data was utilized
to sort the genes into functional categories. A large number of genes were involved in either
transcription or intracellular signaling pathways
pigmentary disorder, binds to EDN3 (Hofstra et al. 1996) and activates an intracel-
lular signaling cascade via the G-protein GNA11 (Van Raamsdonk et al. 2004).
Mouse knockout studies revealed that EDNRB impacts MITF expression and
function (Hou and Pavan 2008). Recent studies have determined that the novel
gene targets identified in our screen that are topologically similar to EDNRB impact
MITF expression and are components of pathways downstream of EDNRB
(Ho et al. 2010). Complementation strategies validated that these novel genes
were components of the EDNRB pathway in MNT-1 melanoma cells and normal
melanocytes (Ho et al. 2010). Together, these studies provide compelling evidence
that topologically similar proteins are more likely components of the same molecu-
lar pathway, providing an additional method to identify pathway components from
RNAi datasets.
The studies outlined in chapter ‘RNAi Suppression and its Application’ by authors
Xiaoping Yi and Rui Lu sought to use novel approaches to improve the identifica-
tion of score cutoffs in high-throughput RNAi screens. Before we could investigate
the utility of these approaches in the context of our screen, we first needed to
validate hits identified in our screen experimentally to determine the false positive
rate of our approach. To facilitate the identification of genes that significantly
impacted melanogenesis within our dataset, we utilized the values obtained for
tyrosinase siRNA, the enzyme that is the rate-limiting step in melanogenesis, to
determine a cutoff to identify hits in our analysis. In our initial analysis, tyrosinase
siRNA had a normalized pigment ratio 2.5 standard deviations below the mean.
Therefore, an arbitrary threshold of 2 standard deviations below the mean was
utilized to identify siRNAs that significantly impact pigment production. This
cutoff identified 96 novel genes that impact pigment production in MNT-1 cells
(Table 1). Extensive validation studies revealed that a significant proportion of
these genes were true regulators of pigment production. Initial retesting of 35 of the
96 siRNA pools identified in the screen revealed a false positive rate of 12.1%. The
high true positive rate of these validation studies indicate that our cutoff threshold
was too stringent, indicating that many other valid targets exist between Z-score
2 and Z-score 1. Indeed, siRNAs directed towards several known regulators of
melanogenesis had Z-scores between 2 and 1.
A pool deconvolution strategy was utilized to identify siRNA phenotypes that
were a consequence of the off-target effects of an individual siRNA. Briefly, we
retested each of the four siRNAs from each siRNA pool to determine if it had a
significant impact on pigment production. Relative pigmentation was assessed
spectrophotometrically, and the relative inhibition of pigment production was deter-
mined relative to tyrosinase siRNA (normalized percent inhibition). These studies
revealed that for each gene, two or more siRNAs were able to recapitulate the
242 H. Ho et al.
phenotype of the pooled siRNA. These findings led us to conclude that none of the
phenotypes observed in our screen were a consequence of siRNA off-target effects.
The low off-target rate of our screen is likely secondary to the specificity of our assay,
which was able to effectively exclude siRNAs that would have had off-target effects
on cellular survival pathways. To lend further credence to the specificity of our
approach, we determined if the pooled siRNAs visually inhibited pigment produc-
tion. Many of our pooled siRNAs inhibited pigment production both macroscopically
and microscopically (Ganesan et al. unpublished observations), further demonstrat-
ing the specificity of our results. In summary, while our results clearly demonstrated
that the threshold utilized in this screen to identify hits was enriched in true-positives
(Ganesan et al. 2008), the cutoff also had a high false negative rate given the fact that
many known regulators of pigment production had Z-scores between 1 and 2 and
were not identified as hits. Future studies will develop additional methods to identify
relevant false negatives from our screen dataset.
Analysis of GO annotation data for the list of pigment regulators exposed a wide
variety of cellular processes represented by the validated and candidate hits, but it
did not give a clue as to the direct mechanism by which these genes impact pigment
production. Two areas of pigment regulation are less characterized – the signaling
pathways that lead to upregulation of melanin synthesis and the molecular compo-
nents that regulate the early phases of melanosome biogenesis. Therefore, we
sought to identify genes that impact either melanosome biogenesis or that upregu-
late the transcription of melanogenic enzymes. We first employed a focused
approach to identify signaling intermediates that impact the production of tyrosi-
nase, the rate-limiting enzyme specifying melanogenesis (Kim and Uyama 2005)
among novel validated genes. Relative accumulation of tyrosinase, the key tran-
scription factor MITF, and the melanosomal marker protein Melan-A were exam-
ined 96 h post siRNA transfection. Remarkably, over half of the validated pigment
genes appear to be required for tyrosinase protein accumulation. Of those pigment
genes impacting tyrosinase accumulation, approximately half appear to act at the
level of tyrosinase mRNA accumulation, and most of these also impaired MITF
mRNA accumulation (Ganesan et al. 2008). Given that tyrosinase is an MITF target
gene, the pigmentation genes in this later class likely represent action at the level of
MITF mRNA.
Seven genes were identified with significant impacts on MITF and tyrosinase
expression (Table 2) – Aldh1a1, Aldh9a1, Wipi1, SerpinB2, Plekha1, Itpk1, and
tyrosinase. Aldehyde dehydrogenases are a class of enzymes that detoxify lipid
aldehydes produced as a result of UV stress (Downes et al. 1997). The finding that
these genes impact pigment production provides an additional functional link
between UV stress and melanin production. Mutations in Plekha1 have been
linked to age-related macular degeneration, an ocular condition that is character-
ized by defective melanogenesis in the macula (Leveziel et al. 2007). SerpinB2 is
known to protect retinoblastoma protein, a known regulator of melanogenesis,
from degradation (Tonnetti et al. 2008). Itpk1 is a key integrator of inositol
phosphate pathways, and inositol signaling intermediates are known to impact
melanogenesis (Shears 2009).
Harnessing RNAi-Based Functional Genomics to Unravel the Molecular Complexity 243
Table 2 Genome-wide siRNA screening identifies targets that differentially impact tyrosinase
and MITF expression
Phenotype Symbol MITF Tyrosinase Melan A
▾ TYR and MITF TYR ▾ RNA ▾ RNA NO CHANGE
WIPI1 ▾ RNA ▾ RNA ▾ PROTEIN
ALDH1A1 ▾ RNA ▾ RNA NO CHANGE
ALDH9A1 ▾ RNA ▾ RNA NO CHANGE
PLEKHA1 ▾ RNA ▾ RNA ▾ PROTEIN
RAB4A ▾ RNA ▾ RNA ~PROTEIN
SERPINB2 ▾ RNA ▾ RNA ~PROTEIN
MSRA ▾ RNA ▾ RNA NO CHANGE
NPM3 ▾ RNA ▾ RNA ~PROTEIN
▾ MITF mRNA ARHGEF11 ▾ RNA ▾ PROTEIN NO CHANGE
▾ TYR protein ZDHHC9 ▾ RNA ▾ PROTEIN NO CHANGE
ITPK1 ▾ RNA ▾ PROTEIN ▾ PROTEIN
▾ MITF mRNA AGTR2 ▾ RNA NO CHANGE NO CHANGE
▾ TYR protein PPP1R15A ▾ PROTEIN ▾ PROTEIN NO CHANGE
ZFYVE1 NO CHANGE ▾ PROTEIN NO CHANGE
▾ MITF protein GNG2 ▾ PROTEIN NO CHANGE NO CHANGE
No change in MITF EDNRA NO CHANGE NO CHANGE ~PROTEIN
or TYR SMARCC2 NO CHANGE NO CHANGE NO CHANGE
FLJ1123 NO CHANGE NO CHANGE NO CHANGE
UROD NO CHANGE NO CHANGE NO CHANGE
UEV3 NO CHANGE NO CHANGE NO CHANGE
ARL4A NO CHANGE NO CHANGE NO CHANGE
P66A NO CHANGE NO CHANGE NO CHANGE
OR4F15 NO CHANGE NO CHANGE NO CHANGE
Western blotting and quantitative RT-PCR was used to identify siRNAs that impact tyrosinase,
MITF, and Melan-A protein levels or impact tyrosinase and MITF mRNA levels in MNT-1 cells.
siRNAs that significantly impacted the expression of MITF and tyrosinase mRNA as determined
by quantitative RT-PCR (p < 0.05 by Student’s t-test) and protein as determined by western
blotting, or siRNAs that only impacted protein accumulation as determined by western blotting
(densitometry values less than 50%) are shown. Genes are sorted into several phenotypes: genes
that regulate tyrosinase and MITF protein and mRNA accumulation, genes that regulate MITF
mRNA accumulation but only tyrosinase protein accumulation, genes that regulate MITF mRNA
accumulation but not tyrosinase or melan-a protein accumulation, genes that regulate protein but
not mRNA accumulation of tyrosinase or MITF, and genes that did not impact protein accumula-
tion of tyrosinase or MITF. This table was taken with permission from Ganesan et al. (2008)
Tyr shRNA
Ctl shRNA
Tyr
Actin
DAPI Merge
Fig. 3 Tyrosinase shRNA inhibits melanosome formation and pigment production in pigmented
melanoma cells. MNT-1 melanoma cells were infected with a pGIPZ lentivirions encoding a
tyrosinase shRNA downstream of a IRES-GFP to generate a mixed population of cells that express
or do not express tyrosinase shRNA. In this system, green cells express tyrosinase shRNA while
cells that are not infected. Subsequent to infection, cells were fixed and stained with Pmel17
antibody (a melanosome marker, red) and DAPI and visualized by phase contrast and fluorescence
microscopy at high magnification (63) using equivalent exposure times. shRNA-expressing cells
(green cells, top central panel) express low levels of Pmel17 staining (top left panel, bottom
middle panel) when compared to cells that do not express the shRNA. Additionally, green cells
lack melanosome granules (dark dots on phase contrast microscopy, top right panel). MNT-1 cells
infected with tyrosinase shRNA lentivirions or control shRNA containing lentivirions were
harvested and subjected to immunoblotting with tyrosinase and actin antibodies. While tyrosinase
shRNA samples have a higher amount of protein loaded (see actin loading control), they do not
contain any tyrosinase signal
Zdhhc9
Control
Rab4a
Msra
Tyr
siRNA:
bafilomycin: – + – + – + – + – +
TYR
ERK 1/2
Fig. 4 Identification of melanogenesis regulators that impact protein transport. MNT-1 cells
transfected with siRNAs that impact tyrosinase turnover (50 nM) were incubated in the presence
and absence of bafilomycin A1 for 24 h prior to lyses and analyses of tyrosinase protein accumu-
lation. All results shown are representative of a minimum of three independent experiments. This
figure was taken with permission from Ganesan et al. (2008)
Pigment shade varies widely between and among human ethnic populations.
Genome-wide association studies have determined that only a fraction of these
differences appear to be regulated by differences in the expression of known
melanogenesis regulators (Sulem et al. 2008). Therefore, it is likely that the key
molecular regulators of differences in pigment shade have yet to be identified. To
determine if the novel genes identified in our screen may have impacts on
pigment shade, we sought to examine the impact of novel pigmentation genes,
identified in MNT-1 cells, on pigment production in melanocytes of diverse
genetic backgrounds. Remarkably, the majority of targets that regulated tyrosi-
nase expression in MNT-1 cells also impacted tyrosinase expression when
depleted from darkly pigmented primary melanocytes (see Fig. 5). Approximately
half of these targets also inhibited tyrosinase expression when depleted from
moderately pigmented melanocytes. These results indicate that the primary screen
identified a number of genes that impact pigment production in several different
genetic backgrounds. Selective activity of some of these targets in different
genetic backgrounds suggests that some of these novel regulators of melanogene-
sis may play a role in human phenotypic variation. This phenomenon could be
either secondary to varying activities of these pigment regulators in different
pigment backgrounds or differential expression of these genes in different pig-
ment backgrounds. Future large-scale studies are required to determine how these
genes differentially regulate pigment shade. Additional study is also required to
determine the impact of natural selection and evolution on these novel pigment
regulatory genes.
248 H. Ho et al.
MNT-1 DP-PHM
MSRA
NPM3
ALDH9A1 RAB4A ATGR2
SERPINB2 PLEKHA1 ARHGEF11
ITPK1
TYR
WIPI1
ALDH1A1
ZDHHC9
MP-PHM
Fig. 5 Identification of putative regulators of human pigment variation. The indicated siRNAs,
targeting novel pigmentation genes identified in the MNT-1 screen, were tested for consequences
on tyrosinase accumulation in darkly pigmented primary human melanocyte (DP-PHM) and
moderately pigmented primary human melanocyte (MP-PHM) cultures 6 days posttransfection.
The results presented here is a Venn diagram of the data presented in Ganesan et al. (2008),
demonstrating that we have identified pigment regulators that differentially impact pigment
production in different genetic backgrounds. This figure was taken with permission from Ganesan
et al. (2008)
A fundamental goal of our RNAi screen was to identify gene targets to facilitate the
rationale design of novel depigmenting agents. As a first step, we sought to
determine if our gene list contained “druggable” targets for which known inhibitors
are available. Among our gene targets were two isoforms of aldehyde dehydroge-
nase, ALDH1A1 and ALDH9A1, enzymes that regulate ethanol detoxification
(Edenberg 2007) and response to UV stress (Downes et al. 1997). A number of
chemical inhibitors of these enzymes have been identified (DeMaster et al. 1998),
and several of these agents are clinically utilized to treat alcoholism, presenting an
opportunity to pharmacologically validate our screen findings. Two Aldh inhibitors,
cyanamide and Angeli’s salt (DeMaster et al. 1998), inhibited pigmentation and
tyrosinase protein accumulation in MNT-1 cells at doses that are equivalent to those
required to inhibit Aldh activity in culture (Ganesan et al. 2008). In addition, these
compounds impaired UV-induced tyrosinase expression when tested in primary
Harnessing RNAi-Based Functional Genomics to Unravel the Molecular Complexity 249
melanocytes (Ganesan et al. 2008). Cyanamide has been clinically used for alcohol
detoxification in Europe for decades, and abundant clinical safety data is readily
available for this drug. By coupling existing pharmacologic databases with our
RNAi dataset, we have been able to identify important novel depigmenting agents
that can be rapidly translated from the bench to the bedside. Together, these studies
demonstrate the utility of coupling pharmacologic inhibitor databases and RNAi
datasets to identify novel treatments for clinical disorders and illustrate the potential
translational nature of RNAi technology.
9 Concluding Remarks
References
Anderson TM, vonHoldt BM, Candille SI et al (2009) Molecular and evolutionary history of
melanism in North American gray wolves. Science 323:1339–1343
Aspengren S, Hedberg D, Skold HN et al (2009) New insights into melanosome transport in
vertebrate pigment cells. Int Rev Cell Mol Biol 272:245–302
Basrur V, Yang F, Kushimoto T et al (2003) Proteomic analysis of early melanosomes: identifica-
tion of novel melanosomal proteins. J Proteome Res 2:69–79
Bennett DC, Lamoreux ML (2003) The color loci of mice – a genetic century. Pigm Cell Res
16:333–344, Sponsored by the European Society for Pigment Cell Research and the Interna-
tional Pigment Cell Society
Berson JF, Harper DC, Tenza D et al (2001) Pmel17 initiates premelanosome morphogenesis
within multivesicular bodies. Mol Biol Cell 12:3451–3464
Boissy RE, Beato KE, Nordlund JJ (1991a) Dilated rough endoplasmic reticulum and premature
death in melanocytes cultured from the vitiligo mouse. Am J Pathol 138:1511–1525
250 H. Ho et al.
Boissy RE, Liu YY, Medrano EE et al (1991b) Structural aberration of the rough endoplasmic
reticulum and melanosome compartmentalization in long-term cultures of melanocytes from
vitiligo patients. J Invest Dermatol 97:395–404
Boissy RE, Smyth JR Jr, Fite KV (1983) Progressive cytologic changes during the development
of delayed feather amelanosis and associated choroidal defects in the DAM chicken line.
A vitiligo model. Am J Pathol 111:197–212
Bomirski A, Wrzolkowa T, Arendarczyk M et al (1987) Pathology and ultrastructural character-
istics of a hypomelanotic variant of transplantable hamster melanoma with elevated tyrosinase
activity. J Invest Dermatol 89:469–473
Breathnach AS (1988) Extra-cutaneous melanin. Pigment Cell Res 1:234–237
Brown KR, Jurisica I (2005) Online predicted human interaction database. Bioinformatics
21:2076–2082
Chow CY, Zhang Y, Dowling JJ et al (2007) Mutation of FIG4 causes neurodegeneration in the
pale tremor mouse and patients with CMT4J. Nature 448:68–72
Costin GE, Hearing VJ (2007) Human skin pigmentation: melanocytes modulate skin color in
response to stress. FASEB J 21:976–994
Cui R, Widlund HR, Feige E et al (2007) Central role of p53 in the suntan response and pathologic
hyperpigmentation. Cell 128:853–864
Delevoye C, Hurbain I, Tenza D et al (2009) AP-1 and KIF13A coordinate endosomal sorting and
positioning during melanosome biogenesis. J Cell Biol 187:247–264
Delezoide AL, Vekemans M (1994) Waardenburg syndrome in man and splotch mutants in
the mouse: a paradigm of the usefulness of linkage and synteny homologies in mouse and
man for the genetic analysis of human congenital malformations. Biomed Pharmacother 48:
335–339
DeMaster EG, Redfern B, Nagasawa HT (1998) Mechanisms of inhibition of aldehyde dehydro-
genase by nitroxyl, the active metabolite of the alcohol deterrent agent cyanamide. Biochem
Pharmacol 55:2007–2015
Di Pietro SM, Falcon-Perez JM, Tenza D et al (2006) BLOC-1 interacts with BLOC-2 and the
AP-3 complex to facilitate protein trafficking on endosomes. Mol Biol Cell 17:4027–4038
Double KL (2006) Functional effects of neuromelanin and synthetic melanin in model systems.
J Neural Transm 113:751–756
Double KL, Halliday GM (2006) New face of neuromelanin. J Neural Transm 70:119–123
Downes JE, Swann PG, Holmes RS (1997) A genetic basis for corneal sensitivity to ultraviolet
light among recombinant SWXJ inbred strains of mice. Curr Eye Res 16:539–546
Edenberg HJ (2007) The genetics of alcohol metabolism: role of alcohol dehydrogenase and
aldehyde dehydrogenase variants. Alcohol Res Health 30:5–13
Etchevers HC, Amiel J, Lyonnet S (2006) Molecular bases of human neurocristopathies. Adv Exp
Med Biol 589:213–234
Fowler DM, Koulov AV, Alory-Jost C et al (2006) Functional amyloid formation within mammalian
tissue. PLoS Biol 4:e6
Ganesan AK, Ho H, Bodemann B et al (2008) Genome-wide siRNA-based functional genomics of
pigmentation identifies novel genes and pathways that impact melanogenesis in human cells.
PLoS Genet 4:e1000298
Goding CR (2007) Melanocytes: the new black. Int J Biochem Cell Biol 39:275–279
Gorner H (1994) Photochemistry of DNA and related biomolecules: quantum yields and con-
sequences of photoionization. J Photochem Photobiol B 26:117–139
Grimes P, Nordlund JJ, Pandya AG et al (2006) Increasing our understanding of pigmentary
disorders. J Am Acad Dermatol 54:S255–S261
Halaban R (2000) The regulation of normal melanocyte proliferation. Pigment Cell Res 13:4–14
Ho H, Milenkovic T, Memisevic V et al (2010) Protein interaction network topology uncovers
melanogenesis regulatory network components within functional genomics datasets. BMC Sys
Bio, in press
Harnessing RNAi-Based Functional Genomics to Unravel the Molecular Complexity 251
Hoek K, Rimm DL, Williams KR et al (2004) Expression profiling reveals novel pathways in the
transformation of melanocytes to melanomas. Cancer Res 64:5270–5282
Hoekstra HE (2006) Genetics, development and evolution of adaptive pigmentation in vertebrates.
Heredity 97:222–234
Hofstra RM, Osinga J, Tan-Sindhunata G et al (1996) A homozygous mutation in the endothelin-3
gene associated with a combined Waardenburg type 2 and Hirschsprung phenotype (Shah–
Waardenburg syndrome). Nat Genet 12:445–447
Hou L, Pavan WJ (2008) Transcriptional and signaling regulation in neural crest stem cell-derived
melanocyte development: do all roads lead to Mitf? Cell Res 18:1163–1176
Hou L, Pavan WJ, Shin MK et al (2004) Cell-autonomous and cell non-autonomous signaling
through endothelin receptor B during melanocyte development. Development (Cambridge,
England) 131:3239–3247
Huang J, Klionsky DJ (2007) Autophagy and human disease. Cell Cycle 6:1837–1849
Huizing M, Sarangarajan R, Strovel E et al (2001) AP-3 mediates tyrosinase but not TRP-1
trafficking in human melanocytes. Mol Biol Cell 12:2075–2085
Hume AN, Collinson LM, Rapak A et al (2001) Rab27a regulates the peripheral distribution of
melanosomes in melanocytes. J Cell Biol 152:795–808
Kadekaro AL, Kavanagh RJ, Wakamatsu K et al (2003) Cutaneous photobiology. The melanocyte
vs. the sun: who will win the final round? Pigment Cell Res 16:434–447
Kaplow IM, Singh R, Friedman A et al (2009) RNAiCut: automated detection of significant genes
from functional genomic screens. Nat Methods 6:476–477
Kasamatsu S, Hachiya A, Higuchi K et al (2008) Production of the soluble form of KIT, s-KIT,
abolishes stem cell factor-induced melanogenesis in human melanocytes. J Invest Dermatol
128:1763–1772
Kim YJ, Uyama H (2005) Tyrosinase inhibitors from natural and synthetic sources: structure,
inhibition mechanism and perspective for the future. Cell Mol Life Sci 62:1707–1723
Kushimoto T, Basrur V, Valencia J et al (2001) A model for melanosome biogenesis based on the
purification and analysis of early melanosomes. Proc Natl Acad Sci USA 98:10698–10703
Kwon BS, Kim KK, Halaban R et al (1994) Characterization of mouse Pmel 17 gene and silver
locus. Pigment Cell Res 7:94–397
Lamoreux ML, Wakamatsu K, Ito S (2001) Interaction of major coat color gene functions in mice
as studied by chemical analysis of eumelanin and pheomelanin. Pigment Cell Res 14:23–31
Leveziel N, Souied EH, Richard F et al (2007) PLEKHA1-LOC387715-HTRA1 polymorphisms
and exudative age-related macular degeneration in the French population. Mol Vis
13:2153–2159
Levine B, Kroemer G (2008) Autophagy in the pathogenesis of disease. Cell 132:27–42
Levy C, Khaled M, Fisher DE (2006) MITF: master regulator of melanocyte development and
melanoma oncogene. Trends Mol Med 12:406–414
Lewitt PA (2008) Levodopa for the treatment of Parkinson’s disease. N Engl J Med 359:2468–2476
Liang C, Lee JS, Inn KS et al (2008) Beclin1-binding UVRAG targets the class C Vps complex to
coordinate autophagosome maturation and endocytic trafficking. Nat Cell Biol 10:776–787
Manneville JB, Etienne-Manneville S, Skehel P et al (2003) Interaction of the actin cytoskeleton
with microtubules regulates secretory organelle movement near the plasma membrane in
human endothelial cells. J Cell Sci 116:3927–3938
McGowan KA, Li JZ, Park CY et al (2008) Ribosomal mutations cause p53-mediated dark skin
and pleiotropic effects. Nat Genet 40:963–970
Meyskens FL Jr, Farmer PJ, Yang S et al (2007) New perspectives on melanoma pathogenesis and
chemoprevention. Recent Results Cancer Res 174:191–195
Milenkovic T, Memisevic V, Ganesan AK et al (2010) Systems-level cancer gene identification
from protein interaction network topology applied to melanogenesis-related functional geno-
mics data. J R Soc Interface 7(44):423–437
Milenkovic T, Przulj N (2008) Uncovering biological network function via graphlet degree
signatures. Cancer Inform 4:257–273
252 H. Ho et al.
Setty SR, Tenza D, Sviderskaya EV et al (2008) Cell-specific ATP7A transport sustains copper-
dependent tyrosinase activity in melanosomes. Nature 454:1142–1146
Setty SR, Tenza D, Truschel ST et al (2007) BLOC-1 is required for cargo-specific sorting from
vacuolar early endosomes toward lysosome-related organelles. Mol Biol Cell 18:768–780
Shears SB (2009) Molecular basis for the integration of inositol phosphate signaling pathways via
human ITPK1. Adv Enzyme Regul 49:87–96
Shotelersuk V, Gahl WA (1998) Hermansky–Pudlak syndrome: models for intracellular vesicle
formation. Mol Genet Metab 65:85–96
Slominski A, Tobin DJ, Shibahara S et al (2004) Melanin pigmentation in mammalian skin and its
hormonal regulation. Physiol Rev 84:1155–1228
Smit NP, Kolb RM, Lentjes EG et al (1998) Variations in melanin formation by cultured
melanocytes from different skin types. Arch Dermatol Res 290:342–349
Smit NP, Van der Meulen H, Koerten HK et al (1997) Melanogenesis in cultured melanocytes can
be substantially influenced by L-tyrosine and L-cysteine. J Invest Dermatol 109:796–800
Smith JW, Koshoffer A, Morris RE et al (2005) Membranous complexes characteristic of
melanocytes derived from patients with Hermansky–Pudlak syndrome type 1 are macroauto-
phagosomal entities of the lysosomal compartment. Pigment Cell Res 18:417–426
Stark C, Breitkreutz BJ, Reguly T et al (2006) BioGRID: a general repository for interaction
datasets. Nucleic Acids Res 34:D535–539
Stein WD (1955) Ultra-violet absorption investigation of melanins. Nature 175(4454):472
Steingrimsson E, Copeland NG, Jenkins NA (2004) Melanocytes and the microphthalmia
transcription factor network. Annu Rev Genet 38:365–411
Sugimoto M (2002) Morphological color changes in fish: regulation of pigment cell density and
morphology. Microsc Res Tech 58:496–503
Sulem P, Gudbjartsson DF, Stacey SN et al (2008) Two newly identified genetic determinants of
pigmentation in Europeans. Nat Genet 40:835–837
Theos AC, Tenza D, Martina JA et al (2005) Functions of adaptor protein (AP)-3 and AP-1 in
tyrosinase sorting from endosomes to melanosomes. Mol Biol Cell 16:5356–5372
Theos AC, Truschel ST, Tenza D et al (2006) A lumenal domain-dependent pathway for sorting to
intralumenal vesicles of multivesicular endosomes involved in organelle morphogenesis. Dev
Cell 10:343–354
Tonnetti L, Netzel-Arnett S, Darnell GA et al (2008) SerpinB2 protection of retinoblastoma
protein from calpain enhances tumor cell survival. Cancer Res 68:5648–5657
Valencia JC, Watabe H, Chi A et al (2006) Sorting of Pmel17 to melanosomes through the plasma
membrane by AP1 and AP2: evidence for the polarized nature of melanocytes. J Cell Sci
119:1080–1091
Van Raamsdonk CD, Fitch KR, Fuchs H et al (2004) Effects of G-protein mutations on skin color.
Nat Genet 36(9):961–968
Watabe H, Valencia JC, Yasumoto K et al (2004) Regulation of tyrosinase processing and
trafficking by organellar pH and by proteasome activity. J Biol Chem 279:7971–7981
Whitehurst AW, Bodemann BO, Cardenas J (2007) Synthetic lethal screen identification of
chemosensitizer loci in cancer cells. Nature 446:815–819
Wu X, Rao K, Bowers MB et al (2001) Rab27a enables myosin Va-dependent melanosome
capture by recruiting the myosin to the organelle. J Cell Sci 114:1091–1100
Yamaguchi Y, Brenner M, Hearing VJ (2007) The regulation of skin pigmentation. J Biol Chem
282:27557–27561
Zecca L, Bellei C, Costi P et al (2008) New melanic pigments in the human brain that accumulate
in aging and block environmental toxic metals. Proc Natl Acad Sci USA 105:17567–17572
Zecca L, Zucca FA, Albertini A et al (2006) A proposed dual role of neuromelanin in the
pathogenesis of Parkinson’s disease. Neurology 67:S8–S11
Zhou Z, Li CY, Li K et al (2009) Decreased methionine sulphoxide reductase A expression renders
melanocytes more sensitive to oxidative stress: a possible cause for melanocyte loss in vitiligo.
Br J Dermatol 161(3):504–509
mRNA Structure and its Effects
on Posttranscriptional Gene Silencing
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
2 A Structured Target Site Reduces AON and siRNA Activity In Vitro . . . . . . . . . . . . . . . . . . . 257
3 Analysis of Binding Affinity to mRNA and Rate Dependencies
on Concentration for AON and siRNA Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
4 AON and siRNA Guide Strand Have Equal Affinity for the Target mRNA . . . . . . . . . . . . . 261
5 AON and siRNA Display Apparent First and Zero Order Kinetics . . . . . . . . . . . . . . . . . . . . . . 261
6 For Full In Vitro Activity, siRNA Require Greater
Target Site Accessibility Than AON . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
7 A Double-Stranded Target Site Greatly Reduces In Vitro PTGS Activity . . . . . . . . . . . . . . . 265
8 An AON That is More Effective Than the siRNA Against an Identical
Target In Vitro is Less Effective Against the Same Target In Vivo . . . . . . . . . . . . . . . . . . . . . . 265
9 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
10 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
S.I. Rudnick
Division of Hematology/Oncology, Department of Medicine, University of Pennsylvania School
of Medicine, Philadelphia, PA, USA
Fox Chase Cancer Center, Philadelphia, PA, USA
V. Aishwarya and A.M. Gewirtz (*)
Division of Hematology/Oncology, Department of Medicine, University of Pennsylvania School
of Medicine, Philadelphia, PA, USA
e-mail: gewirtz@mail.med.upenn.edu
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 255
RNA Technologies, DOI 10.1007/978-3-642-12168-5_11,
# Springer-Verlag Berlin Heidelberg 2010
256 S.I. Rudnick et al.
secondary structure. Surprisingly, we found that an AON was less stringent in its
requirement for mRNA target accessibility than the corresponding siRNA. By deter-
mining that the AON and siRNA guide strand have the same apparent KD in the
absence of protein, we show that nucleic acid binding affinity does not explain their
difference in in vivo silencing activity. Rather, it appears that RISC must increase the
binding affinity of the siRNA for the target. Furthermore, RNA binding proteins are
also potent inhibitors of AON activity. We conclude that mRNA secondary and
quaternary structure play important roles in PTGS by significantly affecting the
ability of a siRNA or AON to hybridize with their intended target. Recognition of
these effects will facilitate the design of more efficient antisense molecules for
therapeutically motivated gene silencing and argue for continued mechanistic studies
on AON and siRNA mediated mRNA destruction.
1 Introduction
define mRNA structural characteristics that inhibit siRNA and RISC activity have
yet to be made. Numerous chemical modifications have been made to increase the
binding affinity of AON (Pradeepkumar et al. 2003) and siRNA (Braasch et al. 2003;
Elmen et al. 2005) in order to overcome the energetic barriers imposed by mRNA
self structure, but no study has compared the in vitro activity of the RNase H-based
mechanisms of AON and siRNA by monitoring mRNA cleavage.
In order to determine the intramolecular structures in mRNA that affect PTGS by
AON and siRNA, we have employed an in vitro assay to monitor the sequence-
specific cleavage of mRNA (Rudnick et al. 2008). While maintaining a constant
target site and sequence, we introduced mRNA structure as the only variable in our
experiments by preannealing 20 -O-methyl oligonucleotides (20 OMe ON) at various
sites in the target mRNA. Since it is known that siRNA design can dictate activity
(Ding et al. 2003; Schwarz et al. 2003), we chose to use a single sequence for our
AON and siRNA. Therefore, fluctuations in cleavage activity could not be attributed
to the targeting molecule but instead will be a direct result of manipulating the
mRNA structure. This strategy, targeting a 182 nt segment of firefly luciferase,
reveals that secondary structure upstream or downstream of the target site is almost
inconsequential on AON and siRNA activity. Inducing secondary structure directly
on the target site revealed that AON can target smaller accessible regions in the
mRNA than siRNA, while a fully double-stranded target significantly hinders both
pathways. AON activity was much less pronounced than siRNA activity in a dual
luciferase assay. Perhaps not surprisingly, our data showed that in vitro AON and
siRNA activity does not necessarily correspond to cellular activity. This suggests that
AON activity is suppressed and siRNA activity enhanced when in the proper cellular
context and a further mechanistic understanding will allow for the more effective
development of these molecules as therapeutic agents and investigative tools.
AON and siRNA are both thought to bind their substrate in a diffusion mediated
manner, as opposed to having a protein guided mechanism (Stein 1999; Brown
et al. 2005; Yuan et al. 2005). Therefore, two properties closely tied to binding, and
presumably cleavage activity, should be the relative stoichiometry of the AON and
siRNA to the mRNA target and the relative stability of the resultant AON:mRNA or
siRNA:mRNA duplex.
In general, RNA duplexes are known to have relatively high melting tempera-
tures (TM) and a correspondingly low free energy compared to RNA:DNA and
DNA:DNA duplexes (Lesnik and Freier 1995). The TM for nucleic acid duplexes is
defined as the temperature where half of the molecules are double-stranded and half
are single-stranded. TM is related to the free energy of a duplex because the higher
the TM, the more the energy that is required to dissociate one nucleic acid strand
from another. Therefore, if a potential duplex has a high predicted TM, one expects
the two single strands to have a high affinity, or low KD, for each other. In order to
Table 1 Sequences of 20 -O-methyl oligonucleotides used to model intramolecular mRNA structure. Each group of 20 OMe ON was used in experiments where short
regions of double-stranded character were induced over a large area of the 182 nt luciferase target (Group I), in pairs on the target site with a variable number of free
target bases (Group II), or used to make the entire target site double-stranded with a variable internal loop (Group III). Between the sequence of each Group II 20 OMe
ON, the number of unhybridized target bases is indicated. In Group III 20 OMe ON, the lower case letters indicate bases that are not complementary with the mRNA
target
Group I 20 -O-methyl oligonucleotides Group II 20 -O-methyl oligonucleotides Group III 20 -O-methyl oligonucleotides
Name Sequence (50 –30 ) Sequence 50 –30 – Gap(nt) – 50 –30 Mismatches Sequence (50 –30 )
(nt)
A CCGAACGGACAUUUCGAAG GCCCAUAUCGUUUCA – 2 – 2 UCUGUGAUUUGUAUUacGCCCAUAUCGUUUCA
UCUGUGAU UUGUAUU
B GUUUCAUAGCUUCUGCCAA CCCAUAUCGUUUCAU – 4 – 4 UUCUGUGAUUUGUAUcacuCCCAUAUCGUUUCAU
UUCUGUGA UUUGUAU
C AGCCCAUAUCGUUUCAUAG CCAUAUCGUUUCAUA – 6 – 6 AUUCUGUGAUUUGUAgcacuaCCAUAUCGUUUCAUA
AUUCUGUG AUUUGUA
Target blocker AUUUGUAUUCAGCCCAUAU CAUAUCGUUUCAUAG – 8 – GAUUCUGU GAUUUGU
8 GAUUCUGUGAUUUGUcgcacuaaCAUAUCGUUUCAUAG
D CGAUUCUGUGAUUUGUAUU AUAUCGUUUCAUAGC – 10 – 10 CGAUUCUGUGAUUUGgcgcacuaaaAUAUCGUUUCAUAGC
CGAUUCU GUGAUUUG
E UGCAUACGACGAUUCUGUG UAUCGUUUCAUAGCU – 12 – 12 ACGAUUCUGUGAUUUagcgcacuaaacUAUCGUUUCAUAGCU
ACGAUUC UGUGAUUU
F AAUUGAAGAGAGUUUUCAC AUCGUUUCAUAGCUU – 14 – 14 GACGAUUCUGUGAUUcagcacuaaacgAUCGUUUCAUAGCUU
GACGAUU CUGUGAUU
UCGUUUCAUAGCUUC – 16 – 16 CGACGAUUCUGUGAUccagcgcacuaaacgcUCGUUUCAUAGCUUC
CGACGAUU CUGUGAU
mRNA Structure and its Effects on Posttranscriptional Gene Silencing
259
260 S.I. Rudnick et al.
a
A C D F
5'...cuucgaaauguccguucgguuggcagaagcuaugaaacgAUAUGGGCUGAAUACAAAUcacagaaucgucguagcagugaaaacucucuucaauu...3'
B Target Blocker E
b c
r
ke
ke
oc
oc
Bl
Bl
et
et
rg
rg
Ta
- + A B C D E F
Ta
- + A B C D E F
d
0.4
30s AON Reactions
8.5min siRNA Reactions
5' Cleavage Product (pmol)
0.3
0.2
0.1
0.0
ol ol A B C er D E F
o ntr ontr l o ck
e C C t B
e
tiv tiv rge
ga Posi Ta
Ne
Preannealed 2'-O-Methyl Oligonucleotide
Fig. 1 Analysis of inducing structure over a large area of the 182 nt luciferase mRNA target.
Partial sequence of mRNA target is shown (a) with the relative position of each individual
preannealed 20 OMe ON. The target sequence of the AON and siRNA guide strand is indicated
in all capital letters. Each induced structure of a was used in cleavage assays for 30 s in the
presence of 2.5 mM AON (b) or for 8.5 min in the presence of 2.5 mM siRNA (c) and analyzed on
10% sequencing gels. Negative controls () have no siRNA or AON and positive controls (+) are
cleavage reactions against the unknown, native mRNA structure in the absence of 20 OMe ON.
Cleavage product formation (d) for each reaction was calculated by normalization to the amount of
target in the negative control lanes of each gel, and the mean of at least three independent reactions
is plotted (1 standard deviation)
Source: Rudnick et al, PNAS, 2008
critically compare the activity of the AON and siRNA, their relative affinities for
the mRNA target in the absence of protein must be determined. Then, in order to
evaluate a potential correlation between TM and substrate turnover, AON and
siRNA activities are measured above and below the apparent KD.
mRNA Structure and its Effects on Posttranscriptional Gene Silencing 261
The AON and siRNA guide strand used for these studies were predicted to have
TM’s of 47 C and 59 C, respectively, with the RNA target (www.idtdna.com).
Therefore, in order to determine if oligonucleotide binding affinity and TM play a
role in AON or siRNA activity, we had to determine the apparent binding affinity
under the conditions used to detect mRNA cleavage. Since the mRNA target was
182 nt and the siRNA guide strand and AON are 21 nt, analyzing the mRNA and
the 203 nt duplexes can be accomplished easily by electrophoresis. 100 nM
mRNA was incubated in increasing concentrations of AON or siRNA guide strand
(Fig. 2). From 100 to 500 nM oligonucleotide, the duplexes can clearly be resolved
from the mRNA alone (Fig. 2a). At concentrations greater than or equal to 1 mM,
the mRNA is saturated with oligonucleotide and almost no free mRNA is detect-
able. The percent duplex in each lane of four independent experiments was plotted
against concentration of the oligonucleotide (Fig. 2b) with nearly identical results
for the AON and siRNA guide strand. The apparent KD for the AON and siRNA
under cleavage assay conditions, but in the absence of protein, is 378 37 nM
and 397 28 nM, respectively, suggesting that TM does not affect their affinity
in vitro.
To determine if the apparent KD has impact on AON and siRNA activity, they were
screened in a dose-response fashion while keeping the mRNA at 100 nM. Since the
binding affinity of the oligonucleotides in Fig. 2 had no significant difference, the
average of the two, 388 nM, was used as a center concentration for the activity
screen. AON and siRNA were tested in the in vitro cleavage assay at 25 nM,
194 nM, 388 nM, 587 nM, and 776 nM, which correspond to 0.06, 0.5, 1.0,
1.5, and 2 the average apparent KD.
AON cleavage reactions showed a broad range of activity (Fig. 3a). The
samples at 25 nM (filled circle) and 194 nM (empty triangle) had barely distin-
guishable amount of product formation. Increasing the AON concentration to
388 nM and above gave significant stepwise increases in initial rate and total
amount of product formed. However, there was no significant difference in initial
rates or total product formed for the siRNA when tested at 25 nM and above
(Fig. 3b). The only concentration that displayed multiple turnovers was the 25 nM
(0.125 pmol) siRNA (filled circle) since it generated greater than 0.2 pmol
50 cleavage product.
262 S.I. Rudnick et al.
Co ol
l
ro
ive ontr
nt
P o ve C
15 M
nM
nM
M
5n
M
ti
n
nM
nM
ga
si t
0n
0n
5n
0n
00
00
00
7.
Ne
25
50
10
18
25
37
50
10
20
AON:mRNA
AON
mRNA
ss-siRNA:mRNA
ss-siRNA
mRNA
b 120
siRNA
100 AON
80
Percent Duplex
60
40
20
0
10 100 1000 10000
Oligonucleotide(nM)
Fig. 2 Determination of apparent KD with mRNA for AON and siRNA in vitro. (a) mRNA:AON
duplex (top panel) or single-stranded siRNA:mRNA duplex (bottom panel) separated from mRNA
alone (bottom band) at indicated concentrations of oligonucleotide in a cell free system. Negative
controls are mRNA alone. Positive control has 2 mM oligonucleotide annealed to mRNA by heat
cool. (b) Quantification of plots in a. Points are average standard deviation of four independent
experiments. Data fit to four parameter logistic function
Calculating the initial velocity (Vo) of each reaction and plotting against concen-
tration of oligonucleotide used yielded relationships for both the AON and siRNA.
As expected from the data in Fig. 3, the AON reactions became faster and siRNA
reaction velocity was static, as the concentration of each oligonucleotide increased.
Therefore, in this concentration range, the AON and siRNA behave under apparent
first and zero order kinetics, respectively.
mRNA Structure and its Effects on Posttranscriptional Gene Silencing 263
a b
0.25 0.40
194nm AON
388nm AON 0.35
582nm AON
0.20 776nm AON
0.30
0.15 0.25
0.20
0.10 0.15
25nm siRNA
0.10 194nm siRNA
0.05 388nm siRNA
0.05 582nm siRNA
776nm siRNA
0.00
0 20 40 60 80 100 120 0 20 40 60 80 100 120
Time (min) Time (min)
Fig. 3 In vitro activity of AON and siRNA at varying concentrations. AON (a) and siRNA (b)
cleavage activity at increasing concentrations. The average standard deviation of three inde-
pendent experiments is plotted against time. The scale of the Y-axis used is different in order to
better illustrate differences in AON reactions
Since structure directly at the site of siRNA and AON targeting proved to be a
critical factor in mRNA cleavage, the minimum number of accessible bases for
activity was next examined. In order to create accessible targets of varying sizes
(Fig. 4a), two double-stranded regions were made equidistantly from the expected
siRNA cleavage site by annealing two Group II 20 OMe ON (Table 1). This tiling of
the target site was always done symmetrically, leaving an equal number of accessi-
ble bases on either side of the cleavage site.
