Signalling by Ras and Rho GTPases

8/6/2019 Signalling by Ras and Rho GTPases

http://slidepdf.com/reader/full/signalling-by-ras-and-rho-gtpases 1/16

Signalling by Ras and Rho GTPases

Sean W. Wallace1 and Aron B. Jaffe2

1 University College London, London, UK 2 Memorial Sloan-Kettering Cancer Center, New York, NY, USA

OVERVIEW

RAS and RHO proteins constitute two branches of the RASsuperfamily of small guanosine triphosphatases (GTPases)(Figure 1). Like all GTPases, these proteins cycle betweenan active, guanosine triphosphate (GTP)-bound form, andan inactive, guanosine diphosphate (GDP)-bound form. Twofamilies of proteins, guanine nucleotide exchange factors(GEFs) and GTPase activating proteins (GAPs), activateand inactivate the GTPases, respectively, and the activityof GEFs and GAPs are controlled by a large number of cellular cues. Active GTPases interact with a diverse groupof proteins, termed effectors, which transduce the signal

from the GTPase, resulting in a range of cellular responses.RAS GTPases and components of the signalling pathwayscontrolled by them are frequently mutated in human cancers.RHO GTPases are key regulators of many normal cellularprocesses which go awry during tumour progression.

RAS GTPases

Three closely related members of this family, Harvey-RAS(HRAS), Neuroblastoma-RAS (NRAS), and Kirsten-RAS(KRAS); there are two splice forms, KRAS4A and the more

abundant KRAS4B, that are mutated in a large number of human cancers, resulting in a constitutively active formof the protein. A great deal of the work directed towardsunderstanding how the active forms of RAS contribute tocancer has focused on the downstream signalling pathwaysregulated by these proteins. A large number of effectormolecules, which preferentially bind the active form of theGTPase, have been identified (Mitin et al., 2005); the down-stream pathways that have most convincingly been shownto be involved in cancer are the RAF/MAPK (mitogen-activated protein kinase), phosphoinositide 3-kinase (PI3K),and RAL guanine nucleotide dissociation stimulator (RAL-GDS) pathways (Figure 2). Although all active RAS proteins

can activate these pathways, there are examples of humangenetic disorders that are specifically associated with muta-tions in individual RAS family members. Recent advancesin high-throughput genomics and sequencing have indeeduncovered distinct associations of each RAS isoform withspecific tumour types. Understanding how these biologicaldifferences relate to signalling will be key to our ability totreat cancers with mutations in the different members of thisfamily of GTPases.

DOWNSTREAM SIGNALLING PATHWAYS

RAF

RAF proteins constitute a branch of a large family of mitogen-activated protein kinase kinase kinases (MAP-KKKs). There are three RAF proteins in mammals, A-RAF, B-RAF, and C-RAF (also called RAF-1), whichare part of an evolutionarily conserved signalling cas-cade that transduces signals from the cell surface to thenucleus. Active RAS recruits RAF to the membrane, astep generally believed to be required for RAF activa-tion, although there is at least one study suggesting thatC-RAF can be activated in a RAS-independent manner(Mischak et al., 1996). Active RAF then phosphorylates

and activates the MAPKKs, MEK1 and MEK2, which inturn activate the MAPKs, Extracellular signal-related kinaseERK1 and ERK2. ERK1/2 have a number of cytoplas-mic and nuclear targets that regulate gene transcription, thecytoskeleton, and other intracellular pathways. Not surpris-ingly, therefore, activation of this pathway has been shownto mediate diverse cellular processes, such as cell prolifer-ation, survival, and differentiation. The regulation of RAFactivity has turned out to be more complicated than sim-ply binding to active RAS, and in addition to membranerecruitment, it involves phosphorylation, homo- and het-erodimerization, and interaction with other proteins (Well-brock et al., 2004).

The Cancer Handbook 2nd Edition. Edited by Malcolm R. Alison 2007 John Wiley & Sons, Ltd.



2 THE MOLECULAR AND CELLULAR BASIS OF CANCER

Figure 1 Human RAS superfamily members. Subfamilies of proteins are indicated by coloured arcs: RAS (red), RHO (green), Gα (orange), ARF (yellow),

and RAB (blue). (Modified with permission from Colicelli, J., Sci STKE 2004, re13, Copyright 2004, AAAS.)

RAF proteins have long been thought to be key

players in RAS-mediated tumourigenesis. Many tumoursand tumour cell lines have increased levels of activeERK (Hoshino et al., 1999). More direct evidence has

come from the finding that B-RAF is frequently mutated

in a number of tumour types (Table 1) (Davies et al.,

2002). The vast majority of B-RAF mutations (∼90%)result in the substitution of glutamic acid at position

600 (http://www.sanger.ac.uk/genetics/CGP/cosmic/), result-ing in constitutive kinase activity. Since activating RAS

mutations and activating B-RAF mutations are mutually

exclusive in tumours (Davies et al., 2002), activation of the

RAF/MEK/ERK pathway may be a key event in the for-mation of some tumour types. In contrast to B-RAF , only

one tumour-associated mutation in C-RAF has been reported

(Davies et al., 2005), and none have been reported for A-

RAF . This may reflect a difference in their ability to signal

to MEK and ERK. Cell culture studies indicate that the

B-RAF kinase domain is much more efficient at activat-ing MEK than A-RAF or C-RAF (Pritchard et al., 1995).

Moreover, fibroblasts from A-Raf and C-Raf knockout mice

have relatively normal levels of active Erk, while fibrob-

lasts from B-Raf knockout mice have decreased Erk activity

(Huser et al., 2001; Mercer et al., 2002; Pritchard et al.,2004). Although they have the ability to activate MEK in

various cell culture assays, the primary function of A-RAF

and C-RAF in vivo may be to phosphorylate other substrates.

Alternatively, these two RAF family members may have

kinase-independent functions, such as acting as scaffold pro-

teins – an intriguing possibility supported by the finding thatkinase-dead C-Raf can rescue the morphological and migra-

tion defects seen in C-Raf null fibroblasts (Ehrenreiter et al.,

2005). However, whether kinase-dead C-RAF can fully sub-

stitute for wild-type C-RAF in vivo, is not known.



SIGNALLING BY RAS AND RHO GTPases 3

RAS

RAF PI3K RALGDS

MEK PDK1 RAL

ERK AKT

Figure 2 RAS signalling pathways. Shown are the major signal transduc-tion pathways downstream of RAS that have been implicated in cancer. See

text for details.

The ability of RAF isoforms to heterodimerize has impor-tant consequences on their biological activity, both in nor-mal cells and in tumours with B-RAF mutations. Wild-typeB-RAF can bind to C-RAF in a RAS-dependent manner,thereby activating C-RAF and promoting MEK activation(Garnett et al., 2005). B-RAF kinase activity is requiredfor its ability to activate C-RAF, although it is not knownwhether C-RAF is phosphorylated directly or via anotherkinase (Garnett et al., 2005). Mutant versions of B-RAF

with impaired kinase activity have been identified in humantumours. It appears they activate C-RAF through dimeriza-tion and promote MEK activation. In these kinase-defectivemutants, B-RAF is predicted to be in an “active” or “open”

conformation, but is unable to directly phosphorylate MEKbecause the mutations are in conserved residues necessaryfor kinase activity (Wan et al., 2004). However, this “open”conformation allows B-RAF to bind C-RAF in a RAS-independent manner (Garnett et al., 2005), although how itthen activates C-RAF to signal to MEK is unclear.

The fact that B-RAF kinase activity is highly specific forMEK1 and MEK2, which themselves are only known tophosphorylate ERK1 and ERK2, make targeting this pathwayan extremely attractive approach to fighting tumours with

B-RAF mutations. Indeed, several companies have designedchemical inhibitors to target components of this signaltransduction pathway. One of these inhibitors, CI-1040, aselective MEK inhibitor (Sebolt-Leopold et al., 1999), blocksproliferation of cell lines or tumour xenografts with anactivating B-RAF mutation (Solit et al., 2006). While CI-1040 also blocked ERK activation by MEK in wild-typecells, it had no effect on their proliferation. Interestingly, CI-1040 only partially inhibited growth of cell lines or tumourxenografts with activating RAS mutations (Solit et al., 2006).This indicates that, while MEK activity is required fortumours with activating B-RAF mutations, active RAS issignalling to other downstream pathways to promote tumourformation (see the subsequent text). It would be interesting toknow whether tumours with less common B-RAF mutations,

Table 1 Frequency of tumour-associated mutations in RAS pathway components.

Tissue NRAS HRAS KRAS B-RAF PIK3CA

Adrenal gland 5% (7/137) 0% (1/137) 0% (0/177) 0% (0/4) 0% (0/2)Autonomic ganglia 8% (9/106) 0% (0/65) 2% (1/65) 1% (1/119) 0% (0/27)

Biliary tract 1% (3/213) 0% (0/151) 32% (440/1373) 14% (23/158) 0% (0/5)

Bone 0% (0/135) 2% (3/140) 1% (1/103) 0% (0/45) 0% (0/16)Breast 0% (0/145) 1% (1/198) 4% (11/259) 1% (2/146) 27% (167/613)

CNS 2% (6/277) 0% (0/277) 1% (4/269) 4% (15/342) 5% (18/362)

Cervix 2% (2/132) 8% (23/260) 8% (27/302) 0% (0/57) 12% (1/8)

Endometrium 0% (1/275) 1% (3/291) 14% (197/1392) 1% (2/212) 50% (4/8)Eye 0% (0/32) 0% (0/3) 6% (4/60) 2% (6/280) 0% (0/1)

Gential tract 100% (2/2) 0% (0/2) 6% (1/15) 0% (0/24) 0% (0/2)

Haematopoietic and lymphoid tissue 11% (617/5370) 0% (7/2357) 4% (135/3200) 2% (13/726) 0% (0/227)

Kidney 0% (1/343) 0% (1/277) 1% (4/377) 0% (0/108) 0% (0/23)Large intestine 3% (11/409) 0% (2/506) 31% (3758/12050) 15% (478/3201) 25% (95/387)

Liver 10% (10/104) 0% (0/62) 8% (16/193) 3% (2/60) 18% (15/82)

Lung 1% (20/1834) 1% (7/1322) 18% (1494/7893) 2% (22/1026) 2% (8/394)

Oesophagus 0% (0/119) 1% (1/160) 3% (12/311) 2% (3/117) 0% (0/12)Ovary 3% (3/110) 0% (0/96) 16% (209/1235) 15% (101/648) 6% (12/204)

Pancreas 2% (4/211) 0% (0/187) 58% (2460/4182) 4% (5/130) 0% (0/24)

Placenta 0% (0/4) 0% (0/4) 0% (0/4) 0% (0/4) 0% (0/3)

Prostate 1% (8/539) 5% (29/503) 8% (41/502) 0% (0/46) 0% (0/3)Salivary gland 0% (0/30) 19% (22/111) 2% (1/45) 0% (0/1) 0% (0/1)

Skin 17% (463/2680) 6% (89/1592) 2% (32/1295) 42% (1237/2909) 0% (0/55)

Small intestine 25% (1/4) 0% (0/4) 26% (58/223) 4% (1/23) 0% (0/1)

Soft tissue 7% (10/147) 8% (32/406) 12% (65/543) 4% (11/268) 6% (1/18)Stomach 2% (5/220) 4% (14/389) 6% (138/2048) 1% (10/791) 6% (20/312)

Testis 3% (7/186) 0% (0/100) 5% (11/217) 0% (0/24) 0% (0/4)

Thyroid 7% (129/1902) 5% (90/1817) 3% (61/1763) 26% (800/2967) 0% (0/154)

Upper aerodigestive tract 4% (23/581) 7% (59/788) 4% (31/855) 3% (9/265) 15% (3/20)Urinary tract 3% (3/99) 12% (116/944) 4% (9/209) 0% (0/194) 11% (2/18)

All data was obtained from the Catalogue of Somatic Mutations in Cancer database at the Wellcome Trust Sanger Institute. (http://www.sanger.ac.uk/genetics/CGP/cosmic/).




such as those that promote heterodimerization with C-RAF,are also sensitive to MEK inhibition.

