Regulation of RhoGEF proteins by G_12/13-coupled receptors

Abstract

G protein-coupled receptors (GPCRs) represent a large family of seven transmembrane receptors, which communicate extracellular signals into the cellular lumen. The human genome contains 720–800 GPCRs, and their diverse signal characteristics are determined by their specific tissue and subcellular expression profiles, as well as their coupling profile to the various G protein families (G_s, G_i, G_q, G₁₂). The G protein coupling pattern links GPCR activation to the specific downstream effector pathways. G_12/13 signalling of GPCRs has been studied only recently in more detail, and involves activation of RhoGTPase nucleotide exchange factors (RhoGEFs). Four mammalian RhoGEFs regulated by G_12/13 proteins are known: p115-RhoGEF, PSD-95/Disc-large/ZO-1 homology-RhoGEF, leukemia-associated RhoGEF and lymphoid blast crisis-RhoGEF. These link GPCRs to activation of the small monomeric GTPase RhoA, and other downstream effectors. Misregulated G_12/13 signalling is involved in multiple pathophysiological conditions such as cancer, cardiovascular diseases, arterial and pulmonary hypertension, and bronchial asthma. Specific targeting of G_12/13 signalling-related diseases of GPCRs hence provides novel therapeutic approaches. Assays to quantitatively measure GPCR-mediated activation of G_12/13 are only emerging, and are required to understand the G_12/13-linked pharmacology. The review gives an overview of G_12/13 signalling of GPCRs with a focus on RhoGEF proteins as the immediate mediators of G_12/13 activation.

Activation of RhoGEF by G₁₂-coupled G protein-coupled receptors (GPCRs)

Coupling to G_12/13 proteins is being found for an increasing number of GPCRs, but remains to be elucidated for many others. Known GPCRs are purinergic receptors (P2Y₁, P2Y₂, P2Y₄, P2Y₆), M₁ and M₃ muscarinic acetylcholine receptors, receptors for thrombin [protease-activated receptor (PAR)-1, PAR-2], thromboxane (TXA₂), sphingosine 1-phosphate (S1P₂, S1P₃, S1P₄ and S1P₅), lysophosphatidic acid (LPA₁, LPA₂, LPA₃), angiotensin II (AT₁), serotonin (5-HT_2c and 5-HT₄), somatostatin (sst₅), endothelin (ET_A and ET_B), cholecystokinin (CCK₁), V_1a vasopressin receptors, D₅ dopamine receptors, fMLP formyl peptide receptors, GAL₂ galanin receptors, EP₃ prostanoid receptors, A₁ adenosine receptors, α₁ adrenergic receptors, BB₂ bombesin receptors, B₂ bradykinin receptors, calcium-sensing receptors, KSHV-ORF74 chemokine receptors, NK₁ tachykinin receptors and thyroid-stimulating hormone (TSH) receptors (Katoh et al., 1996; Sauzeau et al., 2000; Komatsuzaki et al., 2001; Thevananther et al., 2001; Greenberg et al., 2003; Zheng et al., 2003; Riobo and Manning, 2005; Hains et al., 2006; Alexander et al., 2008). G_12/13 coupling can be predicted using a hidden Markov model algorithm based on the presence of certain amino acid sequences (Sgourakis et al., 2005). All known G_12/13-linked GPCRs signal additionally through other types of G proteins such as G_i/o and/or G_q/11, but GPCRs uniquely coupling to G_12/13 may have been missed in the past as assay technologies for their readout evolved only recently (Riobo and Manning, 2005; Siehler, 2008). Activation of G_12/13-coupled GPCRs causes activation of heterotrimeric G₁₂ and/or G₁₃ proteins (Figure 1). The activated and guanosine triphosphate (GTP)-bound α subunit is thought to dissociate from the βγ dimer, and its activity is terminated upon GTP hydrolysis (Hepler and Gilman, 1992). Gα₁₂ and Gα₁₃ were discovered in 1991 as a fourth class of Gα proteins. They are ubiquitously expressed and mouse isoforms reveal 67% homology (Strathmann and Simon, 1991; Offermanns, 2003). N terminal palmitoylation is relevant for plasma membrane attachment and receptor interactions. Their slow guanosine diphosphate (GDP)/GTP exchange rate indicates their importance in prolonged signalling (Dhanasekaran and Dermott, 1996).

Figure 1.