After a 30 s cleavage reaction, the AON activity was greater than the positive
control when 8–16 nt of the target site remained accessible (Fig. 4b and d). Multiple
RNase H generated cleavage products were detectable when 14 nt and 16 nt were
free, while fewer products were observed when 8, 10, or 12 nt of the target were free
(Fig 4b). When 20 OMe ON were used to reduce target accessibility to 6 nt or less,
AON activity ranged from 56 to 27% of the positive control where no 20 OMe ON
were annealed.
In contrast to the AON, all siRNA cleavage reactions with a tiled target site generated
a single cleavage product (Fig. 4c). After an 8.5 min cleavage reaction, the siRNA had
enhanced activity with respect to the positive control when 16 nt of the target were left
accessible (Fig. 4c and d). However, when the target site was any size less than 16 nt,
siRNA activity ranged from 52 to 29% of the positive control (Fig. 4c and d).
These data again reveal differences in the way each PTGS pathway processes its
substrate in vitro. Surprisingly, the siRNA proved to be quite sensitive to any
addition of double-stranded character to the target site as demonstrated by the
need of 16 accessible nucleotides for full activity. Equally surprising, the AON
264 S.I. Rudnick et al.
a
5'...uuggcagaagcuaugaaacgAUAUGGGCU*GAAUACAAAUcacagaaucgucguaugca...3'
Unhybridized Nucleotides
b c
Gap in 2'OMe Oligo (nt) Gap in 2'OMe Oligo (nt)
- + 2 4 6 8 10 12 14 16 - + 2 4 6 8 10 12 14 16
d 0.4
30s AON Reactions
8.5min siRNA Reactions
5' Cleavage Product (pmol)
0.3
0.2
0.1
0.0
ol ol 2 4 6 8 10 12 14 16
o ntr ontr
e C ive C
tiv t
e ga Posi
N
Gap Between Preannealed
2'-O-Methyl Oligonucleotides (nt)
Fig. 4 Determination of the minimum number of accessible bases for AON and siRNA guided
cleavage in vitro. (a) Partial sequence of the 182 nt firefly luciferase target mRNA. Two 20 OMe
ON were annealed at the outer edges of the target site (CAPS) and positioned incrementally closer
to the expected siRNA cleavage site (*). Cleavage activity is reported as a function of the number
of unhybridized bases between the two preannealed 20 OME ON. Analysis of the 30 s AON (b) and
8.5 min siRNA (c) reactions was performed on 10% sequencing gels where negative controls ()
have no AON or siRNA and positive controls (þ) target the unknown, native structure of the
mRNA in the absence of 20 OMe ON. The amount of cleavage product (d) was calculated by
normalization to the amount of target in the negative controls of each gel, and the mean of at least
three independent reactions is plotted (1 standard deviation)
Source: Rudnick et al, PNAS, 2008
revealed robust activity so long as 8 nt or more were not bound by 20 OMe ON.
Based on these data, one might anticipate that siRNA would have considerably less
activity than AON in vivo though clearly this is not the case.
mRNA Structure and its Effects on Posttranscriptional Gene Silencing 265
a
5'...uuggcagaagcuaugaaacgAUAUGGGCUGAAUACAAAUcacagaaucgucguaugca...3'
3' GAUACUUUGCUAUAC UGUUUAGUGUCUUAG 5'
AAUCACGC
b c
Mismatches in Bulge (nt) Mismatches in Bulge (nt)
- + 2 4 6 8 10 12 14 16 - + 2 4 6 8 10 12 14 16
d
0.4
30s AON Reactions
8.5min siRNA Reactions
5' Cleavage Product (pmol)
0.3
0.2
0.1
0.0
l l 2 4 6 8 10 12 14 16
tro tro
C on Con
e e
tiv tiv
e ga Posi
N
Bulge Created by Preannealed
2'-O-Methyl Oligonucleotides (nt)
Fig. 5 Analysis of cleavage activity when targeting a double-stranded mRNA with a variable
internal loop. (a) Partial sequence of the 182 nt firefly luciferase target. 20 OMe ON (CAPS) of
differing length was preannealed over the center of the target site (CAPS). Two 15 nt reverse
complementary arms were always annealed to the target to maintain duplex stability. Cleavage
activity is reported as a function of the number of mismatches (CAPS) used to vary the size of the
internal loop. Analysis of the 30 s AON (b) and 8.5 min siRNA (c) reactions was performed on 10%
sequencing gels where negative controls () have no AON or siRNA and positive controls (þ)
target the unknown, native structure of the mRNA in the absence of 20 OMe oligos. The amount of
cleavage product (d) was calculated by normalization to the amount of target in the negative controls
of each gel, and the mean of at least three independent reactions is plotted (1 standard deviation)
Source: Rudnick et al, PNAS, 2008
than the corresponding siRNA. We tested this hypothesis using K562 human
leukemia cells because similar AON and siRNA nucleofection efficiency was
obtained when codelivered with the reporter vectors. In three independent experi-
ments, scrambled AON and scrambled siRNA showed no effect on luciferase
mRNA Structure and its Effects on Posttranscriptional Gene Silencing 267
a b
0 2.5 5 15 30 60 90 120min 0 2.5 5 15 30 60 90120min
T T
c
0.5
2.5µM AON
2.5µM siRNA
5' Cleavage Product (pmol)
0.4
0.3
0.2
0.1
0.0
0 20 40 60 80 100 120
Time (min)
Fig. 6 Cleavage activity as a function of time with double-stranded target. Two hour cleavage
reactions in the presence of 2.5 mM AON (a) or 2.5 mM siRNA (b) are analyzed on 10%
sequencing gels. The double-stranded mRNA target with 6 nt mismatched bulge, primary cleavage
product, and secondary cleavage products are indicated by T, PCP, and SCP, respectively. Primary
cleavage product formation as a function of time (c) where values were calculated by normalizing
product at each time to the amount of target at the initial time point. In the AON reaction (filled
circles), all bands migrating near 100 nt were summed after normalization and plotted while only
one product was seen for the siRNA reaction (empty circles). The mean of at least three indepen-
dent reactions is plotted (1 standard deviation)
Source: Rudnick et al, PNAS, 2008
activity when compared to vector alone (Fig. 7). When firefly luciferase was
targeted, a reduction in the relative luminescence of 35 and 78% for AON and
siRNA was observed, respectively.
These data show that in a cellular context, an AON is not necessarily as robust
with respect to target cleavage as it is in vitro. We found the obverse true as well,
268 S.I. Rudnick et al.
1.4
1.2
1.0
Relative Luminesence
0.8
0.6
0.4
0.2
0.0
Scrambled Scrambled Luciferase Luciferase
AON siRNA AON siRNA
Fig. 7 Dual luciferase assay in K562 cells. Reduction in luminescence was compared for the same
AON and siRNA used in in vitro experiments. For both the scrambled controls and knockdown
experiments, 0.8 nmol of total oligonucleotide was nucleofected into K562 cells. Twenty four hour
postnucleofection, cells were lysed and the ratio of firefly to renilla luciferase was determined and
normalized to that of the samples with luciferase vectors alone. The mean of at least three
independent experiments is plotted (1 standard deviation)
Source: Rudnick et al, PNAS, 2008
i.e., a siRNA with more modest in vitro cleaving activity can be very efficient when
employed in vitro.
9 Discussion
a b
0 2.5 5 15 30 60 90 120min 0 2.5 5 15 30 60 90 120min
T T
c
0.5
5' Cleavage Product (pmol)
0.4
0.3
0.2
0.1
2.5µM siRNA
2.5µM AON
0.0
0 20 40 60 80 100 120
Time (min)
Fig. 8 Comparison of AON and siRNA activity in Drosophila embryo whole cell lysate. A 182 nt
50 cap labeled segment of firefly luciferase mRNA was incubated in Drosophila embryo whole cell
lysate with 2.5 mM AON (a) or 2.5 mM siRNA (b). Aliquots were removed at given time points and
analyzed on 10% sequencing gels where the target mRNA, 104 nt primary cleavage product, and
51 nt secondary cleavage product are indicated by T, PCP, and SCP, respectively. (c) Plot of 50
cleavage product formation as a function of time. The amount of cleavage product was calculated
by normalizing the signal to that of the target at the initial time point, and the mean of at least three
independent reactions is plotted (1 standard deviation). The dotted line at 0.22 pmol product
indicates single time points used in later experiments in order to compare AON and siRNA activity
against structure targets
Source: Rudnick et al, PNAS, 2008
binding (Fig. 8). Clearly, our results suggest that further PTGS mechanistic
studies are warranted.
Since RNA:RNA duplexes are known to have higher TM’s than DNA:RNA
duplexes, it is reasonable to think that this would give an advantage to the siRNA
mRNA Structure and its Effects on Posttranscriptional Gene Silencing 271
in binding the mRNA target. However in our in vitro cleavage assay conditions, the
AON and siRNA displayed the same apparent KD in the absence of their cognate
nuclease. When tested for cleavage activity, the AON cleavage activity reflected
the measured affinity. In other words, the AON generated significant product at the
concentration that should yield half of the mRNA bound, and half free. Consistent
with a simple two reagent binding equilibrium,
KD k
(1) AON + mRNA Duplex Fragments + AON
ð1Þ
RNase H
adding more AON generated more AON:mRNA duplex and an associated increase
cleavage product (Fig. 3). As illustrated in (1) below, it is hypothesized that the rate
of the forward reaction catalyzed by RNase H is dependent on duplex concentra-
tion. However, duplex formation is dependent on the binding affinity of the AON
for the mRNA. Furthermore, protein binding of the mRNA may provide as steric
barrier to AON hybridization. If so, this could certainly pose a mechanism of
inhibition that would block some, but not all AON.
While having the same affinity for the mRNA, the siRNA did not show the same
pattern of activity as the AON. Instead, every concentration tested between 0.06
and 2 the apparent KD gave the same rate of reaction and total product turnover.
Since the siRNA was most efficient at a concentration more than ten times lower
than the apparent KD measured, we hypothesize that RISC increases the binding
affinity of the siRNA for the target mRNA (Fig. 9). Since R2D2 and Dicer are
already known to interact with Ago2 and aid in loading the siRNA, as shown in (2),
perhaps they also aid in association with the mRNA.
Dicer, R2D2
siRNA + Ago2 siAgo2
(2) +
mRNA siAGo2 • mRNA siAGo2 + fragments
ð2Þ
D ic er, R 2D 2,or
un kn ow n protein
Since, as seen in Figs. 2 and 3, adding extra siRNA did not increase mRNA
degradation, we suspect that the in vitro RNAi machinery was saturated. Similar
observations have been reported in vivo and are thought to be from limiting
amounts of exportin-5 (Yi et al. 2003; Grimm et al. 2006). It is unlikely however
that exportin-5 activity is important in a WCL system, and it is probably Ago2
binding that is saturated in vitro.
Based on the in vitro results presented here, one would expect that AON and
RNase H can generally act more rapidly and efficaciously than siRNA when targeting
structured mRNA. Clearly, the body of evidence from in vivo studies does not support
this conclusion. In order for AON technology to be developed to its maximum
therapeutic potential, studies should be conducted to determine the true cellular
barrier to their activity. Such barriers may include poor cellular localization with
272 S.I. Rudnick et al.
0.030
0.025
Reaction Velocity (pmol /min)
0.020
0.015
0.010
0.005
0.000 AON
siRNA
–0.005
0 200 400 600 800 1000
Oligonucleotide Concentration (nM)
respect to mRNA and RNase H, or may be as simple as steric hindrance from proteins
bound to mRNA in the nucleus not encountered in the cytoplasm by RNAi.
Our work also demonstrates that siRNA activity was unexpectedly sensitive to
mRNA structure in vitro. Due to the high success and relative ease by which an
effective siRNA can be designed for in vivo studies, it is suggestive that there exists
an unidentified means of overcoming mRNA structure in the RNAi pathway. Since
many successful studies have been conducted in vitro with purified proteins, any
target recognition mechanism involved in RNAi, however, would not be fully
critical to the basic activity of the minimal RISC complex.
10 Conclusions
It is difficult not to see some irony in the fact that the first, and at least thus far, the only
clinically approved AON was delivered intraocularly for treatment of HIV-associated
CMV retinitis (Roehr 1998). Whether this antisense oligodeoxynucleotide worked by
hybridization with its target is uncertain but its efficacy is clear (Group TVS 2002). A
siRNA molecule targeting VEGF, which was also delivered by intraocular injection
and designed to treat macular degeneration, appeared well on its way to becoming the
first approved siRNA therapeutic but stalled for lack of clear cut efficacy and strong
indications that it worked not by inhibition of VEGF mRNA but by stimulation of
mRNA Structure and its Effects on Posttranscriptional Gene Silencing 273
Toll-like receptor 3 (Rossi et al. 2008; Yang et al. 2008). Still, the fact that nucleic acid
drugs continue to be developed shows that interest in this class of therapeutics remains
strong. Later generation of AON and siRNA are being tested in clinical trials of all
phases to treat a wide range of diseases (de Fougerolles et al. 2007; Graham et al. 2007;
Kamada et al. 2007; Prakash and Bhat 2007). Both technologies face similar, and by
now familiar, challenges of achieving specific, high efficiency gene knockdowns in
patients in the absence of toxic side effects. mRNA structure has long been thought to
act as a barrier to Watson–Crick base-pairing and heteroduplex formation (Mir and
Southern 1999), but most studies investigating this in PTGS applications never clearly
separate structure from other assay variables and rarely know the structures involved
in their target (Holen et al. 2002; Vickers et al. 2003). The work reported in this chapter
is a direct response to this challenge. Having said this, it is also important to state that
even the most perfectly targeted oligonucleotide will be of little use if it cannot be
delivered in a biologically relevant manner, meaning, not only introduced into cells
but also in a form that is or becomes bioavailable for hybridization. So, while
considerable work remains to be done, the prize is well worth the effort to those on
the hunt for effective oligonucleotide drugs.
References
Bartz SR, Zhang Z, Burchard J et al (2006) Small interfering RNA screens reveal enhanced
cisplatin cytotoxicity in tumor cells having both BRCA network and TP53 disruptions. Mol
Cell Biol 26:9377–9386
Berns K, Hijmans EM, Mullenders J et al (2004) A large-scale RNAi screen in human cells
identifies new components of the p53 pathway. Nature 428:431–437
Braasch DA, Jensen S, Liu Y et al (2003) RNA interference in mammalian cells by chemically-
modified RNA. Biochemistry 42:7967–7975
Brown KM, Chu CY, Rana TM (2005) Target accessibility dictates the potency of human RISC.
Nat Struct Mol Biol 12:469–470
de Fougerolles A, Vornlocher HP, Maraganore J et al (2007) Interfering with disease: a progress
report on siRNA-based therapeutics. Nat Rev Drug Discov 6:443–453
Ding H, Schwarz DS, Kenne A et al (2003) Selective silencing by RNAi of a dominant allele that
causes amyotrophic lateral sclerosis. Aging Cell 2:209–217
Elmen J, Thonberg H, Ljungberg K et al (2005) Locked nucleic acid (LNA) mediated improve-
ments in siRNA stability and functionality. Nucleic Acids Res 33:439–447
Gewirtz AM (2007) On future’s doorstep: RNA interference and the pharmacopia of tomorrow.
J Clin Invest 117:1–3
Graham MJ, Lemonidis KM, Whipple CP et al (2007) Antisense inhibition of proprotein
convertase subtilisin/kexin type 9 reduces serum LDL in hyperlipidemic mice. J Lipid Res
48:763–767
Grimm D, Streetz KL, Jopling CL et al (2006) Fatality in mice due to oversaturation of cellular
microRNA/short hairpin RNA pathways. Nature 441:537–541
Group TVS (2002) A randomized controlled clinical trial of intravitreous fomivirsen for
treatment of newly diagnosed peripheral cytomegalovirus retinitis in patients with AIDS.
Am J Ophthalmol 133:467–474
Holen T, Amarzguioui M, Wiiger MT et al (2002) Positional effects of short interfering RNAs
targeting the human coagulation trigger tissue factor. Nucleic Acids Res 30:1757–1766
274 S.I. Rudnick et al.
Contents
1 Antisense RNAs as Regulators of Gene Expression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
2 Regulation of p53 at the mRNA Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
3 Wrap53 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
4 Future Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
Abstract The tumor suppressor p53 triggers cell death by apoptosis in response to
cellular stress. p53 is regulated at the protein level by various posttranslational
modifications, such as phosphorylation and acetylation. However, recent studies
have revealed a critical regulation of p53 at the RNA level. A natural antisense
gene, designated Wrap53, is localized in a head-to-head fashion with p53 on human
chromosome 17p13. Wrap53 mRNA positively regulates steady-state levels of p53
mRNA and p53 protein by targeting the 50 untranslated region of p53 mRNA.
Knockdown of Wrap53 by siRNA results in a significant decrease in p53 mRNA
and suppression of p53 induction upon DNA damage, whereas overexpression of
Wrap53 transcripts containing the antisense overlap region enhances p53 mRNA
and protein levels and sensitizes cells to p53-dependent apoptosis. Antisense
transcription, which occurs widely in mammalian genomes, is thought to play an
important role in regulation of gene expression. Wrap53 antisense RNA is a novel
mechanism for controlling p53 activity and an interesting example of antisense-
mediated gene regulation in human cells.
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 277
RNA Technologies, DOI 10.1007/978-3-642-12168-5_12,
# Springer-Verlag Berlin Heidelberg 2010
278 M. Farnebo and K.G. Wiman
Natural antisense transcripts (NATs) are a group of regulatory RNAs with sequence
complementarity to other cellular RNAs referred to as sense RNAs. The existence
of antisense RNAs has been known for a long time, but their functional relevance
is still relatively unknown. Nevertheless, studies have shown that antisense
RNAs have the ability to modulate expression of their sense RNA. Considering
the widespread occurrence of antisense transcription in mammalian cells, this
mechanism may have a central role in gene regulation. Antisense RNAs can act
in cis or trans, depending on if the antisense RNA is transcribed from the same or a
distant locus with respect to the sense RNA. Around 20% of all human genes
overlap in a cis-antisense fashion, giving rise to cis-antisense RNAs with perfect
complementarity to their sense RNA (Chen et al. 2004; Yelin et al. 2003). The
orientation and length of the overlap varies between pairs. Most commonly, the
pairs overlap in a head-to-head or tail-to-tail orientation, reflecting overlap between
50 or 30 ends of both transcripts, respectively. However, complete overlap between
pairs is also found. In contrast, trans-antisense RNAs generally display imperfect
complementarity to its sense partner. One example of trans-antisense RNAs is
microRNAs.
The mechanisms of cis-antisense-mediated gene regulation are not fully under-
stood. Several modes of action have been proposed:
1. Regulation at the transcriptional level independently of the antisense transcript.
In this model, transcription of the sense RNA is repressed due to a competition
between the sense and antisense genes for transcription factors. Also, RNA
polymerases might collide if both genes are transcribed simultaneously, result-
ing in transcription termination.
2. Epigenetic alterations induced by the antisense transcript. Here, the antisense
transcript regulates expression of the sense RNA through epigenetic alterations
of the sense promoter region, such as DNA methylation or chromatin remodel-
ing. These processes shut down expression of the sense transcript. This type of
regulation has been found to control inactivation of the X chromosome and
genomic imprinting.
3. RNA–RNA interaction between the sense and antisense transcripts. This mode
of action occurs at the posttranscriptional level and involves base-paring
between the sense and antisense RNAs. RNA–RNA interaction can activate
the RNA interference pathway resulting in sense RNA degradation and endo-
siRNA formation. Alternatively, translation of the sense transcript is blocked.
Conversely, RNA–RNA interaction or duplex formation has been found to
Antisense RNA-Mediated Regulation of the p53 Tumor Suppressor 279
enhance sense RNA stability, possibly through masking sequences within the
sense transcript that otherwise could be recognized by RNA degradation factors.
Transcriptome studies show that antisense transcription occurs for up to 40–70%
of all mammalian genes (Katayama et al. 2005; Engstrom et al. 2006). However,
only a minor fraction of all putative sense–antisense pairs have been verified
and even fewer have been characterized. The identification of Wrap53 does not
only reveal a regulatory pathway for p53 but also demonstrates that RNA–RNA
interaction is a mechanism for antisense-mediated gene regulation in human cells
(Mahmoudi et al. 2009).
The p53 tumor suppressor triggers cell cycle arrest, senescence, or apoptosis in
response to DNA damage, oncogene activation, and other stress stimuli (Vogelstein
et al. 2000). This allows elimination of incipient tumor cells. p53 is a transcription
factor that upregulates apoptosis-promoting genes such as Bax, Puma, and Noxa.
Several studies have also highlighted the miR-34a microRNA as a p53 target
(Raver-Shapira and Oren 2007). p53 mutation occurs frequently in human tumors,
resulting in evasion of apoptosis and progression towards more malignant pheno-
types. Most p53 mutations give rise to single amino acid substitutions in the DNA-
binding core domain, resulting in disruption of DNA binding and transcriptional
transactivation of target genes (Soussi and Wiman 2007). Accumulating evidence
suggests that mutant p53 proteins can also acquire novel functions that contribute
to tumor growth, e.g., illegitimate transactivation of tumor-promoting genes such
as c-myc.
Under normal conditions, p53 is expressed at very low levels due to rapid protein
turnover. Cellular stress leads to p53 protein stabilization via phosphorylation that
prevents binding to the p53-induced E3 ligase MDM2 that targets p53 for protea-
some-mediated degradation. However, studies on knock-in mice expressing mutant
p53 proteins with substitutions of residues that are phosphorylated upon stress, e.g.,
Ser18 (corresponding to Ser15 in human p53) and Ser23 (Ser20 in human p53),
have indicated that each of these posttranslational modifications by themselves do
not have any significant effect on either basal p53 levels or induction of p53 upon
cellular stress (Zhang and Chen 2008). This suggests that regulation of p53 is
complex and that multiple mechanisms contribute to modulation of p53 expression
in vivo. Other posttranslational modifications, such as acetylation and sumoylation,
are also important. Interestingly, recent studies have revealed that p53 is regulated
at the RNA level. RNA-binding proteins, such as HuR, L26, RPL26, and nucleolin,
can regulate p53 activity by binding to the 50 or 30 UTR of p53 mRNA and affect
mRNA stability and/or translation (Zhang and Chen 2008). Moreover, the miR-125b
microRNA was shown to target the 30 UTR of p53 and modulate p53-induced
apoptosis during development and stress conditions (Le et al. 2009). Our discovery
280 M. Farnebo and K.G. Wiman
of Wrap53 adds another level of complexity as to the role of regulatory RNAs for
p53 function (Mahmoudi et al. 2009).
3 Wrap53
Wrap53 is a natural antisense transcript of p53 that regulates basal and stress-
induced endogenous p53 mRNA and protein expression by targeting the 50 untrans-
lated region of p53 mRNA. Our discovery of Wrap53 is the first demonstration of
antisense-mediated regulation of p53. We baptized this gene Wrap53 for WD40
encoding RNA antisense to p53, a name that has been approved by the HUGO Gene
Nomenclature Committee as the official name of this gene. Wrap53 is located on
chromosome 17p13 and overlaps the p53 gene in a head-to-head fashion (Fig. 1).
The gene has three alternative start exons, exon 1a, 1b, and 1g. Exon 1a overlaps
the first exon of p53 in an antisense fashion. Wrap53 also encodes a protein with
homology to members of the WD40 protein family, thus the name Wrap53. This is
in contrast to many other regulatory RNAs that are noncoding and exert their effect
only at the RNA level. The Wrap53 protein (also denoted WDR79 and TCAB1)
was recently identified as a Cajal body protein that binds and directs small
Cajal body-specific RNAs (scaRNAs), including telomerase RNA, to Cajal bodies
(Venteicher et al. 2009; Tycowski et al. 2009).
p53
Chromosome - Strand
17p13.1 + Strand
Wrap53
Transcription
p53 mRNA
degradation Wrap53/p53 RNA interaction
protects p53 mRNA from
degradation
p53
protein
protein
Fig. 1 Model for Wrap53-mediated regulation of p53. Wrap53a and p53 are coexpressed in cells.
Interaction between the two transcripts via their complementary regions (IIIIII) protects p53
mRNA from degradation. Knockdown of Wrap53a or blockage of Wrap53/p53 hybrid formation
leads to reduced p53 mRNA levels. Conversely, overexpression of Wrap53a increases p53 mRNA
levels, and consequently, p53 protein expression. Thus, Wrap53a will potentiate p53-induced cell
cycle arrest and/or apoptosis in response to cellular stress
Antisense RNA-Mediated Regulation of the p53 Tumor Suppressor 281
4 Future Perspectives
Loss of p53 function is a key step during tumor development, allowing evasion
of apoptosis and accelerated tumor progression. The control of p53 is complex with
tight regulation at both the posttranscriptional and posttranslational levels. The
identification of Wrap53 as a novel regulator of p53 adds important new insights
into the impact of regulatory RNAs on p53 activity. It also opens the possibility
that dysfunction of Wrap53 itself may contribute to cancer. Could lack of Wrap53a
RNA, or insufficient levels of expression, represent a novel mechanism for p53
inactivation in wild type p53-carrying tumors? Future studies should examine
expression of Wrap53 transcripts in primary tumors of various origin. Moreover,
it will be important to determine if manipulation of Wrap53 expression to either
enhance wild type p53 levels or inhibit expression of mutant p53 and putative
gain-of-function activities may be a useful strategy for cancer treatment.
Acknowledgements We thank the Swedish Cancer Society (Cancerfonden), the Swedish Childhood
Cancer Society (Barncancerfonden), and the King Gustaf V Jubilee Fund for generous financial
support.
References
Balzer M, Wagner R (1998) Mutations in the leader region of ribosomal RNA operons cause
structurally defective 30 S ribosomes as revealed by in vivo structural probing. J Mol Biol
276:547–557
Besancon W, Wagner R (1999) Characterization of transient RNA–RNA interactions important
for the facilitated structure formation of bacterial ribosomal 16S RNA. Nucleic Acids Res
27:4353–4362
Borsani O, Zhu J, Verslues PE et al (2005) Endogenous siRNAs derived from a pair of natural
cis-antisense transcripts regulate salt tolerance in Arabidopsis. Cell 123:1279–1291
284 M. Farnebo and K.G. Wiman
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
2 Antisense Oligonucleotides in Cancer Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
2.1 BCL2 (B-cell CLL/lymphoma 2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
2.2 XIAP (X-Linked Inhibitor of Apoptosis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
2.3 Survivin, BIRC5 (Baculoviral IAP Repeat-Containing 5) . . . . . . . . . . . . . . . . . . . . . . . . . . 291
2.4 CLU (Clusterin) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
2.5 TGFB2 (Transforming Growth Factor, b 2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
3 Application of Antisense Oligonucleotides in Noncancerous Diseases . . . . . . . . . . . . . . . . . . . 294
3.1 Asthma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
3.2 Cardiovascular Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
3.3 Duchenne Muscular Dystrophy – Exon-Skipping Therapy . . . . . . . . . . . . . . . . . . . . . . . . . 295
3.4 Virus Infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
4 Specificity of Antisense-Mediated Gene Silencing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 285
RNA Technologies, DOI 10.1007/978-3-642-12168-5_13,
# Springer-Verlag Berlin Heidelberg 2010
286 D. Kunze et al.
Abbreviations
RNA polymerase
DNA STOP
a DNA
Transcription ASO
RNA Triplex formation
Pre-mRNA
Post-transcriptional
modification Impairment of mRNA
mRNA processing b
RNase H activation
Ribosome
STOP mRNA c
d Steric blockade
Protein
Fig. 1 Mechanism of ASO action: (a) interference with transcription via triplex formation with
complementary DNA; (b) inhibition of splicing or destabilization of mRNA by hindrance of
polyadenylation and 50 -capping after hybridization to pre-mRNA; (c) mRNA degradation after
induction of RNase H activity; (d) translational arrest following binding of the ASO in the region
of the initiation codon of the target mRNA and sterical blocking of the ribosome
1 Introduction
O B O B
1st
O H O H
– –
O P O O P S
O O
Unmodified 2nd PS
O B O B
3rd O O CH3 O O
– – O CH3
O P O O P O
O O
2'-OMe 2'-MOE
O B
O B
B N
N
O O O
O
– O P N
O P O O NH
O O
Fig. 2 Examples for the three generations of chemically modified ASOs: phosphorothioate (PS)
backbone; 20 -O-methyl-(20 -OMe)- and 20 -O-methoxyethyl-(20 -MOE)-RNA-substitutions; locked
nucleic acid (LNA), peptide nucleic acid (PNA), and phosphoroamidate morpholino (PMO)
modifications. B – bases (adenine, thymine, guanine, cytosine)
The main field of application for ASOs is the treatment of diseases that involve
the overexpression of a detrimental gene, e.g., tumor growth or viral infections.
Additionally, oligonucleotides containing unmethylated cytosine–guanine dinucle-
otide (CpG) motifs can stimulate the mammalian immune system (Vollmer and Krieg
2009). Hence, they are tested as vaccine adjuvants for cancer, asthma, and allergies.
Since the discovery of ASOs in the late 1970s, numerous in vitro and in vivo studies
have been performed showing the possibilities and limitations of these constructs.
In 1998, fomivirsen, a 21-mer PS–ASO, inhibiting viral proteins from the major
immediate early transcriptional unit, received FDA approval for treatment of
cytomegalovirus-induced retinitis in AIDS patients. Up to now, no other ASO has
reached the clinic although promising candidates are in development.
At present, there are 16 phase I/II and 6 phase III clinical trials with G3139,
either alone or in combination with various anticancer agents, in hematologic
malignancies and solid tumors ongoing (www.clinicaltrials.gov)1. In 2000, the
first phase III study started comparing G3139 plus dacarbazine versus dacarbazine
alone in 771 patients with advanced melanoma (Bedikian et al. 2006). Median
overall survival in the combination group was longer (9.0 vs. 7.8 months) but did
not reach statistical significance. However, the study showed a correlation between
pretreatment serum lactate dehydrogenase (LDH) level and outcome after G3139
therapy. Survival and overall response significantly improved in the G3139 group
in participants with normal LDH, whereas patients with elevated LDH did not
benefit from G3139 treatment. Therefore, AGENDA, a second randomized phase
III trial in chemo-naı̈ve patients with advanced melanoma and low baseline LDH
was initiated comparing dacarbazine G3139 treatment (www.clinicaltrials.gov).
Altogether, 241 patients with relapsed or refractory chronic lymphocytic leukemia
(CLL) were enrolled in a phase III study with fludarabine plus cyclophosphamide
G3139 (O’Brien et al. 2007). Response rate (complete plus nodular partial
response) was significantly higher (17% vs. 7%) and longer in the G3139 combina-
tion arm. In a phase III study enrolling 503 patients with acute myeloid leukemia
(AML), the addition of G3139 to chemotherapy with cytarabine and daunorubicin
did not change patients’ outcome (Marcucci et al. 2007). Likewise, no significant
difference in time to tumor progression was found after dexamethasone G3139
treatment in a phase III trial involving 224 patients with relapsed or refractory
multiple myeloma (Chanan-Khan et al. 2009). Up to now, no results regarding the
phase III trial examining G3139 in combination with docetaxel in patients with
nonsmall cell lung cancer (NSCLC) have been published.
Most important adverse events reported from the phase III trials so far are grade
3 and 4 neutropenia and thrombocytopenia (Bedikian et al. 2006; O’Brien et al.
2007). Furthermore, catheter-related problems, which are due to the continuous
intravenous infusion of the PS–ASO, increased in the G3139 groups. Alternative
administration routes like short intravenous infusions or subcutaneous injections
might produce relief (Lin et al. 2007; www.genta.com). Safety and maximum
tolerated dose of G3139 administered once or twice per week as 2 h intravenous
infusion are currently examined in a phase I clinical trial in patients with solid
tumors (www.clinicaltrials.gov).
SPC2996, an LNA-modified ASO directed at BCL2, is clinically tested in CLL
(www.santaris.com). In the first dose-escalating study, all six patients in the treat-
ment group with maximum drug concentration (4 mg/kg/dose) showed a reduction
in lymphocyte count (Tilly et al. 2007). A second phase I/II clinical trial examining
new dosing regimens is completed (www.santaris.com). So far, however, no results
have been published.
1
All websites mentioned in the text were viewed on July 15th, 2009.
Antisense Oligonucleotides: Insights from Preclinical Studies and Clinical Trials 291
Up to now, approximately 300 patients have been treated with OGX-011 in six
phase I and II clinical trials (Chi et al. 2008). In the first phase I dose-escalation
study, 25 patients with high-risk localized PCa were treated with 40–640 mg
OGX-011 given seven times as a 2 h intravenous infusion prior to radical prosta-
tectomy. Besides the safety of this ASO, the biological activity was demonstrated,
too. The study showed that OGX-011 treatment is well tolerated, with adverse
events limited to grade 1 or 2 (e.g., leucopenia, thrombocytopenia, fever, fatigue)
(Chi et al. 2005). Furthermore, a statistically significant dose-dependent reduction
of CLU expression in PCa and lymph node tissue was detected after OGX-011
treatment in comparison to control tissues taken from a tumor bank.
In ongoing clinical studies, OGX-011 is applied in combination with docetaxel,
mitoxantrone, or gemcitabine–cisplatin in patients with prostate, lung, or breast
cancer (www.clinicaltrials.gov). Preliminary results indicate that combined treat-
ment with OGX-011 plus standard chemotherapy is well tolerated (reviewed in Chi
et al. 2008). The mostly single-armed study design hampers the interpretation of
response rates and survival data. However, in a randomized phase II trial, encour-
aging improvement of median survival (median follow-up 32 months) was seen in
82 patients with metastatic castration-resistant PCa who have been treated with
docetaxel þ prednisone þ OGX-011 (27.5 months) compared to docetaxel þ pred-
nisone alone (16.9 months), though the primary endpoint, 50% PSA decline from
baseline, was achieved in both treatment arms (Chi et al. 2009; www.oncogenex.
com). Currently, two randomized phase III trials with OGX-011 plus docetaxel with
overall survival or durable pain palliation as primary end points in patients with
castrate-resistant PCa are in preparation (http://www.isispharm.com).
3.1 Asthma
TPI ASM8 is a drug for asthma treatment, which consists of two PS–ASOs targeted
at human chemokine receptor 3 (CCR3) and the common b-chain of IL-3, IL-5, and
GM-CSF receptors, respectively. A study in monkeys showed safety and tolerability
of TPI ASM8 after 14 days of inhalation of up to 2.5 mg/kg/day (Guimond et al.
2008). ASOs were mainly localized in the pulmonary tract with only limited
distribution in plasma, liver, and kidney.
In humans with mild atopic asthma, inhalation of 1.5 mg TPI ASM8 on four
consecutive days reduced influx of sputum inflammatory cells by 46% (Gauvreau
et al. 2008). Furthermore, TPI ASM8 treatment decreased the early asthmatic
response and mRNA levels of allergen-induced targets in sputum-derived cells
without serious adverse events. A new phase II study examining efficacy and safety
of escalating dose regimens of TPI ASM8 in patients with allergic asthma has been
started in April 2009 (www.clinicaltrials.gov).
structural protein of low density lipoprotein cholesterol (LDL-c) (Yu et al. 2007).
The inhibition of elevated LDL-c levels in patients is an attractive approach, since
high LDL-c is a risk factor for atherosclerosis. In particular, ASO-mediated apoB-
100 inhibition seems to be promising, because apoB-100 is synthesized in the liver,
the organ where ISIS 301012 predominantly concentrates after intravenous or
subcutaneous application (Yu et al. 2007).
In a phase I study with 36 healthy volunteers, a dose-dependent reduction of
serum apoB-100 and LDL-c was measured after injections of 50–400 mg ISIS
301012 (Kastelein et al. 2006). No serious adverse events were reported. Similar
results were obtained in patients with high levels of cholesterol in the blood
(reviewed in Yu et al. 2009). Mipomersen can be safely administered in combina-
tion with other LDL-c lowering drugs like simvastatin or ezetemibe (Yu et al.
2009). It is currently examined in three phase II and four phase III clinical trials
(www.clinicaltrials.gov). Recently, a placebo-controlled phase III clinical trial in
51 patients with homozygous familial hypercholesterolemia (FH), a genetic dis-
order characterized by high LDL-c level leading to increased risk of cardiovascular
disease, met its primary and secondary endpoints (www.isispharm.com, press
release 05/20/09). FH patients, being on standard lipid-lowering therapy, were
treated weekly with 200 mg ISIS 301012 for 26 weeks. LDL-c, apoB-100, and
total cholesterol were significantly decreased. Reported adverse events were injec-
tion site reactions, flu-like symptoms, and elevations in liver transaminases.
injected weekly or biweekly for 5–22 weeks. Two weeks after treatment, increased
dystrophin expression and reduced inflammatory signals were detected.
AVI-4658 is a PMO designed to skip exon 51 of the DMD gene. It is currently
examined in two phase I/II clinical trials (www.clinicaltrials.gov). First results
demonstrated that injection of 0.09 or 0.9 mg AVI-4658 into the exterior digitorum
brevis muscle has been well tolerated and has significantly increased dystrophin
synthesis in DMD patients (www.avibio.com, press release 01/21/09). The second
trial, which was started in January 2009, will evaluate effects of intravenous
drug application. Furthermore, AVI-5038, a PPMO for skipping of exon 50, is in
preclinical development (www.avibio.com).
PRO051 is a 20 OMe-PS–ASO mediating exon 51 skipping. In a first study, four
DMD patients received a single injection of 0.8 mg PRO051 into the tibialis anterior
muscle (van Deutekom et al. 2007). Twenty-eight days later, biopsy samples from
the patients were obtained showing induction of dystrophin expression.
Currently, only few antiviral ASOs are in stage of development. Three PMO–ASOs
targeted at VP35, VP24, and RNA polymerase L mRNAs of Ebola Zaire virus,
respectively, showed as single agents and in combination, profound virus inhibition
in mice (Warfield et al. 2006). In rhesus macaques, only the combination was
effective and protected three out of four primates from lethal Ebola virus dose.
ASOs against VP35 and VP24 are now tested in the drug AVI-6002. AVI-6003,
also based on PMO chemistry, is analyzed in the treatment of Marburg virus. It
prevented 100% of treated monkeys from virus related death (www.avibio.com).
The number of treated animals, however, is not mentioned.
A novel approach in antisense therapy is the inhibition of microRNAs (miRNAs),
small noncoding RNAs that posttranscriptionally regulate expression of estimated
30% of the protein-coding genes (reviewed in Wiemer 2007). MicroRNA-122
(miR-122) is a liver-specific miRNA that is implicated in the replication of hepatitis
C virus (reviewed in Niepmann 2009). Mice were treated on three consecutive
days with 2.5–25 mg/kg of a 16-mer LNA-modified ASO targeted at miR-122.
Intravenous ASO injection reduced miR-122 expression in murine livers in a
concentration-dependent manner (Elmén et al. 2007). Effects were strongest
24 h after treatment and disappeared after 2 weeks. No hepatotoxicity was induced.
Further studies in monkeys showed safety, drug uptake in hepatocytes, and
consequent miR-122 decrease after systemic treatment with 3 or 10 mg/kg ASO
(Elmén et al. 2008). In May 2008, a clinical phase I study in healthy men was
started to evaluate the safety of SPC3649, an LNA targeted at miR-122 (www.
clinicaltrials.gov). A phase II study examining SPC3649 in patients with hepatitis
C is in planning stages (www.santaris.com).
Antisense Oligonucleotides: Insights from Preclinical Studies and Clinical Trials 297
5 Conclusion
References
Wang Y, Liu X, Chen L et al (2009) Tumor delivery of antisense oligomer using trastuzumab within
a streptavidin nanoparticle. Eur J Nucl Med Mol Imaging. doi:10.1007/s00259-009-1201-2
Warfield KL, Swenson DL, Olinger GG et al (2006) Gene-specific countermeasures against Ebola
virus based on antisense phosphorodiamidate morpholino oligomers. PLoS Pathog 2:e1
Wasan EK, Waterhouse D, Sivak O et al (2002) Plasma protein binding, lipoprotein distribution
and uptake of free and lipid-associated BCL-2 antisense oligodeoxynucleotides (G3139) in
human melanoma cells. Int J Pharm 241:57–64
Wiemer EA (2007) The role of microRNAs in cancer: no small matter. Eur J Cancer
43:1529–1544
Yokota T, Lu QL, Partridge T et al (2009) Efficacy of systemic morpholino exon-skipping in
duchenne dystrophy dogs. Ann Neurol 65:667–676
Yu RZ, Kim TW, Hong A et al (2007) Cross-species pharmacokinetic comparison from mouse
to man of a second-generation antisense oligonucleotide, ISIS 301012, targeting human
apolipoprotein B-100. Drug Metab Dispos 35:460–468
Yu RZ, Geary RS, Flaim JD et al (2009) Lack of pharmacokinetic interaction of mipomersen
sodium (ISIS 301012), a 20 -O-methoxyethyl modified antisense oligonucleotide targeting
apolipoprotein B-100 messenger RNA, with simvastatin and ezetimibe. Clin Pharmacokinet
48:39–50
Zellweger T, Miyake H, Cooper S (2001) Antitumor activity of antisense clusterin oligonucleotides
is improved in vitro and in vivo by incorporation of 20 -O-(2-methoxy)ethyl chemistry.
J Pharmacol Exp Ther 298:934–940
Zellweger T, Chi K, Miyake H et al (2002) Enhanced radiation sensitivity in prostate cancer by
inhibition of the cell survival protein clusterin. Clin Cancer Res 8:3276–3284
Zhang Q, Zhou W, Kundu S et al (2006) The leader sequence triggers and enhances several
functions of clusterin and is instrumental in the progression of human prostate cancer in vivo
and in vitro. BJU Int 98:452–460
What can the New Hammerhead Ribozyme
Structures Teach us About Design?
