G Protein Alpha 12 and 13
Heterotrimeric guanine nucleotide-binding proteins (G proteins) are composed of three distinct gene products: an α subunit that binds GTP or GDP, plus β and γ subunits. Positioned at the cytoplasmic face of the plasma membrane, this trimer associates with G protein-coupled receptors (GPCRs) harboring seven transmembrane spans. Extracellular GPCR stimulation causes the intracellular G protein α subunit to release GDP. The vacated nucleotide-binding site is filled by GTP from an abundant cytoplasmic pool, causing the α subunit to dissociate from the βγ dimer and drive signaling through a variety of downstream effector proteins. Intrinsic GTPase activity returns the α subunit to a GDP-bound state, a process accelerated by regulators of G protein signaling (RGS proteins), restoring the inactive αβγ heterotrimer. Approximately 18 different α subunits are encoded in mammals and are grouped into the subfamilies Gs, Gi, G12/13, and Gq. The G12/13 subfamily comprises the α subunits Gα12 and Gα13 in most vertebrates and a single ancestral α subunit in all invertebrates. Gα12 and Gα13 arose from gene duplication prior to the evolutionary divergence of lampreys and gnathostomes, and share 67% amino acid identity (Krishnan et al. 2015). Gα12 and Gα13 were discovered by low-stringency PCR, using oligonucleotides derived from Gs and Gi α subunits to amplify similar coding sequences from a mouse cDNA library (Strathmann and Simon 1991). Independent of these studies, a genetic screen for female sterile mutations in fruit flies identified the concertina gene encoding a G protein α subunit required maternally for gastrulation in embryonic development. Concertina showed ~54% amino acid identity with both mammalian G12/13 homologs (Parks and Wieschaus 1991). Gα12 and Gα13 are expressed in essentially all tissues as determined by Northern blot and immunoblot analyses. The importance of this subfamily in cell growth and tumorigenesis was revealed when Chan et al. (1993) screened cDNA libraries from Ewing’s sarcoma and synovial tumor-derived cells to identify proteins capable of inducing aberrant growth. The cDNA with highest oncogenic potency in transfected fibroblasts encoded human Gα12. Overexpressed Gα13 also drove fibroblast transformation and stimulated these cells to form tumors when implanted in immunocompromised mice. Subsequent studies of GTPase-deficient, constitutively active mutants of Gα12 and Gα13 demonstrated their acute transforming activity in several cell types. The ability of Gα12 and Gα13 to drive serum-dependent oncogenic transformation through overexpression, absent of an activating amino acid mutation, marked the G12/13 subfamily as unique among G proteins.
Biochemistry and Structure of Gα12 and Gα13
Antibodies raised against the predicted C-terminus of mouse Gα13 allowed its initial chromatographic purification from bovine brain, and binding to immobilized G protein β and γ subunits was employed as a final chromatographic step to yield a protein of the predicted 43-kDa mass. For larger scale purification, recombinant Gα13 was coexpressed with the subunits β2 and γ2 in a baculovirus/Sf9 (Spodoptera frugiperda) cell system (Singer et al. 1994). The first purification of Gα12 was performed by Kozasa and Gilman (1995) who introduced baculovirus-expressed Gα12 to Sf9 cells along with β1 and hexa-histidine tagged γ2, allowing capture of the heterotrimer on nickel resin. Activation of Gα12 by AlF4¯ triggered its elution from the immobilized βγ dimer, and ion-exchange chromatography yielded the purified α subunit. These initial purification schemes allowed the first biochemical characterization of Gα12 or Gα13. Compared to the well-studied Gs and Gi subfamilies, G12/13 α subunits harbor key differences including a slow rate of guanine nucleotide exchange, exceptionally slow rate of GTP hydrolysis, and inability to serve as a substrate for ADP-ribosylation by cholera toxin or pertussis toxin.
A variety of posttranslational modifications regulate Gα12 and Gα13 signaling. Similar to other G proteins, G12/13 α subunits are palmitoylated at key cysteine residues near the N-terminus. Covalent attachment of this 16-carbon fatty acid facilitates α subunit localization at the inner face of the plasma membrane and may mediate its interaction with specific target proteins. Gα12 also is a substrate for protein kinase C, an uncommon property among G protein α subunits. Phosphorylation of Gα12 in vitro inhibited its binding to a βγ dimer, suggesting this modification prolongs the activated state of Gα12. Studies in platelets revealed both Gα12 and Gα13 as protein kinase C targets, although this modification has not been demonstrated for Gα13 using purified components (Suzuki et al. 2009). Recently, Gα12 and Gα13 were found to be modified by the Pasteurella multocida toxin, which causes deamidation of a key glutamine residue conserved in the switch II region of G protein α subunits. This modification generates glutamic acid at this position and results in a GTPase-deficient, constitutively active α subunit (Orth et al. 2013).