PI3K

There are multiple forms of the PI3K family of lipidkinases in mammalian cells; the class I PI3Ks phosphory-late phosphatidylinositol-4,5-bisphosphate (PIP2) to producephosphatidylinositol-3,4,5-triphosphate (PIP3) (Wymann andMarone, 2005). Class I PI3Ks exist as heterodimers com-posed of a regulatory subunit and a catalytic subunit. Acti-vated growth factor receptors recruit PI3K to the cell surfaceby binding the regulatory subunit either directly or indirectlythrough adapter proteins. RAS can also recruit PI3K to themembrane by binding to the catalytic subunit, though thisinteraction alters the structure of PI3K, and could there-fore affect activity directly through an allosteric mechanism(Pacold et al., 2000).

The generation of PIP3 at the cell surface promotesthe recruitment of pleckstrin homology (PH) domain-containing proteins, such as phosphoinositide-dependentkinase 1 (PDK1) and AKT. PDK1 phosphorylates and acti-vates AKT, and AKT can then phosphorylate a numberof proteins to regulate diverse cellular processes, such asgrowth, proliferation, and survival (Brazil et al., 2004). PI3Kalso mediates RAS-induced cytoskeletal changes and pro-motes cancer cell migration and invasion via the RHO familyGTPase RAC (Rodriguez-Viciana et al., 1997; Shaw et al.,1997). The mechanism by which PI3K activates RAC isunclear; the regulatory subunit of PI3K can bind to a het-erotrimeric protein complex containing the GEF SOS-1, andtogether with the production of PIP3 by the catalytic sub-unit, stimulate SOS-1 GEF activity towards RAC (Innocentiet al., 2003). Other RAC GEFs, such as VAV and Tiam1, canbe recruited to the membrane and/or activated by binding toPIP3 via their PH domain (Han et al., 1998; Stam et al.,1997). Because RAC plays a critical role in cell migration(discussed in the subsequent text), PI3K is well positioned tomediate cross talk between Ras activation and RHO GTPasesduring RAS-mediated cancer progression.

Chemical inhibition of PI3K inhibits RAS-mediated trans-formation of immortalized mouse fibroblasts (Rodriguez-Viciana et al., 1997), as well as RAS-induced proliferationand morphological changes in primary rat pancreatic ductalepithelial cells (Agbunag and Bar-Sagi, 2004). In additionto being required for the initial events in RAS-inducedtumour formation, constitutive PI3K activation is sufficientfor tumour maintenance in mice (Lim et al., 2005).

Recently, the gene encoding one of the catalytic subunitsof PI3K, PI3KCA, was found to be frequently mutated in anumber of human tumours, Table 1 and Samuels et al., 2004.The vast majority of PIK3CA mutations (over 80%) clusterin one of two “hot spot” regions, which encode the helicaldomain and kinase domain of the protein (Samuels andVelculescu, 2004). Many tumours with PI3KCA mutationsalso contain KRAS mutations, which may indicate thatthe two gene products signal through distinct pathways inthese tumours (Samuels and Velculescu, 2004). Alternatively,

active KRAS could increase membrane recruitment of mutantPI3KCA, thereby potentiating its oncogenic potential. In

support of this, mutations in PIK3CA and mutations inthe gene encoding the lipid phosphatase, phosphatase and

tensin homologue (PTEN), a negative regulator of thePI3K pathway, frequently coexist in endometrial carcinomas(Oda et al., 2005), although they are mutually exclusive in

breast carcinoma and glioblastoma (Broderick et al., 2004;Saal et al., 2005). Perhaps, some tumours require higher

levels of PI3K signalling. No tumour-associated mutations

in the other isoforms of the PI3K catalytic domains havebeen reported, although their overexpression is sufficient

to transform chicken embryo fibroblasts (CEFs), suggestingthat merely aberrant expression levels of other PI3Ks may

contribute to human cancer (Kang et al., 2006).A number of studies have examined the biochemical and

biological activity of the most common tumour-associated

mutant forms of the PI3KCA protein and shown that thesehave increased lipid kinase activity in vitro (Kang et al.,

2005; Samuels et al., 2004). Overexpressing the tumour-

associated PIK3CA mutants causes transformation of primaryCEFs, as judged by focus formation assays (Kang et al.,

2005), and immortalized human mammary epithelial cells,as judged by the growth in soft agar (Isakoff et al., 2005).

Studies using colorectal cancer (CRC) cell lines that areheterozygous for a “hot spot” mutation in either the helicaldomain or the kinase domain found that the mutant copy

of PIK3CA is responsible for the ability of these cellsto proliferate independently of growth factors, and for

their ability to form metastases in mice (Samuels et al.,2005). Together, these experiments indicate that mutations inPIK3CA are, at least in some cases, necessary and sufficient

for some of the cellular alterations associated with cancer.One of the most intriguing aspects of these studies is that

“hot spot” mutations in either the helical domain or thekinase domain appear to have equivalent biochemical andcellular effects. While mutations in a kinase domain would

be predicted to increase its enzymatic activity, it is unclearhow mutations in the helical domain of PIK3CA would result

in the same effect.

The signalling events downstream of PIK3CA leading totumour formation are beginning to be uncovered, although

there may be some differences among tumour types. Whilestudies generally agree that PIK3CA mutants promote con-

stitutive activation of AKT, the isoform(s), as well as thepathway(s) activated downstream differ depending on themodel system used (Kang et al., 2005; Samuels et al., 2005).

In CRCs, one clue about the relevant signalling componentscomes from a study in which the sequence of 340 kinases in

over 200 human tumour samples was analysed; genes encod-

ing 3 kinases in the PI3K signalling pathway had mutations(PDK1, AKT2, and PAK4 ) (Parsons et al., 2005). Further

analysis with other PI3K signalling components revealedthat almost 40% of CRCs have an alteration in one of the

PI3K-pathway genes, and that the alterations were essentiallymutually exclusive among tumours (Parsons et al., 2005),suggesting that these gene products signal through the same

pathway. Intriguingly, no alterations were found in genes




encoding the other two AKT isoforms, AKT1 and AKT3, sug-gesting signalling specificity downstream of mutant PIK3CAin CRCs. There is at least one report for different contribu-tions of different AKT isoforms in cell migration and cell

morphology, in mammary epithelial cells cultured in three-dimensional basement matrix (Irie et al., 2005). It will beimportant to know which pathways are downstream of acti-vated PIK3CA in different tumours to provide more potentialtargets for therapy.

RalGDS

RALGDS is a member of a family of GEFs for thesmall GTPases RALA and RALB, which are in the RASbranch of the RAS superfamily (Fig. 1). There are fourRAL GEFs in humans with a C-terminal RAS associationdomain (http://smart.embl-heidelberg.de/). RALGDS bindsto active RAS, resulting in its translocation to the plasmamembrane and subsequent activation of RAL (Matsubara,et al. 1999). There are five known RAL effectors: thetranscription factor ZO-1-associated nucleic acid bindingproteins (ZONAB); the RAC/CDC42 GAP RALBP1; twosubunits of the exocyst complex, SEC5 and EXO84; and theactin-binding protein FILAMIN (Cantor et al., 1995; Frankelet al., 2005; Moskalenko et al., 2003; Ohta et al., 1999).Ral also binds phospholipase D (PLD) constitutively, andregulates a number of transcription factors by mechanisms,which are still unidentified (Feig, 2003).

Although tumour-associated mutations in RALGDS or itssignalling pathway components have not yet been identi-fied, a great deal of evidence indicates that they play a keyrole in RAS-mediated tumourigenesis. RALGDS or activeRALA can cooperate with activated RAF or limiting amountsof RAS in mouse fibroblast focus formation assays (Uranoet al., 1996; White et al., 1996). Studies using dominant-negative forms of RAL and RAS effector loop mutants thatsignal through RAF, PI3K or RALGEFs, indicate a rolefor signalling through RAL in RAS-mediated transformation(Rodriguez-Viciana et al., 1997; Urano et al., 1996; Whiteet al., 1996). More convincing evidence comes from recentstudies using mouse knockouts and ribonucleic acid inter-ference (RNAi) of components of this signalling pathway.Mice lacking RalGDS have reduced tumour incidence, size,and progression, in a skin cancer model that results in acti-vating mutations in HRas (Gonzalez-Garcia et al., 2005).Importantly, RalGDS knockout mice are viable and fertile,and have no obvious morphological defects, indicating it isdispensable for normal development (Gonzalez-Garcia et al.,2005) and making it an attractive target for therapy.