William G. Scott
Contents
1 Introduction to the Hammerhead Ribozyme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
1.1 The Genomic Ribozymes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
1.2 What is a Hammerhead Ribozyme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
1.3 Minimal and Full-Length Hammerhead Ribozymes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
1.4 Expanding Biological Context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
2 Hammerhead Ribozyme Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
2.1 Three-Dimensional Structure of Minimal Hammerhead Ribozymes . . . . . . . . . . . . . . . 312
2.2 Three-Dimensional Structures of Full-Length Hammerhead Ribozymes . . . . . . . . . . . 313
3 Structure and Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
3.1 Acid–Base Catalysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
3.2 Metal Ions? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 318
3.3 Substrate Binding and Specificity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
4 Hammerhead Structure, Function, and Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
4.1 Minimal Hammerheads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
4.2 Full-Length Hammerheads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
W.G. Scott
Department of Chemistry and Biochemistry and The Center for the Molecular Biology of RNA,
University of California at Santa Cruz, 228 Sinsheimer Laboratories, Santa Cruz, CA 95064, USA
e-mail: wgscott@chemistry.ucsc.edu
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 305
RNA Technologies, DOI 10.1007/978-3-642-12168-5_14,
# Springer-Verlag Berlin Heidelberg 2010
306 W.G. Scott
have taken place since 2003, rendering previous assumptions about therapeutic
hammerhead ribozyme design largely obsolete. The requirement for a tertiary
contact between Stems I and II to be present in order to achieve a highly active
ribozyme in vivo is described, and design requirements that enable straightforward
incorporation of the tertiary contact are explicitly described. This analysis is only
possible with crystal structures of two classes of full-length natural hammerhead
ribozymes that became available in 2006 and 2008.
Prior to the 1980s, all enzymes were thought to be proteins. RNA was thought to
play a mostly subservient role in cellular biochemistry. tRNAs were merely
adapter molecules that were employed by the ribosome translational apparatus
to read the genomic message, immortalized in DNA, from an intermediary mRNA
transcript, and translate it into the protein sequence corresponding to the DNA
sequence. The ribosome itself was recognized to be an RNA–protein complex, but
conventional wisdom suggested that the ribosomal rRNA was merely scaffolding
that enabled the required collection of ribosomal proteins to assemble. RNA
viruses were considered a rare exception to the Central Dogma of Molecular
Biology, and genetic regulation via RNA interference mechanisms would remain
unimagined for decades. Every step of each of a myriad of biochemical reactions
that comprised the complex entangled web of metabolic pathways was catalyzed
by an enzyme, as were DNA replication and transcription. These enzymes were
always proteins.
The discovery that RNA, like proteins, also can have catalytic activity was
therefore a complete surprise. The Group I intron was shown to have self-splicing
activity in the absence of protein cofactors (Kruger et al. 1982), and the RNA
subunit of the RNA–protein complex enzyme RNase P was shown to be the
catalytic subunit of this precursor-tRNA processing enzyme (Guerrier-Takada
et al. 1983). Both of these ribozymes were comprised of RNA sequences that
were several hundred nucleotides in length and were believed (correctly) to have
rather complex secondary and tertiary structures and catalytic mechanisms. The
third ribozyme to be discovered was rather more simple and compact; it was the
hammerhead self-cleaving motif found in the genome of the satellite RNA of
tobacco ringspot virus (Prody et al. 1986). Subsequently, several other self-splicing
and self-cleaving RNAs have been discovered (Fedor 2009), and the most profound
discovery, both in terms of cellular biochemistry and evolutionary biology, was the
realization that the peptidyl transferase activity within the ribosome is a ribozyme
(Noller et al. 1992; Steitz and Moore 2003).
What can the New Hammerhead Ribozyme Structures Teach us About Design? 307
The hammerhead self-cleaving RNA was the first of several self-cleaving RNAs, or
ribozymes, to be found in the context of RNA genomes (Cochrane and Strobel
2008). The hammerhead motif, first discovered in the satellite RNA of tobacco
ringspot virus (Prody et al. 1986), has been found in a number of other plant satellite
virus RNAs, viroid RNAs, and related genomic elements. In every case, the ham-
merhead RNA is involved in the rolling circle replicative mechanism of RNA
genome replication (Fig. 1). In addition, several other self-cleaving ribozymes
with different sequences have subsequently been discovered, and all catalyze the
same chemical reaction in the same sort of biological context. This self-cleavage
reaction is not a hydrolysis reaction but rather a phosphodiester isomerization
reaction, wherein nucleophilic attack of the 20 -OH upon the adjacent phosphate
results in backbone cleavage, leaving 20 ,30 -cyclic phosphate and 50 -OH termini
(Fig. 2).
The hammerhead RNA sequence within satellite RNA genomes occurs at the
interface of two monomeric segments of a linear concatamer following rolling-
circle replication (Fig. 1). Although it is, in that context, a single self-cleaving
strand of RNA that is capable of catalyzing only a single, albeit highly specific,
cleavage reaction, the hammerhead RNA can be artificially engineered to create a
true multiple-turnover ribozyme simply by separating the molecule into discrete
enzyme and substrate strands. The latter constructs are typically studied in vitro and
also correspond to hammerhead ribozyme sequences that have been used for
targeting other RNAs. Minimal hammerhead ribozymes have typical Km values
of 10 mm, and turnover rates of about 1 substrate molecule/minute, whereas full-
length hammerhead ribozymes have a similar Km but may be 1,000-fold faster.
Soon after the discovery of the hammerhead self-cleaving motif, the minimal
sequence required for self-cleavage activity was identified (Uhlenbeck 1987; Haseloff
and Gerlach 1989). The minimal hammerhead sequence consists of a central core
region of 13 mostly invariant nucleotides flanked by three A-form Watson–Crick
base-paired helical sequences whose detailed sequence is comparatively less impor-
tant (Fig. 3a). The highly conserved central region for the most part is incapable of
forming canonical Watson–Crick base-pairs and was identified as likely giving rise to
a tertiary structure that enabled the RNA to possess catalytic activity.
308 W.G. Scott
Self-cleavage reactions
N-N
Minimal N-N Phosphodiester bond
Hammerhead N-N isomerization
self-cleaving N-N
RNA sequence C-G
A-U
A
A C
NNNNNN G NNNN
NNNNNN A NNNN
C
G U
U AG
O O
G12
O O
N N O G8 N
NH2 N O
NH2
N G H+
N O NH2 N G NH2
H+ N O
H H
–O N O N
O O– N H -O N
H
O O P O O O P O
O
O O O
O -O O
OH OH
N N
O O N N O O
N1.1
H2N
O H2N O
C17
Enzyme-Substrate Complex Transition-State
O
N O
N NH2
N G NH2
HO
N
H
O -O H N
O N
P O
O O
O O
O
OH
N N O O
H2N O
Enzyme-Product Complex
Martick and Scott 2006; Chi et al. 2008), the first of any ribozyme, have been
determined. The crystal structures were only capable of reconciling a subset of
the biochemical experiments designed to probe the catalytic mechanism, and
considerable discord plagued the hammerhead ribozyme biochemical community
(Blount and Uhlenbeck 2005).
All ribozymes, including the hammerhead ribozyme, were originally believed to
be metalloenzymes (Pyle 1993), requiring an obligate Mg2+ for catalysis (Dahm
and Uhlenbeck 1991; Dahm et al. 1993; Peracchi et al. 1997). It has subsequently
been revealed, however, that the hammerhead, in addition to other small self-
cleaving ribozymes, does not strictly require divalent cations for catalysis. Instead,
if a sufficiently high concentration of even nonmetallic monovalent salt is present,
permitting the RNA to fold correctly, it will remain catalytically active, even in a
310 W.G. Scott
Fig. 3 A schematic secondary structure of (a) the minimal and (b) the full-length hammerhead
ribozyme. The conserved residues in the catalytic core are shown explicitly in each case, and the
cleavage site is indicated with a red arrow. The tertiary contact in (b) is indicated in the grey
portion of the schematic diagram. This figure was kindly supplied by Christian Hammann
high concentration of EDTA (Murray et al. 1998a, b). Hence it appeared that the
RNA itself, rather than functioning as a passive scaffold to bind metal ions, instead
must be an active participant in its own chemical catalysis (Scott 1999). This
renewed focus upon the RNA structure itself. However, the crystal structures of
the minimal hammerhead could not be reconciled with this conclusion; none of the
invariant residues were positioned in a way that made their role in catalysis at all
obvious, and the substrate itself was not bound to the enzyme in a way that would
permit the known required in-line attack geometry to be stabilized, as one would
expect from an enzyme (McKay 1996; Blount and Uhlenbeck 2005).
In 2003, two papers appeared (De la Peña et al. 2003; Khvorova et al. 2003) that
had essentially the same conclusion: the minimal hammerhead construct lacked a
tertiary contact between helices I and II that had a rather profound effect upon
hammerhead ribozyme catalysis, despite being distant from the cleavage site
(Fig. 3b). This contact appeared to have little if any sequence conservation between
the large number of natural hammerhead ribozyme sequences that had been identi-
fied to date and had thus escaped notice. However, when the tertiary contact
sequences were included, these natural, full-length hammerheads were observed
to be up to 1,000-fold more active than their minimal counterparts (Khvorova et al.
2003; Canny et al. 2004). Clearly, the tertiary contact imparted some change at the
active site that stimulated catalytic activity.
Fig. 4 The hammerhead ribozyme embedded within the 30 -UTR of the clec2d mRNA transcript,
immediately downstream from the stop codon. The “enzyme” part of the strand is highlighted in
blue, and the “substrate” is highlighted in orange, with the cleavage site indicated
The first structures of an RNA appeared in 1974 with the publication of the yeast
tRNAPhe crystal structures (Robertus et al. 1974; Kim et al. 1974). These structures
revealed that RNA possesses the propensity to fold into comparatively compact
globular protein-like three-dimensional structures, and it also illustrated the impor-
tance of tertiary contacts in stabilizing complex RNA backbone folds (Klug et al.
1974). Another 20 years passed before another complex RNA structure emerged. In
1994, the first structure of a ribozyme was published by McKay and coworkers
(Pley et al. 1994); it consisted of a minimal hammerhead enzyme strand hybridized
with a DNA substrate analog. A second structure of a minimal hammerhead, this
time with an all-RNA substrate (Scott et al. 1995), appeared shortly thereafter,
corroborating the initial structural work.
312 W.G. Scott
The resolution of this vexing conundrum finally came in 2006 with a 2.2 Å resolu-
tion crystal structure of the full-length hammerhead ribozyme from Schistosoma
314 W.G. Scott
mansoni (Fig. 6). C17 is now positioned for in-line attack, and the invariant residues
C3, G5, G8, and G12 all appear involved in vital interactions relevant to catalysis.
Moreover, the A9 and scissile phosphates are observed to be 4.3 Å apart, which is
consistent with the idea that, when modified, these phosphates could bind a single
thiophilic metal ion. The structure also reveals how two invariant residues, G-12 and
G-8, are positioned within the active site – consistent with their previously proposed
role in acid–base catalysis. G12 is within hydrogen bonding distance to the 20 -O of
C17, the nucleophile in the cleavage reaction, and the ribose of G8 hydrogen bonds
to the leaving group 50 -O, while the nucleotide base of G8 forms a Watson–Crick
pair with the invariant C3. This arrangement suggests that G12 is the general base in
the cleavage reaction, and that the G8 ribose may function as the general acid
(Fig. 7). The crystal structure of the full-length hammerhead ribozyme thus clearly
addresses the major concerns that appeared irreconcilable with the previous crystal
structures (Nelson and Uhlenbeck 2006, 2008).
In addition to the rearranged cleavage site, one of the most prominent features of
the full-length hammerhead ribozyme structure is the Stem II loop/Stem I bulge
interaction that appears to induce the structural organization of the catalytic core.
The loop/bulge interaction is composed of an intricate network of interhelical
What can the New Hammerhead Ribozyme Structures Teach us About Design? 315
noncanonical base pairs and stacks interdigitating Stem II loop into Stem I, kinking
Stem I in such a way as to coaxially align its distal helix on top of the Stem II–Stem
III coaxial arm. The tertiary contacts between the loop and bulge regions induce
structural changes affecting the catalytic core, specifically via the relative under-
winding of Stem I. This interaction imparts a severe bend in the distal part of the
Stem I helix and a pronounced kink in the backbone of the substrate strand at the
cleavage site. These distortions appear to accommodate G-8 and U-7 in the catalytic
pocket and in turn stabilize the rearrangement of the augmented Stem II helix that
enables G-8 to form the Watson–Crick base pair with C-3 in the catalytic pocket.
Concurrently, an overwinding or right-handed twist of Stem II positions the con-
served G-12, A-13, and A-14 precisely against the catalytic-site C-17, helping to
lock the latter in a catalytically active conformation in which C-17 is oriented for
in-line attack (Martick and Scott 2006).
The Satellite RNAs all possess a different type of tertiary contact compared to the
Schistosomal hammerhead. In 2008, crystal structures of an unmodified hammer-
head ribozyme derived from the satellite RNA of tobacco ringspot virus (sTRSV)
was published (Chi et al. 2008), permitting comparison of the two types of ham-
merhead (Scott et al. 2009). Briefly, the active site is nearly identical in both types
of hammerhead ribozyme, but the tertiary contacts, though imparting the same net
structural effect upon the active site, are distinctly different in the two cases.
The two classes of tertiary contacts are shown as secondary structural represen-
tations that reflect the tertiary structures in Fig. 8. The only structural feature the
two types of tertiary contacts have in common is the presence of an AU Hoogsteen
316 W.G. Scott
Fig. 8 The secondary structures of two classes of full-length hammerhead ribozymes, with those
represented by satellite virus hammerheads shown in a, and the Schistosoma-like hammerheads,
shown in b. The tertiary contacts are highlighted in light green. Little sequence homology in the
tertiary contact region in apparent
The structure of the GNRA tetraloop capping Stem II in the sTRSV hammerhead
is unusual as well. The GNRA tetraloop, where the 50 nucleotide is always a G, the
second can be any nucleotide (N), the third is a purine (R), and the fourth is always
an A, is a canonically stable RNA secondary structural motif. The G pairs with the
A via a single hydrogen bond, and the N, R, and A all stack upon one another on the
30 -side of the loop. However, in the context of the sTRSV hammerhead tertiary
contact, the structure of the GNRA tetraloop rearranges so that the above-noted
interactions disappear and the A rearranges to form the Hoogsteen base pair with
U1.7 from Stemloop I.
(Fig. 9, light blue) accepts a proton from G12. G12 must ionize to function as the
general base, and the proton is replaced by that from the 20 -OH of C17 (Fig. 9,
black). The original G12 proton can then be relayed directly to the 20 -OH of G8 to
replace a proton that must be donated to the 50 -O leaving group of C1.1 (black) as
the phosphodiester backbone is cleaved. This mechanism conserves the number of
protons during the phosphodiester isomerization. It is testable, in that it predicts that
altering the pKa of either the purine base at position 12 or the 20 -OH at position
8 will alter the cleavage rate without inducing gross structural perturbations. There
are also opportunities for transition-state stabilization of the accumulating negative
charges in the pentacoordinated oxyphosphorane. We suggest that either the
exocyclic amine of A9 or a divalent cation can perform this function.
including molar concentrations of NH4+, the strict requirement for Mg2+ may be
dispensed with (Murray et al. 1998a, b). Nevertheless, the hammerhead ribozyme
appears to be reliant upon Mg2+ under in vivo conditions, and is typically assayed
in vitro in the presence of 10 mM Mg2+, whereas physiological concentrations of
Mg2+ are closer to 1 mM. The apparent Km for Mg2+ in minimal hammerheads
ranges from 10 to 100 mM (Stage-Zimmermann and Uhlenbeck 1998).
Originally, Mg2+ was believed to play a direct role in acid–base catalysis in the
hammerhead ribozyme (Dahm et al. 1993), and although as early as 1998, it was
known that the hammerhead did not require Mg2+ for catalysis, it was only with
the publication of the full-length hammerhead ribozyme structures, which
revealed RNA functional groups positioned for acid/base chemistry, that the
participation of Mg2+ in acid–base catalysis could be ruled out. An additional
potential role for Mg2+, however, is transition-state charge stabilization (Martick
et al. 2008a, b; Lee et al. 2007, 2008). For this, any high concentration of positive
charge should suffice, so the suggestion that Mg2+ or monovalent cations aid in
folding as well as transition-state stabilization appears to be the most consistent
with all of the data, and accounts for the rather high apparent Km values for
Mg2+. What is apparent is that under low ionic strength in vitro assay conditions,
the hammerhead ribozyme needs at least ten times the total physiological con-
centration of Mg2+ to cleave efficiently. Therefore, it is very unlikely that the
minimal hammerhead sequence will be able to fold and cleave efficiently in vivo,
unless folding is assisted by some compensatory mechanism (such as an associated
RNA-binding protein).
The full-length hammerhead ribozyme, even under low ionic strength in vitro
assay conditions, requires only micromolar concentrations of Mg2+ (Khvorova et al.
2003), which is far more consistent with in vivo requirements for activity. It is
likely that the requirement for 10–100 mM Mg2+ for optimal activity of the minimal
hammerhead is partially compensating for the lack of the tertiary contact that
stabilizes the active site. Hence, the full-length hammerhead, which includes a
naturally occurring tertiary contact, would be by far the most preferable starting
point for the design of an in vivo RNA cleaving reagent.
One of the most attractive features of the minimal hammerhead ribozyme to those
hoping to design a specific RNA cleavage reagent is that almost all of the enzyme–
substrate binding specificity can be understood in terms of simple Watson–Crick
base-pairing rules. One does not need to know anything about the hammerhead
ribozyme’s tertiary structure in order to design a hammerhead ribozyme to cleave
an RNA substrate of a given sequence. The only sequence restriction that exists, in
terms of choosing a target substrate, is that a RUH nucleotide triplet be present. H is
the cleavage-site nucleotide. It is typically a C but can be any nucleotide apart from
320 W.G. Scott
The above considerations lead quite naturally to explicit requirements for hammer-
head ribozyme design.
Minimal hammerhead ribozymes have been employed for many years as antisense
RNA cleaving agents in an attempt to target pathological mRNAs and RNA viruses
(For a recent review, please see Tedeschi et al. 2009). An advantage of the minimal
hammerhead is that the only sequence restriction imposed on the target substrate is
at the cleavage site, where a sequence of the type RUH is required (R is either G or
A at position 16.2, U must be uracil at position 16.1, H can be anything except G at
position 17, the cleavage site nucleotide). The sequences of Stems I and III in a
minimal hammerhead enzyme strand, apart from these restrictions, can then be
tailored to base-pair with any target sequence. Unfortunately, their practical utility
as therapeutic agents has been limited by the apparent need for nonphysiological
concentrations of Mg2+ and their slow (1/min) turnover rates.
The much greater activity of the full-length hammerhead ribozymes makes them a
more attractive alternative as in vivo RNA cleavage agents. In addition, the discov-
ery of naturally occurring full-length hammerheads in mammalian mRNAs that
downregulate gene expression (Martick et al. 2008a, b), including some that appear
to work intermolecularly, provide a convincing argument that full-length hammer-
heads should be viable in vivo ribozyme nucleases. But because of the requirement
What can the New Hammerhead Ribozyme Structures Teach us About Design? 321
Fig. 10 A schematic
representation of the
Schistosomal hammerhead
ribozyme, with the enzyme
strand shown in blue, and the
substrate strand shown in
orange. The restrictions
imposed on the sequence are
such that any RNA strand
with a sequence of the form ...
NRUHNNNNNNYN... can
be targeted, although the
optimal sequence will have
C17 and U1.7. The crystal
structures reveal that a U (or
possibly C) at position 1.7 is
the only additional sequence
restriction imposed by the
full-length hammerhead
tertiary contact region
...NRUHNNNNNNYN...
between Stems I and II
GU C
16.2 16.1 17
1.1 U1.7
1.2
1.4
1.3
for the tertiary contact, whose sequence requirements before now have been rather
obscure, full-length hammerhead cleavage agents have not been pursued with vigor.
Comparison of the sTRSV and Schistosomal full-length hammerhead crystal
structures reveals that only one base-pairing interaction between enzyme and
substrate strands in the tertiary contact region is conserved. The AU Hoogsteen
pair requires U1.7 to be present (although C might be able to substitute for U at this
position; cf. Fig. 10). The remainder of the specific tertiary interactions appears in
both cases to lie entirely within the enzyme strands of Stem I and Stem-loop II.
Hence the sequence restrictions imposed by the full-length hammerhead on possi-
ble RNA targets is simply ...NRUHNNNNNNYN... where N is any nucleotide, R
can be G or A at position 16.2, U must be uracil at 16.1, H can be any residue at
position 17, the cleavage site, although C is preferred, and Y at position 1.7, (i.e.,
seven nucleotides downstream of the cleavage site), must be U or possibly C.
Examination of the structures therefore strongly suggests that an mRNA
sequence that possesses the ...NRUHNNNNNNYN... motif can be targeted and
cleaved efficiently by designing a hammerhead enzyme strand complementary to
the target sequence. If cleavage is insufficiently efficient, one can subsequently use
in vitro selection to further optimize the enzyme sequence by selecting for variants
in the tertiary contact region.
322 W.G. Scott
References
Blount KF, Uhlenbeck OC (2005) The structure–function dilemma of the hammerhead ribozyme.
Annu Rev Biophys Biomol Struct 34:415–440
Canny MD, Jucker FM, Kellogg E et al (2004) Fast cleavage kinetics of a natural hammerhead
ribozyme. J Am Chem Soc 126:10848–10849
Chi YI, Martick M, Lares M et al (2008) Capturing hammerhead ribozyme structures in action by
modulating general base catalysis. PLoS Biol 6:e234
Cochrane JC, Strobel SA (2008) Catalytic strategies of self-cleaving ribozymes. Acc Chem Res
41:1027–1035
Dahm SC, Uhlenbeck OC (1991) Role of divalent metal ions in the hammerhead RNA cleavage
reaction. Biochemistry 30:9464–9469
Dahm SC, Derrick WB, Uhlenbeck OC (1993) Evidence for the role of solvated metal hydroxide
in the hammerhead cleavage mechanism. Biochemistry 32:13040–13045
Fedor MJ (2009) Comparative enzymology and structural biology of RNA self-cleavage. Annu
Rev Biophys 38:271–299
Ferbeyre G, Bourdeau V, Pageau M et al (2000) Distribution of hammerhead and hammerhead-
like RNA motifs through the GenBank. Genome Res 10:1011–1019
Ferbeyre G, Smith JM, Cedergren R (1998) Schistosome satellite DNA encodes active hammer-
head ribozymes. Mol Cell Biol 18:3880–3888
Forster AC, Davies C, Sheldon CC et al (1988) Self-cleaving viroid and newt RNAs may only be
active as dimers. Nature 334:265–267
Guerrier-Takada C, Gardiner K, Marsh T et al (1983) The RNA moiety of ribonuclease P is the
catalytic subunit of the enzyme. Cell 35:849–857
Han J, Burke JM (2005) Model for general acid-base catalysis by the hammerhead ribozyme:
pH-activity relationships of G8 and G12 variants at the putative active site. Biochemistry
44:7864–7870
Haseloff J, Gerlach WL (1989) Sequences required for self-catalysed cleavage of the satellite
RNA of tobacco ringspot virus. Gene 82:43–52
Khvorova A, Lescoute A, Westhof E et al (2003) Sequence elements outside the hammerhead
ribozyme catalytic core enable intracellular activity. Nat Struct Biol 10:708–712
Kim SH, Suddath FL, Quigley GJ et al (1974) Three-dimensional tertiary structure of yeast
phenylalanine transfer RNA. Science 185:435–440
Klug A, Ladner J, Robertus JD (1974) The structural geometry of co-ordinated base changes in
transfer RNA. J Mol Biol 89:11–16
Kruger K, Grabowski PJ, Zaug AJ et al (1982) Self-splicing RNA: autoexcision and autocyclization
of the ribosomal RNA intervening sequence of tetrahymena. Cell 31:147–157
De la Peña M, Gago S, Flores R (2003) Peripheral regions of natural hammerhead ribozymes
greatly increase their self-cleavage activity. EMBO J 22:5561–5570
Lee TS, Silva López C, Giambasu GM et al (2008) Role of Mg2þ in hammerhead ribozyme
catalysis from molecular simulation. J Am Chem Soc 130:3053–3064
Lee TS, Silva-López C, Martick M et al (2007) Insight into the role of Mg 2 in hammerhead
ribozyme catalysis from X-ray crystallography and molecular dynamics simulation. J Chem
Theory Comput 3:325–327
Luzi E, Eckstein F, Barsacchi G (1997) The newt ribozyme is part of a riboprotein complex. Proc
Natl Acad Sci USA 94:9711–9716
Martick M, Scott WG (2006) Tertiary contacts distant from the active site prime a ribozyme for
catalysis. Cell 126:309–320
Martick M, Horan LH, Noller HF et al (2008a) A discontinuous hammerhead ribozyme embedded
in a mammalian messenger RNA. Nature 454:899–902
Martick M, Lee TS, York DM et al (2008b) Solvent structure and hammerhead ribozyme catalysis.
Chem Biol 15:332–342
What can the New Hammerhead Ribozyme Structures Teach us About Design? 323
McKay DB (1996) Structure and function of the hammerhead ribozyme: an unfinished story. RNA
2:395–403
Murray JB, Dunham CM, Scott WG (2002) A pH-dependent conformational change, rather than
the chemical step, appears to be rate-limiting in the hammerhead ribozyme cleavage reaction.
J Mol Biol 315:121–130
Murray JB, Seyhan AA, Walter NG et al (1998a) The hammerhead, hairpin and VS ribozymes are
catalytically proficient in monovalent cations alone. Chem Biol 5:587–595
Murray JB, Szöke H, Szöke A et al (2000) Capture and visualization of a catalytic RNA
enzyme–product complex using crystal lattice trapping and X-ray holographic reconstruction.
Mol Cell 5:279–287
Murray JB, Terwey DP, Maloney L et al (1998b) The structural basis of hammerhead ribozyme
self-cleavage. Cell 92:665–673
Nelson JA, Uhlenbeck OC (2006) When to believe what you see. Mol Cell 23:447–450
Nelson JA, Uhlenbeck OC (2008) Hammerhead redux: does the new structure fit the old biochemical
data? RNA 14:605–615
Noller HF, Hoffarth V, Zimniak L (1992) Unusual resistance of peptidyl transferase to protein
extraction procedures. Science 256:1416–1419
Peracchi A, Beigelman L, Scott EC et al (1997) Involvement of a specific metal ion in the
transition of the hammerhead ribozyme to its catalytic conformation. J Biol Chem
272:26822–26826
Pley HW, Flaherty KM, McKay DB (1994) Three-dimensional structure of a hammerhead
ribozyme. Nature 372:68–74
Prody GA, Bakos JT, Buzayan JM et al (1986) Autolytic processing of dimeric plant virus satellite
RNA. Science 231:1577–580
Pyle AM (1993) Ribozymes: a distinct class of metalloenzymes. Science 261:709–714
Robertus JD, Ladner JE, Finch JT et al (1974) Structure of yeast phenylalanine tRNA at 3
A resolution. Nature 250:546–551
Scott WG (1999) RNA structure, metal ions, and catalysis. Curr Opin Chem Biol 3:705–710
Scott WG (2002) Visualizing the structure and mechanism of a small nucleolytic ribozyme.
Methods 28:302–306
Scott WG, Finch JT, Klug A (1995) The crystal structure of an all-RNA hammerhead ribozyme:
a proposed mechanism for RNA catalytic cleavage. Cell 81:991–1002
Scott WG, Martick M, Chi YI (2009) Structure and function of regulatory RNA elements:
ribozymes that regulate gene expression. Biochim Biophys Acta 1789:634–641
Scott WG, Murray JB, Arnold JR et al (1996) Capturing the structure of a catalytic RNA
intermediate: the hammerhead ribozyme. Science 274:2065–2069
Stage-Zimmermann TK, Uhlenbeck OC (1998) Hammerhead ribozyme kinetics. RNA 4:875–889
Steitz TA, Moore PB (2003) RNA, the first macromolecular catalyst: the ribosome is a ribozyme.
Trends Biochem Sci 28:411–418
Tedeschi L, Lande C, Cecchettini A et al (2009) Hammerhead ribozymes in therapeutic target
discovery and validation. Drug Discov Today 14:776–783
Uhlenbeck OC (1987) A small catalytic oligoribonucleotide. Nature 328:596–600
Wang S, Karbstein K, Peracchi A et al (1999) Identification of the hammerhead ribozyme metal
ion binding site responsible for rescue of the deleterious effect of a cleavage site phosphor-
othioate. Biochemistry 38:14363–14378
Wedekind JE, McKay DB (1998) Crystallographic structures of the hammerhead ribozyme:
relationship to ribozyme folding and catalysis. Annu Rev Biophys Biomol Struct 27:475–502
microRNA Biogenesis and its Impact on RNA
Interference
Contents
1 The microRNA Biogenesis Pathway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
1.1 microRNA Gene Transcription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
1.2 microRNA Editing: Small Changes Affect Many Steps . . . . . . . . . . . . . . . . . . . . . . . . . . 327
1.3 pri-miRNA Cleavage by the Microprocessor Complex . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
1.4 Nuclear Export of the microRNA Precursors by Exportin-5 . . . . . . . . . . . . . . . . . . . . . . 332
1.5 The RISC Loading Complex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333
1.6 Terminal Loop Removal by Dicer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
1.7 Ago2 Jumps the Queue: Generation of the ac-pre-miRNA . . . . . . . . . . . . . . . . . . . . . . . 335
1.8 miRNA Duplex Unwinding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336
1.9 Strand Selection: Who Becomes the Guide? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337
1.10 Mediators of RNA Silencing: The Argonaute Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . 337
1.11 Half-Life and Degradation of microRNA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
2 Implication for RNAi Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
2.1 Potentials and Challenges of siRNAs as a Tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342
2.2 siRNA Versus shRNA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
2.3 shRNA-miR Library: Transferring microRNA Structures to Synthetic shRNAs . . . 343
2.4 Enhancement of RNAi by microRNA Biogenesis Factors . . . . . . . . . . . . . . . . . . . . . . . . . . 344
2.5 microRNA Biogenesis in Health and Disease: Basis for RNAi Therapy . . . . . . . . . . . 346
2.6 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 325
RNA Technologies, DOI 10.1007/978-3-642-12168-5_15,
# Springer-Verlag Berlin Heidelberg 2010
326 S. Grund and S. Diederichs
Abbreviations
miRNAs microRNAs
pri-miRNAs primary miRNA transcripts
RISC RNA-induced silencing complex
RNAi RNA interference
Pol II RNA Polymerase II
Tudor-SN Tudor staphylococcal nuclease
bp basepairs
nt nucleotide
dsRBD double-stranded RNA binding domain
OB-fold oligonucleotide/oligosaccharide binding fold
shRNAs short hairpin RNAs
ac-pre-miRNA Ago2-cleaved pre-miRNA
microRNA Biogenesis and its Impact on RNA Interference 327
Since RNAi was first described in the nematode Caenorhabditis elegans (Fire et al.
1998), it has been found to be a widespread phenomenon in eukaryotic organisms.
Over the past years, hundreds of miRNA genes have been identified in animals,
plants, and viruses, making them one of the largest gene families. Several different
miRNAs can control one mRNA target cooperatively and each miRNA can bind to
many different mRNAs. Thus, more than 30% of all human genes are predicted to
be subject to miRNA regulation (Lewis et al. 2005). Hence, the control of miRNA
expression is by itself a key step in regulating target mRNAs.
Analysis of the genomic position of miRNAs revealed that the majority resides in
intergenic regions (Lagos-Quintana et al. 2003), indicating that they are autono-
mous transcription units. The miRNA genes embedded within known transcripts
are primarily found in intronic regions (Lagos-Quintana et al. 2003; Rodriguez et al.
2004; Kim and Kim 2007), although some are found in exonic locations, such as the
untranslated regions of mRNAs (Rodriguez et al. 2004; Kim and Kim 2007).
Animal miRNA genes are often localized in close proximity to each other forming
clusters (Lau et al. 2001; Lagos-Quintana et al. 2003), which are transcribed
polycistronically (Lee et al. 2002). In contrast, this polycistronic arrangement is
quite rare in plants with few exceptions (Guddeti et al. 2005; Zhang et al. 2008).
Transcription of miRNA genes is mediated by RNA Polymerase II (Pol II) as
primary miRNA transcripts (pri-miRNAs) have been shown to contain the hall-
marks of Pol II transcripts, a cap structure and a poly(A) tail (Cai et al. 2004; Lee
et al. 2004a). Further, expression of miRNAs was decreased by the Pol II-specific
inhibitor a-amanitin, and Pol II has been shown to associate with miRNA promoters
(Lee et al. 2004a; Bortolin-Cavaille et al. 2009). Transcription by Pol II enables
tissue-specific or developmental control by regulatory transcription factors. In
several cases, where the miRNA gene is transcribed together with a protein-coding
gene as a single transcription unit, the regulated expression pattern appears to be
coordinated (Baskerville and Bartel 2005).
The largest human miRNA gene cluster, C19MC, is embedded in Alu repeat
sequences. It was proposed that this cluster is unique in being transcribed by RNA
Pol III (Borchert et al. 2006). However, another study claims that C19MC miRNAs
are encoded within introns of a Pol II transcript (Bortolin-Cavaille et al. 2009).
sequences (Han et al. 2006) (Fig. 1). The hairpin, containing the future miRNA in
either the 50 or 30 half of its stem, is excised by the endonuclease Drosha, which
liberates the 60–70 nt long pre-miRNA with a two nucleotide (nt) overhang at its
30 -end (Lee et al. 2003; Denli et al. 2004). This process, also referred to as
“cropping”, occurs cotranscriptionally and precedes the splicing reaction (Kim
and Kim 2007; Morlando et al. 2008). Notably, the Drosha cleavage reaction is
no hindrance for the spliceosome to execute its function because continuity of the
intron is not a prerequisite (Dye et al. 2006; Kim and Kim 2007).
Drosha is a member of the highly conserved RNase III enzyme family, contain-
ing two RNase III domains, a double-stranded RNA binding domain (dsRBD), and
a long N-terminal region with unknown function (Filippov et al. 2000; Wu et al.
2000). By intramolecular dimerization, the two RNase III domains form a single
processing center with two catalytic sites that each cut one strand of the stem (Han
et al. 2004a). Drosha is present in two different complexes. In the smaller complex,
referred to as the Microprocessor, Drosha binds to its cofactor DGCR8 (DiGeorge
syndrome critical region 8; also known as Pasha in C. elegans and Drosophila
melanogaster) (Denli et al. 2004; Gregory et al. 2004; Han et al. 2004a). This
interaction is essential for the conversion of pri-miRNAs into pre-miRNAs
(Gregory et al. 2004; Han et al. 2004a). While Drosha is crucial for the catalysis,
DGCR8 establishes specificity for pri-miRNAs and determines the cleavage site
11 bp distant from the junction of the stem base and the flanking single-stranded
RNA (ssRNA) (Gregory et al. 2004; Han et al. 2006). Binding of the dsRBDs of
DGCR8 to the pri-miRNA requires the ssRNA regions, which are therefore indis-
pensible for pri-miRNA processing (Zeng and Cullen 2005; Han et al. 2006).
However, several sequence alterations in pri-miRNAs of human tumors lead to
conformational changes without affecting processing efficiency pointing towards a
certain flexibility of the Microprocessor (Diederichs and Haber 2006). Aside from
the necessity of the basal segments, the stem and a loop at the end of the stem are
important for efficient cleavage (Zeng et al. 2005b; Han et al. 2006). After deter-
mination of the cleavage site by DGCR8, Drosha can interact transiently with this
preformed RNA–protein complex and execute the cut.
Drosha-mediated cleavage represents a critical step in miRNA biogenesis since
this initial processing event defines one end of the mature miRNA. Nonetheless, it
is not compulsory for the generation of pre-miRNAs. Short introns can also form
hairpin structures that resemble pre-miRNAs. These alternative precursors, termed
“mirtrons”, can be spliced and debranched into pre-miRNA-like hairpins. Mirtrons
lack the lower stem of pri-miRNAs and therefore omit processing by the Micropro-
cessor. The debranched hairpins are then exported and further processed by the
canonical miRNA biogenesis pathway (Okamura et al. 2007; Ruby et al. 2007).
To control Microprocessor activity, Drosha and DGCR8 regulate each other via a
regulatory feedback circuit. Protein–protein interaction of DGCR8 with Drosha
330 S. Grund and S. Diederichs
ADAR
m7G
transcription & editing
pre-mRNA
m 7G
splicing
DGCR8
Drosha
pri-miRNA
branched pre-mirtron mRNA
AAA
RanGTP
Exp-5
5’ nucleus
pre-miRNA 3’
cytoplasm
Dicer
Dicer
5’ ac-pre-miRNA
3’
Ago2 TRBP
Dicer cleavage
Dicer cleavage
Dicer Dicer
5’ 5’
miRNA duplex 3’ 3’
Ago2 TRBP Ago2 TRBP
duplex unwinding
& RISC formation
Fig. 1 The microRNA biogenesis pathway in vertebrates. MicroRNA (miRNA) genes are tran-
scribed by RNA Polymerase II (Pol II) generating primary transcripts (pri-miRNAs). Cleavage by
the Microprocessor complex consisting of Drosha and DGCR8 results in a 65 nt precursor-
miRNA (pre-miRNA). Intron-derived miRNAs (mirtrons) are generated by splicing and debranch-
ing. The intron resembles the hairpin structure of the pre-miRNA thereby bypassing the Drosha
processing step. After export to the cytoplasm, mediated by Exportin-5 in a Ran-dependent
microRNA Biogenesis and its Impact on RNA Interference 331
stabilizes Drosha. In turn, Drosha cleaves hairpins within the Dgcr8 mRNA result-
ing in destabilization of Dgcr8 mRNA (Han et al. 2009; Triboulet et al. 2009).
Apart from this general mechanism, Drosha-mediated cleavage of specific pri-
miRNAs can be influenced by additional factors. The DEAD-box RNA helicases
p68 (DDX5) and p72 (DDX17), which are present in the large Drosha complex,
seem to be involved in the processing of a subset of pri-miRNAs, as deficiency of
these factors decreases their mature miRNA expression levels (Fukuda et al. 2007).
Some auxiliary processing factors act even on individual miRNAs. Transforming
growth factor-b (TGF-b) and bone morphogenetic factors (BMPs) activate specific
SMAD signal transducers, which then form a complex with the Microprocessor via
the RNA helicase p68. This interaction promotes processing of pri-miR-21 into pre-
miR-21 by Drosha (Davis et al. 2008). Although the terminal loop seems to be of
inferior significance for Microprocessor action per se (Han et al. 2006), it seems to
have a fine-tuning role, providing a binding platform for regulatory proteins. The
heterogeneous nuclear ribonucleoprotein A1 (hnRNP A1) protein, for example,
contributes to the biogenesis of only a single miRNA in the miR-17 cluster,
miR-18a (Guil and Caceres 2007). By binding to the conserved loop of the hairpin,
hnRNP A1 remodels the stem structure and thereby creates a more advantageous
cleavage site for Drosha (Michlewski et al. 2008). In contrast, interaction of a stem
cell-specific regulator, lin-28, with the loop of pri-let-7 prevents Drosha cleavage
(Newman et al. 2008; Viswanathan et al. 2008). The repressive effect of lin-28
seems to be antagonized by the KH-type splicing regulatory protein (KSRP), which
promotes maturation of a let-7 and a subset of miRNA precursors by binding to the
terminal loop (Trabucchi et al. 2009). Thus, it is likely that there are far more
factors to be identified, optimizing recruitment and positioning of the Microproces-
sor thereby controlling processing in a coordinated manner.
Plant miRNAs are also derived from long, primary transcripts – although rather
diverse in structure and with longer hairpins than in animals – in a stepwise process
(Kurihara and Watanabe 2004). Nevertheless, biogenesis in plants holds some
substantial differences compared to metazoans. A key characteristic of plant
miRNA maturation is the lack of a Drosha-like protein, which is highly conserved
Fig. 1 (Continued) manner, the pre-miRNA interacts with the preformed ternary complex of
Dicer, TRBP, and Ago2 forming the RISC loading complex (RLC). Dicer removes the terminal
loop of the pre-miRNA creating a miRNA duplex. Pre-miRNAs with a high degree of comple-
mentarity are cleaved by Ago2 in their passenger strand, producing a nicked hairpin called ac-pre-
miRNA, before Dicer cleavage. After unwinding of the miRNA duplex, the passenger strand is
degraded whereas the functional guide strand is loaded onto Ago2, constituting the RNA-induced
silencing complex (RISC), which silences target mRNAs by translational repression or mRNA
cleavage
332 S. Grund and S. Diederichs
After the initial cropping by Drosha, the precursor has to be further processed to
yield the final miRNA. This second cleavage reaction is mediated by the cytoplas-
mic enzyme Dicer. Owing to the different spatial appearance of the two endonu-
cleases, nucleocytoplasmic transit of the pre-miRNA has a pivotal role in miRNA
biogenesis. Like other noncoding RNAs, pre-miRNAs are exported by a member of
the karyopherin family of nucleocytoplasmic transport receptors. Several studies
demonstrated that Exportin-5 is the main if not only transport factor for nuclear
export of pre-miRNAs and that the transport process – characteristic of karyopherin
mediated export – depends on the Ran cycle (Yi et al. 2003; Bohnsack et al. 2004;
Lund et al. 2004). Binding of Exportin-5 to its cargo requires a stem of at least 16 bp
and a short 30 -overhang, an end structure characteristic of RNase III cleavage.
Improperly processed pre-miRNAs with, e.g., 50 -overhangs prevent binding of
Exportin-5 and thus efficient export, highlighting the importance of precise cleav-
age by Drosha (Lund et al. 2004; Zeng and Cullen 2004). Notably, Exportin-5 is not
only required for nucleocytoplasmic transport of pre-miRNAs but also aids in
stabilizing the relatively unstable precursor (Yi et al. 2003; Zeng and Cullen 2004).
In plants, the role of HASTY, the plant homolog of Exportin-5 is not as well
defined as in animals. Since miRNAs are completely processed in the nucleus,
mature miRNAs are found in both the nucleus and the cytoplasm; however, they are
more abundant in the latter compartment. Although loss-of-function mutants of
HASTY reduce miRNA levels, they do not cause accumulation in the nucleus.
Hence, till date, there is no direct evidence that HASTY is involved in miRNA
microRNA Biogenesis and its Impact on RNA Interference 333
export. Also, the exact form as which miRNAs leave the nucleus – either as miRNA
itself, loaded into the RISC complex, or in complex with their target mRNAs –
remains still unclear (Park et al. 2005).
Once in the cytoplasm, the pre-miRNA still has to undergo several additional
processing steps before the single-stranded mature miRNA can guide the effector
complex known as RNA-induced silencing complex (RISC) to its target mRNA.
Maturation of the pre-miRNA and RISC loading is performed by the RISC loading
complex (RLC). This complex comprises in the human system the RNases and
Dicer and the structurally related dsRNA binding proteins TRBP and PACT
(Chendrimada et al. 2005; Haase et al. 2005; Lee et al. 2006; MacRae et al.