High-resolution structural analysis of G12/13 α subunits posed a challenge to investigators because early purification approaches did not yield high enough concentrations of Gα12 or Gα13 for crystallography. A major advance came from a chimeric Gα13 in which several regions were replaced by sequence from the Gi subfamily member Gαi1, and subsequently a purification strategy was developed using Gα12 and Gα13 chimeras in which the N-terminal regions (residues 1-48 for Gα12, 1-46 for Gα13) were replaced by the N-terminal 28 amino acids of Gαi1. These chimeric α subunits yielded high-resolution structures for inactive, GDP-bound Gα13 and an AlF4¯ complexed form of GDP-bound Gα12 that mimics the activated state (Chen et al. 2005; Kreutz et al. 2006). Although numerous binding partners of the G12/13 subfamily have been discovered (see below), the RGS-homology (RH) domains of RhoGEFs are the only polypeptides for which a complex with Gα13 has been analyzed by X-ray diffraction. Another significant advance was purification of recombinant Gα13 harboring native rather than chimeric amino acid sequence, allowing investigators to examine the full-length α subunit for determinants of effector interaction (Kozasa et al. 2011). These studies identified contact points between Gα13 and its RhoGEF targets, providing a major step forward in the mechanism of G12/13 signaling to Rho.
Receptors Coupled to the G12/13 Subfamily
A diverse group of cell surface receptors serve as activators of Gα12 and/or Gα13. These GPCRs are stimulated by extracellular agents that include phospholipid-derived ligands, peptides and other hormones, and proteases that activate the receptor through cleavage of its N-terminus. As reviewed in Riobo and Manning (2005), coupling of specific receptors to the G12/13 subfamily has been demonstrated through various methods. Because Gs and most Gi α subunits are sensitive to cholera toxin and pertussis toxin, respectively, GPCRs that drove cytoskeletal rearrangements unperturbed by these toxins were suggested to couple to Gq or G12/13 subfamilies. Therefore, an indirect approach for implicating a receptor as G12/13-coupled involved engineering knockout mice for Gq α subunits and then examining embryonic cells for cytoskeletal responses driven by extracellular ligands. Stimulation of 5-HT2C (serotonin), vasopressin V1A, and bradykinin B2 receptors drove actin stress fiber formation in Gαq(−/−)/Gα11(−/−) embryonic fibroblasts, presumably through Gα12 or Gα13. Embryonic cells isolated from mice lacking G12/13 α subunits have allowed scientists to query involvement of this subfamily more directly, and implicated GPCRs for thrombin (e.g., PAR1) and lysophosphatidic acid (LPA receptor) as G12/13-coupled. Germinal center B lymphocytes lacking Gα13 showed impaired signaling downstream of the orphan receptor P2RY8 (Muppidi et al. 2015). In addition, GPCRs have been implicated as G12/13-coupled through their sensitivity to dominant-negative RhoGEFs, which lack the Dbl-homology and pleckstrin-homology domains required for downstream Rho activation but retain the RH region engaged by the upstream α subunit. Such GPCRs include the α1-adrenergic, calcium-sensing, lysophosphatidylcholine G2A, and muscarinic M1 acetylcholine receptors, plus the Kaposi’s sarcoma-associated herpesvirus GPCR, the chemokine receptor CXCR4, and the adhesion receptor Gpr56 (Riobo and Manning 2005; Kozasa et al. 2011; O’Hayre et al. 2013).