G_12/13-RhoGEF-RhoA signalling pathway. GPCR-activated G_12/13-proteins mediate translocation of RhoGEF from the cytosol to the plasma membrane, and stimulate its GEF activity. RhoGEFs oligomerize in their inactive form, and active monomers bind via their RGS domains to active Gα_12/13 at the plasma membrane. RhoGEF reveals GAP activity towards Gα_12/13, and induces downstream activation of the monomeric GTPase RhoA. Active GTP-bound RhoA is released from its inhibitory protein RhoGDI, and anchors to the plasma membrane through its geranylgeranyl-residue. Downstream targets of RhoA include ROCK, which mediates cell contraction through inhibition of MLC phosphatase, and increased SRF-dependent gene transcription. GAP, GTPase-activating protein; GDI, guanine nucleotide dissociation inhibitor; GEF, guanine nucleotide exchange factor; GPCR, G protein-coupled receptor; GTP, guanosine triphosphate; MLC, myosin light chain; RGS, regulator of G protein signaling; ROCK, Rho kinase; SRF, serum response factor.

‘Regulator of G protein signalling’ (RGS) proteins further regulate the activity of G proteins, mostly as GTPase-activating proteins (GAPs) accelerating G protein inactivation, or they can enhance G protein function (Zhong and Neubig, 2001; Xie and Palmer, 2007). RGS proteins for G_12/13 proteins are RhoGEFs, four of which are known to be regulated by G₁₂-type proteins: p115-RhoGEF, PSD-95/Disc-large/ZO-1 homology (PDZ)-RhoGEF, leukemia-associated RhoGEF (LARG) and lymphoid blast crisis (Lbc)-RhoGEF. They function as GAPs towards Gα_12/13, and binding to Gα_12/13 stimulates their guanine nucleotide exchange factor (GEF) activity (Fukuhara et al., 2001; Dutt et al., 2004). RhoGEF catalyses the exchange of GDP for GTP by promoting an active conformation of the small monomeric GTPase RhoA (Sah et al., 2000; Wheeler and Ridley, 2004). Upon activation by Gα_12/13 RhoGEF translocates from the cytosol to the plasma membrane (Meyer et al., 2008). Oligomerization of RhoGEF negatively regulates its activity, which may prevent interaction with RhoA (Chikumi et al., 2004; Baisamy et al., 2005). Activation of RhoA is further regulated by guanine nucleotide dissociation inhibitors (GDIs), which bind to the C terminus of RhoA, where they mask the geranylgeranyl residue responsible for membrane association. RhoGDI release is required for RhoA activation, and it extracts RhoA from membranes for subsequent inactivation. Moreover, RhoGAP proteins inactivate RhoA by accelerating intrinsic GTPase activity (Fukuhara et al., 2001; Wheeler and Ridley, 2004). RhoA regulates multiple downstream effectors including many cytoskeletal proteins, and most of these have not been studied in detail. Research has elucidated RhoA-mediated activation of Rho kinases (ROCK1/2) and certain downstream pathways. ROCK phosphorylates the focal adhesion kinase (FAK), which leads to formation of actin stress fibres, and to activation of the ‘serum response’ transcription factor SRF. SRF binds in the nucleus to serum response element (SRE) sequences in the promoter of specific genes, and regulates their transcription (Narumiya et al., 1997; Treisman et al., 1998; Riento and Ridley, 2003). ROCK causes on the other hand inhibition of myosin light chain (MLC) phosphatase and direct phosphorylation of MLCs. The increase of phosphorylated MLCs in the cytoskeleton induces cell contraction (Narumiya et al., 1997; Riento and Ridley, 2003). Described cellular phenomena caused by the G_12/13-RhoGEF-RhoA pathway of GPCRs hence often involve signalling or contraction through cytoskeletal proteins.

G_12/13-regulated RhoGEF proteins

GPCRs coupled to G_12/13 proteins activate the RhoGEF family members p115-RhoGEF, PDZ-RhoGEF, LARG and Lbc-RhoGEF (Fukuhara et al., 2001; Dutt et al., 2004). RhoGEFs are large proteins, and human p115-RhoGEF, PDZ-RhoGEF and LARG consist of 927, 1522 and 1544 amino acids respectively. Lbc-RhoGEF has multiple splice variants present in various tissues, and the unspliced form consists of 893 amino acids (Table 1). The amino acid identity of human p115-RhoGEF and human PDZ-RhoGEF, LARG and Lbc-RhoGEF is 36%, 39% and 31% respectively. All four RhoGEFs are widely expressed in mammals (Fukuhara et al., 2001; Dutt et al., 2004). High transcript levels of p115-RhoGEF are present in hematopoietic cells, and of PDZ-RhoGEF in the central nervous system (Kuner et al., 2002). The Lbc-RhoGEF splice variant A-kinase anchoring protein (AKAP)-Lbc is highly expressed in the heart. AKAPs bind the regulatory subunit type II of the cyclic adenosine monophosphate-dependent protein kinase A (PKA) anchoring it at specific subcellular sites, but additionally function as scaffold proteins for other signalling complexes (Diviani et al., 2001; 2004;).

