Dispatch
R449
6. Caussinus, E., Colombelli, J., and Affolter, M.
(2008). Tip-cell migration controls stalk-cell
intercalation during Drosophila tracheal tube
elongation. Curr. Biol. 18, 1727–1734.
7. Forster, D., and Luschnig, S. (2012). Src42A-
dependent polarized cell shape changes
mediate epithelial tube elongation in
Drosophila. Nat. Cell Biol., epub ahead of print.
8. Nelson, K.S., Khan, Z., Molnar, I., Mihaly, J.,
Kaschube, M., and Beitel, G.J. (2012).
Drosophila Src regulates anisotropic apical
surface growth to control epithelial tube size.
Nat. Cell Biol., epub ahead of print.
9. Thomas, S.M., and Brugge, J.S. (1997). Cellular
functions regulated by Src family kinases.
Annu. Rev. Cell Dev. Biol. 13, 513–609.
10. Takahashi, F., Endo, S., Kojima, T., and
Saigo, K. (1996). Regulation of cell-cell contacts
in developing Drosophila eyes by Dsrc41,
a new, close relative of vertebrate c-src. Genes
Dev. 10, 1645–1656.
11. Shindo, M., Wada, H., Kaido, M., Tateno, M.,
Aigaki, T., Tsuda, L., and Hayashi, S. (2008).
Dual function of Src in the maintenance of
adherens junctions during tracheal epithelial
morphogenesis. Development 135, 1355–1364.
Tumour Invasion: A New Twist on
Rac-Driven Mesenchymal Migration
Elongated mesenchymal migration of cancer cells is driven by Rac1 activation
mediated by the adaptor NEDD9 and the exchange factor DOCK3. A new study
reports a role for the transcription factor Twist1 in inducing mesenchymal
migration by relieving the suppression of NEDD9 and DOCK3 by the microRNA
let-7i.
Victoria Sanz-Moreno
The epithelial–mesenchymal transition
(EMT) is a highly conserved and
fundamental process that governs
morphogenesis in multicellular
organisms and is thought to promote
metastatic progression of carcinomas
[1]. Some key events typical of EMT are
loss of the cell–cell adhesion molecule
E-cadherin, gain of N-cadherin and
augmented expression of certain
transcription factors. EMT leads to loss
of cell–cell adhesion and increased cell
migration and invasion [1]. Twist1,
a transcriptional regulator, induces
EMT by suppression of E-cadherin [2].
A recent paper by Yang et al. [3] now
reveals that Twist1 is a key regulator of
elongated mesenchymal migration of
cancer cells through regulation
of a microRNA (miRNA) that
controls signalling by the Rho-family
GTPase Rac.
Rho-family GTPases are key
regulators of cell migration through their
actions on actin assembly and
actomyosin contractility. Cells can
migrate as collective groups or as
individual cells. Two modes of individual
12. Wu, V.M., and Beitel, G.J. (2004). A junctional
problem of apical proportions: epithelial
tube-size control by septate junctions in the
Drosophila tracheal system. Curr. Opin. Cell
Biol. 16, 493–499.
13. Chung, S., Vining, M.S., Bradley, P.L.,
Chan, C.C., Wharton, K.A., Jr., and
Andrew, D.J. (2009). Serrano (sano) functions
with the planar cell polarity genes to control
tracheal tube length. PLoS Genet. 5, e1000746.
14. Beitel, G.J., and Krasnow, M.A. (2000). Genetic
control of epithelial tube size in the Drosophila
tracheal system. Development 127, 3271–3282.
15. Takahashi, M., Takahashi, F., Ui-Tei, K.,
Kojima, T., and Saigo, K. (2005). Requirements
of genetic interactions between Src42A,
armadillo and shotgun, a gene encoding
E-cadherin, for normal development in
Drosophila. Development 132, 2547–2559.
16. Aspenstrom, P., Richnau, N., and
Johansson, A.S. (2006). The
diaphanous-related formin DAAM1
collaborates with the Rho GTPases RhoA
and Cdc42, CIP4 and Src in regulating cell
morphogenesis and actin dynamics. Exp.