For many G12/13-coupled receptors, it is unclear whether Gα12 or Gα13 provides the preferential or sole conduit for signaling to downstream effectors. Molecular tools used for examining individual α subunits in a cellular context include inhibitory peptides mimicking the C-terminus of Gα12 or Gα13, constitutively GDP-bound mutants of Gα12 or Gα13 that act in a dominant-negative manner, and microinjected antibodies specific to one α subunit or the other. Direct approaches for individually querying Gα12 and Gα13 examined [32P]GTP-azidoaniline photolabeling or [35S]GTPγS binding of these α subunits. In these experiments, ligand stimulation of isolated membranes was followed by immunoprecipitation of Gα12 or Gα13, and autoradiography was used to assess guanine nucleotide exchange on each α subunit. Coupling to both Gα12 and Gα13 was observed for most receptors tested, including galanin type 2; multiple subtypes of the sphingosine-1-phosphate receptor; and receptors for angiotensin II, endothelin, cholecystokinin-8, thromboxane A2, and thyroid-stimulating hormone. One GPCR that showed exclusive coupling was the 5-HT4 receptor, which activated Gα13 but not Gα12. Results from indirect approaches have suggested preferential coupling by other GPCRs. In cells lacking Gαq and Gα11, dominant-negative Gα12 blocked signaling through the thrombin, endothelin A, and vasopressin V1A receptors, whereas the corresponding mutant of Gα13 had no effect. Conversely, LPA and bradykinin B2 receptors showed sensitivity to dominant-negative Gα13 (Riobo and Manning 2005). In cardiomyocytes, signaling from the angiotensin II type I receptor to an AKAP-Lbc–Rho axis drove pro-fibrotic cellular differentiation, and this effect was hindered by dominant-negative Gα12 (Cavin et al. 2014). Chimeric G12/13 α subunits have shed light on the structural basis for specificity in GPCR coupling. Although the C-terminus plays a key role in receptor recognition, experiments interchanging N-terminal regions of Gα12 and Gα13 indicated this region harbors determinants of Gα12-specific signaling from PAR1, whereas the N-terminus of Gα13 facilitated its selective coupling to the LPA receptor (Suzuki et al. 2009).
Synthetic approaches have allowed scientists to study engagement of G12/13 α subunits by specific GPCR ligands while eliminating input from other GPCRs possibly stimulated by the same extracellular agent. Using the receptor CXCR4 engineered to respond solely to an artificial ligand, Gutkind and colleagues demonstrated Gα13 coupling through bioluminescence resonance energy transfer of proteins fused to the receptor and the α subunit. These studies also revealed the importance of CXCR4 signaling through a G12/13-Rho pathway in triple-negative breast cancer metastasis (O’Hayre et al. 2013). Other studies have revealed GPCR-independent functions for the G12/13 subfamily, notably in pathways driven by receptor tyrosine kinases. Shan et al. (2006) utilized cultured embryonic cells from knockout mice to show requirement of Gα13 in cell migration and lamellipodia formation induced by epidermal growth factor (EGF) and platelet-derived growth factor (PDGF). These signaling events were rescued by wildtype Gα13 but not Gα12, indicating specificity within the G12/13 subfamily for this function. Interestingly, a truncated Gα13 lacking its GPCR-binding C-terminus also rescued growth factor signaling, suggesting this role of Gα13 is unrelated to GPCR coupling. Other reports showed a reciprocal effect, in which a dominant-negative EGF receptor disrupted the ability of Gα13 to mediate LPA receptor-driven cytoskeletal effects (Gohla et al. 1998).
Signaling Effectors of the G12/13 Subfamily
After discovery and initial characterization of Gα12 and Gα13, several years elapsed before their first downstream effector was identified: the Rho-stimulatory guanine nucleotide exchange factor p115RhoGEF. This protein interacted with Gα12 and Gα13 in vitro, and in response to Gα13 binding induced GDP-GTP exchange on Rho (Hart et al. 1998). In addition, p115RhoGEF accelerated GTP hydrolysis by the upstream α subunit, indicating its specificity as an RGS for the G12/13 subfamily. Two related RhoGEFs were discovered as G12/13 effectors: leukemia-associated RhoGEF (LARG) and PDZ-RhoGEF, with the RH domain providing an interacting surface with the activated α subunit. Gα12 and Gα13 employ nonredundant mechanisms in triggering RhoGEFs to activate Rho. For example, Gα12 activation of LARG in vitro required prior phosphorylation of this RhoGEF by Tec kinase, whereas LARG without this modification was activated by Gα13 (Kozasa et al. 2011). In the two decades after discovery of these RhoGEFs, yeast two-hybrid screening along with biochemical and proteomic approaches yielded a diverse collection of G12/13 targets, with at least 30 nonreceptor proteins now reported to interact with Gα12 and/or Gα13. Proteins bound by both α subunits include several cadherins, the actin-binding protein radixin, p120 catenin, protein phosphatase-5, apoptosis signal-regulating kinase-1, endothelial nitric oxide synthase, and Socius, a protein that enhances Rho stimulation by G12/13 α subunits through an undefined mechanism. A nonreceptor activator of G proteins, Ric-8A, binds Gα12 and Gα13 but appears to assist folding of these α subunits rather than serve as a downstream effector (Kelly et al. 2007; Chan et al. 2013).
