Role of G12 proteins in oncogenesis and metastasis

Abstract

The G12 subfamily of heterotrimeric guanine nucleotide-binding proteins consists of two α subunits, Gα12 and Gα13. These proteins mediate signalling via G protein-coupled receptors and have been implicated in various physiological and pathophysiological processes. A number of direct and indirect effectors of Gα12 and Gα13 have been identified that mediate, or have been proposed to mediate, the diverse cellular responses accompanying activation of G12 proteins. This review describes the signalling pathways and cellular events stimulated by G12 proteins, with a particular emphasis on processes that are important in regulating cell migration and invasion, and could potentially be involved in the pathophysiology of cancer metastasis. Experimental findings directly implicating G12 proteins in the spread of metastatic disease are also summarized, indicating the importance of targeted inhibition of G12 signalling as a potential therapeutic option for locally advanced and metastatic disease.

Heterotrimeric guanine nucleotide-binding proteins (G proteins) mediate extracellular signals from transmembrane G protein-coupled receptors (GPCRs) to engage intracellular effector pathways leading to a variety of cellular responses (Marinissen and Gutkind, 2001; Pierce et al., 2002; Oldham and Hamm, 2007). G proteins consist of a guanine nucleotide-binding α subunit and a βγ subunit dimer. In its inactive state, the α subunit binds a GDP molecule. Upon receptor activation by an agonist, the engagement of the liganded receptor with the G protein triggers a conformational change in the α subunit that leads to the exchange of GDP for GTP, and dissociation of the α subunit from the βγ dimer, both of which can then signal to their downstream effectors (Fields and Casey, 1997).

Heterotrimeric G proteins are classified into four subfamilies based on the sequence similarity of the α subunits: Gs, Gi, Gq and G12. The G12 subfamily consists of two α subunits, Gα12 and Gα13. Activation of the G12 proteins impacts on several signalling pathways including those linking G proteins to monomeric GTPases, mitogen-activated protein kinases (MAPKs) and non-receptor tyrosine kinases (non-RTKs), among others. Thus, G12 proteins have been implicated in several physiological and pathophysiological processes (Dhanasekaran et al., 1998; Rohrer and Kobilka, 1998; Offermanns, 2001; Dorsam and Gutkind, 2007). This review details our current understanding of how G12 proteins regulate some of these cellular events, in particular signalling pathways impacting oncogenesis and metastasis.

Biological roles of G12 proteins

Gα12 and Gα13 are expressed in virtually every tissue in the body (Milligan et al., 1992; Spicher et al., 1994). Activation of these proteins impacts such cellular processes as growth and proliferation, cytoskeleton rearrangement, cell polarity, paracellular permeability, cell–cell adhesion, migration and invasion (Kurose, 2003; Kelly et al., 2007). Several direct binding partners as well as indirect downstream effectors of Gα12/13 have been identified that have been implicated, and in some cases directly shown to be involved, in these and other biological events mediated by the G12 subfamily (Kelly et al., 2007). Many of the effects of G12/13 signalling are mediated by the monomeric GTPase Rho. Gα12 and Gα13 activate Rho principally through direct interaction of the activated Gα subunit with Rho-specific guanine nucleotide exchange factors (RhoGEFs) (see Figure 1), which include p115RhoGEF (Hart et al., 1998; Kozasa et al., 1998), PDZ-RhoGEF (Fukuhara et al., 1999) and leukemia-associated RhoGEF (LARG) (Fukuhara et al., 2000a). Other proteins that directly interact with Gα12/13, and hence could serve as effectors include cadherins (Meigs et al., 2001), radixin of the ezrin/radixin/moesin protein family (Vaiskunaite et al., 2000; Liu and Voyno-Yasenetskaya, 2005), non-RTKs (Jiang et al., 1998; Mao et al., 1998; Shi et al., 2000), protein phosphatases (Yamaguchi et al., 2002; Zhu et al., 2004; Zhu et al., 2007), A-kinase anchoring proteins (AKAPs) (Diviani et al., 2001; Niu et al., 2001), the tight junction protein, zonula occludens-1 (Meyer et al., 2002; Sabath et al., 2008), Hsp90 (Vaiskunaite et al., 2001) and regulators of G protein signalling RGS1, RGS16 and axin (Moratz et al., 2000; Johnson et al., 2003; Stemmle et al., 2006), among others (Kurose, 2003; Kelly et al., 2007).