2008). Like Drosha, Dicer is an RNase III enzyme and removes the terminal loop
by a staggered cut. Albeit TRBP and PACT are not required for the Dicer-mediated
processing step per se, they have a stimulating effect on the cleavage reaction,
influence the efficiency of RNA silencing, and are involved in the recruitment of
Ago2 (Haase et al. 2005; Lee et al. 2006). Ago2 (also known as eIF2C2) acts in
the effector phase of RNAi and cleaves the mRNA targeted for destruction by
the complementary miRNA or mediates – as well as other human Ago proteins –
translational inhibition. Ago2 stability and thus also efficiency of RNAi is regulated
by hydroxylation (Qi et al. 2008).
The general paradigm that RNase III enzymes cooperate with dsRBD proteins
holds true for Drosophila Dicer enzymes. Dicer-1 and Dicer-2 interact with Loqua-
cious (Loqs) and R2D2, respectively. While the heterodimer Dicer-1/Loqs processes
pre-miRNAs, the counterpart Dicer-2/R2D2cleaves long dsRNAs into siRNAs
(Lee et al. 2004b). Loqs is required for efficient pre-miRNA processing and confers
substrate specificity for pre-miRNAs to Dicer-1 (Forstemann et al. 2005; Saito et al.
2005), whereas R2D2 is necessary for RISC loading but not for the cleavage reaction
(Liu et al. 2003; Tomari et al. 2004b).
The assembly of the RLC and the final RISC differs in flies and humans.
In humans, the RLC is formed prior to pre-miRNA binding and Dicer is released
after miRNA incorporation into Ago2 (Gregory et al. 2005; Maniataki and Mour-
elatos 2005). On the contrary, in Drosophila, Dcr-2/R2D2 binds first the siRNA
duplex. Thereafter, Ago2 joins the complex and the single-stranded siRNA is
loaded onto Ago2. This final complex is called holo-RISC and still retains Dcr-2/
R2D2 (Pham et al. 2004; Tomari et al. 2004b). For the miRNA pathway, however,
Dicer-1/Loqs was recently shown to be excluded from the RISC. After processing of
the pre-miRNA and loading of the duplex onto Ago1, Dicer dissociates from the
RLC. Ago1, loaded with the mature miRNA, interacts then with GW182, a P body
component with a role in miRNA-mediated silencing, forming the RISC (Miyoshi
et al. 2009).
334 S. Grund and S. Diederichs
Once the pre-miRNA joins the RLC, Dicer cleaves the precursor removing the
terminal loop. The cleavage defines the other end of the mature miRNA and results
in a 22 nt long miRNA duplex with 2 nt 30 -overhangs at each end (Bernstein et al.
2001; Provost et al. 2002). While in invertebrates this step necessitates ATP
(Zamore et al. 2000; Nykanen et al. 2001), dicing is ATP-independent in mamma-
lian cells (Provost et al. 2002; Zhang et al. 2002). Studies with dicer mutants in
different organisms have shown the significance of the Dicer reaction for functional
RNAi and its requirement for embryonic development (Knight and Bass 2001;
Bernstein et al. 2003). Dicer proteins are evolutionary conserved throughout
the eukaryotic kingdom except budding yeast (Bernstein et al. 2001). In some
organisms, even multiple Dicer homologs exist, dedicated to mainly one aspect
of RNA silencing, like generation of siRNAs or miRNAs. In contrast to Arabidopsis
and Drosophila, where two Dicer isoforms can share the duties (Lee et al. 2004b;
Xie et al. 2004), the single Dicer enzymes in nematodes and humans process both
dsRNA and miRNA precursors.
As members of the RNase III family, Dicer enzymes comprise two neighboring
RNase III domains responsible for the catalytic reaction and a dsRBD that interacts
with dsRNA in vitro (Provost et al. 2002). The N-terminus harbors in addition a
DExH/DEAH box RNA helicase domain, a PAZ domain and a domain of
unknown function (DUF283). The RNA-binding PAZ domain, named after the
Piwi, Argonaute, and Zwille proteins, is also found in the Argonaute protein family
and adopts a topology related to the oligonucleotide/oligosaccharide-binding fold
(OB-fold). The binding pocket for dsRNAs with a two nucleotide 30 -overhang
within the PAZ domains enables the anchoring and recognition of pre-miRNAs
processed by Drosha (Song et al. 2003; Lingel et al. 2004; Ma et al. 2004).
An additional loop in the PAZ domain of the Dicer family alters the electrostatic
potential of the surface surrounding the binding pocket compared to the Argonaute
PAZ. Due to this difference, substrate recognition and transfer of the substrate to
other complexes might differ between the two PAZ-domain families (Macrae et al.
2006). Notably, some Dicer proteins, such as Drosophila Dicer-2, Arabidopsis
DCL-4, or Schizosaccharomyces pombe Dicer, do not contain a PAZ domain.
In these species, adaptor molecules might compensate the lack of a PAZ domain
providing an alternative to recognize pre-miRNAs or the organism – like in the case
of S. pombe – does not encode miRNAs.
The crystal structure of the parasite Giardia intestinalis Dicer shed light onto the
question how the Dicer cleavage site is determined. G. intestinalis Dicer is a
minimalist among the Dicer enzymes as it consists only of a PAZ domain and
two consecutive RNase III domains. However, it generates RNA duplexes of
25–27 nt from dsRNA in a similar way to human Dicer (Zhang et al. 2002; Macrae
et al. 2006). Unlike the nuclear RNase III Drosha, which is dependent on the “ruler”
activity of its partner DGCR8/Pasha, Dicer is capable of measuring the distance of
about 25 nt from the 30 -end on its own. This length is defined by the distance
microRNA Biogenesis and its Impact on RNA Interference 335
between the PAZ domain that binds the 30 - and the 50 -end and the RNase IIIa
domain, forming the catalytic center (Zhang et al. 2004; Macrae et al. 2006). The
connecting helix between both domains is predicted to form in all Dicer homologs,
and thus also the mechanism of cleavage site determination might be conserved
(Macrae et al. 2006).
As soon as the miRNA duplex is generated by Dicer, Dicer and its binding partners
TRBP and PACT dissociate from the miRNA. Since only one strand is finally
retained in the active RISC complex, the duplex has to be separated into their
component strands: the guide strand, complementary to the target and mediating
RNAi, and the nonfunctional passenger strand, which is subsequently degraded.
Separation of double-stranded RNA molecules is usually mediated by helicases
using energy derived from ATP hydrolysis. Albeit several helicases have been
shown to associate with proteins that act in RISC formation or activity and
influence RNA silencing, a direct involvement in strand unwinding could till date
not be proven. For example, the DEAD box helicases Gemin3/4, RCK/p54, and the
putative DExD-box helicase MOV10 interact with Argonaute proteins and are
required for posttranscriptional silencing (Mourelatos et al. 2002; Meister et al.
2005; Chu and Rana 2006). MOV10 homologs in flies (Armitage) or plants
(SDE-3) play also a role in RNAi (Cook et al. 2004; Tomari et al. 2004a). RHA/
DHX9, a member of the DEAH-containing family of RNA helicases unwinds
dsRNA (Lee and Hurwitz 1992) and aids in active RISC loading by promoting
the association of siRNA with Ago2 (Robb and Rana 2007). Although this obser-
vation points towards a role of RHA in siRNA duplex unwinding, it is unclear
whether this possible function also applies to miRNAs. For miRNA let-7 unwind-
ing, the ATP-dependent helicase p68/DDX5 is sufficient in vitro. In accordance
with this, the lack of p68/DDX5 inhibits let-7 miRNA function (Salzman et al.
2007). These findings and the multitude of potential factors in the unwinding
process suggest that specific helicases may regulate particular subclasses of miRNAs.
Whether they participate directly in unwinding the duplexes or whether they rather
remodel the RLC to facilitate miRNA loading remains to be investigated. As RISC
assembly and RISC loading can be accomplished in an ATP-independent manner
(Gregory et al. 2005; Maniataki and Mourelatos 2005; MacRae et al. 2008), the
general necessity of helicases in this process is challenged.
Another factor implicated in strand separation is the endonuclease Ago2. In the
siRNA pathway, the effector protein Ago2, which is loaded with the siRNA duplex,
cleaves the passenger strand to reduce the internal stability (Matranga et al. 2005;
Rand et al. 2005). This destabilization is necessary for strand dissociation and
facilitates removal of the passenger strand. Mismatches and unpaired bulges in
some miRNA hairpin stems could render the cleavage-assisted strand separation
mechanism not only feasible but also unnecessary due to their inherent thermody-
namic instability. However, Ago2 is able to cleave the passenger strand of si-like
miRNA precursor stems with a highly complementary sequence like the let-7
miRNA family (Diederichs and Haber 2007) (see above). Thus, Ago2-mediated
passenger strand cleavage and generation of the ac-pre-miRNA could facilitate
strand dissociation and RISC activation also in the miRNA pathway – at least for
miRNA precursors that resemble siRNAs with a high degree of complementarity in
the middle of the hairpin stem. In Drosophila, a novel complex of Translin and
microRNA Biogenesis and its Impact on RNA Interference 337
In theory, unwinding of the miRNA duplex yields two different mature miRNAs
with the potential to become the effector strand. However, the two strands are not
equally competent in entering the RISC. Therefore, only one strand from each
duplex, the so-called guide strand, is predominantly incorporated into the RISC.
The remaining strand, named passenger strand, is primarily targeted for degrada-
tion, although both strands can be functionally active (Ro et al. 2007; Okamura
et al. 2008). Which strand is chosen to participate in RNAi and which is condemned
to be destroyed is predestined by the thermodynamic properties of the base pairs at
the ends of the duplex. The strand whose 50 -end is less stably paired to its
counterpart is loaded into RISC (Khvorova et al. 2003; Schwarz et al. 2003).
In Drosophila, the binding partner of Dicer-2, R2D2, senses the functional asym-
metry of the two siRNA strands and binds the strand with the greater double-
stranded character. R2D2 orients the Dcr-2/R2D2 heterodimer on the siRNA in
such a way that the correct strand is handed to Ago2, thereby facilitating RISC
loading (Liu et al. 2003; Tomari et al. 2004b).
Yuan et al. 2005; Wang et al. 2008b), the PAZ domain (as described above)
provides a binding pocket for dsRNAs with 30 overhangs (Song et al. 2003; Lingel
et al. 2004; Ma et al. 2004; Wang et al. 2008b) suggesting that the miRNA after
generation by Dicer could be directly handed over from the Dicer-PAZ to the
Argonaute-PAZ domain. The PIWI domain structurally resembles an RNase H
fold and provides the endonuclease activity for mRNA target cleavage (Song et al.
2004; Rivas et al. 2005; Yuan et al. 2005). Structural studies with a ternary complex
consisting of a Thermus thermophilus argonaute protein, a guide DNA and a target
RNA revealed the processes within this complex that happen upon binding to the
RNA target (Wang et al. 2008a, 2009). The binding of the target RNA starts with
basepairing to the 50 end of the guide DNA, which is anchored in the MID domain.
Progressing duplex formation liberates the 30 end of the guide DNA from its
anchoring site in the PAZ domain. This release leads to a conformational change,
which brings the cleavage site of the target RNA in close proximity to the catalytic
residues in the PIWI domain.
Although most organisms encode several Argonaute proteins (ranging from one
in S. pombe to 27 in C. elegans), which are functionally not redundant, many of
them are endonucleolytically inactive. In humans, only Ago2, also known as
eIF2C2, is equipped with the so-called “slicer” activity (Meister et al. 2004).
However, Ago2 is not the only Argonaute protein associated with miRNAs. The
remaining three members of the Ago subfamily, Ago1, Ago3, and Ago4, interact
with miRNAs as well (Meister et al. 2004). How miRNAs are partitioned between
effector complexes with different Ago proteins and what are the functional
consequences of this sorting in the mammalian system is currently unknown.
More about sorting of small RNAs in different RISC variants is known in flies
and plants. In Drosophila, siRNAs and miRNAs are generated by two different
Dicer enzymes (Lee et al. 2004b). It was assumed that due to their different
biogenesis, the two classes of small RNAs are predetermined to be loaded into a
particular RISC variant, miRNAs into Ago1-RISCs and siRNAs into Ago2-RISCs
(Okamura et al. 2004; Saito et al. 2005). In contrast, Zamore and colleagues could
show that Argonaute loading is not coupled to the biogenesis pathway of miRNAs
or siRNAs but that sorting in one or the other complex depends rather on the
intrinsic structure of the small RNA (Forstemann et al. 2007; Tomari et al. 2007).
The heterodimer Dcr-2/R2D2 enforces incorporation of perfectly matched short
RNAs (e.g., siRNAs) into Ago2-RISC, while bulged or mismatched miRNAs are
excluded. In analogy, a yet unidentified mechanism sorts nonperfectly matched
miRNAs into Ago1-RISCs. For small RNAs with a structure in between that of a
completely complementary siRNA duplex and a typical miRNA duplex with bulges
and mismatches, both RISCs compete (Tomari et al. 2007). Importantly, sorting
into one or the other RISCs determines the fate of the target mRNA as the
two Argonaute proteins are functionally specialized: Ago1-RISCs only silence
imperfectly matched targets whereas Ago2-RISCs, which have the stronger cata-
lytic capacity, silence targets that are fully complementary to the guide strand
(Forstemann et al. 2007). Notably, only those guide strands – irrespective from
which arm of the pre-miRNA they originate – are capable of associating with
microRNA Biogenesis and its Impact on RNA Interference 339
Drosophila Ago2, which have a 50 -end derived from an accurate cleavage by Drosha
or Dicer (Seitz et al. 2008). In Arabidopsis, sorting of small RNAs is directed by the
50 -terminal nucleotide; e.g., joining AGO1, the Argonaute protein prevailing in the
miRNA pathway, requires a uridine at the 50 -end (Seitz et al. 2008).
Once loaded onto an Argonaute protein, the miRNA guides the complex to a fully
or partially complementary mRNA target. Target recognition is considered to
require primarily full complementarity of nucleotides 2–7, termed the miRNA
“seed” sequence, but other nucleotides have been shown to aid in this process
(Doench and Sharp 2004; Brennecke et al. 2005; Grimson et al. 2007). The shortness
of the seed region enables each miRNA to regulate a large number of target genes.
The extent of complementarity between the miRNA and the target is likely the
key determinant for the mechanism of regulation. Highly complementary target
sites – which particularly occur in plants but rarely in animals – cause slicing and
subsequent decay of the target by 50 - and 30 -exonucleases (Zamore et al. 2000). For
siRNAs, the 50 -end of the guide strand defines the position of the target cleavage
between the nucleotides paired to bases 10 and 11 of the siRNA (Elbashir et al.
2001b,c). Most animal miRNAs form imperfect hybrids with their target mRNA,
which precludes endonucleolytic cleavage, due to central mismatches (bases 9–12)
(Elbashir et al. 2001c). Instead, they promote translational attenuation or exonu-
cleolytic mRNA decay. The underlying molecular mechanism of this silencing
pathway is still under debate. Accounting the multitude of steps that have been
proposed to be targeted by RISC action, it is likely that more than one mechanism is
involved. For example, evidences exist that translation inhibition can be mediated
at the level of translation initiation as well as elongation. Additionally, models for
premature termination or cotranslational protein degradation have been suggested
(Eulalio et al. 2008; Wu and Belasco 2008). By triggering the removal of the target
poly(A) tail and subsequent decapping, miRNAs can cause exonucleolytic mRNA
decay (Behm-Ansmant et al. 2006; Wu et al. 2006; Eulalio et al. 2007b). The
enzymes for deadenylation and decapping localize to processing bodies (P bodies),
cytoplasmic sites of mRNA turnover (Eulalio et al. 2007a). Since Argonaute
proteins, miRNAs, and their targets colocalize with these foci, miRNA targets
could be directed to P bodies to be separated from the translation machinery and
be accessible for the decay components (Liu et al. 2005; Pillai et al. 2005; Sen and
Blau 2005). Targeting of Ago2 to P bodies is regulated by its phosphorylation status
(Zeng et al. 2008). The environment of the P bodies per se is not essential for target
silencing as miRNA function is not or only marginally affected by the loss of
P-bodies. Hence, enrichment of the RISC components in P bodies seems to be the
consequence and not the cause of silencing (Eulalio et al. 2007a).
Considering their core function in translation inhibition by various means, it is
not surprising that Argonaute proteins are mainly localized in the cytoplasm and
enriched in processing bodies. Nevertheless, nuclear functions for Ago proteins
340 S. Grund and S. Diederichs
have been reported in worms, flies, plants, and S. pombe (Lippman and Martienssen
2004; Matzke and Birchler 2005). Evidence is accumulating that human Ago proteins
also function in the nucleus. They were detected in nuclear fractions (Robb et al.
2005), and Ago1 and Ago2 have also been shown to associate with the promoter
region of the progesterone receptor (Janowski et al. 2006). Additionally, siRNAs
directed against promoter regions resulted in Ago1- or Ago2-dependent transcrip-
tional repression (Morris et al. 2004; Janowski et al. 2006; Kim et al. 2006). Importin-
8 is one factor involved in the transport of Ago2 into the nucleus (Weinmann et al.
2009). In addition to the effector proteins, mature miRNAs have also been detected
in the nucleus. The human miR-29b is even predominantly detected in the nucleus
(Meister et al. 2004; Hwang et al. 2007). Nuclear import of miR-29b is mediated
by a 30 -terminal hexanucleotide motif, which acts as a nuclear localization element
(Hwang et al. 2007). These examples of reimport of a miRNA into the nucleus
and the nuclear localization and action of effector proteins raise the possibility
that RISC-dependent gene silencing could occur also in humans at the transcriptional
level.
For steady-state miRNA expression levels, their stability, turnover, and degradation
could be as important as the regulation of miRNA maturation. So far, rather little is
known about this aspect of the miRNA life cycle.
Generally, mature miRNAs are rather stable, as they persist long in the cell after
depletion of miRNA processing factors (Lee et al. 2003; Gregory et al. 2004).
However, miRNA levels can drop rapidly under certain conditions, albeit the
mechanism is yet unidentified (Pedersen et al. 2007). One factor known to play a
role in miRNA homeostasis is Ago2, as miRNA levels depend on the amount of
Ago2 in the cell (Diederichs and Haber 2007). Exoribonucleases responsible for
miRNA degradation have till date only been discovered in two organisms: SDN1 in
Arabidopsis and XRN-2 in C. elegans (Ramachandran and Chen 2008; Chatterjee
and Grosshans 2009). XRN-2 aids in miRNA release from the Argonaute complex
and mediates miRNA turnover. Both separation and degradation are antagonized
by the presence of target RNA implicating a coordination of miRNA and target
RNA levels (Chatterjee and Grosshans 2009). An alternative way to impede miRNA
activity could be the shielding of miRNA-binding sites by interaction with RNA-
binding proteins (Bhattacharyya et al. 2006; Kedde et al. 2007).
The discovery of gene silencing by small RNA molecules, termed RNAi, was a
milestone in biology by providing a novel level of gene expression regulation.
Beyond the scientific importance, RNAi opened up the possibility for experimental
microRNA Biogenesis and its Impact on RNA Interference 341
and therapeutic applications. However, since the small RNA molecules, miRNAs
or siRNAs, provide only the specificity and depend on the endogenous miRNA
pathway to mediate RNAi, a detailed knowledge of miRNA biogenesis is manda-
tory to use RNAi successfully as tool (Fig. 2).
miRNA gene
shRNA-mir
shRNA
m7G
construct
pri-miRNA
AAA
shRNA
5’ nucleus
pre-miRNA 3’
RanGTP
cytoplasm
Exp-5
5’ Ago2
Dicer 3’
Ago2 cleavage
Dicer cleavage
siRNA 5’ ac-pre-miRNA
3’
Dicer
Dicer cleavage
5’ 5’
miRNA duplex 3’ 3’
duplex unwinding
& RISC formation
Fig. 2 RNAi takes advantage of the endogenous miRNA biogenesis machinery and effector
pathway. To use RNAi as a tool, artificial RNAs can be used, which enter the endogenous pathway
at different steps. While shRNAs or shRNA-mirs, which are transcribed from a plasmid by
Polymerase III or II, undergo (almost) the entire processing pathway, synthetic siRNA duplexes
join the pathway only in the cytoplasm. For simplicity, only those factors improving RNAi are
shown. Exp-5; Exportin-5
342 S. Grund and S. Diederichs
be able to repress also targets, which only partially basepair with the siRNA (Jackson
et al. 2003).
Early approaches with shRNAs – when knowledge about miRNA biogenesis was
still rudimentary – used simple short hairpin RNAs of varying stem length
(19–29 bp) and loops of 4–15 nt transcribed under the control of RNA polymerase
III (Brummelkamp et al. 2002; Paddison et al. 2002). Those vectors have been
344 S. Grund and S. Diederichs
successfully used to mediate RNAi. The deeper insights in the action of miRNA
processing factors and the knowledge gained on structural demands for efficient
cleavage and strand incorporation found consideration in the design of shRNA
vectors. The new generation of shRNAs, termed shRNA-miRs, are modeled after
endogenous pri-miRNAs (Dickins et al. 2005; Silva et al. 2005). The starting basis
for such shRNA-miRs is the backbone of the primary miR-30a, which is a well-
studied human miRNA of which the processing sites have been mapped (Zeng and
Cullen 2003). When the miR-30a stem is replaced by heterologous sequences, the
particular miRNA is generated and mediates effective and regulated target gene
inhibition (Zeng et al. 2002, 2005a). Using the backbone of pri-miR-30a, Hannon
and colleagues constructed an improved shRNA-miR library (Silva et al. 2005).
The vector designed for this library contains not only the remodeled pri-miR-30a
but also the 50 - and 30 -flanking regions to optimize ectopic expression (Chen et al.
2004). Mapping the processing sites of the shRNA-miRs allowed the shRNA design
following the thermodynamic asymmetry rule, even when cleavage is inaccurate
and translocated 1 bp (Seitz et al. 2008). One criterion for efficient RNAi is the
amount of siRNA that is available for incorporation into RISC (Siolas et al. 2005).
In comparison with first generation constructs containing simple short hairpins, the
shRNA-miR vectors produced about 12 times more small RNAs (Silva et al. 2005).
The expression of the shRNAs is like in the first generation library driven by a U6
promoter (Pol III dependent), since this gave together with expression from a CMV
promoter (Pol II dependent) the best results regarding efficiency and consistency of
target repression. However, one advantage of the advanced library system is the
possibility to swap the shRNA-miR cassette into other vectors without transferring
the U6 promoter. By recombination, the shRNA-miR can thus be combined with
any other desired promoter (e.g., promoters for inducible or tissue-specific expres-
sion) (Dickins et al. 2005; Stegmeier et al. 2005). With the resulting shRNA-miR
library (Hannon–Elledge library), the majority of human and mouse genes is
covered on average by two shRNA-miRs (Silva et al. 2005), which allows the
functional analysis on a single gene basis or in genome-wide approaches.
Bohnsack et al. 2004; Lund et al. 2004) is expressed at low levels. Indeed, over-
expression of Exportin-5 enhances RNAi triggered by shRNAs (Yi et al. 2005).
However, the effect caused by Exportin-5 overexpression is not absolutely specific.
Reporter constructs are silenced by elevated Exportin-5 levels even in the absence
of a corresponding ectopic miRNA. Further, repression of the transcript also
occurred when the miRNA-binding sites were mutated, making an approach with
Exportin-5 more prone to off-target effects (Diederichs et al. 2008).
Apart from Exportin-5, other factors of the miRNA biogenesis and effector
machinery have been tested for their ability to improve RNAi. While Drosophila
Dicer-2 could enhance RNAi (Dietzl et al. 2007), human Dicer did not affect RNAi
potency in mammalian cells (Diederichs et al. 2008). In contrast, human Ago2
overexpression led to increased let-7a activity towards a reporter constructs con-
taining let-7a binding sites (Diederichs et al. 2008). This effect seems not to be due
to a simultaneous increase of mature let-7a miRNA, which is derived from a
construct encoding the let-7a primary miRNA, as overexpression of the other
Argonaute proteins, Ago1, Ago3, and Ago4, results also in elevated levels of
mature let-7a but not in enhanced RNAi. Importantly, the effect obtained by over-
expression of Ago2 is extremely specific, as it acts only towards perfectly matched
binding sites, and endogenous RNAi activity is not affected. That coexpression of
Ago2 allows reduction of the pri-miRNA construct concentration for transfection
is another asset (Diederichs et al. 2008). Thereby, off-target effects caused by
oversaturation of the endogenous pathway (Grimm et al. 2006) can be reduced.
Especially high-throughput screens using shRNA or siRNA libraries can profit from
these effects. False-positive or false-negative results of such screens could be
reduced by coexpression of Ago2, thereby increasing specificity and efficiency of
RNAi. Moreover, overexpression of Ago2 could be of use for medical applications
based on RNAi technology. The mechanism behind the effect of Ago2 overexpres-
sion on RNAi is likely a combination of several modes of action. First, Ago2
improves the stability of mature miRNAs. However, increased expression levels
of mature miRNAs alone cannot be decisive for a more potent RNAi response.
Overexpression of the other human Argonaute proteins, Ago1, Ago3, and Ago4,
enhances miRNA expression, as well (Diederichs and Haber 2007), but RNAi
efficiency was not altered (Diederichs et al. 2008). Hence, a property unique to
Ago2 seems to account for RNAi enhancement. Ago2 differs from the other argo-
naute proteins by its intrinsic endonuclease activity, which operates at two different
levels: during miRNA processing and in the effector phase. Certain pre-miRNAs
with high complementarity are cleaved by Ago2 in their passenger strand creating an
ac-pre-miRNA intermediate, whereby strand dissociation is thought to be facilitated
(Matranga et al. 2005; Rand et al. 2005; Diederichs and Haber 2007). This function in
processing seems to be important for enhancing RNAi as the effect is more
pronounced when shRNAs, which undergo this cleavage event, are used instead
of siRNAs, which enter the pathway at one of the last steps (Diederichs et al.
2008). In the effector phase, the endonuclease activity of Ago2 mediates cleavage
of the target mRNA paired with the small RNA (Meister et al. 2004). Target mRNA
cleavage and subsequent degradation might result in a more potent RNAi response
346 S. Grund and S. Diederichs
RNAi has not only become an important scientific tool but holds great potential
for therapeutic applications due to its potency and specificity to inhibit gene
expression. Diseases that could be treated by an RNAi approach are, e.g., cancer
and infections caused by viruses such as the human immunodeficiency virus (HIV)
or the hepatitis C virus (HCV). Although initial successful experiments in cell
culture show promise, several substantial hindrances have to be overcome before
the first RNAi based therapeutics will be available.
The first important task is to find a good target. This is particularly problematic
for viral diseases in case the virus has a high mutation rate. Alternatively to
targeting the viral RNA directly, cellular cofactors required for infection could
be downregulated. However, this approach would also target noninfected cells
potentially resulting in adverse effects. Next to the potency and specificity of the
siRNA/shRNA for therapeutic targets, the intracellular delivery of the RNAi drug is
a major challenge, which will be discussed elsewhere.
In general, successful RNAi-based therapy approaches critically depend on our
comprehensive understanding of the miRNA pathway in health and disease. In
cancer cells, a global reduction of mature miRNA levels is observed (Lu et al. 2005;
Lee et al. 2008). One possible reason for this aberrant miRNA pattern might be
defective miRNA processing factors or components of the effector machinery
(Thomson et al. 2006). Consistently, cellular transformation and tumorigenesis
was promoted by knockdown of the miRNA processing factors Drosha, Dgcr8,
and Dicer (Kumar et al. 2007). In cases where cancer results from impaired miRNA
maturation, drugs based on RNAi are probably not suitable, as they require an
intact processing and effector machinery to function effectively. But not only
knowledge of miRNA regulation in pathological cases is crucial for medical
application. RNAi is such a powerful method because it benefits from an endoge-
nous machinery enhancing its potency and specificity. This inevitably means that a
detailed understanding of the complex regulatory network of expression and
maturation of miRNAs in general is a prerequisite for RNAi drug development.
Thus, the biggest advantage of RNAi is also its major challenge.
microRNA Biogenesis and its Impact on RNA Interference 347
Since the initial successful transfer of the naturally occurring miRNA phenomenon
to the practical application of RNAi, our knowledge on the biogenesis pathway of
small RNAs has immensely increased. Our deeper understanding enabled the
development of more efficient approaches by accounting for the requirements of
the endogenous machinery. Also, future progress in RNAi as genetic tool or in
clinical applications will go hand in hand with advances in our understanding of the
molecular mechanisms and the complex regulatory network of miRNA biogenesis
and action as both depend on the same machinery.
References
Davis BN, Hilyard AC, Lagna G et al (2008) SMAD proteins control DROSHA-mediated
microRNA maturation. Nature 454:56–61
Denli AM, Tops BB, Plasterk RH et al (2004) Processing of primary microRNAs by the
Microprocessor complex. Nature 432:231–235
Desterro JM, Keegan LP, Lafarga M et al (2003) Dynamic association of RNA-editing enzymes
with the nucleolus. J Cell Sci 116:1805–1818
Dickins RA, Hemann MT, Zilfou JT et al (2005) Probing tumor phenotypes using stable and
regulated synthetic microRNA precursors. Nat Genet 37:1289–1295
Diederichs S, Haber DA (2006) Sequence variations of microRNAs in human cancer: alterations in
predicted secondary structure do not affect processing. Cancer Res 66:6097–6104
Diederichs S, Haber DA (2007) Dual role for argonautes in microRNA processing and posttran-
scriptional regulation of microRNA expression. Cell 131:1097–1108
Diederichs S, Jung S, Rothenberg SM et al (2008) Coexpression of Argonaute-2 enhances RNA
interference toward perfect match binding sites. Proc Natl Acad Sci USA 105:9284–9289
Dietzl G, Chen D, Schnorrer F et al (2007) A genome-wide transgenic RNAi library for conditional
gene inactivation in Drosophila. Nature 448:151–156
Doench JG, Sharp PA (2004) Specificity of microRNA target selection in translational repression.
Genes Dev 18:504–511
Dye MJ, Gromak N, Proudfoot NJ (2006) Exon tethering in transcription by RNA polymerase II.
Mol Cell 21:849–859
Elbashir SM, Harborth J, Lendeckel W et al (2001a) Duplexes of 21-nucleotide RNAs mediate
RNA interference in cultured mammalian cells. Nature 411:494–498
Elbashir SM, Lendeckel W, Tuschl T (2001b) RNA interference is mediated by 21- and 22-nucleo-
tide RNAs. Genes Dev 15:188–200
Elbashir SM, Martinez J, Patkaniowska A et al (2001c) Functional anatomy of siRNAs for
mediating efficient RNAi in Drosophila melanogaster embryo lysate. EMBO J 20:6877–6888
Eulalio A, Behm-Ansmant I, Izaurralde E (2007a) P bodies: at the crossroads of post-transcriptional
pathways. Nat Rev Mol Cell Biol 8:9–22
Eulalio A, Rehwinkel J, Stricker M et al (2007b) Target-specific requirements for enhancers of
decapping in miRNA-mediated gene silencing. Genes Dev 21:2558–2570
Eulalio A, Huntzinger E, Izaurralde E (2008) Getting to the root of miRNA-mediated gene
silencing. Cell 132:9–14
Fang Y, Spector DL (2007) Identification of nuclear dicing bodies containing proteins for microRNA
biogenesis in living Arabidopsis plants. Curr Biol 17:818–823
Filippov V, Solovyev V, Filippova M et al (2000) A novel type of RNase III family proteins in
eukaryotes. Gene 245:213–221
Fire A, Xu S, Montgomery MK et al (1998) Potent and specific genetic interference by double-
stranded RNA in Caenorhabditis elegans. Nature 391:806–811
Forman JJ, Legesse-Miller A, Coller HA (2008) A search for conserved sequences in coding
regions reveals that the let-7 microRNA targets Dicer within its coding sequence. Proc Natl
Acad Sci USA 105:14879–14884
Forstemann K, Tomari Y, Du T et al (2005) Normal microRNA maturation and germ-line stem cell
maintenance requires Loquacious, a double-stranded RNA-binding domain protein. PLoS Biol
3:e236
Forstemann K, Horwich MD, Wee L et al (2007) Drosophila microRNAs are sorted into function-
ally distinct argonaute complexes after production by dicer-1. Cell 130:287–297
Fukuda T, Yamagata K, Fujiyama S et al (2007) DEAD-box RNA helicase subunits of the Drosha
complex are required for processing of rRNA and a subset of microRNAs. Nat Cell Biol
9:604–611
Gregory RI, Yan KP, Amuthan G et al (2004) The Microprocessor complex mediates the genesis
of microRNAs. Nature 432:235–240
Gregory RI, Chendrimada TP, Cooch N et al (2005) Human RISC couples microRNA biogenesis
and posttranscriptional gene silencing. Cell 123:631–640
microRNA Biogenesis and its Impact on RNA Interference 349
Grimm D, Streetz KL, Jopling CL et al (2006) Fatality in mice due to oversaturation of cellular
microRNA/short hairpin RNA pathways. Nature 441:537–541
Grimson A, Farh KK, Johnston WK et al (2007) MicroRNA targeting specificity in mammals:
determinants beyond seed pairing. Mol Cell 27:91–105
Guddeti S, Zhang DC, Li AL et al (2005) Molecular evolution of the rice miR395 gene family. Cell
Res 15:631–638
Guil S, Caceres JF (2007) The multifunctional RNA-binding protein hnRNP A1 is required for
processing of miR-18a. Nat Struct Mol Biol 14:591–596
Haase AD, Jaskiewicz L, Zhang H et al (2005) TRBP, a regulator of cellular PKR and HIV-1 virus
expression, interacts with Dicer and functions in RNA silencing. EMBO Rep 6:961–967
Hammond SM, Bernstein E, Beach D et al (2000) An RNA-directed nuclease mediates post-
transcriptional gene silencing in Drosophila cells. Nature 404:293–296
Han J, Lee Y, Yeom KH et al (2004a) The Drosha-DGCR8 complex in primary microRNA
processing. Genes Dev 18:3016–3027
Han MH, Goud S, Song L et al (2004b) The Arabidopsis double-stranded RNA-binding protein
HYL1 plays a role in microRNA-mediated gene regulation. Proc Natl Acad Sci USA
101:1093–1098
Han J, Lee Y, Yeom KH et al (2006) Molecular basis for the recognition of primary microRNAs by
the Drosha–DGCR8 complex. Cell 125:887–901
Han J, Pedersen JS, Kwon SC et al (2009) Posttranscriptional crossregulation between Drosha and
DGCR8. Cell 136:75–84
Heo I, Joo C, Cho J et al (2008) Lin28 mediates the terminal uridylation of let-7 precursor
MicroRNA. Mol Cell 32:276–284
Hiraguri A, Itoh R, Kondo N et al (2005) Specific interactions between Dicer-like proteins and
HYL1/DRB-family dsRNA-binding proteins in Arabidopsis thaliana. Plant Mol Biol
57:173–188
Hwang HW, Wentzel EA, Mendell JT (2007) A hexanucleotide element directs microRNA
nuclear import. Science 315:97–100
Jackson AL, Bartz SR, Schelter J et al (2003) Expression profiling reveals off-target gene
regulation by RNAi. Nat Biotechnol 21:635–637
Janowski BA, Huffman KE, Schwartz JC et al (2006) Involvement of AGO1 and AGO2 in
mammalian transcriptional silencing. Nat Struct Mol Biol 13:787–792
Kawahara Y, Zinshteyn B, Chendrimada TP et al (2007a) RNA editing of the microRNA-151
precursor blocks cleavage by the Dicer–TRBP complex. EMBO Rep 8:763–769
Kawahara Y, Zinshteyn B, Sethupathy P et al (2007b) Redirection of silencing targets by
adenosine-to-inosine editing of miRNAs. Science 315:1137–1140
Kawahara Y, Megraw M, Kreider E et al (2008) Frequency and fate of microRNA editing in
human brain. Nucleic Acids Res 36:5270–5280
Kawamata T, Seitz H, Tomari Y (2009) Structural determinants of miRNAs for RISC loading and
slicer-independent unwinding. Nat Struct Mol Biol 16:953–960
Kedde M, Strasser MJ, Boldajipour B et al (2007) RNA-binding protein Dnd1 inhibits microRNA
access to target mRNA. Cell 131:1273–1286
Khvorova A, Reynolds A, Jayasena SD (2003) Functional siRNAs and miRNAs exhibit strand
bias. Cell 115:209–216
Kim YK, Kim VN (2007) Processing of intronic microRNAs. EMBO J 26:775–783
Kim DH, Villeneuve LM, Morris KV et al (2006) Argonaute-1 directs siRNA-mediated transcriptional
gene silencing in human cells. Nat Struct Mol Biol 13:793–797
Knight SW, Bass BL (2001) A role for the RNase III enzyme DCR-1 in RNA interference and
germ line development in Caenorhabditis elegans. Science 293:2269–2271
Kumar MS, Lu J, Mercer KL et al (2007) Impaired microRNA processing enhances cellular
transformation and tumorigenesis. Nat Genet 39:673–677
Kurihara Y, Watanabe Y (2004) Arabidopsis micro-RNA biogenesis through Dicer-like 1 protein
functions. Proc Natl Acad Sci USA 101:12753–12758
350 S. Grund and S. Diederichs
Lagos-Quintana M, Rauhut R, Meyer J et al (2003) New microRNAs from mouse and human.
RNA 9:175–179
Lau NC, Lim LP, Weinstein EG et al (2001) An abundant class of tiny RNAs with probable
regulatory roles in Caenorhabditis elegans. Science 294:858–862
Lee CG, Hurwitz J (1992) A new RNA helicase isolated from HeLa cells that catalytically
translocates in the 30 to 50 direction. J Biol Chem 267:4398–4407
Lee Y, Jeon K, Lee JT et al (2002) MicroRNA maturation: stepwise processing and subcellular
localization. EMBO J 21:4663–4670
Lee Y, Ahn C, Han J et al (2003) The nuclear RNase III Drosha initiates microRNA processing.
Nature 425:415–419
Lee Y, Kim M, Han J et al (2004a) MicroRNA genes are transcribed by RNA polymerase II.
EMBO J 23:4051–4060
Lee YS, Nakahara K, Pham JW et al (2004b) Distinct roles for Drosophila Dicer-1 and Dicer-2 in
the siRNA/miRNA silencing pathways. Cell 117:69–81
Lee Y, Hur I, Park SY et al (2006) The role of PACT in the RNA silencing pathway. EMBO
J 25:522–532
Lee EJ, Baek M, Gusev Y et al (2008) Systematic evaluation of microRNA processing patterns in
tissues, cell lines, and tumors. RNA 14:35–42
Lewis BP, Burge CB, Bartel DP (2005) Conserved seed pairing, often flanked by adenosines,
indicates that thousands of human genes are microRNA targets. Cell 120:15–20
Li J, Yang Z, Yu B, Liu J et al (2005) Methylation protects miRNAs and siRNAs from a 30 -end
uridylation activity in Arabidopsis. Curr Biol 15:1501–1507
Lingel A, Simon B, Izaurralde E et al (2004) Nucleic acid 30 -end recognition by the Argonaute2
PAZ domain. Nat Struct Mol Biol 11:576–577
Lippman Z, Martienssen R (2004) The role of RNA interference in heterochromatic silencing.
Nature 431:364–370
Liu Q, Rand TA, Kalidas S et al (2003) R2D2, a bridge between the initiation and effector steps of
the Drosophila RNAi pathway. Science 301:1921–1925
Liu J, Valencia-Sanchez MA, Hannon GJ et al (2005) MicroRNA-dependent localization of
targeted mRNAs to mammalian P-bodies. Nat Cell Biol 7:719–723
Liu Y, Ye X, Jiang F et al (2009) C3PO, an endoribonuclease that promotes RNAi by facilitating
RISC activation. Science 325:750–753
Lobbes D, Rallapalli G, Schmidt DD et al (2006) SERRATE: a new player on the plant microRNA
scene. EMBO Rep 7:1052–1058
Lu J, Getz G, Miska EA et al (2005) MicroRNA expression profiles classify human cancers.
Nature 435:834–838
Lund E, Guttinger S, Calado A et al (2004) Nuclear export of microRNA precursors. Science
303:95–98
Ma JB, Ye K, Patel DJ (2004) Structural basis for overhang-specific small interfering RNA
recognition by the PAZ domain. Nature 429:318–322
Ma JB, Yuan YR, Meister G et al (2005) Structural basis for 50 -end-specific recognition of guide
RNA by the A. fulgidus Piwi protein. Nature 434:666–670
Ma E, MacRae IJ, Kirsch JF et al (2008) Autoinhibition of human dicer by its internal helicase
domain. J Mol Biol 380:237–243
Macrae IJ, Zhou K, Li F et al (2006) Structural basis for double-stranded RNA processing by
Dicer. Science 311:195–198
MacRae IJ, Ma E, Zhou M et al (2008) In vitro reconstitution of the human RISC-loading complex.
Proc Natl Acad Sci USA 105:512–517
Manche L, Green SR, Schmedt C, Mathews MB (1992) Interactions between double-stranded
RNA regulators and the protein kinase DAI. Mol Cell Biol 12:5238–5248
Maniataki E, Mourelatos Z (2005) A human, ATP-independent, RISC assembly machine fueled by
pre-miRNA. Genes Dev 19:2979–2990
microRNA Biogenesis and its Impact on RNA Interference 351
Provost P, Dishart D, Doucet J et al (2002) Ribonuclease activity and RNA binding of recombinant
human Dicer. EMBO J 21:5864–5874
Qi HH, Ongusaha PP, Myllyharju J et al (2008) Prolyl 4-hydroxylation regulates Argonaute
2 stability. Nature 455:421–424
Ramachandran V, Chen X (2008) Degradation of microRNAs by a family of exoribonucleases in
Arabidopsis. Science 321:1490–1492
Rand TA, Petersen S, Du F, Wang X (2005) Argonaute2 cleaves the anti-guide strand of siRNA
during RISC activation. Cell 123:621–629
Rivas FV, Tolia NH, Song JJ et al (2005) Purified Argonaute2 and an siRNA form recombinant
human RISC. Nat Struct Mol Biol 12:340–349
Ro S, Park C, Young D et al (2007) Tissue-dependent paired expression of miRNAs. Nucleic
Acids Res 35:5944–5953
Robb GB, Rana TM (2007) RNA helicase A interacts with RISC in human cells and functions in
RISC loading. Mol Cell 26:523–537
Robb GB, Brown KM, Khurana J et al (2005) Specific and potent RNAi in the nucleus of human
cells. Nat Struct Mol Biol 12:133–137
Rodriguez A, Griffiths-Jones S, Ashurst JL et al (2004) Identification of mammalian microRNA
host genes and transcription units. Genome Res 14:1902–1910
Ruby JG, Jan CH, Bartel DP (2007) Intronic microRNA precursors that bypass Drosha processing.