Recent experiments indicate that, while RalGEF-Ral sig-nalling is not sufficient for RAS-induced transformation of mouse cells, a RAS mutant that only signals through RAL-GEFs can transform human cells (Hamad et al., 2002), andexpression of active form of RLF, another RALGEF, cantransform human cells (Lim et al., 2005). This activity isdependent on the catalytic activity of RLF, indicating thatsignalling through RAL is sufficient to transform humancells (Lim et al., 2005). RAS-induced transformation of

human cells is inhibited by RNAi knockdown of RALA, andincreased levels of active, GTP-bound RALA are present ina number of pancreatic cancer cell lines (Lim et al., 2005),further highlighting the importance of this signalling path-

way in cancer. Interestingly, the activation state of RALAin pancreatic cell lines does not always reflect the levels of active KRAS, suggesting that RAL is being activated by anindependent mechanism (Lim et al., 2005). It is not knownwhether the cell lines with low levels of active KRAS buthigh levels of active RALA have mutations in one or more

RALGEFs, or in RALA itself.Precisely how RALGEF-RAL signalling contributes to

cancer is not clear, but several studies indicate that RALAand RALB, and perhaps the different RALGEFs, play dis-tinct roles in cancer progression. RalGDS is required fortumour cell survival in a mouse skin cancer model, possiblyvia the jun kinase (Jnk)/MAPK pathway (Gonzalez-Garcia

et al., 2005). RALB, but not RALA, is also required for thesurvival of some human cancer cell lines, as RNAi knock-down induces apoptosis (Chien and White, 2003). Interest-ingly, loss of RALA, but not RALB, prevents anchorage-independent proliferation of the same cell lines (Chien andWhite, 2003). RLF-induced transformation of human cellsis blocked by RNAi knockdown of RALA, while RALBRNAi potentiates it (Lim et al., 2005). These discrepan-cies in the RALB RNAi phenotype may reflect cell typedifferences; however, in both cases the cellular effects of reducing RALA and RALB levels are different and supportthe notion that they make distinct contributions to cancer.Active forms of RALA, but not RALB, can transform human

cells (Hamad et al., 2002; Lim et al., 2005). This activityof RALA depends on its binding to SEC5 and RALBP1,but not PLD (Lim et al., 2005). RALA, but not RALB,promotes basolateral membrane delivery in Madin-Darbycanine kidney (MDCK) epithelial cells, which may be dueto a differential affinity for subunits of the exocyst complex(Shipitsin and Feig, 2004). Together these studies suggestthat at least some of the ability of RALA to mediate tumouri-genesis is through deregulated membrane trafficking via theexocyst complex, which would lead to changes in cell mor-phology and perhaps promote the epithelial-to-mesenchymaltransition associated with cancer progression.

RAS ISOFORM SPECIFICITY: LOCATION,SIGNALLING, AND BIOLOGICAL ACTIVITIES

Mutated RAS genes were first identified in human cancercells through their ability to transform immortilized fibrob-lasts in deoxyribonucleic acid (DNA) transfection assays(Malumbres and Barbacid, 2003). Since then, a large efforthas been made to characterize RAS mutations associatedwith different types of cancer (Table 1). These studies haverevealed that, although the different RAS isoforms share thesame regulators and effector proteins, each isoform is associ-ated with different tumour types. For example, KRAS (whichis the most commonly mutated RAS isoform) is mutant in58% of pancreatic tumours, while NRAS and HRAS aremutant in 2 and 0% of pancreatic tumours, respectively.




In skin cancer, 17% of tumours harbour NRAS mutations,compared with 6% for HRAS and 2% for KRAS. Becausethe three RAS proteins are ubiquitously expressed, thesedata suggest that there are differences among RAS proteins

that account for the varying biological consequences asso-ciated with mutations in each RAS isoform. RAS proteinsare modified post-translationally, resulting in their targetingto cellular membranes (Mor and Philips, 2006). All threeRAS proteins are farnesylated at their C-termini. NRAS andHRAS are further modified with one palmitoyl or two palmi-toyl groups, respectively, while KRAS contains a polybasicstretch of amino acids that is required for its associationwith membranes (Hancock et al., 1990). The differences inpost-translational modification play a large role in localizingeach isoform to distinct subdomains of the plasma membrane(Prior and Hancock, 2001). NRAS and HRAS also localize tothe Golgi apparatus (Apolloni et al., 2000; Choy et al., 1999)

and shuttle between the plasma membrane and the Golgivia a depalmitoylation/palmytoylation cycle (Rocks et al.,2005) (Figure 3a). All three RAS isoforms can be activatedat the plasma membrane (Augsten et al., 2006; Bivona et al.,2003; Chiu et al., 2002). NRAS and HRAS activations havealso been reported to occur on intracellular membranes inresponse to extracellular stimuli (Chiu et al., 2002), althoughanother study using a different method to visualize activeRAS did not observe this (Augsten et al., 2006).

Constitutively active forms of HRAS activate subsetsof signalling pathways depending upon their subcellularlocalization (Figure 3b). In one study, tethering of activeHRAS to the plasma membrane results in activation of ERK,

AKT, and JNK; tethering HRAS to the Golgi results inactivation of ERK and AKT, and only weak activation of JNK, and tethering HRAS to the endoplasmic reticulum(ER) preferentially activates JNK but not ERK or AKT(Chiu et al., 2002). In addition to lipid modification, all threeRas isoforms are subject to other types of post-translationalmodification, affecting their subcellular localization anddownstream signalling. Farnesylated, palmitoylated NRASand HRAS can also be ubiquitinated, and at least in the caseof HRAS, this promotes the association with endosomes (Juraet al., 2006) (Figure 3a). Ubiquitination of HRAS impairsits ability to activate the RAF/MAPK pathway (Jura et al.,2006) (Figure 3b); this is consistent with previous studiesshowing that blocking recycling of HRAS back to theplasma membrane impairs activation of the Raf/MAPK (Royet al., 2002). How ubiquitination of HRAS is regulated, andwhether ubiquitinated HRAS signals to other downstreampathways, is not known. Interestingly, KRAS, which is amore potent activator of RAF than HRAS (Yan et al., 1998)is neither endocytosed (Roy et al., 2002) nor ubiquitinated(Jura et al., 2006).

KRAS can, however, relocalize from the plasma mem-brane to endomembrane compartments. In neurons, KRAS(but not HRAS) relocalizes to the Golgi in response toglutamatergic signalling (Fivaz and Meyer, 2005), while inother cell types KRAS relocalizes to the ER, Golgi, andouter mitochondrial membrane (Bivona et al., 2006; Silviuset al., 2006) upon phosphorylation on S181 in the polybasicregion (Bivona et al., 2006) (Figure 3a). A phospho-mimetic

mutant of KRAS promotes apoptosis; this cellular effectrequires BCL-2 family member BCL-XL, which associatespreferentially with active, phosphorylated KRAS (Bivonaet al., 2006) (Figure 3b). The potential to “switch” activated

mutants of KRAS from tumour promoting to apoptosis pro-moting by modulating the phosphorylation status has obviousappeal as an anticancer therapy strategy. Indeed, treatment of KRAS-induced tumours in nude mice with the protein kinaseC (PKC) agonist bryostatin (which is currently in clinicaltrials) promoted apoptosis, whereas tumours induced by amutant KRAS that cannot be phosphorylated by PKC wereresistant to bryostatin treatment (Bivona et al., 2006).

In addition to post-translational modifications RAS sig-nalling can be influenced by scaffold proteins. The bestcharacterized family of scaffold proteins involved in RASsignalling are the MAPK scaffolds (Morrison and Davis,2003). One intriguing member of this class of scaffold

protein is SEF, which binds to MEK/ERK complexes andlocalizes to the Golgi (Torii et al., 2004). SEF binds toactive MEK and prevents MEK/ERK from dissociating andERK from translocating to the nucleus, thereby restrictingERK signalling to cytoplasmic substrates (Torii et al., 2004).SEF may therefore act to capture MEK activated by Golgi-localized RAS proteins. In agreement with this, SEF1 RNAiresults in more nuclear ERK and increased expression of ERK target genes (Torii et al., 2004). It will be interestingto see whether mutations in SEF, or other MAPK scaffolds,are associated with human cancers.

RHO GTPases

The Rho family of small GTPases consists of 22 members,of which RhoA, Rac1, and Cdc42 are the best characterized.They regulate a range of normal cellular processes includingproliferation, adhesion, polarity, and migration that arerelevant to tumour formation and progression. With theexception of the RHOH gene, which is rearranged in somelymphomas, mutations have not been found in the genesof RHO family members in tumours. However, there area number of reports of overexpression of RHO GTPases, aswell as aberrant expression of some of the regulatory proteins

that control their activity, which could contribute to tumourprogression. Here we give a brief overview of the role of Rho GTPases in the regulation of the cell cycle, epithelialmorphogenesis, and cell migration, followed by a discussionof how the observed changes in expression of RHO GTPasescould contribute to cancer.

Rho GTPases and the Cell Cycle

Progression of cells through G1 phase of the cell cycleis dependent on the activation of cyclin-dependent kinases(CDKs), which phosphorylate the retinoblastoma (RB) fam-ily of proteins to relieve inhibition of the E2F family of transcription factors (Coleman et al., 2004). Microinjectionof constitutively active mutants of RhoA, Rac1, and Cdc42




H R A S

K R A S

N / H R A S

N / H R

A S

PKRAS

P K R AS

PKRAS

D e p a l m i t

o y l a t i o n P

a l m

i t o y

l a t i o

n

U b

H R A S

ER

G

M

E

N

H R A S

U b

H R A S

H R A S

K R A S

BCL-XL

HRAS

PKRAS

HRASER

G

MN

E R K

A K T

J N K

E E R K

Apoptosis

(a)

(b)

Figure 3 RAS isoform specificity. (a) Effect of post-translational modifications on the subcellular localization of the different RAS isoforms. (b) Effect

of RAS subcellular localization on downstream signalling. Intracellular organelles are labeled as: E – endosome, ER – endoplasmic reticulum, G – Golgiapparatus, M–mitochondria, N–nucleus. See text for details.




into quiescent fibroblasts promotes G1 progression, whiledominant-negative mutants or C3 transferase (which inhibitsRho) block serum-induced G1 progression (Olson et al.,1995), showing that all three play a role in regulating cell

cycle progression. CDKs are active when complexed withcyclins and inhibited by binding to cyclin-dependent kinaseinhibitors (CKIs), and Rho GTPases exert their effects onprogression through G1 phase, at least in part, by influencingthe levels of cyclins and CKIs (Figure 4).

Induction of cyclin D is a key event for cells to progressthrough G1 phase, and Rho GTPases can regulate inductionof cyclin D1 in a variety of cell types. Active Rac1 promotesinduction of cyclin D1 expression in fibroblasts, myocytes,and endothelial cells, although different mechanisms havebeen proposed. In fibroblasts, Rac has been found to actvia its effector Pak to regulate cyclin D1 transcription(Westwick et al., 1997), while another study found that it

requires an NF-κB-dependent pathway (Joyce et al., 1999).These pathways may not be mutually exclusive, becauseat least one study has found Rac to signal to NF-κB viaPak (Dadke et al., 2003). In myocytes, active Rac1 alsopromotes cyclin D1 transcription, and in this case seemsto be acting through the nicotinamide adenine dinucleotidephosphate (NADPH) oxidase complex and formation of reaction oxygen species (Page et al., 1999). In endothelialcells, RAC1 is activated downstream of α5β1 integrins whencells are plated on fibronectin, and this leads to RAC-dependent stimulation of CYCLIN D1 translation, ratherthan transcription (Mettouchi et al., 2001). Cdc42 has alsobeen shown to promote transcription from the cyclin D1

promoter in fibroblasts (Gjoerup et al., 1998; Welsh et al.,2001). However, its role in cyclin D1 induction is less clear,

Growth factors/ mitogens

Cell–celladhesion

Rho GTPases

Cell cycleprogression

Cell-matrixadhesion

Cyclin upregulation

CKI downregulation

Cdk activation

Figure 4 Rho GTPases promote cell cycle progression. Activation of

Rho GTPases is required for induction of cyclins and downregulation

of cyclin-dependent kinase inhibitors (CKIs). Rho-family GTPases areactivated downstream of growth factors and adhesion to the extracellular

matrix, while Rho (but not Rac or Cdc42) activity decreases in response to

cell–cell adhesion. See text for details.

as another study found constitutively active Cdc42 stimulatedexpression of cyclin E, and not cyclin D1 (Chou et al., 2003).