Table 1.

Characteristics of G_12/13-regulated RhoGEF Proteins

	p115-RhoGEF	PDZ-RhoGEF	LARG	Lbc-RhoGEF
Expression	Blood cells, wide at low levels	CNS, wide at low levels	Ubiquitous	Wide (AKAP-Lbc in heart)
Size (human)	927 aa	1522 aa	1544 aa	Variable (splice variants)
RGS domain	Yes	Yes	Yes	RGS-like (C-terminal)
PDZ domain	No	Yes	Yes	No
DH domain	Yes	Yes	Yes	Yes
PH domain	Yes	Yes	Yes	Yes
Localization inactive form	Cytoplasm	Cytoplasm or PM (cell type)	Cytoplasm or PM (cell type)	Cytoplasm
Localization active form	Plasma membrane	Plasma membrane	Plasma membrane	Plasma membrane
Gα12-induced GEF activation	No	No	Yes	Yes
Gα13-induced GEF activation	Yes	Yes	Yes	Yes
Regulation of GEF activity	PKC (+), PAK (−)	Tec, FAK (+)	Tec, FAK (+)	PKA (-) for AKAP-Lbc
Oligomerization	Homo-oligomers	Homo- and hetero-oligomers	Homo- and hetero-oligomers	Homo-oligomers (AKAP-Lbc)
Knockout mice	Yes	No	No	No

AKAP, A-kinase anchoring protein; CNS, central nervous system; DH, Dbl-homology; FAK, focal adhesion kinase; GEF, guanine nucleotide exchange factor; LARG, leukemia-associated RhoGEF; Lbc, lymphoid blast crisis; PAK, p21-activated kinase; PDZ, PSD-95/Disc-large/ZO-1 homology; PH, pleckstrin-homology; PKA, protein kinase A; PKC, protein kinase C; PM, plasma membrane; RGS, regulator of G protein signalling; Tec, tyrosine kinase expressed in hepatocellular carcinoma.

All four RhoGEFs contain an RGS domain to directly interact with G_12/13 and to exert GAP activity towards G_12/13 The RGS sequence homology to classical RGS proteins is rather low. Whereas p115-RhoGEF, PDZ-RhoGEF and LARG contain the RGS domain in the N terminal sequence, Lbc-RhoGEF has a C terminal region sharing 39% amino acid identity to the consensus RGS domain and hence called ‘RGS-like’. Nevertheless, the homology of the RGS domain of p115-RhoGEF with those of PDZ-RhoGEF and LARG is also only 32% and 41% respectively (Hart et al., 1996; Kozasa et al., 1998; Wells et al., 2002; Dutt et al., 2004). Two other common motifs of RhoGEF proteins in the C terminal region are the Dbl-homology (DH) and pleckstrin-homology (PH) domains. The DH domain specifically binds to and stabilizes nucleotide- and Mg²⁺-free RhoA transition states to enhance its nucleotide exchange. Cells contain higher levels of GTP as compared with GDP, and the RhoGEF DH domain therefore induces loading of RhoA with GTP, and its transition into an active state (Rossman et al., 2005). The PH domain is essential for full GEF activity, and moreover anchors RhoGEF to other signalling proteins to trigger specific subcellular localizations (Fukuhara et al., 2001; Dutt et al., 2004). PDZ-RhoGEF and LARG additionally contain an N terminal PDZ domain, which enables coupling to cell surface receptors such as plexins, insulin-like growth factor receptors or GPCRs (Fukuhara et al., 2001; Taya et al., 2001; Chikumi et al., 2004; Dutt et al., 2004; Yamada et al., 2005; Grabocka and Wedegaertner, 2007; Kelly et al., 2007). The G protein-coupled LPA₁ and LPA₂ receptors were shown to directly interact through their C terminal PDZ domain-binding motifs with the PDZ domain of LARG and PDZ-RhoGEF to activate RhoA (Yamada et al., 2005), and it remains to be seen whether further GPCRs can directly interact with these RhoGEFs.