Nature 448:83–86
Rybak A, Fuchs H, Smirnova L et al (2008) A feedback loop comprising lin-28 and let-7 controls
pre-let-7 maturation during neural stem-cell commitment. Nat Cell Biol 10:987–993
Ryman K, Fong N, Bratt E et al (2007) The C-terminal domain of RNA Pol II helps ensure that
editing precedes splicing of the GluR-B transcript. RNA 13:1071–1078
Saito K, Ishizuka A, Siomi H et al (2005) Processing of pre-microRNAs by the Dicer-1-Loqua-
cious complex in Drosophila cells. PLoS Biol 3:e235
Salzman DW, Shubert-Coleman J et al (2007) P68 RNA helicase unwinds the human let-7
microRNA precursor duplex and is required for let-7-directed silencing of gene expression.
J Biol Chem 282:32773–32779
Sasaki T, Shiohama A, Minoshima S et al (2003) Identification of eight members of the Argonaute
family in the human genome small star, filled. Genomics 82:323–330
Scadden AD (2005) The RISC subunit Tudor-SN binds to hyper-edited double-stranded RNA and
promotes its cleavage. Nat Struct Mol Biol 12:489–496
Schwarz DS, Hutvagner G, Du T et al (2003) Asymmetry in the assembly of the RNAi enzyme
complex. Cell 115:199–208
Seitz H, Ghildiyal M, Zamore PD (2008) Argonaute loading improves the 50 precision of both
MicroRNAs and their miRNA strands in flies. Curr Biol 18:147–151
Sen GL, Blau HM (2005) Argonaute 2/RISC resides in sites of mammalian mRNA decay known
as cytoplasmic bodies. Nat Cell Biol 7:633–636
Silva JM, Li MZ, Chang K et al (2005) Second-generation shRNA libraries covering the mouse
and human genomes. Nat Genet 37:1281–1288
Siolas D, Lerner C, Burchard J et al (2005) Synthetic shRNAs as potent RNAi triggers. Nat
Biotechnol 23:227–231
Sledz CA, Holko M, de Veer MJ et al (2003) Activation of the interferon system by short-
interfering RNAs. Nat Cell Biol 5:834–839
Song JJ, Liu J, Tolia NH et al (2003) The crystal structure of the Argonaute2 PAZ domain reveals
an RNA binding motif in RNAi effector complexes. Nat Struct Biol 10:1026–1032
Song JJ, Smith SK, Hannon GJ et al (2004) Crystal structure of Argonaute and its implications for
RISC slicer activity. Science 305:1434–1437
Song L, Han MH, Lesicka J et al (2007) Arabidopsis primary microRNA processing proteins
HYL1 and DCL1 define a nuclear body distinct from the Cajal body. Proc Natl Acad Sci USA
104:5437–5442
microRNA Biogenesis and its Impact on RNA Interference 353
Zeng Y, Cullen BR (2003) Sequence requirements for micro RNA processing and function in
human cells. RNA 9:112–123
Zeng Y, Cullen BR (2004) Structural requirements for pre-microRNA binding and nuclear export
by Exportin 5. Nucleic Acids Res 32:4776–4785
Zeng Y, Cullen BR (2005) Efficient processing of primary microRNA hairpins by Drosha requires
flanking nonstructured RNA sequences. J Biol Chem 280:27595–27603
Zeng Y, Wagner EJ, Cullen BR (2002) Both natural and designed micro RNAs can inhibit the
expression of cognate mRNAs when expressed in human cells. Mol Cell 9:1327–1333
Zeng Y, Cai X, Cullen BR (2005a) Use of RNA polymerase II to transcribe artificial microRNAs.
Methods Enzymol 392:371–380
Zeng Y, Yi R, Cullen BR (2005b) Recognition and cleavage of primary microRNA precursors by
the nuclear processing enzyme Drosha. EMBO J 24:138–148
Zeng Y, Sankala H, Zhang X et al (2008) Phosphorylation of Argonaute 2 at serine-387 facilitates
its localization to processing bodies. Biochem J 413:429–436
Zhang H, Kolb FA, Brondani V et al (2002) Human Dicer preferentially cleaves dsRNAs at their
termini without a requirement for ATP. EMBO J 21:5875–5885
Zhang H, Kolb FA, Jaskiewicz L et al (2004) Single processing center models for human Dicer and
bacterial RNase III. Cell 118:57–68
Zhang B, Pan X, Stellwag EJ (2008) Identification of soybean microRNAs and their targets. Planta
229:161–182
MicroRNAs in Epithelial Antimicrobial
Immunity
Jun Liu, Guoku Hu, Rui Zhou, Kristen M. Drescher, and Xian-Ming Chen
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
2 Abundant Expression of miRNAs in Epithelial Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
3 Regulation of miRNA Expression in Epithelial Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
4 MicroRNAs in the Regulation of Epithelial Antimicrobial Defense . . . . . . . . . . . . . . . . . . . . . . 360
4.1 MicroRNAs and Maintenance of Epithelial Barrier Integrity . . . . . . . . . . . . . . . . . . . . . . . 360
4.2 MicroRNAs and Regulation of Epithelial Intracellular Signaling Pathways . . . . . . . . 361
4.3 MicroRNAs and Expression of B7-Costimulatory Molecules in Epithelial Cells . . . 363
4.4 MicroRNAs in the Exosomes Released from Epithelial Cells . . . . . . . . . . . . . . . . . . . . . . 363
4.5 MicroRNAs-Mediated Antivirus Response in Epithelial Cells . . . . . . . . . . . . . . . . . . . . . 363
5 Conclusion and Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 355
RNA Technologies, DOI 10.1007/978-3-642-12168-5_16,
# Springer-Verlag Berlin Heidelberg 2010
356 J. Liu et al.
Abbreviations
1 Introduction
Epithelial cells at skin and mucosal sites represent the host’s first line of defense
against microbial infection. Beyond their role in creating a physical barrier to
potential disease agents, epithelial cells are critical in the initiation, regulation,
and resolution of both innate and adaptive immune responses to infection (Sansonetti
2004; Viswanathan and Hecht 2000). Epithelial cells express several pathogen-
sensing receptors, such as Toll-like receptors (TLRs) and NOD-like receptors
(NLRs). Through these receptors, epithelial cells recognize pathogen infection
MicroRNAs in Epithelial Antimicrobial Immunity 357
and evoke diverse responses including antimicrobial peptide and cytokine release
(Akira and Takeda 2004). Epithelial cells also express a wide range of proteins
critical to both the innate and adaptive immune responses, including major histo-
compatibility complex (MHC) class I and II, B7 costimulatory molecules, chemo-
kines, cytokines, and prostaglandins, which work in concert to eliminate microbes
(Viala et al. 2004; Menendez and Brett 2007). The epithelial immune response to
infection is finely controlled and reflects a delicate balance between effector
functions and the potential of the immune response to cause damage to healthy
tissue (Han and Ulevitch 2005; Shibolet and Podolsky 2007; Michael et al. 2006).
The importance of microRNAs (miRNA) in regulating normal cellular functions
is becoming increasingly clear as more miRNA targets are discovered and the
molecular mechanisms underlying miRNA gene regulation becomes unraveled
(Bartel 2009). Recent studies implicate specific miRNAs in controlling regulation
of cellular differentiation, determination of cell fate (cell death and proliferation),
initiation and regulation of antimicrobial immunity, control of inflammatory
responses, and activation of intracellular signaling pathways in epithelial cells
(Moschos et al. 2007; Chen et al. 2007; Pedersen et al. 2007; Gong et al. 2009;
Zhai et al. 2008; Yi et al. 2006; Andl et al. 2006; Harris et al. 2006; Hino et al. 2007;
Lu et al. 2007; Yi et al. 2008; Otsuka et al. 2007; Jopling et al. 2005; Sarasin-
Filipowicz et al. 2009). Because these functions are also involved in the host’s
response to infection, it is likely that miRNAs modify the epithelial immune
response to permit optimal responses to pathogens. This chapter summarizes recent
advances in the identification and expression of epithelial cell miRNAs and highlights
the functional significance of miRNA expression in immunoregulation of epithelial
antimicrobial defense, including maintenance of epithelial barrier integrity, regulation
of intracellular signaling pathways, and epithelial antiviral defense. Major findings in
this emerging field are summarized in Table 1.
Studies on numerous human cell lines derived from normal epithelial cells demon-
strated the abundant expression of miRNA molecules in epithelial cells (Chen et al.
2007; Mattick and Makunin 2005; Liu et al. 2004; Krichevsky et al. 2003). As an
example, over 70 miRNAs were detected in H69 cells, a line of human biliary
epithelial cells (Chen et al. 2007). As expected, the miRNA expression profiles
are distinct among epithelial cells from different tissues as demonstrated by micro-
array technology and validated by quantitative real-time PCR approach
(Krichevsky et al. 2003). MicroRNA-122, a tissue-specific miRNA in hepatocytes
(the major epithelial cell type in the liver), has been found to play essential roles in
diverse hepatic functions including tumorigenesis, antiviral responses, and choles-
terol biosynthesis (Girard et al. 2008). Unique miRNA expression profiles have also
been described in normal epithelial cells from lung, breast, stomach, prostate,
colon, and pancreas (Calin and Croce 2006; Lu et al. 2005; Volinia et al. 2006).
358 J. Liu et al.
Additionally, miRNA signatures in cancer cell lines of epithelial origin are distinct
from those in normal epithelial cells. The unique expression patterns of these
powerful gene regulators in a tissue- and differentiation-specific manner further
solidifies the critical nature of miRNAs in host homeostasis and defense from
pathogens.
While the previous section described the impact of immune stimulation via either
pathogen-associated molecular patterns (PAMPs) or host immunomodulators on
miRNA expression, miRNAs also appear to impact signaling pathways directly.
Several recent studies reported that miRNAs control TLRs protein expression in
epithelial cells under certain conditions. MicroRNA-105 modulates TLR2 expres-
sion in human oral keratinocytes and let-7 targets TLR4 expression in human
cholangiocytes (Chen et al. 2007; Benakanakere et al. 2009). Cryptosporidium
parvum infection induced let-7 downregulation via the TLR/MyD88/NF-kB path-
way activation and enhanced TLR4 expression. Experimental manipulation
of let-7i expression caused reciprocal alterations in the infection dynamics of
C. parvum in vitro (Chen et al. 2007). Since TLRs recognize PAMPs and are key
modulators of epithelial cell immune responses to microbial infection, these data
support the critical role of miRNAs to host-cell regulatory responses to microbial
362 J. Liu et al.
The role of epithelial cell miRNAs in the control of microbial infections has
recently been investigated. Otsuka et al. demonstrated that Dicer knockout mice
364 J. Liu et al.
were highly susceptible to vesicular stomatitis virus (VSV) infection (Otsuka et al.
2007). To study the mechanism of this increase in VSV susceptibility, segments of
the VSV genome were fused to the 30 UTR of a luciferase reporter gene and the
resulting plasmid transfected into RAW 264.7 cells. Three VSV genome sequences
were identified that decreased reporter gene expression in these cells. Further
studies demonstrated that four miRNAs (miR-24, miR-93, miR-146, and miR-378)
were highly expressed by the RAW 264.7 cells with the potential to target VSV.
None of these specific miRNAs were detected in macrophages isolated from Dicer
knockout mice. Transfection of the RAW 264.7 cells with either anti-miR-24 or anti-
miR-93 resulted in 4- to 5-fold increase in virus titer, suggesting that miR-24 and
miR-93 are involved in controlling VSV replication in the host. Although it remains
to be determined if these miRNAs inhibit VSV replication in the virus’s natural hosts,
the abundant expression of these specific miRNAs at the site of VSV replication in
the epithelial layer suggests that miR-24 and miR-93 may participate in defense of the
epithelial barrier.
In hepatocytes, eight IFN-inducible miRNAs (miR-1, miR-30, miR-128, miR-
196, miR-296, miR-351, miR-431, and miR-448) have been shown to have near
perfect complementarities between their seed sequences and the hepatitis C virus
(HCV) RNA genome (Pedersen et al. 2007). Transfection of HCV replicon-containing
hepatocytes with precursors of these eight miRNAs decreased the levels of HCV
RNA accumulation in the cells. Functional inhibition of these particular miRNAs
abrogated the inhibitory effect of IFN on HCV replication in hepatocytes. IFN-b
treatment also decreased miR-122 expression, a liver specific miRNA essential
for HCV replication in hepatocytes (Jopling et al. 2005). The downregulation of
miR-122 in response to IFN-b further enhanced the antiviral effects of this cytokine.
Together, the data suggest a novel mechanism involving miRNA-mediated gene
targeting to fight HCV infection in hepatocytes (Pedersen et al. 2007).
The initial data derived from these in vitro studies are not compatible with some
in vivo data thus far. A very recent research examined miR-122 levels in liver
biopsies from 42 patients with chronic hepatitis C (CHC) undergoing IFN treatment
(Sarasin-Filipowicz et al. 2009). Pretreatment levels of miR-122 in nonresponders
were several times lower than miR-122 levels in responders. Given the results from
in vitro studies suggesting that miR-122 is crucial for efficient replication of HCV
in hepatocytes, this finding in CHC patients is unexpected and suggests that the
impact of miR-122 on HCV replication may be less pronounced in vivo than it is
in vitro, probably a result of the complex in vivo interactions that are difficult to
model in tissue culture.
The study of miRNAs is flourishing in the decade after their discovery. It is clear
that miRNAs have the potential to affect every aspect of cellular function, from cell
differentiation and proliferation to apoptotic death. miRNA appear to regulate a
MicroRNAs in Epithelial Antimicrobial Immunity 365
References
Akira S, Takeda K (2004) Toll-like receptor signaling. Nat Rev Immunol 4:499–511
Andl T, Murchison EP, Liu F et al (2006) The miRNA-processing enzyme dicer is essential for the
morphogenesis and maintenance of hair follicles. Curr Biol 16:1041–1049
Baetz A, Frey M, Heeg K et al (2004) Suppressor of cytokine signaling (SOCS) proteins indirectly
regulate toll-like receptor signaling in innate immune cells. J Biol Chem 279:54708–54715
Bartel DP (2009) MicroRNAs: target recognition and regulatory functions. Cell 136:215–233
Benakanakere MR, Li Q, Eskan MA et al (2009) Modulation of TLR2 protein expression by a
miR-105 in human oral keratinocytes. J Biol Chem 284:23107–23115
Bracken CP, Gregory PA, Khew-Goodall Y et al (2008) MicroRNAs as regulators of epithelial–
mesenchymal transition. Cell Cycle 7:3112–3118
B€uning J, von Smolinski D, Tafazzoli K et al (2008) Multivesicular bodies in intestinal epithelial
cells: responsible for MHC class II-restricted antigen processing and origin of exosomes.
Immunology 125:510–521
Calin GA, Croce CM (2006) MicroRNA signatures in human cancers. Nat Rev Cancer 6:857–866
Chen ZJ (2005) Ubiquitin signalling in the NF-kappaB pathway. Nat Cell Biol 7:758–765
Chen XM, Splinter PL, O’Hara SP et al (2007) A cellular micro-RNA, let-7i, regulates Toll-like
receptor 4 expression and contributes to cholangiocyte immune responses against Cryptospo-
ridium parvum infection. J Biol Chem 282:28929–28938
Gebeshuber CA, Zatloukal K, Martinez J et al (2009) MiR-29a suppresses tristetraprolin, which is
a regulator of epithelial polarity and metastasis. EMBO Rep 10:400–405
Girard M, Jacquemin E, Munnich A et al (2008) MiR-122, a paradigm for the role of microRNAs
in the liver. J Hepatol 48:648–656
Gong AY, Zhou R, Hu G et al (2009) MicroRNA-513 regulates B7-H1 translation and is involved
in IFN-gamma-induced B7-H1 expression in cholangiocytes. J Immunol 182:1325–1333
Gottipati S, Rao NL, Fung-Leung WP (2008) IRAK1: a critical signaling mediator of innate
immunity. Cell Signal 20:269–276
Han J, Ulevitch RJ (2005) Limiting inflammatory responses during activation of innate immunity.
Nat Immunol 6:1198–1205
Harris KS, Zhang Z, McManus MT et al (2006) Dicer function is essential for lung epithelium
morphogenesis. Proc Natl Acad Sci USA 103:2208–2213
366 J. Liu et al.
Tetsuro Hirose
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370
2 Intracellular Behaviors of ncRNAs Distinct from Those of mRNAs . . . . . . . . . . . . . . . . . . . . . 370
3 Unique Pathways for Long ncRNA Biogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372
4 ncRNA Functions in the Regulation of Gene Expression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374
4.1 Regulation of Transcription Factor Activity by Long ncRNAs . . . . . . . . . . . . . . . . . . . . . 374
4.2 ncRNA Transcription Affects Adjacent Gene Expression . . . . . . . . . . . . . . . . . . . . . . . . . . 376
4.3 ncRNA Recruits or Modulates Epigenetic Factors on the Chromosome . . . . . . . . . . . . 378
4.4 ncRNAs Regulate Gene Expression at Posttranscriptional Steps . . . . . . . . . . . . . . . . . . . 380
5 Structural Roles of ncRNAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
6 ncRNAs in Biomedical Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
7 Future Directions for ncRNA Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
Abstract In the postgenomic era, we have learned that large numbers of RNAs that
do not code for proteins, the so-called noncoding RNAs (ncRNAs), are transcribed
from large portions of the intergenic regions in mammalian genomes. The
biological significance of these ncRNAs remains elusive. Although the research
is still limited, recent progress has revealed that several ncRNAs play important
roles in various steps of gene expression, including epigenetic chromatin regula-
tion, transcription, RNA processing, protein assembly, and transport. Novel ncRNA
functions in the organization of intracellular structures have also been reported.
Major ncRNA subsets are expressed in a tissue-specific manner and some are
induced by external stimuli. The expression of numerous ncRNAs is drastically
changed in some types of cancer cells, suggesting that ncRNAs may be involved in
disease as well as in physiological events. Thus, understanding ncRNA functions
T. Hirose
Biomedicinal Information Research Center, National Institute of Advanced Industrial Science
and Technology (AIST), 2-4-7 Aomi, Koutou, 135-0064, Tokyo, Japan
e-mail: tets-hirose@aist.go.jp
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 369
RNA Technologies, DOI 10.1007/978-3-642-12168-5_17,
# Springer-Verlag Berlin Heidelberg 2010
370 T. Hirose
and mechanisms of action will open new opportunities for developing RNA-based
technologies for pharmaceutical application.
1 Introduction
The transcriptome has primarily been analyzed using the sequences of full-length
cDNAs that were constructed from transcripts possessing both a cap structure at the
50 terminus and a poly(A) tail at the 30 terminus; therefore, the ncRNAs that
emerged through these analyses are likely to be products of RNA polymerase II.
The biogenesis pathway of protein-coding mRNAs that are produced by RNA
polymerase II has been studied for a long time and is well characterized (Fig. 1)
(Dreyfuss et al. 2002). It is an amazingly organized multistep process in which
quality control mechanisms eliminate aberrantly synthesized mRNAs that may
eventually produce harmful polypeptides. In contrast, the biogenesis pathways of
RNA polymerase II-produced ncRNAs remain to be investigated. To what extent
are ncRNAs subject to the rules of mRNA biogenesis (Fig. 1)? One form of quality
control in mRNA biogenesis, nonsense-mediated decay (NMD), recognizes aberrant
mRNAs to be eliminated by detecting the presence of a premature termination
Emerging Roles of Long Noncoding RNAs in Gene Expression and Intracellular 371
mRNA ncRNA
AAAAA
AAAAA
Nuc
Cyt
AAAAA
Functional polypeptide
Fig. 1 Comparison of the defined expression pathway for protein-coding genes and a putative
pathway for noncoding genes. The biogenesis pathway and intracellular localization of most
ncRNAs remain to be investigated. The mechanisms of ncRNA biogenesis that are distinct from
those for mRNA will be the focus of studies in the near future
codon in mRNAs (Chang et al. 2007; Isken and Maquat 2007). However, canonical
ncRNAs lack meaningful ORFs and instead have plenty of termination codons,
which is why they are called “noncoding” RNAs. Indeed, two ncRNAs, gas5 and
UHG, which are known to be host genes of small nucleolar RNAs (snoRNAs) are
rapidly degraded by the NMD pathway in mammalian cells (Tycowski et al. 1996;
Smith and Steitz 1998; Ideue et al. 2007). Because mammalian snoRNAs are
processed from excised introns, some of the spliced host genes are thought to be
nonfunctional RNAs that can be rapidly degraded in the cytoplasm. The degrada-
tion is arrested by either cycloheximide treatment or knockdown of a NMD factor,
UPF1, indicating that these ncRNAs are natural NMD targets (Tycowski et al.
1996; Smith and Steitz 1998; Ideue et al. 2007). Recently, a genome-wide search of
NMD targets in Arabidopsis plants revealed that 20% of known ncRNAs are
natural NMD targets, and the majority of these 20% are natural antisense ncRNAs
(Kurihara et al. 2009). The majority of NMD-insusceptible ncRNAs may avoid
entering the NMD pathway because ncRNA genes lack introns. ncRNAs are
retained in the nucleus and are never exported to the cytoplasm, and ncRNAs
would not be recognized by ribosomes even if they were transported to the
cytoplasm. A genome-wide analysis of gene structures for mRNAs and ncRNAs
in mouse revealed an apparent structural difference, in that 72% of the mapped
ncRNA sequences were uninterrupted by an intron, whereas only 18% of the
mRNAs were unspliced (Ravasi et al. 2006). As described below, numerous long
ncRNAs that have been functionally annotated play important biological roles in
nuclear events, suggesting that an ncRNA subpopulation is retained in the nucleus
and therefore would not encounter the NMD machinery in the cytoplasm. Indeed,
our data on 70 long ncRNAs selected from a human full-length cDNA database
(Sasaki et al. 2007) showed that >70% of them are predominantly localized to the
nuclear fraction and are therefore insusceptible to NMD (our unpublished results).
372 T. Hirose
Some long ncRNAs mature via unique RNA processing events that have not been
observed in the well-characterized mRNA biogenesis pathway (Srikantan et al.
2000; Dreyfuss et al. 2002; Kiyosawa et al. 2005; Ganesan and Rao 2008; Wilusz
et al. 2008). All eukaryotic mRNAs possess poly(A) tails at their 30 termini, with the
exception of histone mRNAs, which lack the poly(A) tail at their 30 termini and
instead possess conserved stem–loop structures that contribute to their cell cycle-
dependent regulation. The two long ncRNA genes for Malat-1 (multiple endocrine
neoplasia a [MENa]) and MENb adjoin each other at a chromosome 11 locus in
human and on the syntenic locus on chromosome 19 in mouse (Guru et al. 1997;
Hutchinson et al. 2007). The 30 terminus of the major Malat-1 transcript lacks a poly
(A) tail (Fig. 2a). The evolutionarily conserved tRNA-like structure located down-
stream of the 30 terminus is recognized and processed by the tRNA processing
enzyme RNase P in vitro, resulting in the creation of a nonpolyadenylated 30
terminus in the Malat-1 ncRNA (Fig. 2a) (Wilusz et al. 2008). Subsequently, a
similar tRNA-like structure was found in the downstream region of the 30 terminus
of the MENb ncRNA, which is transcribed from the chromosome locus adjacent to
the Malat-1 gene. The tRNA-like structure is also recognized and processed by
RNase P, creating a nonpolyadenylated 30 terminus in the MENb ncRNA (Sunwoo
et al. 2009). The cleaved downstream portion of the Malat-1 precursor containing
the tRNA-like structure is further processed by RNaseZ and CCA-adding enzyme
to produce a tRNA-like small RNA (mascRNA) that is subsequently transported to
the cytoplasm. The tRNA-like portion of the MENb precursor accumulates at low
levels, probably due to its unstable structure. The mechanism protecting the non-
polyadenylated 30 termini from RNA degradation has remained elusive. A U-rich
region located upstream of the 30 terminus of Malat-1 is critical for Malat-1’s stable
accumulation, raising the possibility that the U-rich region interacts with the
genomically encoded A-rich tract located at the 30 terminus (Wilusz et al. 2008).
This model is consistent with that proposed for the stable nuclear accumulation of
Kaposi’s sarcoma-associated herpesvirus polyadenylated nuclear RNA (Conrad
et al. 2006, 2007). Although the significance of these nonpolyadenylated 30 termini
of Malat-1 and MENb ncRNAs, as well as the function of mascRNA, remains
explainable, their presence is noteworthy as it represents a novel mechanism in
which the 30 termini of long ncRNAs that are RNA polymerase II products are
processed by a tRNA-processing enzyme to produce small RNAs. Because the long
ncRNAs have been mainly identified by large-scale transcriptome analyses in
which full-length cDNA clones derived from polyadenylated RNAs were sequenced,
long ncRNAs with nonpolyadenylated 30 termini may be missing from transcriptome
lists. A genome-wide analysis of natural antisense transcripts revealed that vast
amounts of antisense noncoding transcripts of various sizes are expressed in
mice, including large numbers of nonpolyadenylated and nuclear-localized ncRNAs
(Kiyosawa et al. 2005). Although we cannot rule out the possibility that the
detected nonpolyadenylated ncRNAs are transcriptionally nascent RNAs, the further
Emerging Roles of Long Noncoding RNAs in Gene Expression and Intracellular 373
a
AAAAAAAAA
RNase P
AAAAA AAAAA
Fig. 2 New pathways for ncRNA biogenesis. (a) 30 end processing of Malat1 and MENb ncRNAs by
cleavage with RNase P. The downstream tRNA-like structure acts as a processing signal that is
recognized by RNase P. The tRNA-like region cleaved from the Malat-1 precursor is stably
accumulated as mascRNA and transported to the cytoplasm, whereas the mature Malat1 is exclu-
sively localized in the nucleus. (b) Posttranscriptional processing and capping of mature mRNAs and
ncRNAs. Extensive analysis of CAGE tags revealed the existence of small RNAs produced from
mature mRNAs and long ncRNAs, some of which have been spliced. The small RNAs may be
endonucleolytically processed, followed by the addition of a cap structure at their 50 termini
cap structure (Fig. 2b) (Fejes-Toth et al. 2009). Although many CAGE tags mark
transcription start sites, significant numbers were found in exonic regions and, in
some cases, even crossed splice junctions, indicating that they must have arisen
from at least partially processed mRNAs. Therefore, it has been proposed that
mature long transcripts, including both mRNAs and long ncRNAs, can be pro-
cessed posttranscriptionally to yield small RNAs, which are subsequently modified
by the addition of a cap structure (Fejes-Toth et al. 2009) (Fig. 2b).
region (Feng et al. 2006; Kohtz and Fishell 2004). Evf-1 is a 2.7-kb polyadenylated
RNA, and its expression is developmentally regulated (Kohtz and Fishell 2004).
The Evf-2 ncRNA (3.8 kb) specifically cooperates with the homeodomain protein
Dlx-2 to increase the transcriptional activity of the Dlx-5/6 enhancer region.
Interestingly, a stable complex containing the Evf-2 ncRNA/Dlx-2 homeodomain
protein forms in the nucleus (Feng et al. 2006). These data suggest that the Evf-2/
Dlx-2 complex stabilizes the interaction between Dlx-2 and its target Dlx-5/6
enhancer sequences to increase transcriptional activity. The possible role of Evf-2
produced from the genomic ultraconserved region suggests that other ultracon-
served regions may function to produce ncRNA regulators for key developmental
processes.
The NRON (noncoding repressor of NFAT) ncRNA was identified in an exten-
sive RNAi screening of about 500 long ncRNAs that were selected as evolutionarily
conserved ncRNAs from the FANTOM mouse full length cDNA database
(Willingham et al. 2005). The NRON ncRNA is alternatively spliced to produce
0.8–3.7 kb isoforms. Knockdown of NRON results in a dramatic activation of Ca++-
dependent NFAT (nuclear factor of activated T-cell) activity, suggesting that
NRON represses NFAT function. Characterization of NRON-binding proteins by
RNA affinity purification was used to identify the mechanism of NRON action and
members of the importin-b family, which likely function in NFAT nuclear trafficking,
were identified (Fig. 3a). Taken together, these findings show that NRON represses
NFAT function by capturing nuclear transporter proteins to prevent the nuclear
import of NFAT (Fig. 3a) (Willingham et al. 2005).
The heat-shock RNA-1 (HSR1) ncRNA was identified as a necessary factor for
activation of the heat-shock transcription factor 1 (HSF1) upon heat shock (Fig. 3b)
(Shamovsky et al. 2006). HSF1 has an important role in the heat-shock response in
vertebrates by inducing the expression of heat-shock proteins (HSPs) and other
cytoprotective proteins. HSF1 is present in unstressed cells in an inactive mono-
meric form and becomes activated by heat and other stress stimuli. HSF1 activation
involves trimerization to acquire site-specific DNA-binding activity, which is
negatively regulated by interaction with certain HSPs. HSF1 activation by heat
shock is mediated by a ribonucleoprotein complex containing the translation
elongation factor eEF1A and HSR1 ncRNA. Both HSR1 and eEF1A are required
for HSF1 activation in vitro, and the specific knockdown of HSR1 impairs the heat-
shock response, rendering cells thermosensitive (Fig. 3b).
In addition to specific transcription factors, RNA polymerase II is directly
targeted by ncRNAs (Allen et al. 2004; Espinoza et al. 2004; Mariner et al.
2008). Several SINE-encoded RNAs such as B2 in mouse and Alu in human
directly bind to RNA polymerase II during heat shock and globally inhibit
mRNA transcription. Thus, ncRNAs are involved in the various aspects of tran-
scriptional regulation by modulating the activities of transcription factors. The
ncRNAs described above are effectors for specific transcription factors, suggesting
that many more ncRNAs will be discovered that regulate the activity of hundreds of
transcription factors in mammalian cells.
376 T. Hirose
a b p23
Immunophilin
HSP90
HSF
Heat shock
p23
HSP90 Immunophilin
Ca++ signal
HSF
eEF1A
NFAT
P
NFAT
HSR1
HSF
calcineurin
NRON NFAT
importinb
HSF Active
c PRC2 complex
d
Long ncRNAs
IRES
Recruitment and
histone methylation
spliceosome
K27K27K27 IRES
3Me3Me3Me
ribosome
Fig. 3 Diverse functions of long ncRNAs involved in various processes of gene expression. (a)
NRON ncRNA captures importin-b and negatively controls the nuclear transport of the NFAT
transcription factor upon Ca++ signaling. (b) NSR1 ncRNA and the associated eEF1A facilitate the
trimerization of the HSF1 transcription factor upon heat shock. (c) An ncRNA subset including
Hotair, RepA, and Kcnqot1 act to recruit the histone modification complex to the specific target
region on the chromosome. (d) The antisense ncRNA regulates pre-mRNA splicing to produce the
translatable mRNA isoform possessing an IRES
a b
Basic transcription
factors
Transcription factors
DHFR
Triple helix
c d
TLS/FUS I. Retention and function in cis
p300 CBP
CCND
Fig. 4 Promoter-associated ncRNAs regulate transcription through distinct mechanisms. (a) Tran-
scriptional interference. The transcription of SRG ncRNA from the flanking region inhibits SER3
gene transcription in cis. (b). The transcript from the flanking promoter region of a DHFR gene
forms a triple helix with promoter DNA that leads to interference with the transcriptional initiation
of the DHFR gene. The ncRNAs act both in cis and in trans. (c). The ncRNAs that are transcribed
from the CCND promoter region bind to an RNA-binding protein, TLS/FUS, which inhibits CBP/
p300 histone acetyl transferase activity and is allosterically regulated by the bound ncRNAs.
(d). Three basic mechanisms of action of promoter-associated ncRNA. First, the produced ncRNA
is retained at the transcription site and functions only in cis. Second, transcription interference
inhibits transcription downstream. The transcription event itself is enough to inhibit the down-
stream gene in cis. Third, the produced ncRNA may diffuse and act on the promoter region both in
cis and in trans
DHFR promoter and directly interacting with TFIIB, which results in the disruption
of the preinitiation complex at the DHFR promoter (Fig. 4a) (Martianov et al. 2007).
In contrast, ncRNA transcription positively regulates the expression of adjacent
genes. Recently, it was reported that transcription of long ncRNAs upstream of the
Schizosaccharomyces pombe fbp1+ locus induces chromatin remodeling that is
critical for transcriptional activation of the downstream protein-coding gene (Hirota
et al. 2008). In this case, ncRNA transcription is initiated in a stepwise manner from
multiple sites in the fbp1+ promoter, causing chromatin opening to proceed pro-
gressively toward the mRNA transcription start site. The insertion of a transcrip-
tional terminator within these ncRNA regions prevents downstream chromatin
remodeling, resulting in inefficient transcriptional induction because of the reduced
recruitment of transcription factors to the fbp1+ promoter. Noncoding transcription
also plays a role in activation of the S. cerevisiae PHO5 gene. A 2.4-kb antisense
ncRNA transcribed from near the 30 end of the yeast PHO5 gene acts to evict
histones from the PHO5 gene promoter during the repression under high phosphate
conditions, making it possible to respond rapidly to the signal for gene activation
(Uhler et al. 2007). In Drosophila, intergenic transcription through cis-acting
negative regulatory elements in the promoter regions prevents the silencing of
certain Hox genes by Polycomb group (PcG) proteins (Bender and Fitzgerald
2002; Hogga and Karch 2002; Schmitt et al. 2005; Preker et al. 2008). This seems
to be the opposite of transcriptional interference, where the transcription of ncRNAs
prevents binding of the positive transcription factors to the promoter regions.
An important question is whether chromatin remodeling is induced by the
transcription events that produce ncRNAs or by the produced ncRNAs themselves.
To answer this question, researchers examined whether the ncRNA can act in trans,
or whether knockdown of the accumulated ncRNAs disturbs the possible ncRNA’s
regulatory function. At least at the yeast Ser3, IME4, and PHO5 loci, and Drosophila
Ubx locus, it appears to be the act of noncoding transcription rather than the ncRNA
itself that contributes to the regulation of the expression of protein-coding genes
because these cognate ncRNAs cannot act in trans (Hongay et al. 2006; Uhler et al.
2007). On the other hand, at least at the human DHFR locus, the ncRNA does work
in trans (Martianov et al. 2007). Therefore, depending on the gene locus, ncRNA
transcription can act either negatively or positively. In some cases, the act of tran-
scription is sufficient to have functional consequences, but it is likely that many of the
ncRNAs play diverse regulatory roles with currently unknown mechanisms.
in specific genomic locations (Silva et al. 2003; Petruk et al. 2006). There are
precedents for ncRNA involvement in X-chromosome inactivation (XCI) in female
mammalian cells. The XCI center harbors two major long ncRNA genes, the 17-kb
Xist and its antisense repressor, the 40-kb Tsix (Masui and Heard 2006; Peters and
Robson 2008). Tsix blocks Xist expression and prevents the recruitment of silenc-
ing factors in cis on the future active X chromosome. In contrast, on the future
inactive X chromosome, Tsix is downregulated, leading to Xist expression and the
spread of Xist RNA along the chromosome (Masui and Heard 2006; Peters and
Robson 2008). The accumulation of Xist transcripts correlates with chromatin
alteration (Masui and Heard 2006; Payer and Lee 2008), but how Xist directs
these alterations is unknown. Recently, the catalytic subunit of Polycomb-repressive
complex 2 (PRC2) was found to directly bind an independent, shorter ncRNA derived
from Xist RNA called RepA (Zhao et al. 2008). The RepA ncRNA is transcribed
from the Repeat A region of the Xist gene and has been proposed to play a key role in
the early stages of mammalian XCI (Zhao et al. 2008).
Hotair is the only long ncRNA that recruits the PcG complex to another
chromosome locus in trans (Rinn et al. 2007). Hotair, a 2.2-kb ncRNA that is
transcribed from the HOXC locus, regulates the HOXD locus in trans by recruiting
the PRC2 complex (Rinn et al. 2007). Recently, Guttman et al. performed a
genome-wide search for histone H4K36 trimethylation marks in intergenic regions
and found that large numbers of large intergenic ncRNAs (lincRNAs) (Guttman
et al. 2009, Khalil et al. 2009) further showed that numerous lincRNAs are
associated with PRC2 and negatively regulate thousands of protein-coding genes.
Therefore, the interaction of a long ncRNA with the PRC2 complex is a general
epigenetic regulatory mechanism for recruiting chromosomal silencing factors to
specific chromosomal locations (Fig. 3c).
Trithorax group (TrxG) proteins counteract the silencing functions of PcG
proteins to maintain active transcription states. Certain ncRNAs from the Hox
loci were shown to interact directly with the histone methyltransferase Ash1
in vitro and were proposed to target TrxG proteins to chromatin (Sanchez-Elsner
et al. 2006). In fact, ectopic expression of these ncRNAs in trans was found to
activate Ubx gene expression (Sanchez-Elsner et al. 2006). On the other hand,
another report failed to observe a similar association between ectopic expression of
these ncRNAs and transcriptional activation, and instead suggested that Ubx gene
expression is inhibited by transcriptional interference from ncRNAs from the
flanking promoter region (Petruk et al. 2006; Payer and Lee 2008). Although
there are still some discrepancies in the literature, it appears that certain ncRNAs
play key roles in maintaining the active or inactive state of the chromosome by
modulating the recruitment of PcG and TrxG proteins.
Numerous long ncRNAs have been suggested to function as key players in
uniparental expression due to genomic imprinting (Petruk et al. 2006; Royo and
Cavaillé 2008). In the mouse placenta, the 108-kb nuclear-retained Airn ncRNA
is required for the paternal-specific silencing in cis of a 400-kb region that includes
the Slc22a2, Slc22a3, and Igf2r genes (Sleutels et al. 2002; Seidl et al. 2006).
Although Airn covers the imprinted locus on the paternal chromosome, Airn
380 T. Hirose
50 untranslated region (UTR) containing the IRES is spliced out of the mature
mRNA; however, upon the signal for the epithelial–mesenchymal transition, the
antisense RNA that is complementary to the 50 splice site of this intron is induced
and blocks the splicing of this intron (Fig. 3d). The resultant mature mRNA is able
to be translated into Zeb2/Sib1 protein through the IRES in the 50 UTR (Fig. 3d)
(Beltran et al. 2008).
Pseudogene transcripts act to regulate the levels of their homologous coding
mRNAs. The transcriptional reduction caused by transgene integration into the
vicinity of the expressed pseudogene of Makorin1 destabilizes the Makorin1
mRNA in trans via an RNA sequence element within the 50 region of Makorin1
that is homologous between Makorin1 and its pseudogene (Hirotsune et al. 2003).
These findings demonstrate a novel and specific regulatory role for an expressed
pseudogene as well as demonstrating additional functional significance for
ncRNAs.
Currently, it is recognized that miRNAs and the related small RNAs broadly
impact gene expression in various developmental and physiological conditions.
Long ncRNAs have been reported to modulate the function or biogenesis of such
small RNAs. Caenorhabditis elegans rncs-1 is an 800-nt, starvation-induced
ncRNA that is highly base-paired (Hellwig and Bass 2008). Rncs-1 ncRNA is
not a substrate for Dicer because of branched structures at its termini. Rncs-1
RNA inhibits Dicer cleavage of a second dsRNA in vitro, and the expression of
rncs-1 leads to varying mRNA levels of several Dicer-regulated genes in vivo
(Hellwig and Bass 2008). Arabidopsis thaliana IPS1 (Induced by Phosphate
Starvation 1) contains a motif with sequence complementarity to the phosphate
(Pi) starvation-induced miRNA miR-399, but the pairing is interrupted by a
mismatched loop at the expected miRNA cleavage site (Franco-Zorrilla et al.
2007). The IPS1 ncRNA is not cleaved, but instead sequesters miR-399. IPS1
overexpression results in increased accumulation of the miR-399 target, PHO2
mRNA, and concomitantly leads to reduced shoot Pi content. That is, IPS1 acts by
‘target mimicry’ to inhibit miRNA activity (Franco-Zorrilla et al. 2007). Thus,
rncs-1 and IPS1 ncRNAs both have the potential to act as molecular decoys by
competing with the actual RNA substrates. In the transcriptome data, large
numbers of ncRNA-like transcripts that overlap with mRNA 30 -UTRs have been
registered; these ncRNAs may be decoys for specific regulatory factors that
interact with the 30 UTR.
The mammalian cell nucleus contains more than ten membrane less suborganelles
that serve specialized functions (Lamond and Spector 2003; Misteli 2005). Recent
works suggest that long ncRNAs serve as key structural components in some of
these suborganelles. Paraspeckles are a relatively newly discovered subnuclear
structure with unknown function (Fox et al. 2002). Paraspeckles are observed
382 T. Hirose
as 10–20 granular foci in the interphase cell nucleus and contain numerous RNA-
binding proteins, including PSP1, PSP2/CoAA, p54/nrb, PSF, and the 68-kDa
subunit of cleavage factor I m (Fig. 5a) (Fox et al. 2002, 2005 Dettwiler et al.
2004). Interestingly, RNase A treatment disrupts the structural integrity of para-
speckles (Fox et al. 2005), suggesting that RNA is a critical component of these
nuclear structures. Recently, three groups independently identified MENe and -b as
the critical RNA components of paraspeckles (Clemson et al. 2009; Sasaki et al.
2009; Sunwoo et al. 2009). Two MEN isoforms, MENe/NEAT1 (3.7 kb) and
MENb (23 kb), are transcribed from a single RNA polymerase II promoter, but
differ in the location of their 30 end. The MENe/b depletion phenotype was
examined in human and mouse cells by knockdown with chimeric antisense
oligonucleotides (Sasaki et al. 2009; Sunwoo et al. 2009) or siRNA (Clemson
et al. 2009). MENe/b knockdown resulted in the disruption of the paraspeckles
but not other nuclear bodies, indicating that these long ncRNAs are required for
paraspeckle establishment and maintenance (Fig. 5a) (Clemson et al. 2009; Sasaki
et al. 2009; Sunwoo et al. 2009). Experiments examining the interaction between
MENe/b ncRNAs and the known paraspeckle proteins, and RNAi knockdown of
paraspeckle proteins, revealed the significance of the interaction of MENe/b
ncRNAs with at least two paraspeckle RNA-binding proteins, p54/nrb and PSF,
in the organization of paraspeckle structure (Fig. 5b) (Clemson et al. 2009;
a
knockdown control
b
PSF
assembly
MENb
PSP1
p54
MENe
Fig. 5 Structural roles of ncRNAs. (a) MENe/b ncRNA (magenta) is specifically localized to the
nuclear paraspeckles where the RNA-binding protein PSF (green) is colocalized (merged). Upon
knockdown of MENe/b with ASO, the paraspeckle structures disintegrate (panels labeled “knock-
down”). (b) A model of paraspeckle organization achieved by cooperative association of ncRNAs
and RNA-binding proteins. MENb, the longer isoform, interacts with p54/nrb and PSF proteins to
form the core structure, then MENe, the shorter isoform, and PSP1 protein may bind to construct
the intact paraspeckle structure
Emerging Roles of Long Noncoding RNAs in Gene Expression and Intracellular 383
Sasaki et al. 2009; Sunwoo et al. 2009). Physiological roles of paraspeckles have
not been well established, although it has been reported that the paraspeckles serve
as the nuclear-retention sites of mRNA subsets. The CTN-RNA is specifically
localized to paraspeckles and in response to external stimuli, it is endonucleolyti-
cally cleaved at its long 30 -UTR to produce a shorter mCat2 mRNA that is
transported to the cytoplasm, where it is then translated (Prasanth et al. 2005).