Cell Res. 312, 2180–2194.
cell migration have been characterized
in a number of systems. An elongated
(‘mesenchymal-like’) mode is
characterized by cell polarization,
a requirement for extracellular
proteolysis [4], and low actomyosin
contractility [5], and is driven by the
formation of membrane protrusions that
result from a localized activation of the
Rac GTPase [6]. Rounded ‘amoeboid’
modes of cell migration are driven by
high levels of actomyosin contractility
regulated by Rho–ROCK signalling [5,7]
or the Cdc42 GTPase [8]. In rounded
moving cells, high hydrostatic forces
drive cell movement resulting in
membrane blebbing [9]. Interestingly,
Cdc42 is capable of regulating
elongated mesenchymal or rounded
contractile movement via usage of
different guanine nucleotide exchange
factors (GEFs) [8].
Elongated mesenchymal movement
can be driven by activation of Rac1
through the GEF DOCK3 complexed
with the adaptor protein NEDD9 [5].
Rac1 promotes elongation through
WAVE2-mediated actin polymerization
[5,10,11]. Furthermore, Rac1 signalling
suppresses rounded movement
17. Han, B., Bai, X.H., Lodyga, M., Xu, J.,
Yang, B.B., Keshavjee, S., Post, M., and Liu, M.
(2004). Conversion of mechanical force into
biochemical signaling. J. Biol. Chem. 279,
54793–54801.
18. Wang, Y., Botvinick, E.L., Zhao, Y.,
Berns, M.W., Usami, S., Tsien, R.Y., and
Chien, S. (2005). Visualizing the mechanical
activation of Src. Nature 434, 1040–1045.
19. Marx, M., Warren, S.L., and Madri, J.A. (2001).
pp60(c-src) modulates microvascular
endothelial phenotype and in vitro
angiogenesis. Exp. Mol. Pathol. 70, 201–213.
20. Sweeney, W.E., Jr., von Vigier, R.O., Frost, P.,
and Avner, E.D. (2008). Src inhibition
ameliorates polycystic kidney disease. J. Am.
Soc. Nephrol. 19, 1331–1341.
Biozentrum der Universität Basel, CH-4056
Basel, Switzerland.
*E-mail: Markus.Affolter@unibas.ch
DOI: 10.1016/j.cub.2012.04.033
[5,11,12] through decreasing
actomyosin contractility [5,12]. In an
effort to identify new functions of
Twist1, Yang et al. [3] screened for
possible new targets with a particular
focus on miRNAs. They reasoned that
Twist1 cooperates with the Polycomb
group protein BMI1 [2] and they
selected head and neck squamous
cell carcinoma (HNSCC) cell lines that
would have different expression
levels of the Twist1–BMI1 pairing to
compare their migratory behaviour.
The authors performed microarray
analysis to identify which miRNAs
were co-regulated by both Twist1 and
BMI1 and found that these proteins
co-repressed let-7i miRNA [3]. The
authors confirmed that loss of let-7i
induced a morphological switch into
a mesenchymal program of invasion
[3], with long-lived protrusions that are
typical of a Rac-driven phenotype.
They went on to find that let-7i
downregulated NEDD9 and DOCK3
[3], both of which are activators of
Rac and drive mesenchymal
movement in several systems
[3,5,10,12]. Furthermore, Twist1
overexpression induced Rac1
activation in HNSCCs, as a result of
increased expression of both NEDD9
and DOCK3 [3] (Figure 1).
Interestingly, other miRNAs have been
shown to regulate mesenchymal
migration: miR-200 family members
can regulate the plasticity of tumour
cell movement [13] and the miR-200c
target MARCKS is capable of
regulating mesenchymal invasion by
driving cell protrusions [13].
Current Biology Vol 22 No 11
R450

a rounded morphology. One should be

careful when addressing amoeboid
αv phenotypes as several types of
β3
amoeboid migration have been
Src FAK described to date [18]. In the case of
NEDD9 DOCK3 Cdc42 GEFx cancer cells, rounded amoeboid
Rho contractile movement is characterized
Effector x by intensive blebbing. Blebs form when
Let7i the plasma membrane detaches focally
Y722 Rac1
from the underlying actin filament
ROCK
cortex [19], allowing cytoplasmic flow
to push the membrane outwards
rapidly due to hydrostatic pressure in
Twist BMI 1 WAVE2
the cell interior. This high blebbing is
regulated by Rho–ROCK signalling [7,9]
Actomyosin and is characterized by high levels of
contractility
phosphorylated myosin light chain II
Actin assembly: (MLC) [5,12,13]. The round cells
Membrane blebbing cell protrusions described in the study from Yang et al.