p115-RhoGEF localizes throughout the cytosol, and rapidly translocates to the plasma membrane upon GPCR-activation of Gα_12/13, or expression of constitutively active Gα_12/13 mutants (Bhattacharyya and Wedegaertner, 2003a; Meyer et al., 2008). An E229L mutation of Gα₁₃ disrupted p115-RhoGEF interaction and recruitment to the plasma membrane (Grabocka and Wedegaertner, 2005). E27 and E29 in the acidic-rich N terminus of p115-RhoGEF are besides the RGS and PH domains required for binding to activated Gα₁₃, but not for plasma membrane translocation (Bhattacharyya and Wedegaertner, 2003a,b;). LARG is reported to be distributed throughout the cytoplasm in most cells, except of Madin-Darby canine kidney cells, in which LARG is localized at lateral membranes. In fibroblasts, LARG was reported to localize to the microtubule-organizing centre and along microtubule tracks to contribute to cell polarity. PDZ-RhoGEF localizes in some cell types also in the cytosol, but in others at the cell periphery at or near the plasma membranes where it interacts with cortical actin. In polarizing neutrophils, PDZ-RhoGEF localizes to the back of the cells. Activation of PDZ-RhoGEF and LARG by Gα_12/13 induces their translocation to the plasma membrane (Banerjee and Wedegaertner, 2004; Wong et al., 2007; Goulimari et al., 2008; Meyer et al., 2008). The subcellular localization of LARG and PDZ-RhoGEF might depend on PDZ-interacting proteins present in a cell. Gα₁₂ and Gα₁₃ specifically interact with distinct splice variants of Lbc-RhoGEF, and cause redistribution of Lbc-RhoGEF from the cytosol to the plasma membrane (Diviani et al., 2001; 2004;). The localization of all RhoGEFs specifically controls the spacial distribution of RhoA activity in cells.

Gα₁₃, but not Gα₁₂, was found to stimulate the GEF activity of p115-RhoGEF and PDZ-RhoGEF (Hart et al., 1998; Tanabe et al., 2004). Gα₁₂ is reported to enhance the GEF activity of tyrosine-phosphorylated LARG, but not of unphosphorylated LARG. Gα₁₃ on the other hand enhances more weakly both phosphorylated and unphosphorylated forms of LARG (Suzuki et al., 2003). Gα_12/13 enhance the GEF activity of the Lbc-RhoGEF splice variant AKAP-Lbc through triggering the release of PKA, which inhibits its GEF activity by phosphorylation and concomitant recruitment of 14-3-3 (Diviani et al., 2001; Kurose, 2003; Diviani et al., 2004). Stimulation of RhoGEF activities by kinases include tyrosine phosphorylation of PDZ-RhoGEF and LARG by FAK or Tec (Fukuhara et al., 2001; Suzuki et al., 2003; Chikumi et al., 2004), and serine phosphorylation of p115-RhoGEF by protein kinase C (PKC)α (Holinstat et al., 2003). p21-activated kinase (PAK)1 can bind to the DH/PH domains of p115-RhoGEF (but not of LARG or PDZ-RhoGEF), and mediate Rac-induced inhibition of RhoA activation (Rosenfeldt et al., 2006).

RhoGEFs were recently found to oligomerize through their C terminus, which inhibits their GEF activity and interaction with RhoA. Oligomerization was not affected by interaction with Gα_12/13. p115-RhoGEF and AKAP-Lbc form homooligomers, whereas PDZ-RhoGEF and LARG can form homo- and heterooligomers. Blockade of AKAP-Lbc oligomerization abolished the inhibitory effect of PKA and 14-3-3 (Chikumi et al., 2004; Baisamy et al., 2005). RhoGEFs can compete with other RGS proteins for binding to Gα_12/13, which can prevent RhoA activation (Kelly et al., 2007). Examples are axin which blocks Gα₁₂-p115-RhoGEF interaction, and RGS16 that prevents Gα₁₃-p115-RhoGEF interaction (Johnson et al., 2003; Stemmle et al., 2006). Mutant RhoGEFs lacking GEF activity function as dominant-negative inhibitors of GPCR- or G_12/13-mediated responses, whereas constitutively active p115-RhoGEF, lacking the N and C terminal regulatory domains, activates RhoA (Mao et al., 1998; Fukuhara et al., 1999; Lee et al., 2004). G_12/13-regulated RhoGEFs all activate RhoA, but are not completely redundant. Using small interfering ribonucleic acid, the thrombin receptor was found to signal through LARG to activate RhoA, whereas the LPA receptor utilized PDZ-RhoGEF (Wang et al., 2004).

Only limited genetic mouse models and genetic linkages are known for the four G_12/13-regulated RhoGEFs. Knockout mice are described for p115-RhoGEF, but not yet for PDZ-RhoGEF, LARG and Lbc-RhoGEF. The murine homologue of human p115-RhoGEF, Lsc, is solely expressed in hematopoietic cells. Lsc-knockout mice reveal impaired B/T lymphocyte proliferation, and B cell homing (Francis et al., 2006). Moreover, a mutant form of LARG has been found in some human cancers such as acute myeloid leukemia (Kourlas et al., 2000). In the worm species Caenorhabditis elegans only one RGS-containing RhoGEF exists, RHGF-1, that is a homologue of the mammalian p115-RhoGEF. RHGF-1 is regulated by the Gα₁₂ homologue GPA-12, and causes activation of the RhoA homologue RHO-1. RHO-1 has been characterized to regulate neuronal activity such as acetylcholine release and locomotion (McMullan and Nurrish, 2007).