Constitutively active RhoA was found to have no effect oncyclin D1 induction in fibroblasts in some studies (Gjoerup

et al., 1998; Welsh et al., 2001), while other studies foundthat it does promote cyclin D1 expression in fibroblasts(Westwick et al., 1997) and epithelial cells (Liberto et al.,2002). Regardless of this discrepancy, Rho activity is clearlynecessary for proper cyclin D1 induction as the use of C3 transferase to inhibit endogenous Rho blocks cyclin D1expression in fibroblasts (Danen et al., 2000), epithelial cells(Liberto et al., 2002), and hepatocytes (Hansen and Albrecht,1999). In the case of fibroblasts, Rho has been shown tocontrol the timing of cyclin D1 expression by promotingsustained Erk activation necessary for mid-G1 cyclin D1expression and by inhibiting a Rac-dependent cyclin D1expression pathway (Welsh et al., 2001). Here, inhibition

of Rho did not prevent cyclin D1 expression, but insteadresulted in early expression of cyclin D1 via a Rac-dependentpathway and premature S-phase entry. Differences in thebackground levels of active Rac might explain why in somecases inhibition of Rho blocks cyclin D1 induction ratherthan hastening it (Danen et al., 2000). Regulating the relativelevels of active Rho and Rac might therefore provide amechanism to control the timing of cyclin D1 expressionand therefore the length of the G1 phase of the cell cycle.

Downregulation of the CKIs p21 and p27 is required forprogression through G1 phase, and this is also regulated byRho GTPases. Active Rho inhibits transcription from the p21promoter (Olson et al., 1998), while inhibition of endogenous

Rho with C3 transferase results in accumulation of p21 andprevents entry into S phase, thereby blocking proliferation inresponse to mitogenic stimulation or expression of oncogenicRas (Danen et al., 2000; Olson et al., 1998). Rac and Cdc42have been shown to downregulate p21; however, they do soby proteasome-dependent degradation rather than inhibitingp21 expression (Bao et al., 2002). The mechanism by whichRho represses p21 transcription is not clear. One studyshowed that overexpression of constitutively active Rho-associated kinase (ROCK) suppresses transcription from thep21 promoter (Lai et al., 2002). However, it seems unlikelythat ROCK is the Rho effector mediating p21 downregulationin transformed cells as inhibition of ROCK (in contrast toinhibition of Rho) does not lead to p21 accumulation or block proliferation in transformed fibroblasts (Sahai et al., 2001).

Rho may also play a role in downregulating p27 expres-sion, although there is conflicting evidence. Where Rho hasbeen shown to be involved it seems the mechanism is post-transcriptional. Rho is required for p27 degradation in RatFRTL-5 thyroid cells and in IIC9 Chinese hamster embryocells following stimulation of cell cycle progression (Hiraiet al., 1997; Hu et al., 1999; Weber et al., 1997). Anotherstudy has shown Rho can repress p27 mRNA translation(Vidal et al., 2002). However, Rho inhibition had no effecton p27 levels in transformed fibroblasts or human colon car-cinoma cell lines (Sahai et al., 2001).

The requirement for Rho GTPase activity for progressionthrough the cell cycle allows cells to respond to their environ-ment and regulate their proliferation accordingly (Figure 4).




For example, Rho GTPases are activated downstream of integrins, and untransformed cells require adhesion to theextracellular matrix to proliferate. Transformed cells, on theother hand, often show anchorage-independent proliferation,

which is thought to contribute to their ability to overprolifer-ate and form tumours. Similarly Rho GTPases are activateddownstream of growth factor receptors, and transformed cellsoften have altered requirements for growth factors for pro-liferation. Overexpression of active forms of Rho GTPasesin quiescent cells can, therefore, promote cell cycle progres-sion in the absence of serum (Olson et al., 1995). In epithelialcells RhoA activity is inhibited downstream of E-cadherin ascells become confluent and this might contribute to the phe-nomenon of contact inhibition of proliferation (Aijaz et al.,2005; Noren et al., 2003; Noren et al., 2001). Deregulationof Rho GTPases and their signalling pathways could there-fore contribute to the loss of normal cellular controls on

proliferation seen in cancer.

Rho GTPases and Epithelial Morphogenesis

Epithelial cells form continuous sheets that function as reg-ulated barriers between the different environments of a mul-ticellular organism. Formation of a functional epitheliumdepends on the formation of adhesive cell– cell contacts,polarization of proteins and lipids to facilitate selective anddirectional permeability, and differentiation. Epithelial cellsare characterized by the presence of intercellular junctions,which maintain the functional integrity of the epithelium.These junctions are protein complexes formed by the interac-tion of cell adhesion molecules on neighbouring cells, whichlink via their cytoplasmic domains to the cytoskeleton andsignalling complexes. In mammalian epithelial cells, tight junctions formed by the transmembrane proteins claudins,occludin, and junctional adhesion molecules (JAMs), are themost apical of the junctions. They act as a permeability bar-rier to regulate diffusion of solutes across the epithelium andmaintain cell polarity by inhibiting diffusion between theapical and basolateral membrane domains. Adherens junc-tions are the principal adhesion junctions and form basallyto tight junctions by the transmembrane proteins cadherinsand nectins (Figure 5).

Although mature adherens junctions and tight junctionsare spatially separated and perform distinct functions, theirassembly is interdependent. During the initial stages of cell– cell contact formation in epithelial cells components

EMT

Loss of cell junctions

Loss of polarity

Dedifferentiation

EpithelialMesenchymal

Figure 5 Loss of cell junctions and cell polarity contribute to epithe-

lial-to-mesenchymal transition (EMT). Green–tight junctions, Blue–adhe-

rens junctions, Red–actin cytoskeleton. See text for details.

of both adherens junctions and tight junctions colocalize atnascent contacts as puncta at the end of actin-rich membrane

protrusions (Suzuki et al., 2002; Vasioukhin et al., 2000).Subsequent remodelling of the actin cytoskeleton to form

the characteristic junctional actin belt and separation of tight junctions and adherens junctions along the lateral membraneresults in formation of a polarized cell with mature junc-

tions and distinct apical and basolateral membrane domains.These nascent cell –cell contacts are sites of active actin

polymerization, which seals apposing membranes together

(Vasioukhin et al., 2000). Studies in various epithelial celltypes have revealed a requirement for Rho GTPases in junc-

tion formation (Braga et al., 1997; Kodama et al., 1999;Noritake et al., 2004; Nusrat et al., 1995; Takaishi et al.,

1997), and it seems likely that this is primarily through theireffects on the actin cytoskeleton (Fujita and Braga, 2005).

As well as being required for junction formation, RhoGTPases are also involved in the disassembly of junctionsas cell undergo epithelial-to-mesenchymal transition (EMT)

(Figure 5). Loss of cell junctions and loss of polarity con-

tribute to the process of dedifferentiation during which cellslose their epithelial characteristics and become more migra-

tory and invasive, which correlates with the development of cancers from benign to malignant.

Addition of hepatocyte growth factor (HGF)/scatter factorto MDCK cells results in breakdown of cell–cell contacts

and cell scattering, and dominant-negative Rac blocks theseeffects (Ridley et al., 1995). Rac1 is transiently activated byHGF (Zondag et al., 2000), further supporting a role in HGF-

induced scattering. RAC also seems to mediate the effectof transient expression of active RAS on junctions in ker-atinocytes, as dominant-negative Rac blocks RAS-induced

disassembly of adherens junctions (Braga et al., 2000). Reg-ulation of Rho activity is important for transforming growth

factor (TGF)-β-induced EMT. In NMuMG mouse epithe-lial cells, TGF-β transiently activates RhoA and dominant-

negative Rho blocks TGF-β-induced adherens junction dis-assembly and migration, suggesting that Rho activity isrequired (Bhowmick et al., 2001). However, another study in

the same cells showed that TGF-β mediates RhoA degrada-

tion at tight junctions downstream of the E3 ubiquitin ligaseSmad ubiquitin regulatory factor (Smurf) and that this degra-

dation is necessary for EMT (Ozdamar et al., 2005). Theprecise role of Rho in TGF-β-mediated EMT is therefore

unclear.MDCK cells stably transformed with oncogenic Ras

undergo a morphological EMT in which cells adopt a fibrob-lastic phenotype with loss of cadherin-dependent cell– cellcontacts (Zondag et al., 2000). These morphological effects

are associated with decreased expression of the Rac GEF

Tiam1, decreased Rac1 activity, and increased RhoA activ-ity (Zondag et al., 2000), and restoring expression of Tiam1

in these cells reverts them to an epithelial morphology withintact cell–cell contacts and decreased invasiveness (Hordijk et al., 1997). In fibroblasts, activation of Rac1 by Tiam1leads to downregulation of RhoA activity and formation of

cadherin-dependent cell–cell contacts and an epithelial mor-phology (Sander et al., 1999). The relative levels of Rac and




Rho thus seem to control whether cells adopt an epithe-lial or mesenchymal phenotype. Consistent with this idea,subconfluent epithelial cells, which have a more fibroblas-tic morphology than confluent cells, have elevated levels

of active RhoA and decreased levels of active Rac1 andCdc42. Furthermore, RhoA activity is directly reduced bycadherin homophilic ligation during cell–cell contact forma-tion (Noren et al., 2001).

In addition to being required for junction disassem-bly downstream of transforming signals, activation of RhoGTPases is sufficient to disrupt cell junctions and cellpolarity. Overexpression of RHOA and RHOC in humancolon carcinoma cells disrupts adherens junctions (Sahaiand Marshall, 2002), while overexpression of active RAC1disrupts adherens junctions in keratinocytes (Braga et al.,2000). Overexpression of active RhoA, Rac1, or Cdc42 inMDCK cells perturbs tight junction structure and function

(Jou et al., 1998; Rojas et al., 2001). These effects are likelyto arise from alterations in the actin cytoskeleton. Disrup-tion of adherens junctions by RHO is dependent on ROCKand acto myosin contractility (Sahai and Marshall, 2002),while disruption of tight junctions occurs concomitantly withchanges in the actin cytoskeleton (Jou et al., 1998; Rojaset al., 2001). However, other mechanisms for junction dis-assembly involving downregulation of junctional proteinsby endocytosis or cleavage have been proposed (Lozanoet al., 2003). Alterations in both apical and basolateral pro-tein trafficking pathways, which could contribute to loss of polarity, have also been observed in cells expressing con-stitutively active Rac1 and Cdc42 (Joberty et al., 2000; Jouet al., 1998).