[3] lack blebs and instead display short
Actomyosin
contractility protrusions. Furthermore, the levels of
Rho-GTP or phospho-MLC in these
round cells are no different, or in fact
Rounded/amoeboid Elongated/mesenchymal
‘Rac-driven’ invasion lower, than in the elongated cells
‘contractile’ invasion
Current Biology considered in their study [3]. It is likely
that the round cells described in [3] are
different from round ‘amoeboid-like’
Figure 1. Rac GTPase signalling in elongated mesenchymal cancer cell migration and
invasion.
migrating cells [5,7–9,12,13].
Nevertheless, the authors show in very
In elongated mesenchymal migrating cells, Twist1 and BMI1 mediate suppression of the
miRNA let-7i, which results in NEDD9 and DOCK3 overexpression in HNSCC [3]. NEDD9 over- thorough manner that manipulations of
expression promotes elongated mesenchymal motility in HNSCC [3], melanoma [5,12] and Twist1 and let-7i can lead to opposite
breast cancer [16]. NEDD9 and DOCK3 have been reported to cooperate in order to activate results in terms of how they regulate
Rac1 and drive mesenchymal migration in HNSCC [3] and melanoma [5,10,12]. WAVE2, acting elongated mesenchymal migration [3].
downstream of Rac, drives mesenchymal migration by promoting actin assembly to form Twist1 positively impinges on cell
cellular protrusions in melanoma [5,10], glioblastoma [11] and fibrosarcoma [11]. In melanoma
elongation and invasion, and let-7i
cells, NEDD9 acts through integrin b3 and Src to promote mesenchymal migration, as Src
inhibits ROCK2-dependent rounded amoeboid invasion [12]. Cdc42 activation controls elon- suppresses it by inhibiting DOCK3 and
gated mesenchymal migration in melanoma [8] via an unknown guanine nucleotide exchange NEDD9 expression. Another important
factor (GEF). finding in this study is the fact that
suppression of let-7i promotes
HNSCC tends to remain localized at mesenchymal marker, a proposal that tumour-initiating capabilities without
the primary site and regional lymph will now need to be explored in more affecting EMT [3]. It would be
nodes. Severe local tissue destruction malignancies that use mesenchymal interesting to assess whether NEDD9,
with a low frequency of distal organ invasion strategies. DOCK3 and Rac1 are involved in
metastasis is observed in HNSCC The question that arises is how tumour initiation or whether it is an
patients [14]. To translate their NEDD9 might function in invasion and independent biological effect derived
observations to the clinical setting, metastasis. It is a member of the from let-7i suppression that does not
Yang et al. [3] made the interesting p130Cas family and has multiple require the Rac signalling axis.
observation that decreasing amounts partners [17], so it is likely to have Yang et al. [3] have described how
of let-7i are present in a cohort of multiple roles, in addition to the mechanisms of HNSCC dissemination
HNSCC patients with tumours that regulation of Rac activation through rely greatly on mesenchymal invasion
have invaded adjacent tissues. DOCK3. Recent work has reported and its key players NEDD9, DOCK3
Yang et al. [3] find that the that NEDD9 drives elongation and and Rac1 [3], insights that could be
Twist–let-7i–NEDD9–DOCK3 axis has mesenchymal invasion in melanoma useful in the treatment of HSNCC
prognostic value for HNSCC [3]. via engagement of integrin avb3 and patients. Targeting just one invasive
NEDD9 had already been found to recruitment of Src kinase [12]. A key strategy could be an approach against
be overexpressed during human role of NEDD9-dependent Src malignancies that do not metastasize
melanoma progression [15] and was signalling is to suppress ROCK2, so readily and that only use
found to be a marker for elongated which otherwise would drive high mesenchymal migration, as in the case
mesenchymal motility in human actomyosin contractility [12] (Figure 1) of HSNCC. Interestingly Yang et al. [3]
melanoma samples [12]. NEDD9 has that would in turn suppress did not find any correlation with let-7i
also been found to regulate cellular Rac-dependent mesenchymal expression in human metastatic
protrusive activity in breast cancer movement [5,10,12]. HNSCC specimens, suggesting that
models [16]. All of these studies point In the course of their experiments HNSCC metastasis may depend on
to NEDD9 being a bona fide Yang et al. [3] describe cells with different mechanisms compared with
Dispatch
R451
HNSCC local invasion. Tumours like
melanomas, which are highly
metastatic [20], may use both
elongated mesenchymal and rounded
amoeboid-like contractile invasion
strategies in order to disseminate more
efficiently. This plasticity could allow
the tumour cell to cope with different
environments using a larger repertoire
of invasive strategies. Following this
line of argument, melanoma patients
should be treated with a combination
of drugs that inhibit both rounded
amoeboid and elongated
mesenchymal types of movement
[5,7,8,12]. Other tumour types, such as
glioblastomas and fibrosarcomas,
have been reported to show similar
plasticity [11]; therefore, blocking both
strategies would also be necessary in
order to stop their invasion and/or
metastasis [11]. The challenge for the
next few years will be to validate
tumour invasion signatures as
prognosis markers and to find good
therapeutic targets within such
signatures.