Downstream signalling of RhoGEF proteins

The G_12/13-regulated RhoGEFs, p115-RhoGEF, PDZ-RhoGEF, LARG and Lbc-RhoGEF, cause activation of RhoA. Three closely related isoforms of Rho exist: RhoA, RhoB and RhoC, which are ubiquitously expressed, share 85% amino acid sequence identity, and have a C terminal geranylgeranyl residue. RhoA plays a key role in the regulation of the actin cytoskeleton, cell shape, cell polarity, microtubule dynamics, membrane transport pathways, gene transcription, cell adhesion, cell migration, neurite extension/retraction and cell growth. The roles of RhoB and RhoC are less clear. Activation of RhoA by a G_12/13-linked GPCR induces its redistribution to the plasma membrane, which is reversed by injection of RhoGDI (Michaelson et al., 2001; Bhattacharya et al., 2004; Wheeler and Ridley, 2004; Wennerberg and Der, 2004; Yonemura et al., 2004; Jaffe and Hall, 2005; Rossman et al., 2005). Three RhoGDI proteins are described, and only RhoGDIα is ubiquitously expressed and has the highest affinity for RhoA. Phosphorylation of RhoGDIα by PKCα causes activation of RhoA, whereas phosphorylation by PKA causes RhoA inhibition (Mehta et al., 2001; Jaffe and Hall, 2005). Binding of GTP to RhoA causes its conformational change, and subsequent interaction with effector proteins (Bhattacharya et al., 2004). C3 toxin specifically ADP-ribosylates RhoA at N41 in the effector domain and inactivates RhoA. It is utilized to detect involvement of RhoA in signalling (Jaffe and Hall, 2005). The RhoA-S19N dominant-negative mutant has decreased affinity for GTP, and increased affinity for RhoGEFs. The GTPase activity of RhoA is impaired by G14V or G63L mutations, and these result in a constitutively active mutant (Seasholtz et al., 1999). Mutant RhoA forms do not bind RhoGDI, and localize at the plasma membrane and internal membranes (Jaffe and Hall, 2005).

Many effector proteins of RhoA have been found, but only a few have been further studied. The best characterized RhoA effectors, ROCK1/2, are ubiquitously expressed serine/threonine kinases, which translocate from the cytosol to the plasma membrane in the presence of active RhoA (Narumiya et al., 1997). Binding to RhoA disrupts the interaction between the ROCK kinase domain and its autoinhibitory C terminal domain. ROCK-mediated effects can be prevented by the selective and adenosine triphosphate-competitive ROCK inhibitor fasudil (Y27632) (Sah et al., 2000). ROCK phosphorylates many substrates such as FAK, c-Jun N-terminal kinase (JNK), MLC phosphatase, ezrin/radixin/moesin (ERM), LIM kinase (LIMK), Diaphanous (mDia), rhophilin, rhotekin, citron kinase and microtubule-associated Tau. In neurons, ROCK inactivates collapsin response mediator protein-2 (CRMP-2), which leads to blockade of microtubule assembly and hence neurite retraction (Sah et al., 2000; Jaffe and Hall, 2005). Phosphorylated FAK enhances formation of focal extracellular matrix adhesions of cells, complexes additionally containing paxillin, talin, α-actinin and vinculin, among other proteins. Focal adhesions link actin stress fibres to certain integrins at the inner surface of the plasma membrane, and p115-RhoGEF, PDZ-RhoGEF and LARG activation were demonstrated to be involved in focal adhesion formation and movement (Dubash et al., 2007; Iwanicki et al., 2007). Phosphorylated active JNK causes downstream phosphorylation of the transcription factors Jun and ATF2, and hence regulates gene transcription (Kurose, 2003). As previously described, ROCK inhibits MLC phosphatase by phosphorylation of the regulatory myosin-binding subunit of myosin phosphatase (MYPT1), which results in enzymatic inhibition. The resulting increased levels of the phosphorylated regulatory light chain MLC20 of myosin II enhances actomyosin crossbridging which leads to cell contraction in different cell types (Narumiya et al., 1997). The ERM protein family links plasma membrane proteins with the actin cytoskeleton. ERM proteins are relevant in cell adhesion, cell migration and cell division, and are stabilized by ROCK through phosphorylation (Kurose, 2003; Kelly et al., 2007).