Some target proteins show preferential binding within the G12/13 subfamily. However, it is important to note that rigorous comparisons of affinity for Gα12 versus Gα13 have not been performed in many cases. Proteins that exhibit specificity for Gα12 include axin, polycystin-1, heat-shock protein 90, RasGAP1, N-ethylmaleimide-sensitive factor-associated protein (αSNAP), regulator of G protein signaling-1 (RGS1), the scaffolding Aα subunit of protein phosphatase-2A (PP2A), the nonreceptor tyrosine kinases Tec, Bmx, and Bruton’s tyrosine kinase, the tight junction proteins zonula occludens-1 and -2, and the A-kinase-anchoring protein AKAP-Lbc (Kelly et al. 2007; Hewavitharana and Wedegaertner 2012). Another Gα12-specific target is the protein kinase ARAF, which connects LPA receptor activation of Gα12 to mTOR complex 2, driving PKCδ phosphorylation and cell migration (Gan et al. 2012). Proteins that show Gα13-specific interaction include AKAP110, RGS16, the cytoskeleton-associated protein Hax-1, and the ARF6-specific guanine nucleotide exchange factor EFA6 (Gomathinayagam et al. 2014; Hashimoto et al. 2016). Other Gα13 targets are Rgnef, a RhoGEF that activates focal adhesion kinase in colon carcinoma cells, and the nonreceptor tyrosine kinase PYK2 that mediates Gα13 activation of NF-κB. Gα13 also binds proteins involved in cell-substrate adhesion, specifically the integrin subunits β1, β2, and β3 and the integrin-binding protein talin (Masia-Balague et al. 2015; Shen et al. 2013).
The wide range of G12/13 binding partners poses a challenge to scientists in connecting specific G protein-effector interactions to specific Gα12- or Gα13-mediated cellular events. It is well established that the RhoGEF-Rho signaling axis is necessary for the G12/13 subfamily to drive events such as cytoskeletal changes, growth signaling, and oncogenic transformation. However, other G12/13 targets such as radixin and Hsp90 have been implicated in growth and tumorigenic signaling, and it is increasingly evident that some responses mediated by Gα12 or Gα13 involve multiple downstream effectors. Interaction with epithelial (E-) or vascular endothelial (VE-) cadherin provides a mechanism for G12/13 α subunits to regulate cell-cell adhesion, cell proliferation, and embryogenesis, and Gα13 binding to integrin β subunits allows the G12/13 subfamily to participate in cell-substrate adhesion. Gα12 binding to zonula occludens proteins and Hsp90 are key events in tight junction assembly and regulation of paracellular permeability. Gα12 interaction with Bruton’s tyrosine kinase, a protein involved in B lymphocyte maturation, suggests a role for Gα12 in pathways that regulate this process (Worzfeld et al. 2008). Rather than serve as classical downstream effectors (i.e., interacting with the GTP-bound form of Gα12 or Gα13 to propagate a signal), some G12/13-interacting proteins may regulate the folding, trafficking, or anchoring of the α subunit, provide a scaffold for multiple signaling proteins, activate the α subunit in a receptor-independent manner, or serve other functions. The mechanisms utilized by Gα12 and Gα13 to engage their various effectors are far from understood, and characterization of these mechanisms promises to shed light on this G protein subfamily in cellular signaling, organismal physiology, and various diseases.