Figure 1.

Schematic of Gα12/13 effectors and signalling pathways impacting cell growth and transformation, migration and invasion. β-cat, beta-catenin; ATF2, activating transcription factor-2; ERK, extracellular signal-regulated kinase; GPCR, G protein-coupled receptor; jAP1, c-jun AP1-like response element; JNK, c-Jun N-terminal kinase; LPA, lysophosphatidic acid; MEK, mitogen-activated protein kinase/extracellular signal-regulated kinase kinase; RhoGEF, Rho-specific guanine nucleotide exchange factor; ROCK, Rho kinase; SDF-1, CXC chemokine stromal cell-derived factor-1α; SRE, serum response element; SRF, serum response factor.

Perhaps the most extensively studied roles of G12/13 are in cell growth and proliferation, and cell migration (described later). The G12 proteins also influence several other important cellular functions. G12/13 signalling is required for agonist-induced smooth muscle contraction (Gohla et al., 2000; Hersch et al., 2004). GPCRs that mediate vasoconstriction such as angiotensin II, endothelin-1 and thromboxane A₂ couple to both the Gq/11 and G12/13 subfamilies to stimulate myosin light chain (MLC) phosphorylation via Ca²⁺-dependent activation of MLC kinase, and Rho/Rho kinase-mediated inhibition of myosin phosphatase, respectively (Gohla et al., 2000). Phosphorylation of MLC leads to its interaction with actin and the generation of contractile force. A recent study utilizing conditional knockout of Gα12/13 in smooth muscle cells in mice demonstrated that G12/13 signalling is required for the development of salt-induced hypertension but not for the maintenance of basal blood pressure (Wirth et al., 2008). G13-mediated signalling is important for platelet activation in hemostasis and thrombosis (Moers et al., 2003). Platelets lacking Gα13 show defective shape change and aggregation in vitro and these defects are accompanied by reduced activation of RhoA and subsequently decreased MLC phosphorylation. Selective deletion of Gα13 in mouse platelets in vivo results in severely increased bleeding times and lack of formation of arterial thrombi in an experimentally induced thrombosis model (Moers et al., 2003). In addition, G12/13 mediate receptor-induced cardiac hypertrophic responses by activating a Gα12/13-Rho-c-Jun N-terminal kinase (JNK) pathway (Maruyama et al., 2002). JNK is a MAPK that is an indirect effector of G12/13 signalling, usually activated downstream of Rho GTPases (Goldsmith and Dhanasekaran, 2007). The G12 proteins also mediate protease-activated receptor-1 (PAR-1)-induced vascular endothelial barrier permeability via Rho and Rho kinase (ROCK) activation (McLaughlin et al., 2005).

While the G12 proteins clearly play a role in cell growth and proliferation (discussed below), they have also been implicated in apoptotic pathways. Constitutively activated Gα13 triggers apoptosis via a pathway involving Rho activation (Althoefer et al., 1997). Gα12 appears to regulate apoptosis in epithelial cells by activating JNK and protein phosphatase 2A (PP2A) leading to loss of expression of the anti-apoptotic protein, Bcl-2 (Yanamadala et al., 2007). Gα12 and Gα13 have also been shown to stimulate apoptosis via two MAPK pathways, one by activating apoptosis signal regulating kinase-1 (ASK-1) and the other by activating MAPK kinase kinase 1 (MEKK1), both leading to JNK activation (Berestetskaya et al., 1998). Interestingly, Gα13 forms a complex with ASK-1 and regulates apoptosis by reducing the rate of ASK-1 degradation (Kutuzov et al., 2007). Formation of the Gα13-ASK-1 complex is enhanced by coexpression of JNK-interacting leucine zipper protein (JLP) suggesting that JLP may be acting as a scaffolding protein to form a macromolecular complex (Kutuzov et al., 2007). JLP was identified earlier as physically interacting with Gα13 and leading to increased Gα13-mediated JNK activation (Kashef et al., 2005). In this regard, it is also important to note that Gα13 signals via p115RhoGEF and activates JNK to regulate primitive endoderm formation in murine embryonic carcinoma cells (Jho et al., 1997; Lee et al., 2004). An interaction between Gα13 and JLP is required for this process (Kashef et al., 2006).