Serum response factor-dependent gene transcription is controlled by two RhoA-dependent pathways: (i) activated ROCK/LIMK stabilize F-actin by blocking the depolymerizing factor cofilin; and (ii) active mDia1/vasodilator-stimulated phosphoprotein enhance F-actin assembly by promoting filament nucleation. The resulting G-actin depletion induces activation of SRF. The transcription factor SRF binds as a dimer to SREs together with a co-transcription factor ['ternary complex factor' (TCF) like e.g. Elk-1] regulated by the Ras/Raf/extracellular signal-regulated kinase pathway. SRF controls expression of immediate-early and muscle-specific genes, including the cytoskeletal proteins β-actin and vinculin (Marinissen and Gutkind, 2005). A SRF-dependent reporter gene assay has been invented by mutation of the TCF-binding site in the c-fos SRE, and has been widely used to indirectly detect RhoA activation. The canonical SRE core sequence CC[A/T]₆GG represents the DNA-binding site for SRF. Another SRF cofactor, MAL, dissociates from cytosolic actin monomers to translocate to the nucleus upon RhoA activation. The transcription of different SRF target genes requires specific SRF cofactors (Treisman et al., 1998; Jaffe and Hall, 2005). SRF is activated not only by G_12/13-linked GPCRs, active Gα_12/13 proteins and RhoA, but also by overexpression of p115-RhoGEF, PDZ-RhoGEF and LARG (Seasholtz et al., 1999; Chikumi et al., 2004; Wang et al., 2004).

Gene deletion studies in mice revealed differences of G₁₂ versus G₁₃ (Offermanns, 1999; Kelly et al., 2007); however, no effector has been identified that is uniquely activated by either G protein. A few GPCRs apparently couple only to either G protein, such as 5-HT₄ and LPA₂ receptors linked to G₁₃, or the PAR-1 receptor linked to G₁₂. PAR-1 in another cell type, however, coupled also to G₁₃, and these apparent specificities might be cell type or assay-dependent (Gu et al., 2002; Moers et al., 2003; Yamaguchi et al., 2003; Riobo and Manning, 2005). Differences in ligand potencies and efficacies dependent on G₁₂ versus G₁₃ were reported for the ‘thromboxane’ TXA₂ receptor in a Gα subunit-specific guanosine-5′-O-(3′-[³⁵S]thio)-triphosphate ([³⁵S]GTPγS) binding assay. Ligands were only weakly active or inactive at the TXA₂ receptor upon coexpression or receptor fusion of Gα₁₂, but potent in the presence of Gα₁₃ (Zhang et al., 2006). Not only G₁₂ versus G₁₃ activation, but also RhoGEF and downstream effector activation differences might be explained by ligand-effector trafficking, i.e. selective ligand potencies towards specific effector pathways.

Measurement of G_12/13-dependent RhoGEF activation

Multiple assay technologies are available since many years to measure G_s-, G_i- or G_q-type signalling of GPCRs. In contrast, readouts to measure modulation of the G_12/13-RhoGEF-RhoA pathway emerged only recently (Siehler, 2008). Initially, RhoA antibodies have been used to immunoprecipitate RhoA, and to measure [³⁵S]GTPγS or [³²P]GTP binding. This was hampered by the lack of specificity of available RhoA antibodies versus other small RhoGTPases, and by the small signals due to a high GTP hydrolysis rate of RhoA (Seasholtz et al., 1999). Similar assays were developed to directly detect activation of G_12/13 by using specific antibodies. Stimulated cells are lysed, and [³⁵S]GTPγS binding is measured either at immunoprecipitated Gα_12/13, or at Gα_12/13 immunocaptured with antibody-linked scintillation proximity assay (SPA) beads (Barr et al., 1997; Milligan, 2003; DeLapp, 2004) (Figure 2). All these methods were either not quantitative, or in the case of the [³⁵S]GTPγS SPA assay the assay window was only limited. Another method to detect RhoA activation uses the Rho binding domain (RBD) of the effector Rhotekin to specifically extract active RhoA-GTP, and to detect levels by immunoblot detection (Ren and Schwartz, 2000). The rapid GTP hydrolysis rate and immunodetection do not allow a true quantification of signals to characterize GPCR-related pharmacology.

Figure 2.

Quantitative Bioassays for Activation of the G_12/13-RhoGEF-RhoA Axis. Activation of G_12/13 proteins is reflected by enhanced [³⁵S]GTPγS binding to their α subunits, which are captured by antibody-linked SPA beads. GPCR-activated G_12/13 mediate redistribution of RhoGEF from the cytosol to the plasma membrane, which can be visualized by tagging of RhoGEF with, for example, EGFP (or by immunostaining). Fluorescence subcellular imaging allows quantification of cytosolic and plasma membrane-localized RhoGEF using suitable algorithms. Fluorescent biosensors contain CFP-labelled RhoA fused to the Rho binding domain (RBD) of an effector protein labelled with YFP. FRET signals between the donor CFP and acceptor YFP occur upon binding of the effector RBD to activated (GTP-bound) RhoA, and can be quantified using fluorescence imaging. Activation of RhoA can be indirectly measured in an SRF-dependent reporter gene assay. CFP, cyan fluorescent protein; EGFP, enhanced green fluorescent protein; FRET, fluorescence resonance energy transfer; GEF, guanine nucleotide exchange factor; GPCR, G protein-coupled receptor; GTP, guanosine triphosphate; GTPγS, guanosine-thio triphosphate; SPA, scintillation proximity assay; SRF, serum response factor; YFP, yellow fluorescent protein.