Rho GTPases and Cell Migration

Cell migration is essential for the development of multicel-lular organisms, and is required in the adult to respond toinfection and injury. It also contributes to the progressionof cancers through the process of metastasis, during whichtransformed cells migrate from the source of tumour forma-tion and colonize new tissues.

The activity of Rho GTPases is required for cell migra-

tion. In chemotaxing macrophages expression of dominant-negative Rac or C3 transferase to inhibit Rho blocks cellmovement (Allen et al., 1998). Similarly inhibition of Racblocks migration of fibroblasts in various migration assays(Anand-Apte et al., 1997; Banyard et al., 2000; Nobes andHall, 1999). Mice lacking Rac1 are not viable and die dur-ing embryogenesis, but primary epiblast cells have beenextracted from mutant embryos and found to have impairedmigration, confirming the essential role for Rac in this pro-cess (Roberts et al., 1999). Rho GTPases are well knownregulators of the actin and microtubule cytoskeletons, andthis plays a major part in their involvement in cell migration(Raftopoulou and Hall, 2004).

Studies using fluorescence resonance energy transfer(FRET)-based techniques have allowed the visualization of active GTPases in live cells, and have revealed the presence

of active Rac at the front of migrating fibroblasts and neu-trophils (Gardiner et al., 2002; Kraynov et al., 2000). Active

Rac promotes actin polymerization and lamellipodia forma-tion to drive the front of the cell forwards, primarily through

the Arp2/3 complex (Raftopoulou and Hall, 2004). As wellas promoting lamellipodial extension, Rac also regulates theformation and turnover of focal complexes at the leading

edge of migrating cells, which are important for anchoringthe extending lamellipodium to the underlying matrix (Rid-

ley, 2001). Interestingly some RAC activity has also been

observed at the rear of migrating cells where it seems toplay a role in tail retraction, although the details of how this

might be regulated are not known (Gardiner et al., 2002).Rho activity is important for mediating cell body contrac-

tion and rear-end retraction during migration. This effect ismediated by the Rho effector ROCK, which through inhi-bition of myosin light chain (MLC) phosphatase, promotes

MLC phosphorylation and activation, leading to the assem-bly and contraction of actin filaments. Inhibition of Rho or

ROCK in single cells does not prevent lamellipodial exten-

sion but does prevent cell migration because the cell bodycannot move forward (Alblas et al., 2001; Allen et al., 1998).

However, not all cells require Rho in this way. Movementof a fibroblast monolayer in response to wounding does not

require the Rho-ROCK pathway and in fact inhibition of ROCK leads to increased migration speed, although Rhoactivity is still required to maintain adhesion to the underly-

ing matrix (Nobes and Hall, 1999).While Rac and Rho play a role in the mechanics of

cell movement by regulating the actin cytoskeleton, Cdc42has been implicated in directional sensing and polarizedcell migration. Inhibition of Cdc42 in macrophages does

not prevent movement per se but does prevent directionalmovement, as cells fail to respond to a gradient of a

chemotactic stimulus (Allen et al., 1998). Similarly Cdc42is required in fibroblasts to polarize in response to a wound,but is not absolutely required for migration (Nobes and

Hall, 1999). One mechanism by which Cdc42 controlspolarity is by restricting Rac-dependent protrusions to the

front of migrating cells (Cau and Hall, 2005), while also

promoting RhoA degradation at the front (Wang et al., 2003).Another important aspect of polarity in migrating cells is

reorientation of the microtubule-organizing centre (MTOC)to a position in front of the nucleus with concomitant

polarization of the microtubule network and reorientation of the Golgi apparatus to allow polarized delivery of vesicles

to the leading edge, and these processes are also dependenton Cdc42 (Etienne-Manneville and Hall, 2001; Etienne-Manneville and Hall, 2003).

The involvement of Rho GTPases in regulating cell migra-

tion has led to interest in their potential role in cancer.Metastasis involves the formation of secondary tumours as

a result of migration of cells from the site of primarytumour formation and invasion of new tissues. Active Rho

GTPases promote invasion and metastasis in both in vitro

and in vivo assays and could contribute to cancer pro-

gression (Del Peso et al., 1997; Keely et al., 1997). Moststudies of cell migration have been performed using 2D




systems in which cells migrate over planar substrates, how-ever, recent studies on invasion of cells through 3D matri-ces have highlighted some important differences between2D migration and 3D invasion. Whereas cells migrating

over a 2D substrate typically exhibit an elongated polar-ized morphology with actin-rich protrusions at the front,some cancer cells invade 3D substrates with a roundedmorphology, showing bleb-like actin structures around thecell membrane (Sahai and Marshall, 2003). Cells using thisrounded mode of invasion have elevated levels of endoge-nous active RHOA and their invasion is blocked by inhibitorsof RHO or ROCK, while migration of the same cells ona 2D planar substrate occurs with a typical polarised mor-phology and is not dependent on RHO-ROCK signalling.Interestingly, overexpression of RHOC, or constitutivelyactive RHOA or ROCK, causes cells to switch from anelongated to a rounded mode of invasion (Sahai and Mar-

shall, 2003). Furthermore, overexpression of RhoC has beenshown to promote invasion of epithelial cells (van Golenet al., 2000) and promotes metastasis in vivo (Clark et al.,2000), while depletion of RhoC by RNAi or genetic knock-out inhibits invasion and metastasis (Hakem et al., 2005;Simpson et al., 2004). RAC has also been looked at in thecontext of invasion, and while RNAi-mediated depletion of either RAC1 or RAC3 inhibits invasion of glioblastoma cells,depletion of RAC3 does not have a significant effect onmigration in a wound-healing assay (Chan et al., 2005). Inaddition to regulating cell motility, Rho GTPases regulatethe expression, activation, and secretion of matrix metallo-proteinases, which cleave matrix components to facilitate

invasion of tissues (Lozano et al., 2003), and this pro-vides another potential mechanism for the promotion of invasion by active forms of Rho GTPases (see Invasion and

Metastasis).

Rho GTPases and Cancer

Cancer progression involves overproliferation, dediffer-entiation with loss of cell– cell adhesion and cell polarity,and invasion of transformed cells to new sites (Figure 6).RHO GTPases regulate the cell cycle, morphogenesis, andcell migration, and are therefore likely to be key playersin the progression of cancer. In contrast to RAS, activat-ing mutations of RHO GTPases have not been found intumours. However, RHOA and RHOC are overexpressed intumours from a number of different tissues, while RAC1 andCDC42 are overexpressed in some breast tumours (Table 2).In several cases expression levels correlate with the stage of the tumour, suggesting that overexpression could contributeto tumour progression. In principle, increased levels of aGTPase could lead to increased activity, although this hasnot yet been demonstrated. As discussed in previous sectionsactivation of RHO GTPases can promote deregulated cellcycle progression, loss of epithelial cell–cell adhesion andpolarity, and increased invasiveness, and so could contributeto cancer progression. Tissue culture studies have revealedthat overexpression of wild-type RHOC, or constitutivelyactive RHOA, in epithelial cells results in disruption of cell

Overproliferation

Dedifferentiation

Invasion

Metastasis

Deregulated cell cycleprogression

Loss of cell junctionsLoss of polarityEMT

Cell migrationECM degradation

Figure 6 Tumour progression is characterized by overproliferation,

dedifferentiation and invasion. RHO GTPases regulate proliferation,

adhesion, polarity, and migration, processes relevant for tumour pro-gression. See text for details. EMT–epithelial-to-mesenchymal transition;

ECM– extracellular matrix.

junctions and loss of epithelial morphology via a ROCKand actomyosin contractility-dependent pathway (Sahai andMarshall, 2002), and indeed ROCK itself is overexpressedin a number of metastatic cancers (Kamai et al., 2003).Another study found that overexpression of RHOC promotes

anchorage-independent growth and increased migration andinvasion in vitro (van Golen et al., 2000). RHO-ROCK sig-nalling promotes invasion of tumour cells by a mechanisminvolving a rounded morphology, which is independent of matrix metalloproteinase activity, and which is likely tofacilitate invasion of tissues in vivo (Sahai and Marshall,2003). Increased expression of RHOC enhances metastasisof melanoma cells in vivo (Clark et al., 2000), while tumoursin RhoC knockout mice show decreased metastatic potential(Hakem et al., 2005).

RHOG is overexpressed in some breast cancers (Jianget al., 2003), while genetic alterations have been observedin the RHOH gene in several lymphomas, including pointmutations in the 5’UTR, which might affect its expression(Pasqualucci et al., 2001; Preudhomme et al., 2000). How-ever, the functions of these RHO family members have not




Table 2 Overexpression of Rho GTPases in human tumours.

Rho GTPase

overexpressed Tumour type Reference

RHOA Testicular germ cell Kamai et al. (2001)Breast Fritz et al. (1999)Colon Fritz et al. (1999)

Lung Fritz et al. (1999)

Head and neck squamous

cell

Abraham et al.

(2001)Bladder Kamai et al., (2003)

Ovary Horiuchi et al.

(2003)

RHOC Pancreatic ductaladenocarcinoma

Suwa et al. (1998)

Breast Kleer et al. (2002);

van Golen (2000)

Bladder Kamai et al., (2003)Ovary Horiuchi et al.

(2003)

Hepatocellular carcinoma Wang et al. (2004)Non – small cell lung Shikada et al.

(2003)

Gastric carcinoma Shikada et al.

(2003)RHOG Breast Jiang et al., (2003)

RAC1 Breast Fritz et al. (1999)

RAC1b Colon Jordan et al., (1999)

Breast Schnelzer et al.,(2000)

RAC2 Head and neck squamous

cell

Abraham et al.

(2001)

CDC42 Breast Fritz et al. (1999)

been well characterized and further work is needed to addressthe significance of these alterations.

A splice variant of RAC1, named RAC1b, is overexpressedin breast and colon tumours (Jordan et al., 1999; Schnelzeret al., 2000). This isoform has high intrinsic nucleotideexchange activity and exists predominantly in the GTP-bound state. Expression of wild-type RAC1b transformsfibroblasts, similar to a constitutively active RAC1 mutant(Singh et al., 2004). Furthermore, Rac1b promotes EMTand migration of mouse mammary epithelial cells in vitro,apparently as a result of increased production of reactiveoxygen species and changes in gene expression (Radiskyet al., 2005).