Role of Gα12 and Gα13 in Cell Growth, Oncogenic Transformation, and Apoptosis
The wildtype forms of Gα12 and Gα13 were discovered as highly transforming cDNAs when introduced to cultured fibroblasts, revealing the G12/13 subfamily as distinct from other G proteins in driving aberrant cell growth as overexpressed but not mutationally activated proteins. G12/13-induced transformation was serum-dependent, but amino acid substitutions that abolished GTPase activity allowed Gα12 and Gα13 to drive this response in the absence of serum. The mechanisms of G12/13-mediated proliferation proved elusive before binding partners of the activated α subunit were identified. However, expression of mutationally activated Gα12 or Gα13 in cells stimulated the activity of several proteins, including mitogen-activated protein kinases and small GTPases of the Ras superfamily. An early G12/13-mediated readout was activation of c-Jun NH2-terminal kinase (JNK), which participates in stress responses and apoptotic signaling as well as growth and tumorigenesis (Collins et al. 1996). Also, Gα12 activation drove phosphorylation of Shc, an adaptor protein involved in Ras activation, and stimulated an increase in E2F-mediated transcription. Subsequent studies revealed several G12/13-driven transcriptional responses related to growth and transformation. Thrombin stimulation of the receptor PAR1 triggered DNA synthesis and activating protein-1 (AP-1)-mediated transcription in a Gα12-dependent and Ras-dependent manner. In addition, activated Gα12 stimulated G1-to-S phase progression in a manner dependent on Ras, Rho, and Rac. Activated Gα13 was shown to drive transcription of the primary response genes egr-1 and c-fos, and a key mechanistic finding was that the transcriptional activator serum response factor (SRF) is stimulated downstream of G12/13 to bind the serum response element (SRE) within the c-fos promoter. Gα12 and Gα13 are now well characterized as driving Rho-dependent translocation of myocardin-like transcription factor-A (a.k.a. MAL/MKL1) to the nucleus, where it serves as a coactivator for SRF (Yu and Brown 2015). In addition, Gα12 mediates signaling from the LPA receptor to S-phase kinase-associated protein 2, which plays a role in cell cycle progression and is overexpressed in multiple cancers (Radhakrishnan et al. 2010).
Intracellular localization of G12/13 α subunits is crucial to their growth and oncogenic signaling. Gα12 and Gα13 are palmitoylated near the N-terminus, and a single cysteine in Gα12 and two closely positioned cysteines in Gα13 provide attachment points for this fatty acid. Mutation of this residue in Gα12 disrupted attachment of radiolabeled palmitate and abolished ability of Gα12 to transform cultured fibroblasts, as measured by growth in soft agar. Similarly, a Gα13 mutant lacking these Cys residues was disrupted in signaling to SRF. Engineering of the nonpalmitoylated Gα12 mutant to allow attachment of the fatty acid myristate rescued its transforming ability, suggesting that tethering of Gα12 at the plasma membrane is a crucial factor in its proliferative signaling. A distinct property of Gα12 is its requirement for lipid raft association in order to drive growth signaling. Gα12 localization to these membrane domains requires its palmitoylation and its binding to Hsp90, whereas Gα13 is not detected in lipid raft fractions (Jones and Gutkind 1998; Hewavitharana and Wedegaertner 2012). Because this lipid modification of Gα12 and Gα13 is required for oncogenic signaling, acyltransferases that catalyze palmitate attachment may provide targets for therapeutic manipulation of G12/13 function in cancers. G12/13-mediated growth signaling may also require receptor tyrosine kinase pathways, as well as autocrine signaling loops. Inhibitors of the PDGF receptor and the protein tyrosine kinase JAK3 blocked mutationally activated Gα12 from stimulating cell proliferation, and dominant-negative variants of the PDGF receptor and STAT3 hindered Gα12-mediated transformation of cultured fibroblasts. Gα12 also drives expression of cyclooxygenase-2 (COX-2), and specific inhibitors of COX-2 negatively regulated Gα12-mediated DNA synthesis. Interestingly, Gα12 stimulates phospholipase A2 to catalyze arachidonic acid release from cell membranes, an event predicted to increase biosynthesis of prostaglandins. Taken together, these results suggest Gα12 regulates prostaglandin signaling at multiple points in order to stimulate S phase progression and cell proliferation (Kelly et al. 2007).
Rho is a crucial cog in the mechanism of G12/13-mediated cell proliferation. Inhibition of Rho signaling by bacterial toxins or dominant-negative RhoGEFs disrupted Gα12 and Gα13 activation of SRF and oncogenic transformation of cultured fibroblasts. Other G12/13-mediated responses dependent on Rho include activation of phospholipase D and STAT3, expression of the PDGF receptor, and angiotensin II receptor-mediated activation of JNK and p38 MAPK. However, mutationally activated Rho lacks the potency of activated Gα12 or Gα13 in cellular transformation assays, suggesting these α subunits utilize additional, Rho-independent pathways in signaling for growth and tumorigenesis (Hewavitharana and Wedegaertner 2012). For example, Gα12-mediated stimulation of ERK5 was unaffected by Rho inhibition. In colon carcinoma cells impaired in regulated degradation of the transcriptional coactivator β-catenin, interaction of Gα12 or Gα13 with the cytoplasmic tail of cadherins caused β-catenin to dissociate from this cadherin domain, localize to the nucleus, and activate genes involved in cell proliferation. Gα12 interaction with axin, a protein involved in proteolysis of β-catenin, provided an additional point at which the G12/13 subfamily may regulate β-catenin-mediated cell growth. Also, point mutations in Gα12 disrupted its binding to Hsp90 while allowing normal RhoGEF binding, and these Gα12 variants were impaired in stimulating SRF-mediated transcription (Kelly et al. 2007; Montgomery et al. 2014). Small GTPases besides Rho have been implicated in G12/13-mediated growth signaling. Mutationally activated Gα12 stimulated Ras activation in human embryonic kidney cells. A dominant-negative form of Rac blunted the ability of Gα12 to transform cultured fibroblasts and hindered Gα12 activation of JNK in several cultured cell lines. Dominant-negative Ras blocked JNK activation by Gα12 but only in fibroblasts, suggesting G12/13 signaling for growth and transformation engages signaling pathways in a cell type-specific manner (Collins et al. 1996; Juneja and Casey 2009).