G12 proteins in physiological cell migration

The role of G12 proteins in regulating physiological cell migration was first identified in studies in Drosophila. Genetic ablation of concertina, the single ortholog of Gα12 and Gα13 in Drosophila, impairs cell shape changes that underlie mesoderm internalization during gastrulation in flies (Parks and Wieschaus, 1991). An essential role of Gα12/13 has also been demonstrated in cell shape changes and migration events that occur during gastrulation in zebrafish (Lin et al., 2005). In mice, deletion of Gα13 has been shown to disrupt organization of the vascular system, resulting in lethality at approximately day 10.5 of embryogenesis (Offermanns et al., 1997; Ruppel et al., 2005). Embryonic fibroblasts cultured from these mice display a reduced chemokinetic response to thrombin, the agonist for PARs. Thus, this defect in cell migration may be responsible for the failed angiogenesis (Offermanns et al., 1997). Quite recently, a role for Gα12/13 was demonstrated in the development of the central nervous system (Moers et al., 2008). Gα12/13 appear to mediate stop signals which are required for the proper positioning of migrating cortical plate neurons and Purkinje cells during development. Conditional knockout of Gα12 and Gα13 in the nervous system of mice results in neuronal ectopia of the cerebral and cerebellar cortices due to overmigration of these cells (Moers et al., 2008). Another study showed that the orphan GPCR, GPR56, is highly expressed in neural progenitor cells (NPCs) and negatively regulates NPC migration via a Gα12/13-Rho pathway (Iguchi et al., 2008). Interestingly, loss of the mouse GPR56 gene causes neuronal ectopia in the cerebral cortex (Li et al., 2008). Thus, GPR56 may couple to G12/13 to regulate the migration of cells during cortical development.

Changes in cytoskeletal dynamics required for cell migration are coordinated in large part by Rho GTPases (Ras, Rac, Cdc42 and Rho) (Raftopoulou and Hall, 2004), and signalling by G12 proteins is responsible for many of the effects on cell movement that accompany Rho activation. Polymerization of actin and assembly of focal adhesions are important for cell shape changes and cell contraction during migration, and Gα12 and Gα13 stimulate the formation of stress fibres and focal adhesions in a Rho-dependent manner (Buhl et al., 1995; Gohla et al., 1998; 1999;). Such Rho-dependent signalling pathways appear to be important during gastrulation in Drosophila, as noted above a process dependent on concertina function (Barrett et al., 1997; Nikolaidou and Barrett, 2004), and during the corresponding process in zebrafish (Lin et al., 2005; Solnica-Krezel, 2005). Rho activation by Gα12/13 is also required, as mentioned above, for regulating NPC migration (Iguchi et al., 2008).

Besides their role in regulating cell migration during embryonic development, the G12 subfamily is also involved in controlling the migration of lymphocytes, neutrophils and vascular smooth muscle cells (VSMCs) (Xu et al., 2003; Rieken et al., 2006; Tan et al., 2006; Takashima et al., 2008) (see Table 1). The CXC chemokine stromal cell-derived factor-1α (SDF-1) induces cell migration in T lymphocytes by binding to the GPCR, CXCR4, to activate a Gα13-Rho signalling axis (Tan et al., 2006). On the other hand, Gα12/13 likely regulate the number of splenic marginal zone B (MZB) cells, a specialized population of B lymphocytes, by exerting an inhibitory effect on sphingosine-1-phosphate (S1P)-induced migration of these cells. Mice that lack the Gα12/13 subunits show significantly reduced numbers of MZB cells suggesting that, besides affecting peripheral MZB cell maturation, loss of Gα12/13 causes disinhibition of S1P-induced promigratory signalling (Rieken et al., 2006). G12 proteins dictate morphologic polarity in neutrophils that is necessary for their migration in uniform concentrations of attractants. This function of G12 and G13 is mediated by activation of Rho with subsequent stimulation of the Rho-dependent protein kinase, p160-ROCK and myosin II to generate myosin-based contraction of the trailing edge of the migrating cell (Xu et al., 2003). Gα12/13 subunits also control cell polarity and directed migration of fibroblasts in a Rho-dependent manner (Goulimari et al., 2005; 2008;). In VSMCs, S1P activates G12/13 signalling via the S1P(2) receptor, leading to activation of Rho and inhibition of both platelet-derived growth factor (PDGF)-induced Rac activation and migration of VSMCs (Takashima et al., 2008). These disparate effects of Rho activation reflect the intricacy of Rho signalling mechanisms impacting on cell migration (Sahai and Marshall, 2002).