The CTN-RNA contains a long inverted repeat sequence in its 30 UTR that is A-to-I
edited. Artificial reporter mRNAs containing inverted repeated Alu elements in
their 30 UTRs are bound by a p54/nrb-containing complex, which prevents their
export to the cytoplasm and tends to localize them to the paraspeckles (Chen et al.
2008). The detailed mechanism of nuclear retention of mRNAs through the para-
speckles remains elusive, but a recent report showed that the MENe/b ncRNA-
dependent formation of intact paraspeckle structure is required for the extents of
nuclear retention of mRNA subsets that are usually retained in the nucleus (Chen and
Carmichael 2009). It would be most intriguing to pursue the roles of MENe/b
ncRNAs in mRNA nuclear retention.
Thermal and chemical stresses induce the formation of transient nuclear struc-
tures called nuclear stress bodies (nSBs) (Biamonti 2004). These contain HSF1 and
a specific subset of pre-mRNA splicing factors. nSBs are assembled on specific
pericentromeric heterochromatic domains containing satellite III (SatIII) DNA. In
response to stress, these domains change their epigenetic status from heterochro-
matin to euchromatin and are transcribed into polyadenylated SatIII RNAs that
remain associated with nSBs. Downregulation of SatIII RNAs significantly affects
the recruitment of RNA splicing factors to nSBs without altering the association of
HSF-1 with these structures. Thus, SatIII RNAs have a major role in the formation
of nSBs (Valgardsdottir et al. 2005).
RNA also has a structural role in the organization of the cytoskeleton and the
mitotic spindle. In Xenopus oocytes, the Xlsirts ncRNA and the VegT mRNA are
integrated within the cytoskeleton and are required for proper organization of the
cytokeratin cytoskeleton (Kloc et al. 2005, 2007). Downregulation of either tran-
script disrupts the cytokeratin network, but not the actin cytoskeleton. VegT mRNA
may act as an RNA itself, because blocking its translation had no effect on the
cytokeratin network (Heasman et al. 2001; Kloc et al. 2005). Meanwhile, the
mitotic spindles were found to associate with various RNAs, including ribosomal
RNAs and a number of uncharacterized transcripts (Blower et al. 2005). RNase A
treatment, but not translation inhibitors, disrupts spindle assembly and causes the
spindle to collapse, indicating that these RNA species play a role in spindle
assembly in M phase (Blower et al. 2005). A number of subcellular structures are
involved in RNA biogenesis and metabolism, and contain RNA components.
Malat-1 is localized exclusively in the splicing speckle, along with numerous
splicing factors (Hutchinson et al. 2007), and Gomafu, a neuron-specific long
ncRNA, is localized in a novel subnuclear structure that remains to be characterized
(Sone et al. 2007). Therefore, further examples will appear that long ncRNAs play
architectural roles in organizing intracellular structures.
384 T. Hirose
Among the thousands of long ncRNAs whose sequences have been deposited in
databases, limited numbers have been functionally annotated. Future ncRNA
research should include the categorization of these RNA species based on common
features in their structures, functions, binding partners, or other characteristics.
Various bioinformatic approaches have been used to identify RNA sequence motifs
that may characterize ncRNA subsets. However, no sequence motifs that could be
used to categorize ncRNAs have been identified so far. Identification of the binding
partners of specific ncRNAs would also be a useful approach to categorize the
ncRNAs; however, there have been technical difficulties in the analysis of RNA–
protein interactions because of nonspecific associations in vitro that are distinct
from what is seen in DNA–protein interactions. Even in standard immunoprecipita-
tion experiments, the reassociation of RNA-binding proteins with RNA after cell
lysis can lead to misidentification of binding partners (Mili and Steitz 2004). To
circumvent these technical difficulties, a new method called CLIP has been devel-
oped. In this method, the in vivo RNA–protein (RNP) complexes are covalently
crosslinked by UV irradiation, followed by immunoprecipitation of the crosslinked
RNP complex (Ule et al. 2005). The CLIP method has been combined with high-
throughput DNA sequencing (HITS–CLIP) to globally identify the target RNA
sequences of certain RNA-binding proteins (Licatalosi et al. 2008; Chi et al. 2009).
For categorization of long ncRNAs, HITS–CLIP analysis will provide useful
Emerging Roles of Long Noncoding RNAs in Gene Expression and Intracellular 385
information about ncRNA species that bind to a common RNA-binding protein and
about the sequences of the target motifs.
Another goal of ncRNA research is to identify physiological phenomena asso-
ciated with the action of a particular ncRNA. RNA interference (RNAi), currently
the most popular method for investigating RNA function, is used to knock down a
specific RNA, after which any phenotypic alterations are examined. Because
numerous long ncRNAs seem to function in nuclear processes, the knockdown
should be targeted to nuclear-localized ncRNA molecules. However, in mammalian
cells, the RNAi machinery is believed to exclusively localize in the cytoplasm,
making RNAi a poor choice to knock down nuclear ncRNAs, with a few exceptions
(Fig. 6a). Instead, antisense deoxyoligonucleotides (ASOs), which are usually
modified to increase their stability, are recognized as the most effective method
for nuclear RNA knockdown. The introduced ASOs form DNA–RNA hybrids that
are specifically recognized by endogenous RNase H, which degrades the RNA
strand of the hybrid (Fig. 6a and b). In the past, the introduction of ASO into the
nucleus was inefficient, resulting in ASO effectiveness that was often low, but
recently it was reported that ASO can be efficiently delivered into the nucleus
by using the nucleofection method, resulting in efficient depletion of various
nuclear-localized ncRNAs (Fig. 6c) (Ideue et al. 2009). As mentioned above, it is
a b
RNAi ASO
RNA
RNaseH
mRNA
Cytoplasmic ncRNA
decay
Nuclear ncRNA c
ASO ASO siRNA
FP 84 FP 4
G Δ G Δ8
6 24 6 24 6 24 6 24
New method to
Transcriptional interference
arrest transcription
U84
Fig. 6 The experimental system for the functional analysis of nuclear ncRNAs. (a) Canonical
RNAi is thought to occur exclusively in the cytoplasm of mammalian cells; therefore, an antisense
oligonucleotide (ASO) is utilized to knock down nuclear ncRNAs. Knockdown of the posttran-
scriptionally accumulated ncRNA may not inhibit transcriptional interference. A new method to
arrest the transcription of specific ncRNAs would be required to explore their roles in transcrip-
tional interference. (b). The principle of ASO action in the nucleus. The administered ASO forms
an RNA–DNA hybrid with the target RNA, in which the RNA strand is specifically cleaved by
endogenous RNase H. (c) An example of the knockdown of a nuclear ncRNA. U84 snoRNA is
efficiently knocked down with ASO but is not susceptible to siRNA
386 T. Hirose
References
Allen TA, Von Kaenel S, Goodrich JA et al (2004) The SINE-encoded mouse B2 RNA represses
mRNA transcription in response to heat shock. Nat Struct Mol Biol 11:816–821
Beltran M, Puig I, Peña C et al (2008) A natural antisense transcript regulates Zeb2/Sip1
gene expression during Snail1-induced epithelial–mesenchymal transition. Genes Dev 22:
756–769
Bender W, Fitzgerald DP (2002) Transcription activates repressed domains in the Drosophila
bithorax complex. Development 129:4923–4930
Beniaminov A, Westhof E, Krol A (2008) Distinctive structures between chimpanzee and human
in a brain noncoding RNA. RNA 14:1270–1275
Bertone P, Stolc V, Royce TE et al (2004) Global identification of human transcribed sequences
with genome tiling arrays. Science 306:2242–2246
Biamonti G (2004) Nuclear stress bodies: a heterochromatin affair? Nat Rev Mol Cell Biol
5:493–498
Birney E, Stamatoyannopoulos JA, Dutta A et al (2007) Identification and analysis of functional
elements in 1% of the human genome by the ENCODE pilot project. Nature 447:799–816
Emerging Roles of Long Noncoding RNAs in Gene Expression and Intracellular 387
Prasanth KV, Spector DL (2007) Eukaryotic regulatory RNAs: an answer to the ‘genome com-
plexity’ conundrum. Genes Dev 21:11–42
Preker P, Nielsen J, Kammler S et al (2008) RNA exosome depletion reveals transcription
upstream of active human promoters. Science 322:1851–1854
Ravasi T, Suzuki H, Pang KC et al (2006) Experimental validation of the regulated expression of
large numbers of non-coding RNAs from the mouse genome. Genome Res 16:11–19
Rinn JL, Kertesz M, Wang JK et al (2007) Functional demarcation of active and silent chromatin
domains in human HOX loci by noncoding RNAs. Cell 129:1311–1323
Royo H, Cavaillé J (2008) Non-coding RNAs in imprinted gene clusters. Biol Cell 100:149–166
Sanchez-Elsner T, Gou D, Kremmer E et al (2006) Noncoding RNAs of trithorax response
elements recruit Drosophila Ash1 to Ultrabithorax. Science 311:1118–1123
Sasaki YT, Sano M, Ideue T et al (2007) Identification and characterization of human non-coding
RNAs with tissue-specific expression. Biochem Biophys Res Commun 357:991–996
Sasaki YT, Ideue T, Sano M et al (2009) MENe/b noncoding RNAs are essential for structural
integrity of nuclear paraspeckles. Proc Natl Acad Sci USA 106:2525–2530
Schmitt S, Prestel M, Paro R (2005) Intergenic transcription through a polycomb group response
element counteracts silencing. Genes Dev 19:697–708
Seidl CI, Stricker SH, Barlow DP (2006) The imprinted Air ncRNA is an atypical RNAPII
transcript that evades splicing and escapes nuclear export. EMBO J 25:3565–3575
Seila AC, Calabrese JM, Levine SS et al (2008) Divergent transcription from active promoters.
Science 322:1849–1851
Shamovsky I, Ivannikov M, Kandel ES et al (2006) RNA-mediated response to heat shock in
mammalian cells. Nature 440:556–560
Silva J, Mak W, Zvetkova I et al (2003) Establishment of histone h3 methylation on the inactive X
chromosome requires transient recruitment of Eed-Enx1 polycomb group complexes. Dev Cell
4:481–495
Sleutels F, Zwart R, Barlow DP (2002) The non-coding Air RNA is required for silencing
autosomal imprinted genes. Nature 415:810–813
Smith CM, Steitz JA (1998) Classification of gas5 as a multi-small-nucleolar-RNA (snoRNA) host
gene and a member of the 50 -terminal oligopyrimidine gene family reveals common features of
snoRNA host genes. Mol Cell Biol 18:6897–6909
Sone M, Hayashi T, Tarui H et al (2007) The mRNA-like noncoding RNA Gomafu constitutes a
novel nuclear domain in a subset of neurons. J Cell Sci 120:2498–2506
Srikantan V, Zou Z, Petrovics G et al (2000) PCGEM1, a prostate-specific gene, is overexpressed
in prostate cancer. Proc Natl Acad Sci USA 97:12216–12221
Sunwoo H, Dinger ME, Wilusz JE et al (2009) MENe/b nuclear-retained non-coding RNAs are
up-regulated upon muscle differentiation and are essential components of paraspeckles.
Genome Res 19:347–359
Taft RJ, Pheasant M, Mattick JS (2007) The relationship between non-protein-coding DNA and
eukaryotic complexity. Bioessays 29:288–299
Tycowski KT, Shu MD, Steitz JA (1996) A mammalian gene with introns instead of exons
generating stable RNA products. Nature 379:464–466
Uhler JP, Hertel C, Svejstrup JQ (2007) A role for noncoding transcription in activation of the
yeast PHO5 gene. Proc Natl Acad Sci USA 104:8011–8016
Ule J, Jensen K, Mele A et al (2005) CLIP: a method for identifying protein-RNA interaction sites
in living cells. Methods 37:376–386
Valgardsdottir R, Chiodi I, Giordano M et al (2005) Structural and functional characterization of
noncoding repetitive RNAs transcribed in stressed human cells. Mol Biol Cell 16:2597–2604
Wang X, Arai S, Song X et al (2008) Induced ncRNAs allosterically modify RNA-binding proteins
in cis to inhibit transcription. Nature 454:126–130
Willingham AT, Orth AP, Batalov S et al (2005) A strategy for probing the function of noncoding
RNAs finds a repressor of NFAT. Science 309:1570–1573
Emerging Roles of Long Noncoding RNAs in Gene Expression and Intracellular 391
Wilusz JE, Freier SM, Spector DL (2008) 3’ end processing of a long nuclear-retained noncoding
RNA yields a tRNA-like cytoplasmic RNA. Cell 135:919–932
Yamada K, Kano J, Tsunoda H et al (2006) Phenotypic characterization of endometrial stromal
sarcoma of the uterus. Cancer Sci 97:106–112
Yan MD, Hong CC, Lai GM et al (2005) Identification and characterization of a novel gene Saf
transcribed from the opposite strand of Fas. Hum Mol Genet 14:1465–1474
Yu W, Gius D, Onyango P et al (2008) Epigenetic silencing of tumour suppressor gene p15 by its
antisense RNA. Nature 451:202–206
Zhao J, Sun BK, Erwin JA et al (2008) Polycomb proteins targeted by a short repeat RNA to the
mouse X chromosome. Science 322:750–756
Zhao X, Patton JR, Davis SL et al (2004) Regulation of nuclear receptor activity by a pseudouridine
synthase through post transcriptional modification of steroid receptor RNA activator. Mol Cell
15:549–558
Noncoding RNAs as Therapeutic Targets
Contents
1 RNA-Dependent Regulation of Gene Expression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
1.1 Gene Regulation Through Epigenetic Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
1.2 Controlling Transcription Machinery Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
1.3 Posttranscriptional Regulatory Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
2 The Medical Perspective: Noncoding RNAs
in Human Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
2.1 MicroRNAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
2.2 mRNA-Like ncRNAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404
2.3 Other Transcripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406
3 NcRNA-Based Therapeutic Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
Abstract Noncoding RNAs are key players in the regulation of complex cellular
processes. Over the years, it has been demonstrated that their abnormal expression
is associated with many human pathologies including developmental and neurobe-
havioral disorders, diabetes, obesity, and cancer. A wide spectrum of activities and
a large number of genes that are regulated by RNA-dependent mechanisms makes
the ncRNA molecules attractive targets for developing next generation therapeutic
agents. This chapter outlines the principles of RNA regulation, the involvement
of various RNAs in human diseases, and the strategies of application of
ncRNA-targeted therapeutic approaches.
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 393
RNA Technologies, DOI 10.1007/978-3-642-12168-5_18,
# Springer-Verlag Berlin Heidelberg 2010
394 M. Szymański and J. Barciszewski
protein-coding ones (Beiter et al. 2009). It has been proposed that the ncRNAs,
including transcripts from the independent transcriptional units as well as those
processed from intronic sequences constitute elements of the regulatory networks
responsible for controlling gene expression. The increase in complexity of regu-
latory networks dependent on ncRNA accompanied by a decrease in the size of
protein-coding portion of the genome would contribute to increased complexity of
the living systems (Mattick 1994, 2001, 2003, 2009a). Thus, the subtle changes in
expression or function of regulatory RNAs may be essential for the evolution and
development of eukaryotes.
The whole output of transcription from the genomic DNA (the transcriptome)
consists of the protein-coding part (open reading frames) serving as template for
protein biosynthesis and noncoding part including untranslated regions of mRNAs,
ncRNA, and introns removed from primary transcripts of both protein-coding and
noncoding genes.
The ncRNA fall into two broad categories: housekeeping or infrastructural
RNAs essential for protein biosynthesis (tRNA, rRNA), RNA processing
(snRNA), and modifications (snoRNA) and other elementary cellular functions
(e.g., telomerase RNA, RNase P RNA). Expression of these RNAs is constitutive
and their levels in the cells do not change.
In the last decade, it was realized that there also exists a large group of ncRNA
molecules performing regulatory functions affecting expression of numerous genes.
Unlike the housekeeping RNAs that are essential for the fundamental life functions
and thus expressed in every cell, the regulatory ncRNAs are expressed in a very
specific strictly regulated manner, and their repertoire varies depending on tissue or
cell type, developmental stage, or the changes of biotic or abiotic environmental
conditions (Mercer et al. 2009; Szymanski and Barciszewski 2006).
Although in recent years a significant progress was made in identifying new
ncRNA species, there are still many unsolved questions concerning the mechanisms
underlying their role in regulatory pathways. Various ncRNAs act on different
stages of gene expression. They are involved in processes affecting chromatin
structure and its transcriptional competence, activity of transcription factors, and
the fate of RNAs on the posttranscriptional level.
ncRNA play a key role in epigenetic processes associated with silencing of genes
within imprinted clusters (Mohammad et al. 2009), X chromosome inactivation
(Payer and Lee 2008), DNA methylation (Imamura et al. 2004a), and heterochromatin
formation at the centromeres (Iida et al. 2008). Genomic imprinting is unique to
mammals and represents epigenetic modification directing expression of imprinted
alleles according to parent of origin (Mohammad et al. 2009). A common feature of
clusters of imprinted genes is expression of ncRNAs that are required for transcrip-
tional inactivation of protein-coding genes (Morison et al. 2005). The precise role of
396 M. Szymański and J. Barciszewski
most of these imprinted ncRNAs in silencing is unknown, but their expression was
shown to be essential for this process. Aberrant expression of imprinted genes on
human chromosome 11p15 is associated with Beckwith–Wiedemann syndrome
(BWS) and several human cancers (Enklaar et al. 2006). In most human tissues,
the paternal allele produces a 91 kb long nuclear antisense ncRNA (LIT1/Kcnq1ot1)
originating from an unmethylated CpG island (KvDMR1) within intron 10 of the
maternally expressed KCNQ1 (KvLQT1) gene (Niemitz et al. 2004). Methylation
status of KvDMR1 and expression of LIT1 RNA constitute the key elements
controlling imprinted expression of the BWS associated genes. Demethylation of
the maternal KvDMR1 allele or its deletions are among the most frequent causes of
the BWS (DeBaun et al. 2002). It has been demonstrated that LIT1 RNA expression is
necessary for transcriptional inactivation of several maternally expressed genes on the
paternal chromosome (Horike et al. 2000). The silencing effect spreads bidirectionally
(Thakur et al. 2004), affecting genes located downstream and upstream from
KvDMR1, and its extent depends on the length of LIT1 RNA (Mancini-Dinardo
et al. 2006; Kanduri et al. 2006). In the mouse embryo, the region affected by Lit1
RNA spans 400 kb and is extended to over 780 kb in the placenta (Terranova et al.
2009). Transcriptional repression depends on Polycomb group complex Eed-Ezh2
and repressive methylations of histone H3 (Umlauf et al. 2004; Terranova et al. 2009).
Moreover, it has been shown that Lit1 RNA participates in establishing a nuclear
domain comprising exclusively silenced genes (Terranova et al. 2009).
Similar bidirectional silencing induced by ncRNA occurs at the imprinted
cluster on mouse chromosome 17. Differential methylation of the imprinting
control region within intron 2 of insulin-like growth factor type-2 receptor (Igf2r)
regulates expression of maternally expressed genes (Wutz et al. 1997). An anti-
sense, unspliced 108 kb long Air RNA (antisense Igf2r) overlapping 30 kb of Igf2r
gene is transcribed from the unmethylated paternal allele. Transcriptional silencing
affects the sense Igf2r gene and two other genes, Slc22a2 and Slc22a3, 110 and
155 kb downstream from Igf2r, respectively (Sleutels et al. 2003).
The Air RNA was shown to directly interact with the chromatin at Slc22a3
promoter, which correlated with the presence of repressive modifications of histone
H3. The silencing depended on the activity of histone methyltransferase G9a and a
whole length Air transcript (Nagano et al. 2008).
Epigenetic gene silencing induced by noncoding transcripts is not restricted to
imprinted genes clusters. An antisense transcript was found to downregulate
expression of the cyclin-dependent kinase inhibitor p15 that plays a role of tumor
suppressor gene. In leukemia, it was demonstrated that increased transcription of
the p15 antisense RNA (p15AS) results in reduced expression of the sense tran-
script. It has been found that the effect was due to epigenetic mechanism-inducing
heterochromatin formation. The repression of p15 was maintained even in the
absence of antisense transcript (Yu et al. 2008). Although there are limited data
available, it has been suggested that the RNA-induced gene silencing may represent
a more general mechanism and could account for regulatory potential of other
natural antisense transcripts (Pauler et al. 2007) that may be associated with
approximately 70% of loci in the mammalian genome (Katayama et al. 2005).
Noncoding RNAs as Therapeutic Targets 397
Apart from the induction of heat shock proteins, there is a general inhibition of
transcription by RNA polymerase II. This effect is brought about by short inter-
spersed elements (SINEs) transcripts, binding the core PolII and inhibiting tran-
scription at the initiation step. In mouse, the regulatory RNAs are transcribed from
B2 SINEs (Allen et al. 2004), while in humans, the same role is played by Alu
transcripts (Mariner et al. 2008). At the molecular level, inhibition is achieved
by ncRNAs preventing contacts between polymerase and the promoter DNA
(Yakovchuk et al. 2009). Reactivation of transcription occurs either through disso-
ciation of ncRNA or its removal by as yet unknown factors (Espinoza et al. 2004).
from separate transcriptional units. It has been reported that long antisense tran-
scripts originating from pseudogenes can form duplexes with their corresponding
mRNAs. The double-stranded RNAs are further processed by Dicer to generate
endogenous siRNAs that participate in RISC-mediated cleavage of protein-coding
transcripts and downregulation of the gene (Tam et al. 2008). This observation
demonstrated that although pseudogenes were mostly regarded as nonfunctional
remnants of past events, they may also play an important role in regulatory
pathways.
The largest and best understood class of posttranscriptional regulators is micro-
RNAs that in recent years attracted most attention from researchers. This is
primarily due to their wide spectrum of activities and their involvement in the
regulation of many genes playing critical role in development and differentiation
(Stefani and Slack 2008). The ability to posttranscriptionally modulate expression
of many genes makes microRNAs, together with transcription factors, the primary
determinants of unique gene expression patterns in eukaryotic cells (Chen and
Rajewsky 2007; Hobert 2004).
Most of the animal microRNAs are involved in translational regulation depend-
ing on binding partially complementary sequences within target mRNAs’ 30 -UTRs
(Bartel 2004). Mechanism of translational repression depends on the presence of
7-methyl guanosine cap (Pillai et al. 2005) and one of the Argonaute family proteins
(Ago2) that interferes with cap binding by the translation initiation factor eIF4E on
microRNA-bound mRNA (Kiriakidou et al. 2007). Another possibility is binding of
initiation factor eIF6 by the microRNA–mRNA complex that prevents translation
by blocking association of the ribosomal subunits (Chendrimada et al. 2007). In
addition to translational inhibition, microRNAs can induce deadenylation of
mRNA, thus decreasing its stability (Fabian et al. 2009; Beilharz et al. 2009).
Not all animal microRNAs act as translational repressors. Some are involved
in posttranscriptional control by directing cleavage of target mRNAs by Dicer
nuclease as in the case of mir-196 targeting Hox-B8 mRNA (Yekta et al. 2004).
However, microRNA-induced hydrolysis requires full complementarity between
target mRNA and microRNA.
Apart from regulating translation and stability of protein coding genes, micro-
RNAs can also be involved in posttranscriptional regulation of ncRNA. Comple-
mentary sites for miR-155 and miR-24-1 were found within UCRs differentially
expressed in chronic lymphocytic leukemias (CLLs) suggesting that there may exist
links between microRNAs and expression of other ncRNAs (Calin et al. 2007).
Localized translation of mRNA in the neurons of rodents and primates is
regulated by PolIII transcripts BC1 and BC200 (Martignetti and Brosius 1995)
identified in ribonucleoprotein particles in cell bodies and in dendrites (Tiedge et al.
1991, 1993). BC200 was also found in complexes with SYNCRIP (synaptotagmin-
binding cytoplasmic RNA interacting protein), a component of mRNA transport
granules involved in localized protein synthesis at postsynaptic sites. (Duning et al.
2008). It has been shown that BC1 interferes with the formation of a stable 48S
preinitiation complex. BC1 specifically inhibits activity of eIF4A helicase and forms
stable complexes with the poly(A)-binding protein (PABP) (Wang et al. 2002;
Noncoding RNAs as Therapeutic Targets 401
Lin et al. 2008). At synapses, BC1 and BC200 RNAs were also found in complexes
with FMRP (fragile X mental retardation associated protein) participating in trans-
lational repression of a subset of FMRP-regulated mRNAs (Zalfa et al. 2003, 2005).
NcRNAs can also regulate alternative splicing as demonstrated in the case
o serotonin receptor 5-HT(2C)R. Splice site selection in exon Vb of 5-HT(2C)R
is controlled by HBII-52 snoRNA that binds a silencing element. In Prader–Willi
syndrome patients, lacking expression of HBII-52 the 5-HT(2C)R mRNA isoforms
differ from those found in healthy individuals (Kishore and Stamm 2006). Recently,
the involvement of MBII-52, the mouse homolog of HBII-52, in alternative splicing
has been confirmed for five other pre-mRNAs containing alternative exons. It has
also been shown that there is an alternative, shorter isoform associate with hnRNPs,
differing in structure from the canonical C/D box snoRNA that is responsible for
this process (Kishore et al. 2010).
Posttranscriptional regulation by ncRNAs may also involve subcellular locali-
zation of proteins. A calcium-responsive transcription factor NFAT (nuclear factor
of activated T cells) controls expression of genes involved T-cell-mediated immune
response and in the development of vascular and nervous system and muscles. It has
been shown that the nuclear localization of NFAT is regulated by an NRON ncRNA
(noncoding repressor of NFAT), which by interactions with nuclear import factors
specifically inhibits nucleocytoplasmic trafficking of NFAT, but not of other
transcription factors (Willingham et al. 2005).
The diversity of ncRNA functions and their involvement in processes linked to cell
growth, differentiation, and development makes them very susceptible targets for
mutations resulting in various pathologies. In fact, the number of ncRNAs, belong-
ing to various classes of transcripts, implicated in human diseases is constantly
growing. Although, there is not always a clear-cut distinction between a cause and
an outcome, aberrant expression of many ncRNAs has been observed in human
cancers and neurobehavioral and developmental disorders.
2.1 MicroRNAs
Regulation of cell growth and apoptosis has been proposed for miR-16 family
members that target genes responsible for cell cycle progression from G0/G1 to S
phase, including regulators of G1 phase CDK6, CDC27, an activator of NF-kB
signaling CARD10 (Linsley et al. 2007; Cloonan et al. 2008), and BCL2 that plays
a role of antiapoptotic factor (Cimmino et al. 2005).
MicroRNAs also play an important role in viral infections. Human cytomegalo-
virus (HCMV) encodes a microRNA hcmv-mirR-UL112 targeting major histocom-
patibility complex class I-related chain B (MICB) involved in the recognition and
killing of virus-infected cells by natural killer cells (Stern-Ginossar et al. 2007).
MicroRNAs involved in reprogramming endothelial cells by targeting the cellular
transcription factor MAF (musculoaponeurotic fibrosarcoma oncogene homolog)
were identified in Kaposi sarcoma herpesvirus (KSHV) (Hansen et al. 2010).
MicroRNAs have been implicated in the origins of certain neurological disorders
including Alzheimer’s disease and Parkinson’s disease. One of the mechanisms
controlling expression of amyloid precursor responsible for Alzheimer’s depends
on several miR-20a family microRNAs (miR-20a, miR-17-5p and miR-106b),
and decreased levels of miR-29a/b-1 and miR-106b have been reported in brains
of Alzheimer’s disease patients (Hebert et al. 2008a; Hebert et al. 2008b). In
Parkinson’s disease patients, there are reduced levels of miR-133b essential for
development and functions of midbrain dopaminergic neurons (Kim et al. 2007).
MicroRNA miR-181b overexpressed in the cortex of schizophrenia patients may
contribute to etiology of this disorder. The putative miR-181b targets include genes
essential for neuronal functions (e.g., receptors involved in synaptic transmission),
development of nervous system and cell differentiation, and the experimental
verification indicated two mRNAs encoding ionotropic glutamate receptor (GRIA2)
and calcium sensor gene visinin-like 1 (VSNL1). GRIA2 and VSNL1 are crucial for
neuronal functions involved in signal transduction and synaptic plasticity and have
been earlier linked to schizophrenia (Beveridge et al. 2008).
Pathologies linked to microRNAs may also result from mutations changing
microRNA-binding sites within 30 -UTRs of target genes. One of the genes regulated
by let-7 microRNA is HMGA2 (High Mobility Group A2) protein showing elevated
expression in several human tumors and involved in remodeling of chromatin.
Chromosomal rearrangements or mutations disrupting let-7 binding result in onco-
genic transformation (Mayr et al. 2007; Lee and Dutta 2007; Klemke et al. 2010). In
Parkinson’s disease, an overexpression of the fibroblast growth factor 20 (FGF20) is
due to a single nucleotide substitution within the 30 -UTR of FGF20 mRNA that
affects miR-433 target site resulting in increased expression of FGF20 protein
in vitro and in vivo (Wang et al. 2008). A polymorphism at the microRNA target
site within SLITRK1 (SLIT and Trk-like 1) gene responsible for growth of neurons is
responsible for the origin of the Touret’s syndrome (Abelson et al. 2005).
The list of diseases and molecular pathways that have been linked to changes in
microRNA expression is growing. MicroRNAs have been implicated in the origin
of such conditions as diabetes (Tang et al. 2008), obesity (Xie et al. 2009),
cardiovascular diseases (Mishra et al. 2009), autoimmunological disorders (Pauley
et al. 2009), and many others.
404 M. Szymański and J. Barciszewski
Although currently there are limited data concerning precise functions of most of
ncRNAs, there is a growing body of evidence that most of these transcripts are in
fact functional and participate in the regulation of a wide array of cellular processes
(Mattick 2009b). It is, therefore, not surprising that the aberrant expression of many
of them was found to be associated with pathological states.
Deregulated expression of imprinted genes including those encoding ncRNAs is
linked to severe congenital, developmental disorders. At least in some cases, the
RNA is responsible for establishing and maintaining monoallelic expression of
imprinted genes by inducing epigenetic changes as demonstrated for LIT1 or Air
RNAs (for details see Sect. 1.1).
Abnormal expression of imprinted ncRNAs has been observed in various forms
of human cancer. H19 RNA is a product of a maternally expressed gene on
chromosome 11p15.5 located downstream from the paternally expressed IGF2
(insulin like growth factor 2) gene. H19 RNA is expressed primarily during
embryonic development in most fetal tissues (Gabory et al. 2006, 2009). Biallelic
expression of either gene due to epigenetic changes in the differentially methylated
region (DMR) upstream of the H19 can cause malignant cell growth (Manoharan
et al. 2004). Because the effect of H19 RNA differs in various forms of cancer, it
was described either as a tumor suppressor reducing tumorigenicity and growth
(Hao et al. 1993) or as an oncogene promoting cancer progression (Fellig et al.
2005; Berteaux et al. 2005; Lottin et al. 2005). It has been proposed that H19 RNA
may have different effect depending on its alternative splicing or interactions with
RNA-binding proteins (Matouk et al. 2004; Ioannidis et al. 2004). The biological
effect may also depend on microRNA miR-675 for which H19 plays a role of a host
gene (Cai and Cullen 2007). The same imprinted genes cluster produces a paternal-
expressed PEG8/IGF2AS RNA (paternally expressed gene 8, IGF2 antisense) that
shows elevated expression in several fetal cancers (Okutsu et al. 2000). In colorectal
and esophageal cancers, the epigenetic changes altering the imprinting status
and the expression of the LIT1 RNA were observed (Nakano et al. 2006; Soejima
et al. 2004).
An involvement of imprinted ncRNA in tumor suppression was described for the
transcript from the MEG3 (maternally expressed gene 3), highly expressed in
human pituitary whose expression is lost in pituitary adenomas and most human
cancer cell lines. Expression of MEG3 RNA in cancer cells stimulates expression of
p53 followed by activation of p53 downstream targets. It has been demonstrated
that the effect depends on proper folding of MEG3 RNA that is required for p53
activation and growth suppression (Zhang et al. 2009).
Long mRNA-like ncRNAs were often identified as transcripts showing elevated
expression in malignant cells. One of the molecular markers of metastasis in lung
adenocarcinoma is MALAT-1 RNA (metastasis associated in lung adenocarcinoma
transcript 1) (Ji et al. 2003). MALAT-1 is expressed in normal human and mouse
tissues (Ji et al. 2003; Hutchinson et al. 2007) and its overexpression is observed in
Noncoding RNAs as Therapeutic Targets 405
human breast, pancreas, lung, colon, prostate, and liver carcinomas (Lin et al. 2006).
Several cancer-associated translocations were shown to disrupt the MALAT-1 locus
(Davis et al. 2003; Kuiper et al. 2003; Rajaram et al. 2007). MALAT-1 was also
identified as a nuclear-enriched abundant transcripts (NEAT2) present in SC35
nuclear speckles (Hutchinson et al. 2007).
In over 90% of prostate cancers, there is a significant overexpression of several
alternatively spliced isoforms of DD3/PCA3 RNA (prostate cancer antigen 3)
(Bussemakers et al. 1999; de Kok et al. 2002). Another prostate-specific transcript
with markedly elevated levels in malignant cells is PCGEM1 RNA (Srikantan et al.
2000) expression, of which correlates with increased cell proliferation and colony
formation (Petrovics et al. 2004). Another effect of PCGEM1 observed in LNCaP
cell line was a resistance to doxorubicin-induced apoptosis and delayed induction
of p53 and p21 (Fu et al. 2006).
An ncRNA, CUDR (cancer up-regulated drug resistant), was identified in sev-
eral human cancer cell lines including hepatocellular, breast, colon, and lung
carcinomas as well as HeLa cells. CUDR expression correlates with resistance to
apoptosis-inducing drugs doxorubicin and etoposide. Although the molecular basis
of drug-resistance is not known, it has been proposed that CUDR RNA is involved
in the downregulation of caspase 3 that shows reduced levels in CUDR-expressing
cells (Tsang et al. 2007). As a very strong and specific ncRNA marker of cancer-
ogenesis, CUDR may represent a good target for the development of anticancer
drugs.
Several other mRNA-like ncRNAs have also been identified as cancer-associated
transcripts. The NCRMS (noncoding RNA in rhabdomyosarcoma (RMS)) shows
elevated expression in alveolar RMS, neuroblastoma, and synovial sarcoma (Chan
et al. 2002). OCC-1 RNA (overexpressed in colon carcinoma 1) is a specific marker
of colon carcinoma (Pibouin et al. 2002).
The number of novel ncRNAs with altered expression in cancers is growing
each year. A thorough analysis of expressed sequence tags (ESTs) from head, neck,
and thyroid allowed identification of new intronic, antisense RNAs (Reis et al.
2005). Another source of novel ncRNAs are the UCRs discovered by comparing
mammalian genomes (Bejerano et al. 2004) that were shown to be highly tran-
scribed and produce ncRNAs. An analysis of expression of 481 UCRs in CLL
demonstrated that they are differentially expressed in normal and malignant human
cells (Calin et al. 2007). Moreover, specific expression profiles of UCRs were
consistent with poor and good prognosis.
Human endogenous retroviruses (HERVs) were also implicated in cancerogenesis
(Galli et al. 2005; Mangeney et al. 2005; de Parseval et al. 1999). In patients with
bladder transitional cell carcinoma (TCC), an ncRNA UCA1 (urothelial carcinoma
associated 1) belonging to a HERV-H family was identified as a specific, highly
expressed molecular marker (Wang et al. 2006).
Abnormal expression of long ncRNAs is not limited to cancer cells alone.
A specific ncRNA PRINS RNA (psoriasis susceptibility-related RNA gene induced
by stress) is overexpressed in response to UV-B irradiation or viral infection in
patients with psoriasis (Sonkoly et al. 2005). MIAT RNA is associated with
406 M. Szymański and J. Barciszewski
Certain human pathologies are also associated with abnormal expressions of other
classes of ncRNAs including snoRNA and RNA polymerase III transcribed BC200
RNA. Imprinted expression abnormalities of the genes within the human chromo-
some 15q11–q13 are associated with Angelman (AS) and Prade–Willi syndromes
(PWS) that are caused by maternal and paternal deficiencies, respectively
(Horsthemke and Wagstaff 2008). Duplications of this region have been also
found in about 1% of autistic patients (Koochek et al. 2006). The paternal allele
specifies a large neuron-restricted transcript several hundred kilobases long encoding
splicesomal protein (SNRPN), several small nucleolar RNAs and the antisense
transcript to an ubiquitin ligase (UBE3A-AS) expressed from the maternal allele
Noncoding RNAs as Therapeutic Targets 407
(Horsthemke and Wagstaff 2008). Although the deficiency of one of the paternally
expressed snoRNAs (HBII-85) is regarded as a primary cause of PWS, there is
still little known about the mechanisms underlying it (Sahoo et al. 2008; de Smith
et al. 2009).
The deletion involving a region in 6q14–q16 frequently found in breast and
prostate cancer patients was found to contain a gene encoding U50 snoRNA
(Verhagen et al. 2002; Dong et al. 2009). Transcriptional downregulation and
genomic deletions of the U50 was confirmed in a number of breast cancer cell
lines. U50 snoRNA can be considered a tumor suppressor as its expression resulted
in the inhibition of colony formation in breast cancer cell lines (Dong et al. 2009).
Another snoRNA-encoding gene that might be involved in regulation of tumor
growth is GAS5 RNA (growth arrest-specific transcript 5), a snoRNA host gene that
induces growth arrest and apoptosis (Mourtada-Maarabouni et al. 2009).
Acknowledgments This work was supported by grants from the Polish Ministry of Science and
Higher Education.
References
Abelson JF, Kwan KY, O’Roak BJ et al (2005) Sequence variants in SLITRK1 are associated with
Tourette’s syndrome. Science 310:317–320
Abrajano JJ, Qureshi IA, Gokhan S et al (2009) REST and CoREST modulate neuronal subtype
specification, maturation and maintenance. PLoS One 4:e7936
Allen TA, Von Kaenel S, Goodrich JA et al (2004) The SINE-encoded mouse B2 RNA represses
mRNA transcription in response to heat shock. Nat Struct Mol Biol 11:816–821
Anckar J, Sistonen L (2007) Heat shock factor 1 as a coordinator of stress and developmental
pathways. Adv Exp Med Biol 594:78–88
Annilo T, Kepp K, Laan M (2009) Natural antisense transcript of natriuretic peptide precursor A
(NPPA): structural organization and modulation of NPPA expression. BMC Mol Biol 10:81
Bandrés E, Cubedo E, Agirre X et al (2006) Identification by Real-time PCR of 13 mature
microRNAs differentially expressed in colorectal cancer and non-tumoral tissues. Mol Cancer
5:29
Barboric M, Lenasi T, Chen H et al (2009) 7SK snRNP/P-TEFb couples transcription elongation
with alternative splicing and is essential for vertebrate development. Proc Natl Acad Sci USA
106:7798–7803
Bartel DP (2004) MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 116:281–297
Beilharz TH, Humphreys DT, Clancy JL et al (2009) microRNA-mediated messenger RNA
deadenylation contributes to translational repression in mammalian cells. PLoS One 4:e6783
Beiter T, Reich E, Williams RW et al (2009) Antisense transcription: a critical look in both
directions. Cell Mol Life Sci 66:99–112
Bejerano G, Pheasant M, Makunin I et al (2004) Ultraconserved elements in the human genome.
Science 304:1321–1325
Beltran M, Puig I, Pena C et al (2008) A natural antisense transcript regulates Zeb2/Sip1 gene
expression during Snail1-induced epithelial–mesenchymal transition. Genes Dev 22:756–769
Berteaux N, Lottin S, Monté D et al (2005) H19 mRNA-like noncoding RNA promotes breast
cancer cell proliferation through positive control by E2F1. J Biol Chem 280:29625–29636
Beveridge NJ, Tooney PA, Carroll AP et al (2008) Dysregulation of miRNA 181b in the temporal
cortex in schizophrenia. Hum Mol Genet 17:1156–1168
Blenkiron C, Goldstein LD, Thorne NP et al (2007) MicroRNA expression profiling of human
breast cancer identifies new markers of tumour subtype. Genome Biol 8:R214
Bond AM, Vangompel MJ, Sametsky EA et al (2009) Balanced gene regulation by an embryonic
brain ncRNA is critical for adult hippocampal GABA circuitry. Nat Neurosci 12:1020–1027
Bussemakers MJ, van Bokhoven A, Verhaegh GW et al (1999) DD3: a new prostate-specific gene,
highly overexpressed in prostate cancer. Cancer Res 59:5975–5979
410 M. Szymański and J. Barciszewski
Davis IJ, His B, Arroyo JD et al (2003) Cloning of an alpha-TFEB fusion in renal tumors
harboring the t(6;11)(p21;q13) chromosome translocation. Proc Natl Acad Sci USA 100:
6051–6056
Davis S, Lollo B, Freier S et al (2006) Improved targeting of miRNA with antisense oligonucleo-
tides. Nucleic Acids Res 34:2294–2304
De Biase I, Chutake YK, Rindler PM et al (2009) Epigenetic silencing in Friedreich ataxia is
associated with depletion of CTCF (CCCTC-binding factor) and antisense transcription. PLoS
One 4:e7914
de Kok JB, Verhaegh GW, Roelofs RW et al (2002) DD3(PCA3), a very sensitive and specific
marker to detect prostate tumors. Cancer Res 62:2695–2698
de Napoles M, Mermoud JE, Wakao R et al (2004) Polycomb group proteins Ring1A/B link
ubiquitylation of histone H2A to heritable gene silencing and X inactivation. Dev Cell
7:663–676
de Parseval N, Alkabbani H, Heidmann T (1999) The long terminal repeats of the HERV-H human
endogenous retrovirus contain binding sites for transcriptional regulation by the Myb protein.