In addition to overexpression of RHO GTPases, there areseveral examples of alterations in the expression or activityof regulators of the GTPase cycle, which would be expectedto lead to changes in RHO GTPase activity. Chromosomaltranslocations in acute myeloid leukaemia result in the MLL-

LARG fusion, which is expected to lead to misregulation of expression of the RHO GEF LARG (Kourlas et al., 2000).Interestingly, overexpression of LARG has been shown tocooperate with RAF to transform fibroblasts (Reuther et al.,2001). Chromosomal translocations found in chronic myeloidleukaemia result in formation of a BCR-ABL fusion, inwhich the RHO GEF BCR retains its catalytic Dbl homology(DH) domain, although it is not clear whether this stillhas GEF activity (Kin et al., 2001). The RAC-specific GEFTIAM1 is overexpressed in some breast tumours (Adamet al., 2001), and some renal carcinoma cell lines contain

a point mutation in the PH domain, which increases its in

vitro transforming ability (Engers et al., 2000). However,the function of Tiam1 in cells is complicated, and it hasbeen shown to promote both cell migration and cell– cell

adhesion, essentially opposite processes (Sander et al., 1998).Interestingly, mice lacking Tiam1 are resistant to tumourformation, which can be attributed to increased apoptosisof transformed cells, but once tumours have formed theyare more metastatic (Malliri et al., 2002), perhaps as aresult of loss of Tiam1-dependent cell– cell adhesion. Thesignificance of alterations in the TIAM1 gene observed inthese human tumours warrants further study. The RACand CDC42 GEF β-PIX is overexpressed in some breastcancers (Ahn et al., 2003), while VAV1 is overexpressedin some neuroblastomas (Hornstein et al., 2003); both of these alterations might lead to increased activation of RHOGTPases.

Several RHO GAPs are also disrupted in various cancers,which might lead to increased activity of RHO GTPases. Inacute myeloid leukaemia the RHO GAP GRAF is fused to

MLL as a result of a chromosomal translocation that resultsin disruption of the GRAF GAP domain (Borkhardt et al.,2000). The DLC-1 gene, encoding a protein with homologyto RHO GAP proteins, is lost in some liver tumours (Yuanet al., 1998), while P190RHOGAP is lost in some glioblas-tomas and astrocytomas (Tikoo et al., 2000). Interestinglyexpression of P190RHOGAP blocks Ras-induced transfor-mation of fibroblasts (Wang et al., 1997), and P190RHOGAPplays a role in downregulating RHOA activity in epithe-lial cells in response to cadherin-dependent cell–cell adhe-

sion, which might be important for contact-inhibition of cell growth (Noren et al., 2003). Loss of P190RHOGAP

might therefore contribute to overproliferation and tumourformation. Finally, expression of RHOGDI , another negativeregulator of RHO GTPases, is reduced in some breast cancers(Jiang et al., 2003).

Overexpression of RHO GTPases and changes in theexpression or activities of proteins that regulate the GTPasecycle could lead to increased activity of RHO GTPasesand deregulation of their downstream signalling pathways,and this could play a role in cancer progression by themechanisms discussed in the preceding text. An interestingquestion is whether the changes in RHO GTPases observed

in cancers are playing a causative role in cancer formationand progression or whether they are selected for in tumoursas a result of their regulatory effects on cellular processesrequired for cancer formation and progression. In any case,Rho GTPases and the pathways they regulate are attractivetargets for anti-cancer treatments.

ACKNOWLEDGEMENT

We thank Alan Hall for comments and suggestions onthe manuscript, and John Colicelli for Figure 1. SeanW. Wallace is supported by a Medical Research CouncilPh.D. fellowship. Aron B. Jaffe is generously supported byMemorial Sloan-Kettering Cancer Center.




REFERENCES

Abraham, M. T., et al. (2001). Motility-related proteins as markers for headand neck squamous cell cancer. Laryngoscope, 111, 1285–1289.

Adam, L., et al. (2001). Tiam1 overexpression potentiates heregulin-induced lymphoid enhancer factor-1/beta -catenin nuclear signaling inbreast cancer cells by modulating the intercellular stability. The Journal

of Biological Chemistry, 276, 28443– 28450.Agbunag, C. and Bar-Sagi, D. (2004). Oncogenic K-ras drives cell

cycle progression and phenotypic conversion of primary pancreatic ductepithelial cells. Cancer Research, 64, 5659–5663.

Ahn, S. J., et al. (2003). Overexpression of betaPix-a in human breast cancertissues. Cancer Letters, 193, 99–107.

Aijaz, S., et al. (2005). Binding of GEF-H1 to the tight junction-associatedadaptor cingulin results in inhibition of Rho signaling and G1/S phasetransition. Developmental Cell, 8, 777–786.

Alblas, J., et al. (2001). Activation of Rhoa and ROCK are essential fordetachment of migrating leukocytes. Molecular Biology of the Cell, 12,2137–2145.

Allen, W. E., et al. (1998). A role for Cdc42 in macrophage chemotaxis.

Journal of Cell Biology, 141, 1147–1157.Anand-Apte, B., et al. (1997). Platelet-derived growth factor and

fibronectin-stimulated migration are differentially regulated by the Rac

and extracellular signal-regulated kinase pathways. The Journal of

Biological Chemistry, 272, 30688– 30692.Apolloni, A., et al. (2000). H-ras but not K-ras traffics to the plasma

membrane through the exocytic pathway. Molecular and Cellular

Biology, 20, 2475–2487.Augsten, M., et al. (2006). Live-cell imaging of endogenous Ras-GTP

illustrates predominant Ras activation at the plasma membrane. EMBO

Reports, 7, 46–51.Banyard, J., et al. (2000). Motility and invasion are differentially modulated

by Rho family GTPases. Oncogene, 19, 580– 591.Bao, W., et al. (2002). Cell attachment to the extracellular matrix

induces proteasomal degradation of p21(CIP1) via Cdc42/Rac1 signaling.

Molecular and Cellular Biology, 22, 4587–4597.Bhowmick, N. A., et al. (2001). Transforming growth factor-beta1 mediatesepithelial to mesenchymal transdifferentiation through a RhoA-dependentmechanism. Molecular Biology of the Cell, 12, 27–36.

Bivona, T. G., et al. (2003). Phospholipase Cgamma activates Ras on theGolgi apparatus by means of RasGRP1. Nature, 424, 694– 698.

Bivona, T. G., et al. (2006). PKC regulates a farnesyl-electrostatic switchon K-Ras that promotes its association with Bcl-XL on mitochondria andinduces apoptosis. Molecular and Cellular , 21, 481–493.

Borkhardt, A., et al. (2000). The human GRAF gene is fused to MLL ina unique t(5;11)(q31;q23) and both alleles are disrupted in three casesof myelodysplastic syndrome/acute myeloid leukemia with a deletion 5q.Proceedings of the National Academy of Sciences of the United States of

America, 97, 9168–9173.Braga, V. M., et al. (1997). The small GTPases Rho and Rac are required for

the establishment of cadherin-dependent cell-cell contacts. The Journal

of Cell Biology, 137, 1421–1431.Braga, V. M., et al. (2000). Activation of the small GTPase Rac is

sufficient to disrupt cadherin-dependent cell-cell adhesion in normalhuman keratinocytes. Molecular Biology of the Cell, 11, 3703–3721.

Brazil, D. P., et al. (2004). Advances in protein kinase B signalling: AKTionon multiple fronts. Trends in Biochemical Sciences, 29, 233–242.

Broderick, D. K., et al. (2004). Mutations of PIK3CA in anaplasticoligodendrogliomas, high-grade astrocytomas, and medulloblastomas.

Cancer Research, 64, 5048–5050.Cantor, S. B., et al. (1995). Identification and characterization of Ral-

binding protein 1, a potential downstream target of Ral GTPases.

Molecular and Cellular Biology, 15, 4578–4584.

Cau, J. and Hall, A. (2005). Cdc42 controls the polarity of the actinand microtubule cytoskeletons through two distinct signal transductionpathways. Journal of Cell Science, 118, 2579–2587.

Chan, A. Y., et al. (2005). Roles of the Rac1 and Rac3 GTPases in human

tumor cell invasion. Oncogene, 24, 7821–7829.Chien, Y. and White, M. A. (2003). RAL GTPases are linchpin modulators

of human tumour-cell proliferation and survival. EMBO Reports, 4,800–806.

Chiu, V. K., et al. (2002). Ras signalling on the endoplasmic reticulum andthe Golgi. Nature Cell Biology, 4, 343–350.

Chou, M. M., et al. (2003). Cdc42 promotes G1 progression through p70S6 kinase-mediated induction of cyclin E expression. The Journal of

Biological Chemistry, 278, 35241– 35247.

Choy, E., et al. (1999). Endomembrane trafficking of ras: the CAAX motif targets proteins to the ER and Golgi. Cell, 98, 69–80.

Clark, E. A., et al. (2000). Genomic analysis of metastasis reveals anessential role for RhoC. Nature, 406, 532– 535.

Coleman, M. L., et al. (2004). RAS and RHO GTPases in G1-phase cell-cycle regulation. Nature Reviews. Molecular Cell Biology, 5, 355–366.

Dadke, D., et al. (2003). Activation of p21-activated kinase 1-nuclear factorkappaB signaling by Kaposi’s sarcoma-associated herpes virus G protein-coupled receptor during cellular transformation. Cancer Research, 63,8837–8847.

Danen, E. H., et al. (2000). Dual stimulation of Ras/mitogen-activatedprotein kinase and RhoA by cell adhesion to fibronectin supports growth

factor-stimulated cell cycle progression. The Journal of Cell Biology, 151,1413–1422.

Davies, H., et al. (2002). Mutations of the BRAF gene in human cancer. Nature, 417, 949– 954.

Davies, H., et al. (2005). Somatic mutations of the protein kinase genefamily in human lung cancer. Cancer Research, 65, 7591–7595.

Del Peso, L., et al. (1997). Rho proteins induce metastatic properties invivo. Oncogene, 15, 3047–3057.

Ehrenreiter, K., et al. (2005). Raf-1 regulates Rho signaling and cell

migration. The Journal of Cell Biology, 168, 955– 964.Engers, R., et al. (2000). Tiam1 mutations in human renal-cell carcinomas.

International Journal of Cancer , 88, 369– 376.Etienne-Manneville, S. and Hall, A. (2001). Integrin-mediated activation

of Cdc42 controls cell polarity in migrating astrocytes through PKCzeta.

Cell, 106, 489– 498.

Etienne-Manneville, S. and Hall, A. (2003). Cdc42 regulates GSK-3betaand adenomatous polyposis coli to control cell polarity. Nature, 421,753–756.

Feig, L. A. (2003). Ral-GTPases: approaching their 15 minutes of fame.

Trends in Cell Biology,13

, 419–425.Fivaz, M. and Meyer, T. (2005). Reversible intracellular translocation

of KRas but not HRas in hippocampal neurons regulated byCa2+ /calmodulin. The Journal of Cell Biology, 170, 429–441.

Frankel, P., et al. (2005). RalA interacts with ZONAB in a cell density-dependent manner and regulates its transcriptional activity. The EMBO

Journal, 24, 54–62.Fritz, G., et al. (1999). Rho GTPases are over-expressed in human tumors.