Table 1.

GPCRs that activate G12/13 signalling to regulate the migration and invasion of mammalian cells

GPCR	Cell type	References
PAR-1	Embryonic fibroblasts	Offermanns et al., 1997
	Breast cancer cells	Kelly et al., 2006a
	Prostate cancer cells	Kelly et al., 2006b
LPA receptor	Fibroblasts	Goulimari et al., 2005
		Goulimari et al., 2008
	Ovarian cancer cells	Bian et al., 2006
	Embryonic cortical neurons	Moers et al., 2008
CXCR4	T lymphocytes	Tan et al., 2006
S1P receptor	MZB cells	Rieken et al., 2006
	VSMCs	Takashima et al., 2008
	Glioblastoma cells	Malchinkhuu et al., 2008
	Embryonic cortical neurons	Moers et al., 2008
Thromboxane A2 receptor	Breast cancer cells	Kelly et al., 2006a
	Prostate cancer cells	Kelly et al., 2006b
GPR56	Neural progenitor cells	Iguchi et al., 2008

GPCR, G protein-coupled receptor; LPA, lysophosphatidic acid; MZB, splenic marginal zone B; PAR-1, protease-activated receptor-1; S1P, sphingosine-1-phosphate; VSMC, vascular smooth muscle cell.

Cell migration is a complex process (Friedl and Wolf, 2003; Raftopoulou and Hall, 2004; Van Haastert and Devreotes, 2004), and while in some cell types G12 proteins regulate migration via stimulation of Rho, in other cell types they promote migration via processes independent of Rho signalling (Meigs et al., 2002; Radhika et al., 2004) (Figure 1). In fibroblasts, Gα13-stimulated migration requires interaction of the Gα subunit with Hax-1, a cytoskeleton-associated, cortactin-interacting protein. Coexpression of Hax-1 with constitutively active Gα13 attenuates the ability of Gα13 to stimulate Rho activity and leads to a significant reduction in stress fibres, while at the same time potentiating Rac activity (Radhika et al., 2004). In another example, the ability of Gα12 to negatively regulate the adhesive function of cadherin enhances cell migration, and this function of Gα12 is independent of Rho activation (Meigs et al., 2002). The primary mechanism for the increased migration in this instance appears to be due to the ability of the Gα12-E-cadherin interaction to trigger the release of β-catenin from the cytoplasmic tail of E-cadherin and thereby reverse E-cadherin-mediated suppression of migration (Meigs et al., 2002).

In addition to mediating signals from GPCRs to impact on cell migration, in some cellular contexts G12 proteins may mediate signals from RTKs to regulate migration. An intriguing recent study demonstrated that Gα13 is required for RTK-induced migration of fibroblast and endothelial cells, and this function of Gα13 is independent of GPCR signalling (Shan et al., 2006). The mechanism through which Gα13 might be involved in this process without involvement of a GPCR is still unclear, but the finding suggests that G12 proteins may be coordinating promigratory signals from GPCRs and RTKs.