References
1. Sleeman, J.P., and Thiery, J.P. (2011).
SnapShot: The epithelial-mesenchymal
transition. Cell 145, 162.e1.
2. Yang, M.H., Hsu, D.S., Wang, H.W., Wang, H.J.,
Lan, H.Y., Yang, W.H., Huang, C.H., Kao, S.Y.,
Tzeng, C.H., Tai, S.K., et al. (2010). Bmi1 is
essential in Twist1-induced epithelial-
mesenchymal transition. Nat. Cell Biol. 12,
982–992.
Population Genomics: How Bacterial
Species Form and Why They Don’t
Exist
Two processes suggested to drive bacterial speciation — periodic selection and
recombination — are generally thought to be mutually opposed. Recent work
shows that data taken as evidence supporting the former may be explained by
the latter, raising further problems for the idea of bacterial ‘species’.
W. Ford Doolittle
The concept of species is famously
difficult, especially for bacteria. Mayr’s
‘Biological Species Concept’ — that
species are interbreeding groups
separated from other such groups by
reproductive barriers — would not
apply to bacteria at all if, as once
believed, they are always asexual,
never recombining genetically.
Obviously, though not trivially, there
3. Yang, W.H., Lan, H.Y., Huang, C.H., Tai, S.K.,
Tzeng, C.H., Kao, S.Y., Wu, K.J., Hung, M.C.,
and Yang, M.H. (2012). RAC1 activation
mediates Twist1-induced cancer cell migration.
Nat. Cell Biol. 14, 366–374.
4. Wolf, K., Mazo, I., Leung, H., Engelke, K., Von
Andrian, U.H., Deryugina, E.I., Strongin, A.Y.,
Brocker, E.B., and Friedl, P. (2003).
Compensation mechanism in tumor cell
migration: mesenchymal-amoeboid transition
after blocking of pericellular proteolysis. J. Cell
Biol. 160, 267–277.
5. Sanz-Moreno, V., Gadea, G., Ahn, J.,
Paterson, H., Marra, P., Pinner, S., Sahai, E.,
and Marshall, C.J. (2008). Rac activation and
inactivation control plasticity of tumour cell
movement. Cell 135, 510–523.
6. Knight, B., Laukaitis, C., Akhtar, N.,
Hotchin, N.A., Edlund, M., and Horwitz, A.R.
(2000). Visualizing muscle cell migration in situ.
Curr. Biol. 18, 576–585.
7. Sahai, E., and Marshall, C.J. (2003). Differing
modes of tumour cell invasion have distinct
requirements for Rho/ROCK signalling and
extracellular proteolysis. Nat. Cell Biol. 5, 711–719.
8. Gadea, G., Sanz-Moreno, V., Self, A., Godi, A.,
and Marshall, C.J. (2008). DOCK10-Mediated
Cdc42 activation is necessary for tumour cell
invasion. Curr. Biol. 18, 1456–1465.
9. Lorentzen, A., Bamber, J., Sadok, A.,
Elson-Schwab, I., and Marshall, C.J. (2011). An
ezrin-rich, rigid uropod-like structure directs
movement of amoeboid blebbing cells. J. Cell
Sci. 124, 1256–1267.