International Journal of Cancer , 81, 682– 687.Fujita, Y. and Braga, V. (2005). Epithelial cell shape and Rho small

GTPases. Novartis Foundation Symposium, 269, 144– 155; discussion155–158, 223–230.

Gardiner, E. M., et al. (2002). Spatial and temporal analysis of Racactivation during live neutrophil chemotaxis. Current Biology, 12,

2029–2034.Garnett, M. J., et al. (2005). Wild-type and mutant B-RAF activate C-RAF

through distinct mechanisms involving heterodimerization. Moleculesand Cells, 20, 963– 969.

Gjoerup, O., et al. (1998). Rac and Cdc42 are potent stimulators

of E2F-dependent transcription capable of promoting retinoblastomasusceptibility gene product hyperphosphorylation. The Journal of

Biological Chemistry, 273, 18812– 18818.van Golen, K. L., et al. (2000). RhoC GTPase, a novel transforming

oncogene for human mammary epithelial cells that partially recapitulatesthe inflammatory breast cancer phenotype. Cancer Research, 60,5832–5838.

Gonzalez-Garcia, A., et al. (2005). RalGDS is required for tumor formationin a model of skin carcinogenesis. Cancer Cell, 7, 219–226.

Hakem, A., et al. (2005). RhoC is dispensable for embryogenesis and

tumor initiation but essential for metastasis. Genes and Development ,19, 1974–1979.

Hamad, N. M., et al. (2002). Distinct requirements for Ras oncogenesis in

human versus mouse cells. Genes and Development , 16, 2045–2057.Han, J., et al. (1998). Role of substrates and products of PI 3-kinase in

regulating activation of Rac-related guanosine triphosphatases by Vav.

Science, 279, 558– 560.




Hancock, J. F., et al. (1990). A polybasic domain or palmitoylation isrequired in addition to the CAAX motif to localize p21ras to the plasmamembrane. Cell, 63, 133–139.

Hansen, L. K. and Albrecht, J. H. (1999). Regulation of the hepatocytecell cycle by type I collagen matrix: role of cyclin D1. Journal of Cell

Science, 112(Pt 17), 2971–2981.Hirai, A., et al. (1997). Geranylgeranylated rho small GTPase(s) are

essential for the degradation of p27Kip1 and facilitate the progressionfrom G1 to S phase in growth-stimulated rat FRTL-5 cells. The Journal

of Biological Chemistry, 272, 13–16.Hordijk, P. L., et al. (1997). Inhibition of invasion of epithelial cells by

Tiam1-Rac signaling. Science, 278, 1464–1466.Horiuchi, A., et al. (2003). Up-regulation of small GTPases, RhoA

and RhoC, is associated with tumor progression in ovarian carcinoma. Laboratory Investigation, 83, 861– 870.

Hornstein, I., et al. (2003). The haematopoietic specific signal transducerVav1 is expressed in a subset of human neuroblastomas. Journal of

Pathology, 199, 526–533.Hoshino, R., et al. (1999). Constitutive activation of the 41-/43-kDa

mitogen-activated protein kinase signaling pathway in human tumors.Oncogene, 18, 813– 822.

Hu, W., et al. (1999). RhoA stimulates p27(Kip) degradation throughits regulation of cyclin E/CDK2 activity. The Journal of Biological

Chemistry, 274, 3396–3401.Huser, M., et al. (2001). MEK kinase activity is not necessary for Raf-1

function. The EMBO Journal, 20, 1940–1951.

Innocenti, M., et al. (2003). Phosphoinositide 3-kinase activates Rac byentering in a complex with Eps8, Abi1, and Sos-1. The Journal of Cell

Biology, 160, 17–23.Irie, H. Y., et al. (2005). Distinct roles of Akt1 and Akt2 in regulating

cell migration and epithelial-mesenchymal transition. The Journal of Cell

Biology, 171, 1023–1034.

Isakoff, S. J., et al. (2005). Breast cancer-associated PIK3CA mutationsare oncogenic in mammary epithelial cells. Cancer Research, 65,10992–11000.

Jiang, W. G., et al. (2003). Prognostic value of rho GTPases and rho guanine

nucleotide dissociation inhibitors in human breast cancers. ClinicalCancer Research, 9, 6432–6440.

Joberty, G., et al. (2000). The cell-polarity protein Par6 links Par3 andatypical protein kinase C to Cdc42. Nature Cell Biology, 2, 531–539.

Jordan, P., et al. (1999). Cloning of a novel human Rac1b splice variant withincreased expression in colorectal tumors. Oncogene, 18, 6835–6839.

Jou, T. S., et al. (1998). Structural and functional regulation of tight junctions by RhoA and Rac1 small GTPases. The Journal of Cell Biology,

142, 101– 115.Joyce, D., et al. (1999). Integration of Rac-dependent regulation of cyclin

D1 transcription through a nuclear factor-kappaB-dependent pathway.The Journal of Biological Chemistry, 274, 25245– 25249.

Jura, N., et al. (2006). Differential modification of Ras proteins byubiquitination. Molecular and Cellular , 21, 679–687.

Kamai, T., et al. (2001). Overexpression of RhoA mRNA is associatedwith advanced stage in testicular germ cell cancer. BJU International,

87, 227– 231.Kamai, T., et al. (2003). Significant association of Rho/ROCK pathway with

invasion and metastasis of bladder cancer. Clinical Cancer Research, 9,

2632–2641.Kang, S., et al. (2005). Phosphatidylinositol 3-kinase mutations identified

in human cancer are oncogenic. Proceedings of the National Academy of

Sciences of the United States of America, 102, 802– 807.

Kang, S., et al. (2006). Oncogenic transformation induced by the p110beta,-gamma, and -delta isoforms of class I phosphoinositide 3-kinase.Proceedings of the National Academy of Sciences of the United States

of America, 103, 1289–1294.Keely, P. J., et al. (1997). Cdc42 and Rac1 induce integrin-mediated cell

motility and invasiveness through PI(3)K. Nature, 390, 632– 636.

Kin, Y., et al. (2001). The Dbl homology domain of BCR is not a simplespacer in P210BCR-ABL of the Philadelphia chromosome. The Journal

of Biological Chemistry, 276, 39462– 39468.

Kleer, C. G., et al. (2002). Characterization of RhoC expression in benignand malignant breast disease: a potential new marker for small breastcarcinomas with metastatic ability. American Journal of Pathology, 160,

579–584.

Kodama, A., et al. (1999). Involvement of Cdc42 small G protein in cell-

cell adhesion, migration and morphology of MDCK cells. Oncogene, 18,

3996–4006.

Kourlas, P. J., et al. (2000). Identification of a gene at 11q23 encoding a

guanine nucleotide exchange factor: evidence for its fusion with MLL in

acute myeloid leukemia. Proceedings of the National Academy of Sciences

of the United States of America, 97, 2145–2150.

Kraynov, V. S., et al. (2000). Localized Rac activation dynamics visualized

in living cells. Science, 290, 333–337.

Lai, J. M., et al. (2002). Cytosolic retention of phosphorylated extracellular

signal-regulated kinase and a Rho-associated kinase-mediated signal

impair expression of p21(Cip1/Waf1) in phorbol 12-myristate-13- acetate-

induced apoptotic cells. Molecular and Cell Biology, 22, 7581–7592.

Liberto, M., et al. (2002). Rho regulates p21(CIP1), cyclin D1, and

checkpoint control in mammary epithelial cells. Oncogene, 21,

1590–1599.

Lim, K. H., et al. (2005). Activation of RalA is critical for Ras-induced

tumorigenesis of human cells. Cancer Cells, 7, 533– 545.

Lozano, E., et al. (2003). Tumor progression: small GTPases and loss of

cell-cell adhesion. Bioessays, 25, 452– 463.

Malliri, A., et al. (2002). Mice deficient in the Rac activator Tiam1 areresistant to Ras-induced skin tumours. Nature, 417, 867– 871.

Malumbres, M. and Barbacid, M. (2003). RAS oncogenes: the first 30 years.

Nature Reviews. Cancer , 3, 459– 465.

Matsubara, K., et al. (1999). Plasma membrane recruitment of RalGDS is

critical for Ras-dependent Ral activation. Oncogene, 18, 1303–1312.

Mercer, K., et al. (2002). ERK signalling and oncogene transformation are

not impaired in cells lacking A-Raf. Oncogene, 21, 347–355.

Mettouchi, A., et al. (2001). Integrin-specific activation of Rac controls

progression through the G(1) phase of the cell cycle. Molecular and

Cellular , 8, 115–127.

Mischak, H., et al. (1996). Negative regulation of Raf-1 by phosphorylation

of serine 621. Molecular and Cell Biology, 16, 5409–5418.

Mitin, N., et al. (2005). Signaling interplay in Ras superfamily function.

Current Biology, 15, R563–R574.

Mor, A. and Philips, M. R. (2006). Compartmentalized ras/mapk signaling. Annual Review of Immunology, 24, 771– 800.

Morrison, D. K. and Davis, R. J. (2003). Regulation of MAP kinase

signaling modules by scaffold proteins in mammals. Annual Review of

Cell and Developmental Biology, 19, 91–118.

Moskalenko, S., et al. (2003). Ral GTPases regulate exocyst assembly

through dual subunit interactions. The Journal of Biological Chemistry,

278, 51743– 51748.

Nobes, C. D. and Hall, A. (1999). Rho GTPases control polarity, protrusion,

and adhesion during cell movement. The Journal of Cell Biology, 144,

1235–1244.

Noren, N. K., et al. (2001). Cadherin engagement regulates Rho family

GTPases. The Journal of Biological Chemistry, 276, 33305– 33308.

Noren, N. K., et al. (2003). Cadherin engagement inhibits RhoA via

p190RhoGAP. The Journal of Biological Chemistry, 278, 13615– 13618.

Noritake, J., et al. (2004). Positive role of IQGAP1, an effector of Rac1, in

actin-meshwork formation at sites of cell-cell contact. Molecular Biologyof the Cell, 15, 1065–1076.

Nusrat, A., et al. (1995). Rho protein regulates tight junctions and

perijunctional actin organization in polarized epithelia. Proceedings of

the National Academy of Sciences of the United States of America , 92,

10629–10633.

Oda, K., et al. (2005). High frequency of coexistent mutations of PIK3CA

and PTEN genes in endometrial carcinoma. Cancer Research, 65,

10669–10673.

Ohta, Y., et al. (1999). The small GTPase RalA targets filamin to induce

filopodia. Proceedings of the National Academy of Sciences of the United

States of America, 96, 2122–2128.

Olson, M. F., et al. (1995). An essential role for Rho, Rac, and Cdc42

GTPases in cell cycle progression through G1. Science, 269, 1270– 1272.

Olson, M. F., et al. (1998). Signals from Ras and Rho GTPases interact to

regulate expression of p21Waf1/Cip1. Nature, 394, 295– 299.Ozdamar, B., et al. (2005). Regulation of the polarity protein Par6

by TGFbeta receptors controls epithelial cell plasticity. Science, 307,

1603–1609.