G12 proteins in oncogenic transformation and cancer

Soon after their discovery, several studies established a role for both Gα12 and Gα13 in oncogenic transformation (Chan et al., 1993; Xu et al., 1993). Interestingly, the wild-type form of Gα12 was identified as a transforming oncogene due to its ability to promote focus formation in NIH3T3 mouse fibroblasts, revealing the G12 subfamily as the only class of heterotrimeric G proteins that is transforming when over-expressed as a wild-type form (Chan et al., 1993). Subsequent studies utilizing overexpression or mutationally activated forms of Gα12/13 have confirmed the ability of these proteins to transform fibroblasts (Jiang et al., 1993; Xu et al., 1993; Vara Prasad et al., 1994; Voyno-Yasenetskaya et al., 1994). Furthermore, overexpression of GPCRs that couple to G12 proteins such as PAR-1 (Martin et al., 2001) and M1 muscarinic acetylcholine receptor (Fromm et al., 1997), or stimulation of GPCRs by agonists (Aragay et al., 1995; Marinissen et al., 2003; Radhika et al., 2005) has been shown to promote cell growth and transformation through endogenous G12 signalling. These studies led to the hypothesis that GPCRs may signal through G12 proteins to promote tumorigenesis and tumour cell growth (Radhika and Dhanasekaran, 2001). Recently, it was demonstrated that the expression of G12 proteins themselves is significantly upregulated in tissue specimens from patients with adenocarcinoma of the breast and prostate (Kelly et al., 2006a,b;). Interestingly, a growth-promoting effect of G12 proteins was not observed in any of the breast and prostate cancer cell lines examined in this study (Kelly et al., 2006a,b;), suggesting that G12-mediated effects on cell proliferation may be cell type-specific, and much more pronounced in fibroblasts than in epithelial-derived cells.

Importantly, the levels of several G12-coupled receptors are elevated in various cancers and contribute to tumour cell growth and metastasis when activated by circulating or locally produced agonists (Dorsam and Gutkind, 2007). The G12/13-coupled protease-activated receptor, PAR-1, is overexpressed in highly invasive breast carcinoma cell lines and tissue specimens (Even-Ram et al., 1998), and in advanced-stage prostate cancer patient samples (Daaka, 2004). Thrombin stimulates invasion of cancer cells in vitro (Henrikson et al., 1999; Shi et al., 2004; Kelly et al., 2006a,b;), and antisense-mediated downregulation of PAR-1 blocks cancer cell invasion (Even-Ram et al., 1998). Receptors for bio-active lipids, such as lysophosphatidic acid (LPA) and S1P, are involved in cancer cell proliferation and migration (Dolezalova et al., 2003; Mills and Moolenaar, 2003). LPA is secreted by ovarian cancer cells and, by acting on the LPA receptors, sets up an autocrine loop that promotes both growth and survival of the cancer cells (Mills and Moolenaar, 2003). LPA receptors couple to G12/13 (Riobo and Manning, 2005), and it was recently shown that G12 proteins are critical regulators of LPA-stimulated migration of ovarian cancer cells (Bian et al., 2006). In another human cancer cell line, a positive feedback mechanism of Rho/ROCK activation mediated via the Rho effector, Dia1 (a Diaphanous-related formin) and the RhoGEF, LARG, appears to be important for LPA-stimulated invasion (Kitzing et al., 2007). The G12/13-coupled chemokine receptor, CXCR4, is aberrantly overexpressed in many malignant tumours including breast, prostate and lung cancer (Balkwill, 2004). CXCR4 binds to and is activated only by SDF-1. Interestingly, the expression of SDF-1 has been observed to be high in the organs where these tumour cells most frequently metastasize (Muller et al., 2001; Balkwill, 2004). Thus, activation of G12/13 signalling by several GPCRs likely plays an important role in tumour progression. As mentioned earlier, Gα13 has been implicated in transducing promigratory signals from RTKs (Shan et al., 2006). Interestingly, these RTKs, viz, platelet-derived growth factor receptor (PDGFR), epidermal growth factor receptor (EGFR) and vascular endothelial growth factor receptor (VEGFR) are overexpressed in highly metastatic cancers (Ferrara et al., 2003; Jechlinger et al., 2006; Normanno et al., 2006).