A quantitative method was invented using an SRF-specific reporter gene readout with a mutated c-fos promoter SRE that only binds SRF dimers but no co-transcription factor (Mao et al., 1998; Treisman et al., 1998). Nevertheless, the assay window is limited, and the RhoA-dependent activation of SRF is also modulated by other pathways. Fluorescent biosensors were developed to measure RhoA activation by detection of RhoA-GTP. One biosensor consists of four fused proteins, namely yellow fluorescent protein (YFP), the RBD of protein kinase N, truncated RhoA (amino acids 1-189) and cyan fluorescent protein (CFP). Active RhoA within the fusion protein binds the RBD of protein kinase N, and the proximity of the two fluorophores leads to an intramolecular fluorescence resonance energy transfer (FRET) signal that can be measured using subcellular confocal imaging. Another biosensor contains fused Rhotekin-RBD, CFP, YFP and C-terminal full-length RhoA, that is, the fluorophores were placed in the centre to allow RhoA to bind to RhoGDI (Yoshizaki et al., 2003; Pertz et al., 2006). These biosensors have the potential for quantitative FRET measurements using high content screening; however, only single cell analyses have been demonstrated so far.

A quantitative drug screening assay for G_12/13 activation by GPCRs was recently established, and detects translocation of ‘enhanced green fluorescent protein’-tagged p115-RhoGEF. Alternatively, p115-RhoGEF can be visualized by immunostaining using a specific antibody. p115-RhoGEF translocates from the cytosol to the plasma membrane upon activation by G_12/13-linked GPCRs, and signals can be recorded in a kinetic fashion. Translocation signals are measured using subcellular confocal imaging, and signals are quantified using suitable software algorithms for data analysis. All tested G_12/13-coupled GPCRs mediated translocation of p115-RhoGEF. Co-expression of wildtype Gα₁₂ or Gα₁₃ allows to distinguish between G₁₂- versus G₁₃-dependent p115-RhoGEF translocation. The RhoGEF translocation assay enables compound screening and the discovery of GPCR ligands active in a G_12/13-linked disease context (Meyer et al., 2008).

Dysfunction of the G_12/13-RhoGEF pathway

Misregulation of G_12/13 signalling of GPCRs causes multiple pathophysiological conditions, which underlines the important role of the G_12/13-RhoGEF-RhoA pathway in physiological processes (Siehler, 2007; Worzfeld et al., 2008). Some of these are reviewed in more detail. In many types of smooth muscle cells (SMCs) (vascular, coronary, bronchial, cerebral, visceral and gastric) the G₁₂-RhoA-ROCK axis causes sustained contraction through MLC phosphatase inhibition, whereas Ca²⁺-MLC kinase regulates transient contraction. ROCK inhibitors act on vascular SMCs to prevent vasoconstriction, and are in the clinic for arterial hypertension therapy (Watterson et al., 2005; Wettschureck and Offermanns, 2005; Brown et al., 2006). Development of arterial hypertension in a salt-induced mouse model was shown to require G_12/13-LARG signalling in vascular SMCs (Wirth et al., 2008). Moreover, ROCK inhibitors are effective for treatment of cerebral vasospasm, pulmonary arterial hypertension and asthma by preventing contraction of cerebral and bronchial SMCs respectively (Calo and Pessina, 2007). Other data indicate a role of ROCK in coronary artery spasm and hence ischemic heart disease due to constriction of human coronary artery SMCs (Ohmori et al., 2003). In myocardial infarction, RhoA and ROCK expression is up-regulated and causes cardiomyocyte apoptosis, and RhoA activation induces hypertrophy of primary cardiomyocytes (Brown et al., 2006; del Re et al., 2007). Dominant-negative p115-RhoGEF blocks hypertrophic responses in cardiac myocytes (Kurose, 2003). Transgenic expression of active Gα₁₃ or of RhoA in the heart causes cardiac hypertrophy, and p115-RhoGEF is up-regulated in hypertrophic cardiomyocytes (Wettschureck and Offermanns, 2005; Brown et al., 2006; Porchia et al., 2008).