Pacold, M. E., et al. (2000). Crystal structure and functional analysis of

Ras binding to its effector phosphoinositide 3-kinase gamma. Cell, 103,

931–943.

Page, K., et al. (1999). Characterization of a Rac1 signaling pathway to

cyclin D(1) expression in airway smooth muscle cells. The Journal of

Biological Chemistry, 274, 22065– 22071.Parsons, D. W., et al. (2005). Colorectal cancer: mutations in a signalling

pathway. Nature, 436, 792.

Pasqualucci, L., et al. (2001). Hypermutation of multiple proto-oncogenes

in B-cell diffuse large-cell lymphomas. Nature, 412, 341– 346.

Preudhomme, C., et al. (2000). Nonrandom 4p13 rearrangements of the

RhoH/TTF gene, encoding a GTP- binding protein, in non-Hodgkin’s

lymphoma and multiple myeloma. Oncogene, 19, 2023–2032.

Prior, I. A. and Hancock, J. F. (2001). Compartmentalization of Ras

proteins. Journal of Cell Science, 114, 1603–1608.Pritchard, C. A., et al. (1995). Conditionally oncogenic forms of the A-Raf

and B-Raf protein kinases display different biological and biochemical

properties in NIH 3T3 cells. Molecular and Cellular Biology, 15,

6430–6442.

Pritchard, C. A., et al. (2004). B-Raf acts via the ROCKII/LIMK/cofilin

pathway to maintain actin stress fibers in fibroblasts. Molecular and Cellular Biology, 24, 5937–5952.

Radisky, D. C., et al. (2005). Rac1b and reactive oxygen species

mediate MMP-3-induced EMT and genomic instability. Nature, 436,

123–127.

Raftopoulou, M. and Hall, A. (2004). Cell migration: Rho GTPases lead

the way. Developmental Biology, 265, 23–32.

Reuther, G. W., et al. (2001). Leukemia-associated Rho guanine nucleotide

exchange factor, a Dbl family protein found mutated in leukemia,

causes transformation by activation of RhoA. The Journal of Biological

Chemistry, 276, 27145– 27151.

Ridley, A. J. (2001). Rho GTPases and cell migration. Journal of Cell

Science, 114, 2713–2722.

Ridley, A. J., et al. (1995). Regulation of scatter factor/hepatocyte growth

factor responses by Ras, Rac, and Rho in MDCK cells. Molecular and

Cell Biology, 15, 1110–1122.

Roberts, A. W., et al. (1999). Deficiency of the hematopoietic cell-specificRho family GTPase Rac2 is characterized by abnormalities in neutrophil

function and host defense. Immunity, 10, 183– 196.

Rocks, O., et al. (2005). An acylation cycle regulates localization and

activity of palmitoylated Ras isoforms. Science, 307, 1746–1752.

Rodriguez-Viciana, P., et al. (1997). Role of phosphoinositide 3-OH kinase

in cell transformation and control of the actin cytoskeleton by Ras. Cell,

89, 457– 467.

Rojas, R., et al. (2001). Cdc42-dependent modulation of tight junctions and

membrane protein traffic in polarized Madin-Darby canine kidney cells.

Molecular Biology of the Cell, 12, 2257–2274.

Roy, S., et al. (2002). H-Ras signaling and K-Ras signaling are differenti-

ally dependent on endocytosis. Molecular and Cell Biology, 22,

5128–5140.

Saal, L. H., et al. (2005). PIK3CA mutations correlate with hormone

receptors, node metastasis, and ERBB2, and are mutually exclusive

with PTEN loss in human breast carcinoma. Cancer Research, 65,

2554–2559.

Sahai, E. and Marshall, C. J. (2002). ROCK and Dia have opposing

effects on adherens junctions downstream of Rho. Nature Cell Biology,

4, 408– 415.

Sahai, E. and Marshall, C. J. (2003). Differing modes of tumour cell

invasion have distinct requirements for Rho/ROCK signalling and

extracellular proteolysis. Nature Cell Biology, 5, 711–719.Sahai, E., et al. (2001). Cross-talk between Ras and Rho signalling pathways

in transformation favours proliferation and increased motility. The EMBO

Journal, 20, 755– 766.

Samuels, Y. and Velculescu, V. E. (2004). Oncogenic mutations of PIK3CA

in human cancers. Cell Cycle, 3, 1221–1224.

Samuels, Y., et al. (2004). High frequency of mutations of the PIK3CA

gene in human cancers. Science, 304, 554.

Samuels, Y., et al. (2005). Mutant PIK3CA promotes cell growth andinvasion of human cancer cells. Cancer Cell, 7, 561–573.

Sander, E. E., et al. (1998). Matrix-dependent Tiam1/Rac signaling in

epithelial cells promotes either cell-cell adhesion or cell migration and is

regulated by phosphatidylinositol 3-kinase. The Journal of Cell Biology,

143, 1385–1398.Sander, E. E., et al. (1999). Rac downregulates Rho activity: reciprocal

balance between both GTPases determines cellular morphology andmigratory behavior. The Journal of Cell Biology, 147, 1009–1022.

Schnelzer, A., et al. (2000). Rac1 in human breast cancer: overexpression,mutation analysis, and characterization of a new isoform, Rac1b.Oncogene, 19, 3013–3020.

Sebolt-Leopold, J. S., et al. (1999). Blockade of the MAP kinase pathwaysuppresses growth of colon tumors in vivo. Nature Medicine, 5, 810–816.

Shaw, L. M., et al. (1997). Activation of phosphoinositide 3-OH kinase

by the alpha6beta4 integrin promotes carcinoma invasion. Cell, 91,949–960.

Shipitsin, M. and Feig, L. A. (2004). RalA but not RalB enhances polarizeddelivery of membrane proteins to the basolateral surface of epithelialcells. Molecular and Cellular Biology, 24, 5746–5756.

Shikada, Y., et al. (2003). Higher expression of RhoC is related to

invasiveness in non-small cell lung carcinoma. Clinical Cancer Research,9, 5282–5286.

Silvius, J. R., et al. (2006). K-ras4B and prenylated proteins lacking “secondsignals” associate dynamically with cellular membranes. Molecular

Biology of the Cell, 17, 192– 202.Simpson, K. J., et al. (2004). Functional analysis of the contribution of

RhoA and RhoC GTPases to invasive breast carcinoma. Cancer Research,

64, 8694–8701.Singh, A., et al. (2004). Rac1b, a tumor associated, constitutively active

Rac1 splice variant, promotes cellular transformation. Oncogene, 23,9369–9380.

Solit, D. B., et al. (2006). BRAF mutation predicts sensitivity to MEKinhibition. Nature, 439, 358–362.

Stam, J. C., et al. (1997). Targeting of Tiam1 to the plasma membranerequires the cooperative function of the N-terminal pleckstrin homology

domain and an adjacent protein interaction domain. The Journal of

Biological Chemistry, 272, 28447– 28454.Suzuki, A., et al. (2002). aPKC kinase activity is required for the

asymmetric differentiation of the premature junctional complex during

epithelial cell polarization. Journal of Cell Science,115

, 3565–3573.Suwa, H., et al. (1998). Overexpression of the rhoC gene correlates with

the progression of ductal adenocarcinoma of the pancreas. British Journal

of Cancer , 77, 147–152.Takaishi, K., et al. (1997). Regulation of cell-cell adhesion by rac and

rho small G proteins in MDCK cells. The Journal of Cell Biology, 139,1047–1059.

Tikoo, A., et al. (2000). p190-A, a human tumor suppressor gene, maps

to the chromosomal region 19q13.3 that is reportedly deleted in somegliomas. Gene, 257, 23– 31.

Torii, S., et al. (2004). Sef is a spatial regulator for Ras/MAP kinasesignaling. Developmental Cell, 7, 33–44.

Urano, T., et al. (1996). Ral-GTPases mediate a distinct downstreamsignaling pathway from Ras that facilitates cellular transformation. The

EMBO Journal, 15, 810–816.Vasioukhin, V., et al. (2000). Directed actin polymerization is the driving

force for epithelial cell-cell adhesion. Cell, 100, 209– 219.Vidal, A., et al. (2002). Rho activity can alter the translation of p27

mRNA and is important for RasV12-induced transformation in a manner

dependent on p27 status. The Journal of Biological Chemistry, 277,16433–16440.

Wan, P. T., et al. (2004). Mechanism of activation of the RAF-ERKsignaling pathway by oncogenic mutations of B-RAF. Cell, 116,

855–867.Wang, D. Z., et al. (1997). The GTPase and Rho GAP domains of p190,

a tumor suppressor protein that binds the M(r) 120,000 Ras GAP,independently function as anti-Ras tumor suppressors. Cancer Research,

57, 2478–2484.Wang, H. R., et al. (2003). Regulation of cell polarity and protrusion

formation by targeting RhoA for degradation. Science, 302,1775–1779.

Wang, W., et al. (2004). Genomic analysis reveals RhoC as a potential

marker in hepatocellular carcinoma with poor prognosis. British Journalof Cancer , 90, 2349–2355.

Weber, J. D., et al. (1997). Ras-stimulated extracellular signal-related kinase

1 and RhoA activities coordinate platelet-derived growth factor-induced




G1 progression through the independent regulation of cyclin D1 and p27.

The Journal of Biological Chemistry, 272, 32966– 32971.Wellbrock, C., et al. (2004). The RAF proteins take centre stage. Nature

Reviews. Molecular Cell Biology, 5, 875–885.

Welsh, C. F., et al. (2001). Timing of cyclin D1 expression within G1 phase

is controlled by Rho. Nature Cell Biology, 3, 950–957.Westwick, J. K., et al. (1997). Rac regulation of transformation, gene

expression, and actin organization by multiple, PAK-independent

pathways. Molecular and Cell Biology, 17, 1324–1335.White, M. A., et al. (1996). A role for the Ral guanine nucleotide

dissociation stimulator in mediating Ras-induced transformation. The

Journal of Biological Chemistry, 271, 16439– 16442.

Wymann, M. P. and Marone, R. (2005). Phosphoinositide 3-kinase in

disease: timing, location, and scaffolding. Current Opinion in Cell

Biology, 17, 141– 149.

Yan, J., et al. (1998). Ras isoforms vary in their ability to activate Raf-1

and phosphoinositide 3-kinase. The Journal of Biological Chemistry, 273,

24052–24056.Yuan, B. Z., et al. (1998). Cloning, characterization, and chromosomal

localization of a gene frequently deleted in human liver cancer (DLC-1)

homologous to rat RhoGAP. Cancer Research, 58, 2196–2199.Zondag, G. C., et al. (2000). Oncogenic Ras downregulates Rac activity,

which leads to increased Rho activity and epithelial-mesenchymal

transition. The Journal of Cell Biology, 149, 775– 782.

Documents

Signalling by Ras and Rho GTPases