The ability of G12 proteins to transform fibroblasts appears to be primarily dependent on the activation of Rho proteins (Fromm et al., 1997; Martin et al., 2001; Kumar et al., 2006a), although in certain contexts other G12 effectors may be involved (see below; Figure 1). Besides their importance in controlling actin cytoskeleton rearrangement and cell polarity, as noted above, Rho GTPases also regulate microtubule dynamics, transcription factor activity and aspects of cell growth (Etienne-Manneville and Hall, 2002). Stimulation of Rho GTPases by G12 proteins impacts on several signalling pathways that include MAPK signalling cascades, and G12 proteins coordinate the activation and/or inhibition of different MAPKs (Collins et al., 1996; Voyno-Yasenetskaya et al., 1996; Nagao et al., 1999; Arai et al., 2003; Dermott et al., 2004). Gα12 and Gα13 stimulate the activity of the MAPK, JNK (Goldsmith and Dhanasekaran, 2007); a major outcome of JNK activation in cells is the phosphorylation of the transcription factor c-Jun, leading to increased transcription of genes involved in proliferation and survival, cell motility and invasion, among others (Karin et al., 1997). Thus, many of the biological consequences of Gα12/13 activation, including growth, differentiation and cellular transformation, could potentially be due to an impact on c-Jun activity (Jho et al., 1997; Marinissen et al., 2003; Radhika et al., 2005) (Figure 1). Stimulation of Rho by G12 proteins has also been shown to lead to cellular transformation via other signalling pathways such as activation of PDGFα receptor and STAT3 (Kumar et al., 2006a,b;), induction of transcription of COX2 (Dermott et al., 1999; Slice et al., 1999) and Egr-1 (Vara Prasad et al., 1994; Vara Prasad and Dhanasekaran, 1999), and by regulation of transcription from serum response element (SRE) (Fromm et al., 1997). There is some evidence that effector pathways other than Rho GTPases, such as extracellular signal-regulated kinase 5 (ERK5) activation (Fukuhara et al., 2000b), may contribute to the transforming potential of G12 proteins. This is particularly important as Gα12 and Gα13 are more potent stimulators of cellular transformation than overexpressed or mutationally activated RhoA (Fromm et al., 1997). In this regard, activation of Rac is required for Gα13-mediated SRE-dependent gene transcription induced by radixin (Liu and Voyno-Yasenetskaya, 2005). Radixin is involved in cross-linking of the actin cytoskeleton to the plasma membrane, and it has been shown to interact with Gα13 and to play a role in Gα13-induced transformation of Rat-1 fibroblasts (Vaiskunaite et al., 2000). Thus, other G12-effector signalling axes besides the signalling pathways leading to Rho activation may be required for G12-stimulated oncogenic transformation and is an important area for further study.

G12 proteins in cancer invasion and metastasis

G protein-coupled receptors that couple to Gα12/13 as well as downstream effectors of Gα12/13 have been implicated in tumorigenesis and cancer progression (Sahai and Marshall, 2002; Dorsam and Gutkind, 2007). The findings noted above indicating the importance of G12 proteins in cell migration during development, and the fact that G12 protein expression is upregulated in some human cancers, provided hints that the G12 subfamily of heterotrimeric G proteins may have a role in metastasis. Indeed, a direct role of G12 proteins in cancer invasion and metastasis has been recently demonstrated (Kelly et al., 2006a,b;). Expression of constitutively activated Gα12 and Gα13, or activation of Gα12/13 signalling by PAR-1 as well as thromboxane A₂ receptor, induces a striking increase in breast and prostate cancer cell invasion in vitro (Kelly et al., 2006a,b;) (Table 1). Furthermore, blocking downstream signalling via G12 reduces breast cancer metastasis in vivo and results in a significant increase in metastasis-free survival of mice (Kelly et al., 2006a). In these studies 4T1 mouse mammary carcinoma cells were employed; when implanted in the mammary fat pad of recipient mice these cells form tumours and metastasize in a manner similar to human breast cancer (Aslakson and Miller, 1992; Smith et al., 2004). Inhibition of G12 signalling in the 4T1 cells through expression of a dominant-negative form of p115RhoGEF reduces the rate of metastatic dissemination of the cells from the primary tumour following their implantation in the mouse mammary fat pad. Interestingly, when the same cells are introduced directly into the bloodstream, inhibition of G12 signalling has no effect on the ability of these cells to metastasize (Kelly et al., 2006a). These findings indicate that G12 signalling is important in the early steps of the metastasis process and appears to promote cancer metastasis by stimulating tumour cell invasion and entry into the bloodstream.