In the immune system, the G_12/13-RhoA axis is relevant in the regulation of lymphocyte adhesion and migration, and neutrophil chemokinesis and chemotaxis. Lsc/p115-RhoGEF-knockout mice reveal defective B-cell homing and T- and B-cell proliferation (Wettschureck and Offermanns, 2005; Francis et al., 2006). G_12/13 signalling is required for splenic B cell maturation, migration and polarization (Rieken et al., 2006). G₁₃ and RhoA are further required for platelet activation, shape change and aggregation, and hence for wound healing (Moers et al., 2003; Kelly et al., 2007). Constitutively active Gα_12/13 proteins were found to be oncogenes in various cell types. No Gα_12/13 mutations have been identified in various tumours of cancer patients, but instead increased expression of Gα_12/13 and RhoA could be detected breast, prostate and colon tumours. RhoA is involved in carcinogenesis and cancer progression. Cell adhesion is reduced, which promotes cell migration and invasion of breast, prostate and colon cancer cells, and ultimatively leads to metastatic tumour progression (Kurose, 2003; Wettschureck and Offermanns, 2005; Kelly et al., 2006; 2007;). Further literature data support a role of the G_12/13-RhoA axis in ovarian and lung cancer (Kelly et al., 2007; Touge et al., 2007).

The G_12/13-RhoGEF-RhoA pathway of GPCRs therefore is implicated in many diseases including cardiac hypertrophy, cardiac failure, ischemic heart disease, arterial and pulmonary hypertension, bronchial asthma, inflammatory diseases, stroke, cancer progression and metastasis.

Conclusions

G_12/13-signalling of GPCRs represents an emerging research area. Whereas aspects of G_12/13 have been studied in greater detail in cellular systems and various tissues over the past years (Riobo and Manning, 2005; Worzfeld et al., 2008), the coupling potential of many GPCRs to this class of G proteins has often not been examined. Even less clear is the involvement of specific RhoGEF proteins and effectors in downstream signalling events. A few GPCRs were found to couple solely to G₁₂ or G₁₃ (Gu et al., 2002; Moers et al., 2003; Yamaguchi et al., 2003), although these specificities have been neither studied in various cell types, nor at the level of various downstream readouts. All four known G_12/13-regulated RhoGEFs, p115-RhoGEF, PDZ-RhoGEF, LARG and Lbc-RhoGEF, are activated by G_13. p115-RhoGEF and PDZ-RhoGEF on the other hand also translocate in the presence of GPCR-activated G₁₂, but do not reveal any GEF activation by G₁₂ (Hart et al., 1998; Tanabe et al., 2004; Meyer et al., 2008). Other factors such as kinases regulate the activity of RhoGEFs. Besides specificities towards G₁₂ versus G₁₃ regulation of RhoGEFs, two G_12/13-linked GPCRs have been discovered to signal specifically through one RhoGEF versus another (Wang et al., 2004). On the other hand, no specific effector protein uniquely activated by either G₁₂ or G₁₃, or by either RhoGEF protein is known so far.

Understanding the complexity of the specific regulation of G_12/13, RhoGEF proteins and RhoA-linked effectors is required to interpret pharmacological differences of G_12/13-coupled GPCRs in various primary cell systems and tissues. Future research on the G_12/13-RhoGEF-RhoA pathway of GPCRs is needed to further elucidate molecular pathways and their physiological context, which will facilitate the development of innovative therapies for patients.

Glossary

Abbreviations:

AKAP

A-kinase anchoring protein

CFP

cyan fluorescent protein

DH

Dbl-homology

ERM

ezrin/radixin/moesin

FAK

focal adhesion kinase

FRET

fluorescence resonance energy transfer

GAP

GTPase-activating protein

GDI

guanine nucleotide dissociation inhibitor

GDP

guanosine diphosphate

GEF

guanine nucleotide exchange factor

GPCR

G protein-coupled receptor

GTP

guanosine triphosphate

JNK

c-Jun N-terminal kinase

LARG

leukemia-associated RhoGEF

Lbc

lymphoid blast crisis

LIMK

LIM kinase

LPA

lysophosphatidic acid

mDia

Diaphanous

MLC

myosin light chain

PAR

protease-activated receptor

PDZ

PSD-95/Disc-large/ZO-1 homology

PH

pleckstrin-homology

PKA

protein kinase A

PKC

protein kinase C

RBD

Rho binding domain

RGS

regulator of G protein signalling

ROCK

Rho kinase

SMC

smooth muscle cell

SPA

scintillation proximity assay

SRE

serum response element

SRF

serum response factor

TCF

ternary complex factor

YFP

yellow fluorescent protein

Conflict of interest

The author declares no conflict of interest.