J Gen Virol 80:841–845
de Smith AJ, Purmann C, Walters RG et al (2009) A deletion of the HBII-85 class of small
nucleolar RNAs (snoRNAs) is associated with hyperphagia, obesity and hypogonadism. Hum
Mol Genet 18:3257–3265
DeBaun MR, Niemitz EL, McNeil DE et al (2002) Epigenetic alterations of H19 and LIT1
distinguish patients with Beckwith–Wiedemann syndrome with cancer and birth defects. Am
J Hum Genet 70:604–611
Dong XY, Guo P, Boyd J et al (2009) Implication of snoRNA U50 in human breast cancer. J Genet
Genomics 36:447–454
Duning K, Buck F, Barnekow A et al (2008) SYNCRIP, a component of dendritically localized
mRNPs, binds to the translation regulator BC200 RNA. J Neurochem 105:351–359
Ebert MS, Neilson JR, Sharp PA (2007) MicroRNA sponges: competitive inhibitors of small
RNAs in mammalian cells. Nat Methods 4:721–726
Elmén J, Lindow M, Silahtaroglu A et al (2008) Antagonism of microRNA-122 in mice by
systemically administered LNA-antimiR leads to up-regulation of a large set of predicted
target mRNAs in the liver. Nucleic Acids Res 36:1153–1162
Emberley E, Huang GJ, Hamedani MK et al (2003) Identification of new human coding steroid
receptor RNA activator isoforms. Biochem Biophys Res Commun 301:509–515
Enklaar T, Zabel BU, Prawitt D (2006) Beckwith–Wiedemann syndrome: multiple molecular
mechanisms. Expert Rev Mol Med 8:1–19
Esau C, Davis S, Murray SF et al (2006) miR-122 regulation of lipid metabolism revealed by
in vivo antisense targeting. Cell Metab 3:87–98
Espinoza CA, Allen TA, Hieb AR et al (2004) B2 RNA binds directly to RNA polymerase to
repress transcript synthesis. Nat Struct Mol Biol 11:822–829
Fabian MR, Mathonnet G, Sundermeier T et al (2009) Mammalian miRNA RISC recruits CAF1
and PABP to affect PABP-dependent deadenylation. Mol Cell 35:868–880
Faghihi MA, Modarresi F, Khalil AM et al (2008) Expression of a noncoding RNA is elevated in
Alzheimer’s disease and drives rapid feed-forward regulation of beta-secretase. Nat Med
14:723–730
Fellig Y, Ariel I, Ohana P et al (2005) H19 expression in hepatic metastases from a range of human
carcinomas. J Clin Pathol 58:1064–1068
Feng J, Bi C, Clark BS et al (2006) The Evf-2 noncoding RNA is transcribed from the Dlx-5/6
ultraconserved region and functions as a Dlx-2 transcriptional coactivator. Genes Dev
20:1470–1484
Fontana L, Fiori ME, Albini S et al (2008) Antagomir-17-5p abolishes the growth of therapy-
resistant neuroblastoma through p21 and BIM. PLoS One 3:e2236
412 M. Szymański and J. Barciszewski
Frankel LB, Christoffersen NR, Jacobsen A et al (2008) Programmed cell death 4 (PDCD4) is an
important functional target of the microRNA miR-21 in breast cancer cells. J Biol Chem
283:1026–1033
Fu X, Ravindranath L, Tran N et al (2006) Regulation of apoptosis by a prostate-specific and
prostate cancer-associated noncoding gene, PCGEM1. DNA Cell Biol 25:135–141
Gabory A, Ripoche MA, Le Digarcher A (2009) H19 acts as a trans regulator of the imprinted gene
network controlling growth in mice. Development 136:3413–3421
Gabory A, Ripoche MA, Yoshimizu T et al (2006) The H19 gene: regulation and function of a non-
coding RNA. Cytogenet Genome Res 113:188–193
Galli UM, Sauter M, Lecher B et al (2005) Human endogenous retrovirus rec interferes with germ
cell development in mice and may cause carcinoma in situ, the predecessor lesion of germ cell
tumors. Oncogene 24:3223–3228
Hansen A, Henderson S, Lagos D et al (2010) KSHV-encoded miRNAs target MAF to induce
endothelial cell reprogramming. Genes Dev 24:195–205
Hao Y, Crenshaw T, Moulton T et al (1993) Tumour-suppressor activity of H19 RNA. Nature
365:764–767
Hatchell EC, Colley SM, Beveridge DJ et al (2006) SLIRP, a small SRA binding protein, is a
nuclear receptor corepressor. Mol Cell 22:657–668
Hebert SS, Horre K, Nicolai L et al (2008a) MicroRNA regulation of Alzheimer’s amyloid
precursor protein expression. Neurobiol Dis. doi:10.1016/j.nbd.2008.11.009
Hebert SS, Horre K, Nicolai L et al (2008b) Loss of microRNA cluster miR-29a/b-1 in sporadic
Alzheimer’s disease correlates with increased BACE1/beta-secretase expression. Proc Natl
Acad Sci USA 105:6415–6420
Hobert O (2004) Common logic of transcription factor and microRNA action. Trends Biochem Sci
9:462–468
Hon LS, Zhang Z (2007) The roles of binding site arrangement and combinatorial targeting in
microRNA repression of gene expression. Genome Biol 8:R166
Horike S, Mitsuya K, Meguro M et al (2000) Targeted disruption of the human LIT1 locus defines
a putative imprinting control element playing an essential role in Beckwith–Wiedemann
syndrome. Hum Mol Genet 9:2075–2083
Horsthemke B, Wagstaff J (2008) Mechanisms of imprinting of the Prader–Willi/Angelman
region. Am J Med Genet 146A:2041–2052
Hutchinson JN, Ensminger AW, Clemson CM et al (2007) A screen for nuclear transcripts
identifies two linked noncoding RNAs associated with SC35 splicing domains. BMC Geno-
mics 8:39
Iida T, Nakayama J, Moazed D (2008) siRNA-mediated heterochromatin establishment requires
HP1 and is associated with antisense transcription. Mol Cell 31:178–189
Imamura T, Yamamoto S, Ohgane J et al (2004a) Non-coding RNA directed DNA demethylation
of Sphk1 CpG island. Biochem Biophys Res Commun 322:593–600
Imamura T, Miyauchi-Senda N, Tanaka S et al (2004b) Identification of genetic and epigenetic
similarities of SPHK1/Sphk1 in mammals. J Vet Med Sci 66:1387–1393
Imamura T, Ohgane J, Ito S et al (2001) CpG island of rat sphingosine kinase-1 gene: tissue-
dependent DNA methylation status and multiple alternative first exons. Genomics 76:113–125
Ioannidis P, Kottaridi C, Dimitriadis E et al (2004) Expression of the RNA-binding protein CRD-
BP in brain and non-small cell lung tumors. Cancer Lett 209:245–250
Ishii N, Ozaki K, Sato H et al (2006) Identification of a novel non-coding RNA, MIAT, that
confers risk of myocardial infarction. J Hum Genet 51:1087–1099
Ji P, Diederichs S, Wang W et al (2003) MALAT-1, a novel noncoding RNA, and thymosin beta4
predict metastasis and survival in early-stage non-small cell lung cancer. Oncogene 22:8031–8041
Johnson SM, Grosshans H, Shingara J et al (2005) RAS is regulated by the let-7 microRNA family.
Cell 120:635–647
Johnson CD, Esquela-Kerscher A, Stefani G et al (2007) The let-7 MicroRNA represses cell
proliferation pathways in human cells. Cancer Res 67:7713–7722
Noncoding RNAs as Therapeutic Targets 413
Jonkers I, Monkhorst K, Rentmeester E et al (2008) Xist RNA is confined to the nuclear territory of
the silenced X chromosome throughout the cell cycle. Mol Cell Biol 28:5583–5594
Kanduri C, Thakur N, Pandey RR (2006) The length of the transcript encoded from the Kcnq1ot1
antisense promoter determines the degree of silencing. EMBO J 25:2096–2102
Karube Y, Tanaka H, Osada H et al (2005) Reduced expression of Dicer associated with poor
prognosis in lung cancer patients. Cancer Sci 96:111–155
Katayama S, Tomaru Y, Kasukawa T et al (2005) Antisense transcription in the mammalian
transcriptome. Science 309:1564–1566
Kawashima H, Takano H, Sugita S et al (2003) A novel steroid receptor co-activator protein
(SRAP) as an alternative form of steroid receptor RNA-activator gene: expression in prostate
cancer cells and enhancement of androgen receptor activity. Biochem J 369:163–171
Kent OA, Mendell JT (2006) A small piece in the cancer puzzle: microRNAs as tumor suppressors
and oncogenes. Oncogene 25:6188–6196
Kim J, Inoue K, Ishii J et al (2007) A MicroRNA feedback circuit in midbrain dopamine neurons.
Science 317:1220–1224
Kiriakidou M, Tan GS, Lamprinaki S et al (2007) An mRNA m7G cap binding-like motif within
human Ago2 represses translation. Cell 129:1141–1151
Kishore S, Khanna A, Zhang Z et al (2010) The snoRNA MBII-52 (SNORD 115) is processed into
smaller RNAs and regulates alternative splicing. Hum Mol Genet 19(7):1153–1164
Kishore S, Stamm S (2006) The snoRNA HBII-52 regulates alternative splicing of the serotonin
receptor 2C. Science 311:230–232
Klemke M, Meyer A, Hashemi Nezhad M et al (2010) Loss of let-7 binding sites resulting from
truncations of the 30 untranslated region of HMGA2 mRNA in uterine leiomyomas. Cancer
Genet Cytogenet 196:119–123
Kocerha J, Faghihi MA, Lopez-Toledano MA et al (2009) MicroRNA-219 modulates NMDA
receptor-mediated neurobehavioral dysfunction. Proc Natl Acad Sci USA 106:3507–3512
Koochek M, Harvard C, Hildebrand MJ et al (2006) 15q duplication associated with autism in a
multiplex family with a familial cryptic translocation t(14;15)(q11.2;q13.3) detected using
array-CGH. Clin Genet 69:124–134
Krichevsky AM, Gabriely G (2009) miR-21: a small multi-faceted RNA. J Cell Mol Med
13:39–53
Krystal GW, Armstrong BC, Battey JF (1990) N-myc mRNA forms an RNA–RNA duplex with
endogenous antisense transcripts. Mol Cell Biol 10:4180–4191
Kuhn DE, Nuovo GJ, Martin MM et al (2008) Human chromosome 21-derived miRNAs are
overexpressed in down syndrome brains and hearts. Biochem Biophys Res Commun
370:473–477
Kuiper RP, Schepens M, Thijssen J et al (2003) Upregulation of the transcription factor TFEB in t
(6;11)(p21;q13)-positive renal cell carcinomas due to promoter substitution. Hum Mol Genet
12:1661–1669
Kumar MS, Lu J, Mercer KL et al (2007) Impaired microRNA processing enhances cellular
transformation and tumorigenesis. Nat Genet 39:673–677
Kuwabara T, Hsieh J, Nakashima K et al (2004) A small modulatory dsRNA specifies the fate of
adult neural stem cells. Cell 116:779–793
Lander ES, Linton LM, Birren B et al (2001) Initial sequencing and analysis of the human genome.
Nature 409:860–921
Lanz R, Razani B, Goldberg AD et al (2002) Distinct RNA motifs are important for coactivation of
steroid hormone receptors by steroid receptor RNA activator (SRA). Proc Natl Acad Sci USA
99:16081–16086
Lanz RB, McKenna NJ, Onate SA et al (1999) A steroid receptor coactivator, SRA, functions as an
RNA and is present in an SRC-1 complex. Cell 97:17–27
Lee YS, Dutta A (2007) The tumor suppressor microRNA let-7 represses the HMGA2 oncogene.
Genes Dev 21:1025–1030
Leygue E (2007) Steroid receptor RNA activator (SRA1): unusual bifaceted gene products with
suspected relevance to breast cancer. Nucl Recept Signal 5:e006
414 M. Szymański and J. Barciszewski
Mishra PK, Tyagi N, Kumar M et al (2009) MicroRNAs as a therapeutic target for cardiovascular
diseases. J Cell Mol Med 13:778–789
Mohammad F, Mondal T, Kanduri C (2009) Epigenetics of imprinted long noncoding RNAs.
Epigenetics 4:277–286
Morison IM, Ramsay JP, Spencer HG (2005) A census of mammalian imprinting. Trends Genet
21:457–463
Mourtada-Maarabouni M, Pickard MR, Hedge VL et al (2009) GAS5, a non-protein-coding RNA,
controls apoptosis and is downregulated in breast cancer. Oncogene 28:195–208
Munroe SH, Lazar MA (1991) Inhibition of c-erbA mRNA splicing by a naturally occurring
antisense RNA. J Biol Chem 266:22083–22086
Murakami Y, Yasuda T, Saigo K et al (2006) Comprehensive analysis of microRNA expression
patterns in hepatocellular carcinoma and non-tumorous tissues. Oncogene 25:2537–2545
Nagano T, Mitchell JA, Sanz LA et al (2008) The air noncoding RNA epigenetically silences
transcription by targeting G9a to chromatin. Science 322:1717–1720
Nakano S, Murakami K, Meguro M et al (2006) Expression profile of LIT1/KCNQ1OT1 and
epigenetic status at the KvDMR1 in colorectal cancers. Cancer Sci 97:1147–1154
Newall AE, Duthie S, Formstone E et al (2001) Primary non-random X inactivation associated
with disruption of Xist promoter regulation. Hum Mol Genet 10:581–589
Nguyen VT, Kiss T, Michels AA et al (2001) 7SK small nuclear RNA binds to and inhibits the
activity of CDK9/cyclin T complexes. Nature 414:322–325
Niemitz EL, DeBaun MR, Fallon J et al (2004) Microdeletion of LIT1 in familial Beckwith-
Wiedemann syndrome. Am J Hum Genet 75:844–849
Okutsu T, Kuroiwa Y, Kagitani F et al (2000) Expression and imprinting status of human PEG8/
IGF2AS, a paternally expressed antisense transcript from the IGF2 locus, in Wilms’ tumors.
J Biochem 127:475–483
Papagiannakopoulos T, Shapiro A, Kosik KS (2008) MicroRNA-21 targets a network of key
tumor-suppressive pathways in glioblastoma cells. Cancer Res 68:8164–8172
Pauler FM, Koerner MV, Barlow DP (2007) Silencing by imprinted noncoding RNAs: is tran-
scription the answer? Trends Genet 23:284–292
Pauley KM, Cha S, Chan EK (2009) MicroRNA in autoimmunity and autoimmune diseases.
J Autoimmun 32:189–194
Payer B, Lee JT (2008) X chromosome dosage compensation: how mammals keep the balance.
Annu Rev Genet 42:733–772
Peter ME (2009) Let-7 and miR-200 microRNAs: guardians against pluripotency and cancer
progression. Cell Cycle 8:843–852
Petrovics G, Zhang W, Makarem M et al (2004) Elevated expression of PCGEM1, a prostate-
specific gene with cell growth-promoting function, is associated with high-risk prostate cancer
patients. Oncogene 23:605–611
Pibouin L, Villaudy J, Ferbus D et al (2002) Cloning of the mRNA of overexpression in colon
carcinoma-1: a sequence overexpressed in a subset of colon carcinomas. Cancer Genet
Cytogenet 133:55–60
Pillai RS, Bhattacharyya SN, Artus CG et al (2005) Inhibition of translational initiation by Let-7
microRNA in human cells. Science 309:1573–1576
Pizzuti A, Novelli G, Ratti A et al (1999) Isolation and characterization of a novel transcript
embedded within HIRA, a gene deleted in DiGeorge syndrome. Mol Genet Metabol 67:227–235
Plath K, Talbot D, Hamer KM et al (2004) Developmentally regulated alterations in Polycomb
repressive complex 1 proteins on the inactive X chromosome. J Cell Biol 167:1025–1035
Polesskaya OO, Haroutunian V, Davis KL et al (2003) Novel putative nonprotein-coding RNA
gene from 11q14 displays decreased expression in brains of patients with schizophrenia.
J Neurosci Res 74:111–122
Rajaram V, Knezevich S, Bove KE et al (2007) DNA sequence of the translocation breakpoints in
undifferentiated embryonal sarcoma arising in mesenchymal hamartoma of the liver harboring
the t(11;19)(q11;q13.4) translocation. Genes Chromosomes Cancer 46:508–513
416 M. Szymański and J. Barciszewski
Reis EM, Ojopi EP, Alberto FL et al (2005) Large-scale transcriptome analyses reveal new
genetic marker candidates of head, neck, and thyroid cancer. Cancer Res 65:1693–1699
Roldo C, Missiaglia E, Hagan JP et al (2006) MicroRNA expression abnormalities in pancreatic
endocrine and acinar tumors are associated with distinctive pathological features and clinical
behavior. J Clin Oncol 24:4677–4684
Sahoo T, del Gaudio D, German JR et al (2008) Prader–Willi phenotype caused by paternal
deficiency for the HBII-85 C/D box small nucleolar RNA cluster. Nat Genet 40:719–721
Sevignani C, Calin GA, Nnadi SC et al (2007) MicroRNA genes are frequently located near mouse
cancer susceptibility loci. Proc Natl Acad Sci USA 104:8017–8022
Shamovsky I, Ivannikov M, Kandel ES et al (2006) RNA-mediated response to heat shock in
mammalian cells. Nature 440:556–560
Shamovsky I, Nudler E (2009) Isolation and characterization of the heat shock RNA 1. Methods
Mol Biol 540:265–279
Shi Y, Downes M, Xie W et al (2001) SHARP, an inducible cofactor that integrates nuclear
receptor repression and activation. Genes Dev 15:1140–1151
Shiota K (2004) DNA methylation profiles of CpG islands for cellular differentiation and devel-
opment in mammals. Cytogenet Genome Res 105:325–334
Si ML, Zhu S, Wu H et al (2007) miR-21-mediated tumor growth. Oncogene 26:2799–2803
Sleutels F, Tjon G, Ludwig T et al (2003) Imprinted silencing of Slc22a2 and Slc22a3 does not
need transcriptional overlap between Igf2r and Air. EMBO J 22:3696–3704
Soejima H, Nakagawachi T, Zhao W et al (2004) Silencing of imprinted CDKN1C gene expres-
sion is associated with loss of CpG and histone H3 lysine 9 methylation at DMR-LIT1 in
esophageal cancer. Oncogene 23:4380–4388
Sonkoly E, Bata-Csorgo Z, Pivarcsi A et al (2005) Identification and characterization of a
novel, psoriasis susceptibility-related noncoding RNA gene, PRINS. J Biol Chem 280:
24159–24167
Srikantan V, Zou Z, Petrovics G et al (2000) PCGEM1, a prostate-specific gene, is overexpressed
in prostate cancer. Proc Natl Acad Sci USA 97:12216–12221
Stefani G, Slack FJ (2008) Small non-coding RNAs in animal development. Nat Rev Mol Cell Biol
9:219–230
Stern-Ginossar N, Elefant N, Zimmermann A et al (2007) Host immune system gene targeting by a
viral miRNA. Science 317:376–381
Sutherland HF, Wadey R, McKie JM et al (1996) Identification of a novel transcript disrupted by a
balanced translocation associated with DiGeorge syndrome. Am J Hum Genet 59:23–31
Szymanski M, Barciszewski J (2006) RNA regulation in mammals. Ann NY Acad Sci
1067:461–468
Taft RJ, Pheasant M, Mattick JS (2007) The relationship between non-protein-coding DNA and
eukaryotic complexity. Bioessays 29:288–299
Tam OH, Aravin AA, Stein P et al (2008) Pseudogene-derived small interfering RNAs regulate
gene expression in mouse oocytes. Nature 453:534–538
Tang X, Tang G, Ozcan S (2008) Role of microRNAs in diabetes. Biochim Biophys Acta
1779:697–701
Terranova R, Yokobayashi S, Stadler MB et al (2009) Polycomb group proteins Ezh2 and Rnf2
direct genomic contraction and imprinted repression in early mouse embryos. Dev Cell
15:668–679
Thakur N, Tiwari VK, Thomassin H et al (2004) An antisense RNA regulates the bidirectional
silencing property of the Kcnq1 imprinting control region. Mol Cell Biol 24:7855–7862
Tiedge H, Chen W, Brosius J (1993) Primary structure, neural-specific expression, and dendritic
location of human BC200 RNA. J Neurosci 13:2382–2390
Tiedge H, Fremeau RT Jr, Weinstock PH et al (1991) Dendritic location of neural BC1 RNA. Proc
Natl Acad Sci USA 88:2093–2097
Tsang WP, Wong TWL, Cheung AHH et al (2007) Induction of drug resistance and transformation
in human cancer cells by the noncoding RNA CUDR. RNA 13:890–898
Noncoding RNAs as Therapeutic Targets 417
Tufarelli C, Stanley JA, Garrick D et al (2003) Transcription of antisense RNA leading to gene
silencing and methylation as a novel cause of human genetic disease. Nat Genet 34:157–165
Turner JD, Schote AB, Macedo JA et al (2006) Tissue specific glucocorticoid receptor expression,
a role for alternative first exon usage? Biochem Pharmacol 72:1529–1536
Umlauf D, Goto Y, Cao R et al (2004) Imprinting along the Kcnq1 domain on mouse chromosome
7 involves repressive histone methylation and recruitment of Polycomb group complexes. Nat
Genet 36:1296–1300
Verhagen PC, Hermans KG, Brok MO et al (2002) Deletion of chromosomal region 6q14-16 in
prostate cancer. Int J Cancer 102:142–147
Vincent JB, Herbrick JA, Gurling HM et al (2000) Identification of a novel gene on chromosome
7q31 that is interrupted by a translocation breakpoint in an autistic individual. Am J Hum
Genet 67:510–514
Wang G, van der Walt JM, Mayhew G et al (2008) Variation in the miRNA-433 binding site of
FGF20 confers risk for Parkinson disease by overexpression of alpha-synuclein. Am J Hum
Genet 82:283–289
Wang H, Iacoangeli A, Popp S et al (2002) Dendritic BC1 RNA: functional role in regulation of
translation initiation. J Neurosci 22:10232–10241
Wang XS, Zhang Z, Wang HC et al (2006) Rapid identification of UCA1 as a very sensitive and
specific unique marker for human bladder carcinoma. Clin Cancer Res 12:4851–4858
Watanabe M, Yanagisawa J, Kitagawa H et al (2001) A subfamily of RNA-binding DEAD-box
proteins acts as an estrogen receptor a coactivator through the N-terminal activation domain
(AF-1) with an RNA coactivator, SRA. EMBO J 20:1341–1352
Watanabe T, Totoki Y, Toyoda A et al (2008) Endogenous siRNAs from naturally formed dsRNAs
regulate transcripts in mouse oocytes. Nature 453:539–543
Waterston RH, Lindblad-Toh K, Birney E et al (2002) Initial sequencing and comparative analysis
of the mouse genome. Nature 420:520–562
Willingham AT, Orth AP, Batalov S et al (2005) A strategy for probing the function of noncoding
RNAs finds a repressor of NFAT. Science 309:1570–1573
Wutz A, Smrzka OW, Schweifer N et al (1997) Imprinted expression of the Igf2r gene depends on
an intronic CpG island. Nature 389:745–749
Xiao J, Yang B, Lin H et al (2007) Novel approaches for gene-specific interference via manipulat-
ing actions of microRNAs: examination on the pacemaker channel genes HCN2 and HCN4.
J Cell Physiol 212:285–292
Xie H, Sun L, Lodish HF (2009) Targeting microRNAs in obesity. Expert Opin Ther Targets
13:1227–1238
Yakovchuk P, Goodrich JA, Kugel JF (2009) B2 RNA and Alu RNA repress transcription by
disrupting contacts between RNA polymerase II and promoter DNA within assembled com-
plexes. Proc Natl Acad Sci USA 106:5569–5574
Yan MD, Hong CC, Lai GM et al (2005) Identification and characterization of a novel gene Saf
transcribed from the opposite strand of Fas. Hum Mol Genet 14:1465–1474
Yang Z, Zhu Q, Luo K et al (2001) The 7SK small nuclear RNA inhibits the CDK9/cyclin T1
kinase to control transcription. Nature 414:317–322
Yekta S, Shih IH, Bartel DP (2004) MicroRNA-directed cleavage of HOXB8 mRNA. Science
304:594–596
Yik JH, Chen R, Pezda AC et al (2004) A human immunodeficiency virus type 1 Tat-like arginine-
rich RNA-binding domain is essential for HEXIM1 to inhibit RNA polymerase II transcription
through 7SK snRNA-mediated inactivation of P-TEFb. Mol Cell Biol 24:5094–5105
Yu W, Gius D, Onyango P et al (2008) Epigenetic silencing of tumour suppressor gene p15 by its
antisense RNA. Nature 451:202–206
Zalfa F, Adinolfi S, Napoli I et al (2005) Fragile X mental retardation protein (FMRP) binds
specifically to the brain cytoplasmic RNAs BC1/BC200 via a novel RNA-binding motif. J Biol
Chem 280:33403–33410
418 M. Szymański and J. Barciszewski
Zalfa F, Giorgi M, Primerano B et al (2003) The fragile X syndrome protein FMRP associates
with BC1 RNA and regulates the translation of specific mRNAs at synapses. Cell 112:317–327
Zhang X, Rice K, Wang Y et al (2009) Maternally expressed gene 3 (MEG3) noncoding
ribonucleic acid: isoform structure, expression, and functions. Endocrinology 151(3):939–947
Zhao X, Patton JR, Davis SL et al (2004) Regulation of nuclear receptor activity by a pseudour-
idine synthase through posttranscriptional modification of steroid receptor RNA activator. Mol
Cell 15:549–558
Zhu S, Si ML, Wu H et al (2007) MicroRNA-21 targets the tumor suppressor gene tropomyosin1
(TPM1). J Biol Chem 282:14328–14336
Noncoding RNAs at H19/IGF2 Locus:
Role in Imprinting, Gene Expression,
and Associated Pathologies
Contents
1 The H19/IGF2 Locus and the Parental Imprinting Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
1.1 Overview and Description of the 11p15.5 Locus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
1.2 The Insulator Model of Imprinting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424
2 The mRNA-Like Noncoding RNA H19 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
2.1 Properties and Expression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429
2.2 Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
2.3 Regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
3 The Noncoding Antisense RNA 91H . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
3.1 Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
3.2 Hypothesis About 91H Mechanism of Action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
4 H19/IGF2 Locus-Associated Pathologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
4.1 Hormone-Dependent Cancers (Breast, Uterus) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
4.2 Children Syndromes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
V.A. Erdmann and J. Barciszewski (eds.), RNA Technologies and Their Applications, 419
RNA Technologies, DOI 10.1007/978-3-642-12168-5_19,
# Springer-Verlag Berlin Heidelberg 2010
420 N. Berteaux et al.
trans-factors and epigenetic modifications are also required for full expression of
genes. It is well assumed that H19 RNA functions as a riboregulator of which,
expression is developmentally regulated. Elsewhere, the antisense RNA 91H has
been recently discovered as a large and maternally imprinted noncoding RNA. It
plays a role in the paternal IGF2 regulation and is overexpressed in breast cancer.
This original trans-effect may be due to 91H participation in the three-dimensional
organization of the locus, essential for the appropriate expression of genes. In this
chapter, we summarize our current understanding of the molecular and biological
roles of the ncRNAs expressed at the H19/IGF2 domain, both in terms of their
contribution to genomic imprinting control, as well as in terms of cellular targets
they might interact with. We also review knowledge of the locus-associated patho-
logies such as cancers and children syndromes.
The mammalian genome contains a small but growing number of genes that are
subject to genomic imprinting (Edwards and Ferguson-Smith 2007; Verona et al.
2003). Genomic imprinting is a form of epigenetic gene regulation that results in
expression from a single allele in a parent-of-origin-dependent manner. This form of
monoallelic expression is essential to normal mammalian development. While the
precise nature of the initial epigenetic imprint remains an intensively investigated
topic, it is assumed that the parental imprint is set in the germline, when genomes are
in distinct compartments and can be differentially modified. After fertilization, the
parental imprints must survive the reprogramming that takes place in the preimplan-
tation embryo, including DNA demethylation and changes in histones modifications
(Reik et al. 2001). Imprinting is maintained throughout development and then erased
before being reestablished in the next generation’s germline. About 90 genes have
been reported to be imprinted even if some of them probably remain to be discovered
(for a complete list, see http://igc.otago.ac.nz/home.html and http://www.har.mrc.
ac.uk/research/genomicimprinting/maps.html). Despite extensive studies and some
major advancement regarding this intriguing phenomenon, we have not yet fully
characterized the underlying molecular mechanisms of genomic imprinting. Never-
theless, some principal hallmarks of imprinted genes can be listed:
l Gene expression is allele-specific.
l Gene expression is often tissue or stage-specific.
l Many of imprinted genes are found in clusters throughout the genome.
l The clusters contain two or more imprinted genes over a region that can span
1 Mb or more.
Noncoding RNAs at H19/IGF2 Locus: Role in Imprinting, Gene Expression 421
)
i t1
IP AS) )
5- T1
,L
ssc P
as
KC 1C S (T , IM
1-
QT
5
2)
KN -A ssc
(K )
T1 T1
vL
)
AS )
H2
CD A1L (T
M C6
1)
1 D 7K
SL C22 a2)
1 O LQ
C L (T )
AS X A1
R
2 1L
2 ( SS
NQ (p5
SC MT
SL Phld
NQ ( K v
EM P
N
C2 A
F- S
(
3
IG 2-A
TS 4
P H (T
3(
KC Q1
L2
L
CD 4
PM
P1
SC
81
2
F-
RP
N
H
9
S
NA
KC
TH
TR
IG
H1
TS
IN
91
M
P
IC2 IC1
Centromere Telomere
Fig. 1 Representation of the human chromosomic region 11p15.5. This 1 Mb-locus includes nine
imprinted genes and two imprinting centers. It is delimited by the two maternally expressed genes
TSSC3 et H19 flanked by two nonimprinted genes NAP1L4 et MRPL23 (Tsang et al. 1995; Paulsen
et al. 1998). This region is homologous with the distal region of chromosome 7 in mouse. In spite
of the phylogenetic maintenance if the region over species, it exhibits some structural and
functional discrepancies. Indeed, the TSSC4 gene is imprinted in mouse but not in human (Paulsen
et al. 2000). The TRPM5 gene is imprinted in human but not in mouse (Prawitt et al. 2000).
A TSSC5 antisense transcript has been revealed in human but not in mouse and its imprinted status
has not yet been defined (Crider-Miller et al. 1997; Cooper et al. 1998). The CD81 gene is
imprinted in human and the imprinted status of the INS (Insulin gene) gene is not determined
(Maher and Reik 2000). The ASCL2 gene is imprinted with maternal expression in mouse whereas
its expression is biallelic in human (Miyamoto et al. 2002; Westerman et al. 2001). Usual names
are indicated and followed by secondary names in brackets. Name of antisense transcripts are
underlined and arrows indicating the sense of transcription are depicted below the chromosome
AS-SRO PWS-SRO
ICE
Kνlqt1-as, Lit1
Tssc3 Tssc5 Cdkn1c Kνlqt1 Ltrpc5 Tssc4 Cd81 Tssc6 Ascl2
KnDMR1
Fig. 2 Antisense RNA associated to imprinted clusters. Igf-2r/Air, Ube3A/Ube3A-as, and Kcnq1/
Kcnq1ot1 are the three best described examples of long antisense RNA, which take part in
epigenetic modifications and gene silencing within imprinted loci. (a) The region of chromosome
15q11–q13 responsible for the Angelman and Prader–Willi syndromes contains a number of
imprinted genes that are coordinately regulated by an imprinting center that contains two func-
tional elements, the PWS-SRO and the AS-SRO. The 460 kb-long Ube3A-as RNA is initiated in
the imprinting center from the paternal allele. It acts as a host gene for the transcription of several
snoRNA (small nucleolar RNA) and represses the UBE3A gene on the paternal allele (Rougeulle
et al. 1998; Runte et al. 2001; Landers et al. 2004). (b) The IGF2R locus. The imprinting center
(ICE) produces a paternally expressed 108-kb long transcript called Air that is necessary for the
silencing in cis of the genes IGF2R, Slc22a1, Slc22a2, and Slc22a3. Air could be implicated in the
methylation spreading thought the locus. (Rougeulle and Heard 2002; Sleutels et al. 2002).
(c) Within the 11p15.5 region, the Kcnq1 gene contains within its intron 10 the imprinting center
KvDMR1, which harbors bidirectional silencing property. This is linked to a paternal antisense
RNA, Kcnq1ot1 (also called Lit1) initiated from KvDMR1. The Kcnq1ot1 promoter shows a
maternal specific methylation. Kcnq1ot1 transcript has a key role in silencing of genes contained in
the Kcnq1 gene imprinted region and it participates to both silencing activity and methylation
spreading (Pandey et al. 2004; Thakur et al. 2004)
Numerous imprinted genes are associated to long antisense RNAs that overlap
several genes. The best described examples are Igf-2r/Air, Ube3A/Ube3A-as, and
Kcnq1/Kcnq1ot1 (Fig. 2). The first demonstration of a direct implication of an
Noncoding RNAs at H19/IGF2 Locus: Role in Imprinting, Gene Expression 423
antisense RNA concerns the Air transcript. It consists of a 108 kb-long transcript
localized in the imprinting center within the IGF2r gene and is necessary for the
paternal repression of the gene of the locus (Rougeulle and Heard 2002; Sleutels
et al. 2002). In spite of its well-established role in imprinting process, the molecular
mechanism remains unclear and authors propose hypothesis of methylation propa-
gation from the IGF2R gene or of repressive ARN/protein complexes formation.
The 460 kb-long Ube3A-as RNA is initiated in the imprinting center of the
Prader–Willi syndrome. It acts as a host gene for the transcription of several
snoRNA (small nucleolar RNA) and represses the UBE3A gene on the paternal
allele (Rougeulle et al. 1998; Runte et al. 2001; Landers et al. 2004).
Within the 11p15.5 region, the ICR2 is located within the intron 10 of the
Kcnq1 gene and harbors bidirectional silencing property. This feature is linked to
an antisense RNA, Kcnq1ot1 (also called Lit1), of which the promoter is
contained in ICR2. Expression of this transcript is exclusively paternal. Indeed,
the Kcnq1ot1 promoter shows a maternal-specific methylation. This differential
epigenetic mark is lost in patients affected by Beckwith–Wiedemann syndrome
(BWS) with RNA biallelic expression (Mitsuya et al. 1999; Lee et al. 1999;
Du et al. 2004). More recently, Pandey et al. (2004) have documented that the
Kcnq1ot transcript has a key role in silencing of genes contained in the Kcnq1
gene imprinted region and that it participates directly or indirectly to the methyl-
ation but without RNA interference mechanisms. Furthermore, interruption of
Kcnq1ot1 RNA production by the insertion of a polyadenylation sequence down-
stream of the promoter also caused a loss of both silencing activity and methyla-
tion spreading. Thus, the antisense RNA plays a key role in the silencing function
of the ICR (Thakur et al. 2004).
Elsewhere, a noncoding antisense RNA has also been described in the mouse
and human IGF2 genes (Moore et al. 1997; Okutsu et al. 2000). Like IGF2, this
2.2 kb mRNA transcript is maternally imprinted and overexpressed in Wilms’
tumors. IGF2-AS was expressed at levels comparable with IGF2 sense expression
derived from promoters P1 and P2 in normal tissue and in breast, ovarian, and
Wilms’ tumor tissues. It is composed of three exons, which overlap the exons 3 and
4 of the IGF2 gene (Vu et al. 2003). Its function remains unknown and its
involvement in imprinting has not yet been demonstrated, but findings indicate
that it is a good marker of Wilms’ tumor (Okutsu et al. 2000).
Finally, discovery of microRNAs (miRNA) and RNA interference could likely
provide new insights on imprinting mechanism. Then, the group of Cavaille has
identified in mouse a short cluster of maternally expressed miRNA genes (miR-431,
miR-433, miR-127, miR-434, and miR-136) transcribed and processed from a gene
antisense to the paternally expressed Rtl1 gene (Retrotransposon-like gene1) (Seitz
et al. 2003). Rtl1, also called Peg11 in sheep, displays homology with the Ty3/gypsy
retrotransposon family and its function is currently unknown (Charlier et al. 2001;
Youngson et al. 2005). Due to this peculiar sense–antisense organization, the
encoded miRNAs are obviously perfectly complementary to Rtl1 mRNA and thus
were predicted to cleave Rtl1 mRNA via RNAi-like mechanisms (Seitz et al. 2003).
Indeed, the predicted RNAi-mediated cleavage sites in the middle of the RNA
424 N. Berteaux et al.
The H19 gene is one of the first genes proven to be imprinted. In human, it lies
within 200 kbp downstream of the IGF-2 gene (Zemel et al. 1992). The regulation
of H19 and its closely linked and reciprocally imprinted neighbor, IGF2, has been
studied intensively both as a model for understanding imprinting control mechanisms
and because of its role in human diseases. The two genes are imprinted in an opposite
manner, with the paternal IGF-2 and the maternal H19 alleles being reciprocally
expressed (Giannoukakis et al. 1993; Zhang and Tycko 1992).
The H19 silent paternal allele exhibits several characteristics associated to its
transcriptional repression: it is hypermethylated in the promoter region and in the
50 region in embryonic tissues, the promoter shows a compact chromatin structure
(Bartolomei et al. 1993; Ferguson-Smith et al. 1993) and its histone acetylation rate
is lower than the one of the maternal allele (Grandjean et al. 2001).
Surprisingly, the IGF2 promoter region is not methylated and its chromatin
structure is favorable to a biallelic transcription (Sasaki et al. 1992). However,
two other regions preferentially methylated on the expressed paternal allele have
been identified within the gene: the DMR1 located 3 kbp upstream the P1 promoter
acts as a silencer on the maternal allele when it is unmethylated, and the DMR2,
located within exons 5 and 6 is an activator on the paternal allele when it is
methylated (Feil et al. 1994; Murrell et al. 2001; Constancia et al. 2000). It is
Noncoding RNAs at H19/IGF2 Locus: Role in Imprinting, Gene Expression 425
interesting to notice that both regions acquire this differential methylation after
fecundation (it is then a secondary methylation by opposition to primary methylation
that it is established in gamete and allows to distinguish the two parental alleles), and
this implies that it is not responsible for imprinting establishment.
Analysis of histone acetylation state at the murine locus shows that H4 histone
acetylation discrepancies take place within H19 and IGF2 genes in a parental-
specific manner with expressed alleles being more acetylated than silent alleles.
However, the link between DNA methylation, histone hypoacetylation, and gene
expression is established only for the H19 promoter region (Grandjean et al. 2001;
Pedone et al. 1999). Moreover, several studies show that the inhibition of histone
deacetylases deregulates gene expression at the locus with a repression of the H19
active maternal allele, changes in acetylation patterns of the ICR region (Grandjean
et al. 2001), and an IGF2 biallelic expression (Hu et al. 1998; Yang et al. 2003).
Finally, the HDAC recruitment is directly involved in the repression effect of the
insulator protein CTCF described in the next paragraph (Lutz et al. 2000).
and in vivo experiments have shown in mouse that CTCF binds unmethylated ICR
via four consensus sites and the binding is abolished by DNA methylation (Bell and
Felsenfeld 2000; Hark et al. 2000; Kanduri et al. 2000b; Szabo et al. 2000;
Holmgren et al. 2001; Ulaner et al. 2003). In human, there are seven binding sites
but the sixth is only the one to possess a differential methylation (Takai et al. 2001).
The H19/IGF2 insulation model is represented in the Fig. 3. Both genes share a
common set of enhancers located downstream from the H19 gene. In the maternal
allele, the CTCF recruitment on the unmethylated ICR acts as a chromatin bound-
ary and blocks the enhancer access to the IGF2 promoter to prevent its activation.
The H19 gene is then activated (Reik and Murrell 2000; Wolffe 2000).
+ +
× CTCF CTCF M
IGF-2 H19
+ +
×
CTCF P
IGF-2 H19
Fig. 3 Reciprocal imprinting mechanism of H19 and IGF2 genes. Activation of gene expression
is indicated by (þ), repression by (), and inhibition of the enhancer function is represented by a
vertical bar. Relative positions are expressed in kilobase pairs relatively to the H19 transcription
start site. Three principal mechanisms intervene in the regulation: the methylation (represented by
vertical bars), the enhancer activity, and the insulator activity (Hark et al. 2000). Three DNA
regions are differentially methylated according to the allele, the DMR1 and 2 of the IGF2 gene
(violet diamond) and the ICR (blue oval) at 2 kb upstream of the H19 gene. The enhancer
sequences Enh (Enhancer downstream of the H19 gene, green circles) and Huc (enhancers
upstream of the H19 gene, blue circles) represent, respectively, the endodermic and mesodermic
enhancers (Ishihara et al. 2000; Drewell et al. 2002). On the maternal allele, the nonmethylated
ICR contains four consensus CTCF binding sites (Hark et al. 2000). The CTCF DNA binding
produces then a chromatin boundary, which prohibits enhancers to access to the IGF2 gene. The
Huc enhancers may also activate the H19 gene (Drewell et al. 2002). The nonmethylated IGF2
DMR1 (violet diamond) acts as a silencer (Constancia et al. 2000). On the paternal chromosome,
the methylated ICR does not bind any protein but acts as a H19 expression repressor. The Enh
enhancers can then activate the IGF2 promoter and the methylated IGF2 DMR2 also activates
gene expression (Murrell et al. 2001). ICR has a role of transcriptional repressor for the H19 gene
(Srivastava et al. 2000). In 30 of H19, a secondary chromatin boundary independent of the
methylation delimits the imprinted domain (Ishihara and Sasaki 2002)
Noncoding RNAs at H19/IGF2 Locus: Role in Imprinting, Gene Expression 427
On the paternal allele, the ICR methylation does not allow CTCF binding and
leads to activation of the distal IGF2 promoter and gene expression (for a review,
see Lewis and Murrel 2004). Finally, the maintenance of the unmethylated state of
maternal ICR is due to the CTCF protein, which prevents the de novo methylation
in this region (Fedoriw et al. 2004; Lewis and Murrel 2004; Szabo et al. 2004). The
mesodermic enhancer activity recently discovered (Drewell et al. 2002), which
intervene in the regulation can be added to this model. Finally, the in vivo CTCF
binding upstream of the Mrpl23 gene could be a chromatin boundary delimiting the
imprinted domain (Ishihara and Sasaki 2002).
The cis elements described previously have a long-range action. They have to
physically interact with each other or with their target to exert their effects.