As noted above, many of the effects of G12 proteins on cell growth, transformation and migration are mediated by Rho proteins which have a well-described role in tumorigenesis and metastasis. RhoA and RhoC are overexpressed in several tumours including colon, breast, lung and pancreas (Fritz et al., 1999; Sahai and Marshall, 2002) and their expression levels correlate positively with the progression of the tumour (Suwa et al., 1998; Fritz et al., 1999). Thus, activation of Rho by G12 proteins appears to be a crucial signalling mechanism for regulating cancer metastasis. In fact, it has been demonstrated that G12 signalling leading to invasion in several breast and prostate cancer cell lines requires activation of Rho and its downstream effector ROCK (Kelly et al., 2006a,b; unpublished observations) (see Figure 1). Similarly, G12-mediated migration of ovarian cancer cells also depends on activation of Rho/ROCK signalling (Bian et al., 2006). On the other hand, the G12/13-Rho signalling pathway mediates S1P-induced inhibition of migration of glioblastoma cells (Malchinkhuu et al., 2008), revealing the complexity of cell invasion and migration and the different mechanisms by which Rho can regulate these processes.

The involvement of the two subtypes of G12 proteins in cancer cell invasion may be complicated and cell type-specific. For example, activated Gα12 inhibits, rather than promotes, the invasion of the inflammatory breast cancer cell line SUM149 (Patrick Kelly, unpublished observations). Also in this regard, a recent study showed that expression of constitutively active Gα13 or activation of G13-coupled receptors by lysophosphatidylcholine inhibits chemokine-stimulated invasion of melanoma cells in vitro and impairs their metastasis in mice. This effect of activated Gα13 in melanoma cells appears to be mediated by a reduction in the levels of Rho-GTP due to high p190RhoGAP activity (Bartolome et al., 2008). In addition, G12 signalling independent of Rho activation may also potentially lead to invasion and cancer progression. In the T47D breast cancer cell line, expression of a Gα12 mutant that is uncoupled from Rho-mediated signalling (Meigs et al., 2005) induces a small but significant increase in invasion (Kelly et al., 2006a). This finding suggests that, while Rho activation by G12 proteins is necessary for stimulating cancer cell invasion, other effectors of Gα12/13 may also be required for this function. As mentioned above, Gα12 and Gα13 interact with members of the cadherin superfamily of cell adhesion proteins, most notably E-cadherin, and negatively regulate its adhesive function (Meigs et al., 2002). E-cadherin is required for cells to maintain their epithelial character and several studies have shown that loss of E-cadherin function in cancer is associated with a transition to a more aggressive, mesenchymal phenotype (Conacci-Sorrell et al., 2002; Thiery, 2002). Thus, G12 signalling could promote cancer metastasis by inhibiting E-cadherin function. Taken together, these studies suggest that the G12 proteins possibly regulate cancer cell invasion and migration via several mechanisms.

Gα12 and Gα13 appear to similarly affect the migration and invasion of several cancer cell types (Bian et al., 2006; Kelly et al., 2006a,b; Bartolome et al., 2008; Malchinkhuu et al., 2008), suggesting that they have overlapping functions. For example, the effect of separately blocking signalling via G12 and G13 leads to a similar degree of inhibition of LPA-induced migration of ovarian cancer cells as that observed upon expression of dominant-negative p115RhoGEF that blocks downstream signalling by both G12 proteins. (Bian et al., 2006). The functions of G12 and G13, however, are not completely redundant during embryonic development. Gα13-deficient mice are embryonic lethal at around E10, where as Gα12-deficient mice are apparently normal (Gu et al., 2002). Thus, embryos are able to develop normally in the presence of a full complement of wild-type Gα13 alleles even if both Gα12 alleles are disrupted. Embryos lacking both Gα12 and Gα13 die between E8 and E8.5. Although having a single allele each of Gα12 and Gα13 is sufficient for survival, in the absence of a Gα12 allele one allele of Gα13 is not enough for the embryos to survive beyond E9.5 (Gu et al., 2002). This suggests that there is some functional overlap between Gα12 and Gα13 during early development. This is interesting, both in terms of understanding fully the roles of G12 and G13 signalling in cancer, and for designing drug targets to inhibit metastasis.