10. Ladhani, O., Sánchez-Martinez, C., Orgaz, J.L.,
Jimenez, B., and Volpert, O.V. (2011). Pigment
epithelium-derived factor blocks tumor
extravasation by suppressing amoeboid
morphology and mesenchymal proteolysis.
Neoplasia 13, 633–642.
11. Yamazaki, D., Kurisu, S., and Takenawa, T.
(2009). Involvement of Rac and Rho signaling in
cancer cell motility in 3D substrates. Oncogene
28, 1570–1583.
12. Ahn, J., Sanz-Moreno, V., and Marshall, C.J.
(2012). Metastasis gene NEDD9 acts through
integrin b3 and Src to promote mesenchymal
motility and inhibit amoeboid motility. J Cell
Sci., epub ahead of print.
would be no ‘interbreeding’. But some
authors maintain that even asexual,
non-recombining clones can mimic
‘biological species’ in important ways.
Specifically, ecologically differentiated
clonal organisms can maintain
relatively constant within-population
genomic and phenomic similarity
(cohesion) over time, while exhibiting
increasing between-population
divergence. An important driver in this
ecotype model [1] is periodic selection,
13. Elson-Schwab, I., Lorentzen, A., and
Marshall, C.J. (2010). MicroRNA-200 family
members differentially regulate morphological
plasticity and mode of melanoma cell invasion.
PLoS One 5, e13176.
14. Garavello, W., Ciardo, A., Spreafico, R., and
Gaini, R.M. (2006). Risk factors for distant
metastases in head and neck squamous cell
carcinoma. Arch. Otolaryngol. Head Neck Surg.
132, 762–766.
15. Kim, M., Gans, J.D., Nogueira, C., Wang, A.,
Paik, J.H., Feng, B., Brennan, C., Hahn, W.C.,
Cordon- Cardo, C., Wagner, S.N., et al. (2006).
Comparative oncogenomics identifies NEDD9
as a melanoma metastasis gene. Cell 30,
1269–1281.
16. Giampieri, S., Manning, C., Hooper, S.,
Jones, L., Hill, C.S., and Sahai, E. (2009).
Localized and reversible TGFbeta signalling
switches breast cancer cells from cohesive
to single cell motility. Nat. Cell Biol. 11,
1287–1296.
17. Tikhmyanova, N., Little, J.L., and Golemis, E.A.
(2010). CAS proteins in normal and pathological
cell growth control. Cell Mol. Life Sci. 67,
1025–1048.
18. Lammerman, T., and Sixt, M. (2009).
Mechanical modes of amoeboid migration.
Curr. Opin. Cell Biol. 21, 636–644.
19. Charras, G.T., Hu, C.K., Coughlin, M., and
Mitchison, T.J. (2006). Reassembly of
contractile actin cortex in cell blebs. J. Cell Biol.
175, 477–490.
20. Gupta, P.B., Kuperwasser, C., Brunet, J.P.,
Ramaswamy, S., Kuo, W.L., Gray, J.W.,
Naber, S.P., and Weinberg, R.A. (2005). The
melanocyte differentiation program
predisposes to metastasis after neoplastic
transformation. Nat. Genet. 37, 1047–1054.
Randall Division of Cell and Molecular
Biophysics, School of Biomedical and Health
Sciences, New Hunts House, Guy’s Campus,
King’s College London, London SE1 1UL, UK.
E-mail: victoria.sanz_moreno@kcl.ac.uk
DOI: 10.1016/j.cub.2012.04.024
the operation of which seemed to be
favored by earlier data from several
groups, including the MIT labs of Martin
Polz and Eric Alm. But now these
workers offer a serious challenge to the
model [2].
Periodic selection, first understood
through the chemostat experiments
of Kim Atwood [3], is what happens
when, in a finite population of
non-recombining organisms in a stable
niche, a mutant arises that is better able
to use the niche’s resources. Through
selection, all organisms in the
population will eventually be the direct
descendants of this favored mutant
ancestor. And because there is no
recombination, their genomes will bear
at all loci only direct lineal descendants
of those specific alleles the lucky
mutant happened to have in its genome
at the time.
Populations will thus be ‘purged’ of
all allelic diversity accumulated before