Chromosome conformation capture (3C) analysis in mice, which assay for physical
interactions between chromosomal regions, have suggested that CTCF has a critical
role in the epigenetic regulation of high-order chromatin structure and gene silenc-
ing over considerable distances in the genome, but the precise nature and function
of the looping is debated (Engel et al. 2008; Kurukuti et al. 2006; Lopes et al. 2003;
Murrell et al. 2004; Yoon et al. 2007). Kurukuti et al. 2006 reported that on the
paternal allele, enhancers interact with the IGF2 promoters whereas on the mater-
nal, this is prevented by CTCF binding within the H19 ICR. They demonstrated that
the maternal-specific silencing of IGF2 results when the ICR interacts with a matrix
attachment region (MAR3) and a differentially methylated region (DMR1) at the
IGF2 locus to generate a tight loop around the IGF2 gene, thereby physically
impeding Igf2 expression. Moreover, CTCF interacts with the three clustered
IGF2 promoters and recruits polycomb repressive complexes that lead to the
allele-specific methylation at lysine 27 of histone H3 (H3-K27) and to the suppres-
sion of the maternal IGF2 promoters (Li et al. 2008). Elsewhere, Murrell et al.
(2004) reported that on the maternal allele, the unmethylated ICR binds to the
DMR1 of IGF2 resulting in an inactive domain where IGF2 is away from the
enhancers. On the paternal allele, the methylated ICR associates with the methy-
lated IGF2 DMR2 moving IGF2 into the active chromatin domain (Dekker et al.
2002). More recently, it has been shown that on the maternal allele, the enhancers
make contacts throughout the H19 coding unit and promoter (Engel et al. 2008;
Kato and Sasaki 2005).
Figure 4 provides a simplified overview of available data about chromatin loop
structures at the H19/IGF2 locus (Kurukuti et al. 2006; Murrell et al. 2004; Weber
et al. 2003). Additional mechanisms exist for an imprint mark, such as chromatin
composition, organization, and histone acetylation or methylation state (Fuks 2003;
Grandjean et al. 2001), even if DNA methylation is by far the best candidate
(Bestor 2000).
428 N. Berteaux et al.
enhancers
ICR H19
CTCF
R1
M
AR
DM
3
DMR2
Maternal allele
IGF2
H19
enhancers
ICR
ICR
DMR1 DMR2
DMR2 MAR3
Paternal allele
Fig. 4 Chromatin loop structures at the H19/IGF2 locus. Chromosome conformation capture (3C)
assays revealed long-range physical interactions within the H19/IGF2 region. These chromatin
structures are orchestrated by CTCF and regulate epigenetic hallmarks and gene silencing within
the locus. On the maternal allele, CTCF binding to the ICR prevents interaction between enhancers
and IGF2 promoter. ICR interacts with a matrix attachment region MAR3 and with the differen-
tially methylated region DMR1 within the IGF2 region. This leads to the formation of a silencing
loop, which impedes IGF2 expression. Enhancers can interact with the H19 promoter and activate
its expression (Engel et al. 2008; Kato and Sasaki 2005). Moreover, CTCF interacts with the three
clustered IGF2 promoters and recruits polycomb repressive complexes that lead to the allele-
specific histone methylation and to suppression of the maternal IGF2 promoters (Li et al. 2008).
On the paternal allele, enhancers interact with the IGF2 promoters (Kurukuti et al. 2006). The
methylated ICR associates with the methylated IGF2 DMR2 moving IGF2 into the active
chromatin domain (Dekker et al. 2002). ICR methylation spreads to the H19 promoter that impairs
H19 paternal expression
H19 encodes a spliced and polyadenylated RNA that lacks conserved open reading
frames (ORFs) but does have a conserved secondary RNA structure (Juan et al.
2000). Even if extensive deletions and/or point mutations of an ectopic human H19
RNA generate a protein, (Joubel et al. 1996) no endogenous translation product has
so far been identified (Pachnis et al. 1984, 1988). Therefore, it was quickly proposed
that H19 RNA functions as a riboregulator (Brannan et al. 1990) of which
Noncoding RNAs at H19/IGF2 Locus: Role in Imprinting, Gene Expression 429
The H19 gene was discovered in the mouse as a gene under coordinate regulation
with (-feto-protein in the liver (Pachnis et al. 1984). More recently, Juan et al.
(2000) brought evidence for evolutionarily conserved secondary structure in the
H19 RNA from several mammalian species. The H19 gene is an unusual gene, in
that it is transcribed by RNA polymerase II, processed by capping, splicing, and
polyadenylation; but it does not appear to encode a protein. Actually, the particu-
larity of the H19 transcript is its inability to be translated when the 50 untranslated-
region (50 UTR) is not experimentally altered (deletions and/or point mutations)
(Pachnis et al. 1988; Joubel et al. 1996). Furthermore, hypothetical translation of
established sequences from a range of mammalian species shows an absence
of conserved ORFs of any size. Consequently, given the evolutionary conservation
of structure at the RNA level and the absence of conservation at the protein level, it
has been proposed that the mature transcript is the functional product of the H19
gene and that its function requires the ability to fold into a specific secondary
structure. As early as 1990, Brannan et al. (1990) proposed that this gene could act
as a “riboregulator.”
The H19 gene encodes one of the most abundant RNAs in the developing mouse
and human embryo (Pachnis et al. 1984; Brannan et al. 1990). It is expressed at the
blastocyst stage of development and accumulates to high levels in tissues of
430 N. Berteaux et al.
endodermal and mesodermal origins (Poirier et al. 1991; Lustig et al. 1994) as well
as ectodermal origin (Ohlsson et al. 1994, Hemberger et al. 1998). H19 is expressed
in the choroid plexus and leptomeninges of the developing mouse fetus (Svensson
et al. 1995) but not in these tissues during human development (Ohlsson et al.
1994). After birth, the gene is repressed in almost all tissues except skeletal muscle
(Pachnis et al. 1984; Leibovitch et al. 1995; Douc-Rasy et al. 1993; Milligan et al.
2000). In other respects, a basal but significant H19 expression is detected at
adulthood in lung, heart, and thymus (Poirier et al. 1991), mammary gland
(Douc-Rasy et al. 1993; Dugimont et al. 1995; Adriaenssens et al. 1999), adrenal
gland (Liu et al. 1995), and uterus (Adriaenssens et al. 1999; Ariel et al. 1997).
2.2 Functions
Since the first mention of H19 in 1984 by Pachnis et al., its functions have only
begun to emerge. It has been reported that H19 RNA was involved in the repression
of the IGF2 oncogene by affecting its transcription (Wilkin et al. 2000) or its
translation (Li et al. 1998). In addition, we brought evidence that the H19 gene
posttranscriptionally upregulates the thioredoxin level, a key protein of the cellular
redox metabolism (Lottin et al. 2002b). Deletion of the H19 gene in mouse
(KO mice) leads to a size and weight increase of about 10% (Ripoche et al.
1997). But the consecutive biallelic IGF2 expression in these animals does not
allow concluding to an H19 direct effect. However, the group of Surani demon-
strated that induction of a targeted H19 biallelic expression via specific silencer
deletion in transgenic mice leads to smaller animals whereas IGF2 expression is not
affected (Drewell et al. 2000). So, the H19 expression pattern suggests that the
transcript assures a major function during the development.
2.3 Regulation
Beyond epigenetic regulations, the H19 gene is submitted to local regulations (i.e.,
on the level of the promoter), by hormones, growth factors, or other cytokines.
Indeed, in mammary cells, H19 is activated by HGF-SF (hepatocyte growth factor/
scatter factor), which has been identified as one of the main paracrin mediators of
morphogenetic epithelial/mesenchymal interactions. This factor has potent moto-
genic, mitogenic, and morphogenic effects on epithelial cells in culture, and H19
RNA synthesis is related to the migratory phenotype of cultured cells. EGF and
FGF-2 also activates H19 but less, whereas IGF-2, TGFb1, and TNFa have no
effect on H19 expression in these cells (Adriaenssens et al. 2002). In uterus, during
estrus (proliferative phase) and metestrus (early secretory phase) phases, and in
breast, during puberty and pregnancy, morphologic changes are associated to peak
of H19 expression. Moreover, its expression is regulated by steroid hormones
Noncoding RNAs at H19/IGF2 Locus: Role in Imprinting, Gene Expression 431
estrogen and progesterone, which respectively up- and downregulate the gene
(Adriaenssens et al. 1999). Elsewhere, we also demonstrated a negative regulation
of H19 by the tumor suppressor protein P53 (Dugimont et al. 1998).
H19 and AFP (a-fetoprotein) are abundantly transcripted in mammalian fetal
liver but are rapidly repressed after birth. This repression is partly controlled by the
Afr1 locus (a Fetoprotein Regulator 1) (Pachnis et al. 1984). Recently, Perincheri
et al. (2005) have mapped the locus and identified the murine Zhx gene (Zinc
Fingers and Homeoboxes gene), ortholog to the human ZHX2 gene. This factor is
directly responsible for the H19 postnatal repression in liver but also probably in
other organs since it is expressed in ubiquist manner.
In breast cancer, preferential accumulation of H19 transcripts has been observed
at stroma/epithelium interface (Dugimont et al. 1995; Adriaenssens et al. 1998).
Moreover, scattering and morphogenesis of epithelial cells by a conditioned
medium from fibroblasts induces activation of the H19 gene (Adriaenssens et al.
2002). The H19 gene is then finely regulated by environmental factors and it is
involved in the epithelial/mesenchymal crosstalk, essential during tumorigenesis.
Posttranscriptional regulations have been also reported such as RNA stabilization
by not yet identified proteins (Milligan et al. 2000; Jouvenot et al. 1999) Elsewhere,
a direct association of the human and mouse H19 RNA with the IMP protein
family have been demonstrated (In mouse: CRD-BP (cMyc mRNA coding Region
instability Determinant Binding Protein) and in human: IMP1, 2, and 3 (IGF-II
mRNA-binding protein)). These proteins are able to regulate H19 RNA localization
and stabilization and are also able to bind the IGF2 RNA (Tessier et al. 2004; Runge
et al. 2000; Nielsen et al. 2001, 2004; Liao et al. 2005).
3.1 Characterization
´
CTCF
IGF-2 H19 MrpL23
HUC ICR ENH
91H RNA
Limiting
regulatory
factors
?
´
Fig. 5 Model for 91H trans-effect on IGF2 expression. Two sets of enhancers (HUC and ENH
sequences that correspond, respectively, to mesodermic and endodermic enhancers) regulate IGF2
expression. Both would be required for full expression of the gene. In addition, the two IGF2
alleles would be competing for a common limited stock of regulatory elements (methylation/
acetylation/transcription?). On the maternal allele, 91H would block the locus and prevent the
HUC sequences from interaction with any regulatory factors. These latter would be then directed
on the paternal allele, in the HUC region and/or in the IGF2 promoter region, and would cooperate
with the cis endodermic enhancers resulting in IGF2 enhanced expression. (Arrows indicates
positive regulations whereas lines with bars correspond to inhibitions; ICR: Imprinting Control
Region, CTCF: CCTC-binding factor)
It should be emphasized that the role of the H19 gene in cancer is still a matter
of debate. Investigations in tumor development are delicate on account of the H19
noncoding state, the lack of knowledge of its mechanism of action, and its
imprinted status. Gene expression has been studied in numerous cancer types, but
the gene status appears to be contradictory since it can be either tumor suppressor or
oncogene depending on the studied model. Table 1 provides a nonexhaustive
general survey of bibliographic data available and illustrates their heterogeneity.
Indeed, it has been proposed that H19 functions as a tumor suppressor in some
Wilms’ tumors, embryonic rhabdomyosarcoma, and the Beckwith–Wiedmann can-
cer predisposing syndrome (Okamoto et al. 1997; Steenman et al. 1994). Consis-
tently, some studies conclude that it downregulates the IGF2 factor (Wilkin et al.
2000). By contrast, other studies including ours argue in favor of an oncogenic role
of H19 with a positive correlation with cell aggressiveness (Lottin et al. 2002a;
Rachmilewitz et al. 1995). H19 activation has also been reported in various cancer
434 N. Berteaux et al.
Table 1 Overview of bibliographic data about oncogene or tumor suppressor status of the H19
gene
Oncogene Tumor suppressor
Organ, model or Bibliographic references Organ, model or type Bibliographic references
type of cancer of cancer
Transgenic mice Leighton et al. (1995), Transgenic mice Drewell et al. (2000)
Ripoche et al. (1997)
Bladder Ariel et al. (1995, 1997, SHE cells Wiseman et al. (1991),
2000b), Elkin et al. Isfort et al. (1997)
(1995), Cooper et al.
(1996), Ayesh et al.
(2002), Ohana et al.
(2002)
Chorion (JEG-3 Rachmilewitz et al. (1995) G401 cells (WT) Hao et al. (1993)
cells)
Breast Douc-Rasy et al. (1993), JEG-3 cells (Chorion) Li et al. (1998)
Dugimont et al.
(1995), Adriaenssens
et al. (1998), Lottin
et al. (2002a, b)
Lung Kondo et al. (1995) Children liver, Li et al. (1998), Wilkin
Hepatoblastomas et al. (2000)
Uterus Tanos et al. (2004), Lottin Wilms’ tumors Steenman et al. (1994),
et al. (2005) (kidney) Casola et al. (1997)
Esophagus/colon Hibi et al. (1996), Cui Children muscle, Casola et al. (1997)
et al. (2002) rhabdomyosarcoma
Liver Manoharan et al. (2004) Beckwith–Wiedemann Reik et al. (1995)
syndrome
Ovary Tanos et al. (1999), Adrenals Liu et al. (1995),
Chen et al. (2000) Wilkin et al. (2000)
Head/neck Rainho et al. (2001)
Pharynx Ng et al. (2003)
Testicles Verkerk et al. (1997),
Ariel et al. (2000a)
Demonstration of the oncogene/tumor suppressor status in mentioned works is not always clearly
established. Numerous results converge to one or the other of these statuses and from these data,
authors have proposed their hypothesis
tissues: breast (Douc-Rasy et al. 1993; Dugimont et al. 1995; Adriaenssens et al.
1998), bladder (Ariel et al. 1995; Elkin et al. 1995), lung (Kondo et al. 1995), and
esophageal cancers (Hibi et al. 1996). Its oncogenic role has been well documented
in the bladder, since it is considered as an oncodevelopmental marker (Cooper et al.
1996) and regulates genes involved in metastasis and blood vessel development
(Ayesh et al. 2002). These observations support a H19 role in tumor invasion and
angiogenesis. In spite of polemic on the H19 role in cancer, the oncogenic role of
H19 has been clearly established in some models. In bladder and uterus cancers,
H19 is directly associated to tumor progression and is considered as tumor marker
(Ariel et al. 2000b; Lottin et al. 2005). Some authors consider H19 as an oncofetal
RNA (Ariel et al. 1997, 2000a). Ohana et al. (2002) proposed an approach of gene
therapy based on the use of H19 regulatory sequences driving the expression of the
Noncoding RNAs at H19/IGF2 Locus: Role in Imprinting, Gene Expression 435
The demonstration that the H19 transcript is stabilized in breast cancer cells and
overexpressed in human breast tumors led us to propose a link between antisense
transcription and cancer. Furthermore, we know that the deregulation of the H19/
IGF2 locus is associated to several human fetal syndromes such as BWS and
Silver–Russell syndrome (SRS).
BWS is associated with fetal and postnatal overgrowth and is associated with
embryonic tumors such as children kidney tumors named Wilm’s tumors. BWS can
be caused by a range of different defects. Several distinct genetics or epigenetics
errors involving 11p15 have been identified in different BWS patients. Some
patients have maternal chromosomal rearrangements of 11p15, meaning that
there is a disruption of the chromosome in this region. Other patients have paternal
uniparental disomy of 11p15, meaning that the maternal copy of this region is
replaced with an extra paternal copy. Many other patients have abnormal DNA
methylation in different DMR regions of 11p15, meaning that normal epigenetic
marks that regulate imprinted genes in this region are altered. In some cases, the
expression of IGF2 gene is doubled and expression of H19 gene is silenced. BWS is
often associated with H19 epigenetic inactivation due to ICR hypermethylation.
This overexpression of IGF2 is responsible for symptoms linked to the pathology.
Nevertheless, in some cases, the specific defect causing BWS in an affected patient
may remain unknown. In about one-third of BWS patients, the genetic or epigenetic
mutation is unknown.
SRS is an intrauterine growth delay associated to an altered postnatal growth with
facial dysmorphy and corporal asymmetry. Reasons of SRS are varied, chromosome
mosaic, equilibrated translocation (1; 17 or 17; 20), deletion (8q11–q13). In some
436 N. Berteaux et al.
cases, defect in 11p15-5 region has been observed. For numerous patients, this
syndrome is associated with ICR hypomethylation, IGF2 silencing, and H19 biallelic
expression. As BWS, some cases of SRS remain unexplained. The discovery of 91H
gene and its function on gene regulation in 11p15.5 locus can give research lanes to
explore theses unexplained cases of BWS or SRS.
5 Conclusion
There is growing evidence that noncoding transcripts are functional and take part
in most, if not all, complex genetic phenomena in eukaryotes, including RNA
interference-related processes such as transcriptional and posttranscriptional gene
silencing as well as parental imprinting and allelic exclusion. Noncoding RNAs
intervene in general biological processes and the close connection between non-
coding RNAs and epigenetic processes suggests that they compose a hitherto
hidden layer of genomic programming in humans. The next frontier is now the
functional characterization of these molecules to understand the interactions
between the regulatory RNAs and their targets. This is a way to better understand
how an RNA molecule can regulate cell cycle, gene expression, methylation
spreading, or even chromosomal network. The particular 91H antisense transcript
that is produced at the H19/IGF2 locus adds further complexity to the cluster of
imprinted genes in the human imprinted 11p15.5 region. 91H RNA is involved in
the control of paternal IGF2 expression suggesting that a distinct mechanism of
imprinting arises in the locus. This innovating type of regulation in “trans” between
two homologous chromosomes need to be elucidated to understand, at least in part,
the complex regulation that takes place in the imprinted locus (chromosomal
network, methylation. . .). Functional studies would allow using these RNA mole-
cules as tumor markers or therapeutic targets. Particularly, 91H could be a good
candidate to improve the diagnostic and the therapeutic tools for severe children’s
syndromes or adult cancers associated to the 11p15 chromosomal region.
References
Ariel I, Ayesh S, Perlman EJ et al (1997) The product of the imprinted H19 gene is an oncofetal
RNA. Mol Pathol 50:34–44
Ariel I, de Groot N, Hochberg A (2000a) Imprinted H19 gene expression in embryogenesis and
human cancer: the oncofetal connection. Am J Med Genet 91:46–50
Ariel I, Sughayer M, Fellig Y et al (2000b) The imprinted H19 gene is a marker of early recurrence
in human bladder carcinoma. Mol Pathol 53:320–323
Ayesh S, Matouk I, Schneider T et al (2002) Possible physiological role of H19 RNA. Mol
Carcinog 35:63–74
Bartolomei MS, Webber AL, Brunkow ME et al (1993) Epigenetic mechanisms underlying the
imprinting of the mouse H19 gene. Genes Dev 7:1663–1673
Bell AC, Felsenfeld G (2000) Methylation of a CTCF-dependent boundary controls imprinted
expression of the IGF2 gene. Nature 405:482–485
Bell AC, West AG, Felsenfeld G (1999) The protein CTCF is required for the enhancer blocking
activity of vertebrate insulators. Cell 98:387–396
Berteaux N, Lottin S, Monté D et al (2005) H19 mRNA-like non-coding RNA promotes breast
cancer cell proliferation through its positive control by E2F1. J Biol Chem 280:29625–29636
Berteaux N, Aptel N, Cathala G et al (2008) A novel H19 antisense RNA overexpressed in breast
cancer contributes to paternal IGF2 expression. Mol Cell Biol 28:6731–6745
Bestor TH (2000) The DNA methyltransferases of mammals. Hum Mol Genet 9:2395–2402
Brannan CI, Dees EC, Ingram RS et al (1990) The product of the H19 gene may function as an
RNA. Mol Cell Biol 10:28–36
Cai X, Cullen BR (2007) The imprinted H19 noncoding RNA is a primary microRNA precursor.
RNA 13:313–316
Casola S, Pedone PV, Cavazzana AO et al (1997) Expression and parental imprinting of the H19
gene in human rhabdomyosarcoma. Oncogene 14:1503–1510
Charlier C, Segers K, Wagenaar D et al (2001) Human-ovine comparative sequencing of a 250-kb
imprinted domain encompassing the callipyge (clpg) locus and identification of six imprinted
transcripts: DLK1, DAT, GTL2, PEG11, antiPEG11, and MEG8. Genome Res 11:850–862
Chen CL, Ip SM, Cheng D et al (2000) Loss of imprinting of the IGF-II and H19 genes in epithelial
ovarian cancer. Clin Cancer Res 6:474–479
Constancia M, Dean W, Lopes S et al (2000) Deletion of a silencer element in IGF2 results in loss
of imprinting independent of H19. Nat Genet 26:203–206
Cooper MJ, Fischer M, Komitowski D et al (1996) Developmentally imprinted genes as markers
for bladder tumor progression. J Urol 155:2120–2127
Cooper PR, Smilinich NJ, Day CD et al (1998) Divergently transcribed overlapping genes
expressed in liver and kidney and located in the 11p15.5 imprinted domain. Genomics
49:38–51
Crider-Miller SJ, Reid LH, Higgins MJ et al (1997) Novel transcribed sequences within the BWS/
WT2 region in 11p15.5: tissue-specific expression correlates with cancer type. Genomics
46:355–363
Cui H, Onyango P, Brandenburg S et al (2002) Loss of imprinting in colorectal cancer linked to
hypomethylation of H19 and IGF2. Cancer Res 62:6442–6446
Davis E, Caiment F, Tordoir X et al (2005) RNAi-mediated allelic trans-interaction at the
imprinted Rtl1/Peg11 locus. Curr Biol 15:743–749
Dekker J, Rippe K, Dekker M et al (2002) Capturing chromosome conformation. Science
295:1306–1311
Douc-Rasy S, Coll J, Barrois M et al (1993) Expression of the human fetal BAC/H19 gene in
invasive cancer. Int J Oncol 2:753–758
Drewell RA, Brenton JD, Ainscough JF et al (2000) Deletion of a silencer element disrupts H19
imprinting independently of a DNA methylation epigenetic switch. Development 127:3419–3428
Drewell RA, Arney KL, Arima T et al (2002) Novel conserved elements upstream of the H19 gene
are transcribed and act as mesodermal enhancers. Development 129:1205–1213
438 N. Berteaux et al.
Kaffer CR, Srivastava M, Park KY et al (2000) A transcriptional insulator at the imprinted H19/
IGF2 locus. Genes Dev 14:1908–1919
Kanduri C, Holmgren C, Pilartz M et al (2000a) The 5’ flank of mouse H19 in an unusual
chromatin conformation unidirectionally blocks enhancer-promoter communication. Curr
Biol 10:449–457
Kanduri C, Pant V, Loukinov D et al (2000b) Functional association of CTCF with the insulator
upstream of the H19 gene is parent of origin-specific and methylation-sensitive. Curr Biol
10:853–856
Kato Y, Sasaki H (2005) Imprinting and looping: epigenetic marks control interactions between
regulatory elements. Bioessays 27:1–4
Kondo M, Suzuki H, Ueda R et al (1995) Frequent loss of imprinting of the H19 gene is often
associated with its overexpression in human lung cancers. Oncogene 10:1193–1198
Kurukuti S, Tiwari VK, Tavoosidana G et al (2006) CTCF binding at the H19 imprinting control
region mediates maternally inherited higher-order chromatin conformation to restrict enhancer
access to IGF2. Proc Natl Acad Sci USA 103:10684–10689
Landers M, Bancescu DL, Le Meur E et al (2004) Regulation of the large (approximately 1000 kb)
imprinted murine Ube3a antisense transcript by alternative exons upstream of Snurf/Snrpn.
Nucleic Acids Res 32:3480–3492
Lee MP, DeBaun MR, Mitsuya K et al (1999) Loss of imprinting of a paternally expressed
transcript, with antisense orientation to KVLQT1, occurs frequently in Beckwith–Wiedemann
syndrome and is independent of insulin-like growth factor II imprinting. Proc Natl Acad Sci
USA 96:5203–5208
Leibovitch MP, Solhonne B, Guillier M et al (1995) Direct relationship between the expression of
tumor suppressor H19 mRNA and c-mos proto-oncogene during myogenesis. Oncogene
10:251–260
Leighton PA, Ingram RS, Eggenschwiler J et al (1995) Disruption of imprinting caused by deletion
of the H19 gene region in mice. Nature 375:34–39
Lewis A, Redrup L (2005) Genetic imprinting: conflict at the Callipyge locus. Curr Biol 15:
R291–R294
Lewis A, Murrel A (2004) Genomic imprinting: CTCF protects the boundaries. Curr Biol
14:284–286
Li T, Hu JF, Qiu X et al (2008) CTCF regulates allelic expression of IGF2 by orchestrating a
promoter-polycomb repressive complex 2 intrachromosomal loop. Mol Cell Biol 28:6473–6482
Li YM, Franklin G, Cui HM et al (1998) The H19 transcript is associated with polysomes and may
regulate IGF2 expression in trans. J Biol Chem 273:28247–28252
Liao B, Hu Y, Herrick DJ et al (2005) The RNA-binding protein IMP-3 is a translational activator
of insulin-like growth factor II leader-3 mRNA during proliferation of human K562 leukemia
cells. J Biol Chem 280:18517–18524
Ling JQ, Li T, Hu JF et al (2006) CTCF mediates interchromosomal colocalization between Igf2/
H19 and Wsb1/Nf1. Science 312:269–272
Liu J, Kahri AI, Heikkila P et al (1995) H19 and insulin-like growth factor-II gene expression in
adrenal tumors and cultured adrenal cells. J Clin Endocrinol Metab 80:492–496
Lopes S, Lewis A, Hajkova P et al (2003) Epigenetic modifications in an imprinting cluster are
controlled by a hierarchy of DMRs suggesting longrange chromatin interactions. Hum Mol
Genet 12:295–305
Lottin S, Adriaenssens E, Dupressoir T et al (2002a) Overexpression of an ectopic H19 gene
enhances the tumorigenic properties of breast cancer cells. Carcinogenesis 23:1885–1895
Lottin S, Vercoutter-Edouart AS, Adriaenssens E et al (2002b) Thioredoxin post-transcriptional
regulation by H19 provides a new function to mRNA-like non-coding RNA. Oncogene
21:1625–1631
Lottin S, Adriaenssens E, Berteaux N et al (2005) The human H19 gene is frequently over-
expressed in myometrium and stroma during pathological endometrial proliferative events. Eur
J Cancer 41:168–177
440 N. Berteaux et al.
Lustig O, Ariel I, Ilan J et al (1994) Expression of the imprinted gene H19 in the human fetus. Mol
Reprod Dev 38:239–246
Lutz M, Burke LJ, Barreto G et al (2000) Transcriptional repression by the insulator protein CTCF
involves histone deacetylases. Nucleic Acids Res 28:1707–1713
Maher ER and Reik W (2000) Beckwith-Wiedemann syndrome: imprinting in clusters revisited.
J Clin Invest 105:247–252
Manoharan H, Babcock K, Pitot HC (2004) Changes in the DNA methylation profile of the rat H19
gene upstream region during development and transgenic hepatocarcinogenesis and its role in
the imprinted transcriptional regulation of the H19 gene. Mol Carcinog 41:1–16
Milligan L, Antoine E, Bisbal C et al (2000) H19 gene expression is up-regulated exclusively
by stabilization of the RNA during muscle cell differentiation. Oncogene 19:5810–5816
Mitsuya K, Meguro M, Lee MP et al (1999) LIT1, an imprinted antisense RNA in the human
KvLQT1 locus identified by screening for differentially expressed transcripts using mono-
chromosomal hybrids. Hum Mol Genet 8:1209–1217
Miyamoto T, Hasuike S, Jinno Y et al (2002) The human ASCL2 gene escaping genomic
imprinting and its expression pattern. J Assist Reprod Genet 19:240–244
Moore T, Constancia M, Zubair M et al (1997) Multiple imprinted sense and antisense transcripts,
differential methylation and tandem repeats in a putative imprinting control region upstream of
mouse IGF2. Proc Natl Acad Sci USA 94:12509–12514
Murrell A, Heeson S, Reik W (2004) Interaction between differentially methylated regions partitions
the imprinted genes IGF2 and H19 into parent-specific chromatin loops. Nat Genet 36:889–893
Murrell A, Heeson S, Bowden L et al (2001) An intragenic methylated region in the imprinted
IGF2 gene augments transcription. EMBO Rep 2:1101–1106
Nielsen J, Kristensen MA, Willemoës M et al (2004) Sequential dimerization of human zipcode-
binding protein IMP1 on RNA: a cooperative mechanism providing RNP stability. Nucleic
Acids Res 32:4368–4376
Nielsen FC, Nielsen J, Christiansen J (2001) A family of IGF-II mRNA binding proteins (IMP)
involved in RNA trafficking. Scand J Clin Lab Invest Suppl 234:93–99
Ng A, Tang JP, Goh CH et al (2003) Regulation of the H19 imprinting gene expression in human
nasopharyngeal carcinoma by methylation. Int J Cancer 104:179–187
Ohana P, Bibi O, Matouk I et al (2002) Use of H19 regulatory sequences for targeted gene therapy
in cancer. Int J Cancer 98:645–650
Ohana P, Schachter P, Ayesh B et al (2005) Regulatory sequences of H19 and IGF2 genes in
DNA-based therapy of colorectal rat liver metastases. J Gene Med 7:366–374
Ohlsson R, Hedborg F, Holmgren L et al (1994) Overlapping patterns of IGF2 and H19 expression
during human development: biallelic IGF2 expression correlates with a lack of H19 expres-
sion. Development 120:361–368
Okamoto K, Morison IM, Taniguchi T et al (1997) Epigenetic changes at the insulin-like growth
factor II/H19 locus in developing kidney is an early event in Wilms tumorigenesis. Proc Natl
Acad Sci USA 94:5367–5371
Okutsu T, Kuroiwa Y, Kagitani F et al (2000) Expression and imprinting status of human PEG8/
IGF2AS, a paternally expressed antisense transcript from the IGF2 locus, in Wilms’ tumors.
J Biochem (Tokyo) 127:475–483
Pachnis V, Belayew A, Tilghman SM (1984) Locus unlinked to alpha-fetoprotein under the
control of the murine raf and Rif genes. Proc Natl Acad Sci USA 81:5523–5527
Pachnis V, Brannan CI, Tilghman SM (1988) The structure and expression of a novel gene
activated in early mouse embryogenesis. EMBO J 7:673–681
Pandey RR, Ceribelli M, Singh PB et al (2004) NF-Y regulates the antisense promoter, bidirec-
tional silencing, and differential epigenetic marks of the Kcnq1 imprinting control region.
J Biol Chem 279:52685–52693
Paulsen M, Davies KR, Bowden LM et al (1998) Syntenic organization of the mouse distal
chromosome 7 imprinting cluster and the Beckwith-Wiedemann syndrome region in chromo-
some 11p15.5. Hum Mol Genet 7:1149–1159
Noncoding RNAs at H19/IGF2 Locus: Role in Imprinting, Gene Expression 441
Szabo PE, Tang SH, Silva FJ et al (2004) Role of CTCF binding sites in the IGF2/H19 imprinting
control region. Mol Cell Biol 24:4791–4800
Takai D, Gonzales FA, Tsai YC et al (2001) Large scale mapping of methylcytosines in CTCF-
binding sites in the human H19 promoter and aberrant hypomethylation in human bladder
cancer. Hum Mol Genet 10:2619–2626
Tanos V, Prus D, Ayesh S et al (1999) Expression of the imprinted H19 oncofetal RNA in
epithelial ovarian cancer. Eur J Obstet Gynecol Reprod Biol 85:7–11
Tanos V, Ariel I, Prus D et al (2004) H19 and IGF2 gene expression in human normal, hyper-
plastic, and malignant endometrium. Int J Gynecol Cancer 14:521–525
Tessier CR, Doyle GA, Clark BA et al (2004) Mammary tumor induction in transgenic mice
expressing an RNA-binding protein. Cancer Res 64:209–214
Thakur N, Tiwari VK, Thomassin H et al (2004) An antisense RNA regulates the bidirectional
silencing property of the Kcnq1 imprinting control region. Mol Cell Biol 24:7855–7862
Thorvaldsen JL, Duran KL, Bartolomei MS (1998) Deletion of the H19 differentially methylated
domain results in loss of imprinted expression of H19 and IGF2. Genes Dev 12:3693–3702
Thorvaldsen JL, Mann MR, Nwoko O et al (2002) Analysis of sequence upstream of the endoge-
nous H19 gene reveals elements both essential and dispensable for imprinting. Mol Cell Biol
22:2450–2462
Tremblay KD, Duran KL, Bartolomei MS (1997) A 50 2-kilobase-pair region of the imprinted
mouse H19 gene exhibits exclusive paternal methylation throughout development. Mol Cell
Biol 17:4322–4329
Tsang P, Gilles F, Yuan L et al (1995) A novel L23-related gene 40 kb downstream of the
imprinted H19 gene is biallelically expressed in mid-fetal and adult human tissues. Hum
Mol Genet 4:1499–1507
Ulaner GA, Yang Y, Hu JF et al (2003) CTCF binding at the insulin-like growth factor-II (IGF2)/
H19 imprinting control region is insufficient to regulate IGF2/H19 expression in human
tissues. Endocrinology 144:4420–4426
Verkerk AJ, Ariel I, Dekker MC et al (1997) Unique expression patterns of H19 in human
testicular cancers of different etiology. Oncogene 14:95–107
Verona RI, Mann MR, Bartolomei MS (2003) Genomic imprinting: intricacies of epigenetic
regulation in clusters. Annu Rev Cell Dev Biol 19:237–259
Vu TH, Chuyen NV, Li T et al (2003) Loss of imprinting of IGF2 sense and antisense transcripts in
Wilms’ tumor. Cancer Res 63:1900–1905
Webber AL, Ingram RS, Levorse JM et al (1998) Location of enhancers is essential for the
imprinting of H19 and IGF2 genes. Nature 391:711–715
Weber M, Hagège H, Murrell A et al (2003) Genomic imprinting controls matrix attachment
regions in the Igf2 gene. Mol Cell Biol 23:8953–8959
Wilkin F, Paquette J, Ledru E et al (2000) H19 sense and antisense transgenes modify insulin-like
growth factor-II mRNA levels. Eur J Biochem 267:4020–4027
Wilkins JF, Haig D (2003) What good is genomic imprinting: the function of parent-specific gene
expression. Nat Rev Genet 4:359–368
Wiseman RW, Montgomery JC, Hosoi J et al (1991) Identification of genes associated with tumor
suppression in Syrian hamster embryo cells. Environ Health Perspect 93:105–109
Westerman BA, Poutsma A, Looijenga LH et al (2001) The Human Achaete Scute Homolog
2 gene contains two promotors, generating overlapping transcripts and encoding two proteins
with different nuclear localization. Placenta 22:511–518
Wolffe AP (2000) Transcriptional control: imprinting insulation. Curr Biol 10:R463–465
Yoon YS, Jeong S, Rong Q et al (2007) Analysis of the H19ICR Insulator. Mol Cell Biol
27:3499–3510
Yang Y, Hu JF, Ulaner GA et al (2003) Epigenetic regulation of IGF2/H19 imprinting at CTCF
insulator binding sites. J Cell Biochem 90:1038–1055
Yekta S, Shih IH, Bartel DP (2004) MicroRNA-directed cleavage of HOXB8 mRNA. Science
304:594–596
Noncoding RNAs at H19/IGF2 Locus: Role in Imprinting, Gene Expression 443
445
446 Index
Cancer treatment Dicer, 3–17, 20, 21, 62–65, 68, 73, 74,
BCL2 78, 79, 84, 87, 328, 331–339,
bispecific ASOs, 289 345, 346
G3139, 289, 290, 299 Dicer-substrate siRNA, 161–187
SPC2996, 290 DiGeorge syndrome (DGS), 406
clusterin (CLU) Double-stranded RNA (dsRNAs), 61–64, 68,
OGX-011, 292, 293 73–76, 78, 82, 83, 85–87
survivin, BIRC5 Drosha, 328–334, 339, 343, 346
EZN3042 (SPC3042), 292 dsRNAs. See Double-stranded RNA
LY2181308 (ISIS 23722), 291, 292 DUF283, 63
transforming growth factor, b2 (TGFB2)
AP 12009, 293, 294 E
XIAP E3, 77
AEG35156 (GEM640), 291 EGFR. See Epidermal growth factor receptor
Cap analysis of gene expression (CAGE), E3L, 68, 87
373, 374 Endocytosis, 35, 40, 44, 47
Caveosomal endocytosis pathway, 35 Endogenous siRNA (endo-siRNA), 11–12,
Cell-penetrating peptides (CPPs), 31, 64, 78, 82, 85
39–42, 45 Endoplasmic reticulum (ER), 35–37, 134
Cellular uptake, 31–35, 40, 44–47, 50 endo-siRNA. See Endogenous siRNA
Chalcone synthase (CHS), 62 Endosomal escape, 45
Chronic pain, 162, 163, 165, 167, 184, 186 Enhanced RNAi-1 (eri-1), 82
CHS. See Chalcone synthase Epidermal growth factor receptor (EGFR),
Citrus tristeza virus (CTV), 68, 69, 71, 80, 111, 114, 122
84, 85 Epigenetic regulation, 374, 378, 379
Clinical trials, 49, 50 Ash1, 379
CLIP chromatin remodeling, 378
HITS–CLIP, 384 G9a, 380
immunoprecipitation, 384 genomic imprinting, 379
CMV. See Cucumber mosaic virus histone methytransferase, 379
CNV. See Cucumber necrosis virus polycomb group (PcG), 378–380
Coat protein (CP), 65, 69, 71, 77, 80, 84, 85 PRC2, 379
Combinatorial RNAi, 196–198 trithorax group (TrxG), 379
Conditional RNAi, 142 X-chromosome inactivation (XCI), 379
Conformational change, 313 ER. See Endoplasmic reticulum
CPPs. See Cell-penetrating peptides eri-1. See Enhanced RNAi-1
CRV. See Cymbidium ringspot virus Exosomes, 363
CTV. See Citrus tristeza virus Exportin-5, 330, 332–333, 341, 344, 345
CTV CP, 69, 80, 84 Extracellular RNA (exRNA), 34
Cucumber mosaic virus (CMV), 66, 67, 69,
71, 72, 78, 80, 85 F
Cucumber necrosis virus (CNV), 75 Fatty acid oleoylethanolamide, 221
20 ,30 -Cyclic phosphate, 307, 309, 317 Flock house virus (FHV), 66, 67, 72, 73, 86
Cyclophosphamide, 220 FNY2b, 78
Cymbidium ringspot virus (CRV), 75, 86 Functional genomics, 227–249
G
D General acid, 314
Defense, 356–358, 360–364 General base, 314, 318
Delivery, 29–50 Gene silencing, 225–273
DGCR8, 329–331, 334, 346 Gene therapy, 194–196, 198–201
Diabetic peripheral neuropathy, Genomic imprinting, 395
213–214, 217 GFP. See Green fluorescent protein
Index 447
P R
P0, 77, 78, 85 RDRC. See RNA-dependent RNA
P19, 74–76, 83, 86 polymerase complex
P21, 76 RdRP. See RNA-dependent RNA polymerase
P25, 79, 80, 86 RDVI. See RNAi-directed viral immunity
p53, 277–283 Reactive oxygen species (ROS), 214, 217, 219
Paraspeckles, 381–383 Red chili peppers, 221
Parkinson’s disease, 403 Regulator of gene silencing-calmodulin-like
Pathogen-associated molecular pattern protein (rgs-CaM), 82
(PAMP), 81 rgs-CaM. See Regulator of gene silencing-
Pathologies, 433–436 calmodulin-like protein
PAZ, 63, 64, 77, 78 Rheumatoid arthritis, 218, 219
PAZ domain, 5, 7, 9, 17 RISC. See RNA-induced silencing complex
P-body, 18 RITS. See RNA-induced transcriptional
PcG. See Polycomb group silencing complex
PCR, 215 RNA
PDS. See Phytoene desaturase cleavage, 257, 258, 261, 263–265, 269
PFV. See Primate foamy virus regulation, 277–283
Phosphodiester isomerization, 307, 308, silencing, 1–23
317, 318 structure, 256
Phosphorothioate, 313 RNA-binding protein, 377, 380, 382,
Phosphorothioate-stimulated uptake, 34–35 384, 385
Phytoene desaturase (PDS), 65 68-kDa subunit of cleavage factor
Pigmentation, 229–236, 240–242, 246–249 I m, 382
Ping-Pong amplification loop, 9 p54/nrb, 382
piRNA. See Piwi-interacting RNA PSF, 382
PIWI, 63–65 PSP1, 382
PIWI domain, 5 PSP2/CoAA, 382
Piwi-interacting RNA (piRNA), 4, 6, TLS/FUS, 377, 380
8–11, 17, 64 RNA-dependent RNA polymerase (RdRP), 3,
Piwi subfamily, 4, 8 4, 6, 10, 11, 15–18, 63, 64, 71, 75
Polarity, 358, 360, 361 RNA-dependent RNA polymerase complex
Polycations, 46, 47 (RDRC), 16
Polycomb group (PcG), 378–380 RNA-DNA triple helix, 377
Polycomb-repressive complex 2 (PRC2) RNAi-based therapy for HD, 133, 142, 149
PostGolgi transport, 134 RNAi-directed viral immunity (RDVI), 65–68,
Posttranscriptional gene silencing (PTGS), 72, 76
62, 75 RNA-induced silencing complex (RISC), 3, 5,
Potato virus X (PVX), 66, 67, 75, 79, 80, 12–21, 63, 64, 66, 73, 75, 76, 78,
84, 86 328, 331, 333, 336–340, 342, 344
Potato virus Y (PVY), 65, 67, 86 RNA-induced transcriptional silencing
Preclinical testing, 135 complex (RITS), 16
Primate foamy virus (PFV), 68, 87 RNA interference (RNAi), 2–6, 10, 11, 13, 17,
Protein transduction domains (PTDs), 39, 42, 43 18, 21, 22, 31, 32, 59–87, 107–124,
Pseudogene, Makorin1, 381 163–167, 172, 186, 191–202,
Pseudomonas syringae, 81 212–223, 327, 333, 334, 336, 337,
PTDs. See Protein transduction domains 340–347, 375, 382, 385
PTGS. See Posttranscriptional gene silencing RNAi therapy, 346–347
PVX. See Potato virus X shRNA, 94–96, 99, 100, 103
PVY. See Potato virus Y siRNA, 94–96, 99, 100, 103
RNA–protein (RNP) complex, 384
Q RNA protein interactions, 384
Q2b, 78, 80 RNase3, 76, 85
450 Index