Concluding remarks

The G12 subfamily of heterotrimeric G proteins impacts a variety of cellular functions and physiological processes. The dysregulation of some of these processes clearly underlies the pathophysiology of tumour development and progression. The studies highlighted in this review provide compelling evidence that Gα12 and Gα13 play pivotal roles in many aspects of cancer invasion and metastasis. Various downstream effectors that are important in G12/13-induced cell growth and transformation, migration and invasion have been identified. Rho GTPase, in particular, appears to be a principle downstream target of G12/13 signalling leading to these cellular responses, but other effectors of G12 proteins are likely important for many of the consequences of their activation. A complete understanding of the signalling pathways triggered by G12/13 requires the elucidation of specific effectors that directly interact with the Gα proteins and the downstream molecules that these effectors engage. Although Gα12 and Gα13 affect many similar biologies, they may do so via different signalling mechanisms. For example, Hax-1 only interacts with Gα13 and not Gα12, and, as discussed earlier, this interaction is important for Gα13-stimulated migration of fibroblasts (Radhika et al., 2004). In this regard, the identification of mutational variants of Gα12/13 that are selectively uncoupled from specific effectors (Meigs et al., 2005) provides valuable tools to decipher the importance of specific downstream pathways in particular G12-mediated biologies; such information will be required to fully understand the role of G12 proteins in migration and invasion.

Although in vitro experiments provide information about the mechanism and the molecular players involved in the various signalling pathways triggered by G12/13, in vivo experiments are required for accurate modeling of the impact of activation of G12 proteins on physiological and pathophysiological processes. The importance of such studies has been recently demonstrated with the development of Gα12/13 conditional knockout mice (Ruppel et al., 2005; Moers et al., 2008; Wirth et al., 2008), and with the use of the dominant-negative RhoGEF construct in vivo (Kelly et al., 2006a). With these advances in experimental techniques and the application of siRNA technology to G12 signalling (Andreeva et al., 2006; Bartolome et al., 2008), new and more complex functions of G12/13 signalling have begun to emerge. These approaches will also aid in understanding further the role of G12 proteins in oncogenesis and metastasis.

From a clinical perspective, pharmacologic inhibition of G12 signalling could be an effective therapeutic option for controlling cancer metastases. This may be achieved by targeting a specific G12/GPCR interaction in a particular type of cancer, or by inhibiting specific downstream signalling pathway(s) activated by Gα12/13. Therefore, a complete understanding of the biological role of G12 proteins in cell growth, invasion and cancer progression is imperative and will constitute a major aspect of G12 research in the years to come.

Glossary

Abbreviations:

ASK-1

apoptosis signal regulating kinase-1

G protein

heterotrimeric guanine nucleotide-binding protein

GPCR

G protein-coupled receptor

JLP

JNK-interacting leucine zipper protein

JNK

c-Jun N-terminal kinase

LPA

lysophosphatidic acid

MAPK

mitogen-activated protein kinase

MLC

myosin light chain

MZB

splenic marginal zone B

NPC

neural progenitor cell

non-RTK

non-receptor tyrosine kinase

PAR-1

protease-activated receptor-1

PDGF

platelet-derived growth factor

RGS

regulator of G protein signalling

RhoGEF

Rho-specific guanine nucleotide exchange factor

ROCK

Rho kinase

RTK

receptor tyrosine kinase

S1P

sphingosine-1-phosphate

SDF-1

CXC chemokine stromal cell-derived factor-1α

SRE

serum response element

VSMC

vascular smooth muscle cell