The G12/13 subfamily activates programmed cell death pathways, primarily through JNK. Although G12/13-JNK signaling was associated with proliferative rather than apoptotic effects in early studies of cultured cells, later work revealed Gα12 and Gα13 driving apoptosis in a JNK-dependent manner. Gα13 stimulated nuclear fragmentation in Chinese hamster ovary cells and COS-7 fibroblasts, and this effect was blocked by dominant-negative Rho or overexpression of the antiapoptotic protein Bcl-2. Gα12 stimulated apoptosis through a PP2A-dependent pathway that led to Bcl-2 degradation. In COS-7 cells, activated Gα12 and Gα13 induced apoptosis through two distinct pathways that converge on JNK; one through MEKK1 and the other via apoptosis signal-regulating kinase-1 (Juneja and Casey 2009). G12/13 signaling through JNK also governs cellular responses beyond growth and apoptosis. An interesting example involves the opposing roles of Gα12 and Gα13 in regulating Nrf2, a transcription factor activated by oxidative stress. Nrf2 is stimulated through a Gα13-dependent pathway but negatively balanced by Gα12 signaling through JNK, and genetic knockout of Gα12 hindered JNK-dependent proteolysis of Nrf2, allowing its upregulation (Cho et al. 2007). The Gα13-JNK signaling axis plays a role in cell differentiation, mediating retinoic acid-induced formation of endoderm from mouse embryonic carcinoma cells. One protein that modulates this pathway is JLP (JNK-interacting leucine zipper protein), a scaffold that binds Gα12 and Gα13 and anchors the activated α subunit to the JNK signaling complex (Kashef et al. 2011). A full understanding of G12/13 signaling through JNK will require defining the cell type-specific mechanisms that dictate effects of this pathway on cell proliferation, apoptosis, differentiation, and other responses.
Deep sequencing of a range of human cancers revealed a low percentage of samples harboring activating Gα12 or Gα13 mutations, relative to other G protein subfamilies (O’Hayre et al. 2013). However, overexpressed G12/13 α subunits were observed in multiple cancer types including oral, esophageal, breast, prostate, ovarian, colon/rectal, pancreatic ductal adenocarcinoma, and small cell lung carcinoma. In several tumor types, perturbation of these α subunits by small interfering RNA, dominant-negative RhoGEFs, or C-terminal inhibitory peptides hindered cell proliferation and metastatic invasion. The causes of aberrantly high levels of Gα12 and Gα13 in tumors are not known, but clues have emerged from recent studies of micro-RNAs (miRNA) that regulate Gα13 expression posttranscriptionally. Specific miRNAs were shown to bind the 3′ untranslated region of the Gα13 mRNA, and disruption of these miRNAs increased Gα13 levels and invasion of prostate cancer cells (Rasheed et al. 2013). The mechanisms in which overabundant Gα12 or Gα13 contribute to tumor progression are not understood. However, it may be important that several GPCRs coupled to these α subunits are implicated in cell proliferation and cancer progression. PAR1 is overexpressed in advanced breast and prostate tumors and drives cell invasion; Gpr56 is upregulated in glioblastomas, and endothelin receptors are implicated in metastatic progression of prostate tumors. The LPA receptor provides an autocrine mechanism to promote growth of ovarian cancer cells, and the sphingosine-1-phosphate receptor induces proliferation and migration in vascular smooth muscle cells. The receptor CXCR4, which is bound by stromal cell-derived factor 1, is particularly interesting due to its coupling to Gα13 and ability to drive cell migration and metastasis in several cancers (Juneja and Casey 2009; O’Hayre et al. 2013). Because essentially all G12/13-coupled receptors show ability to signal through other G protein subfamilies, an important yet challenging area of research will be to explore interfaces between GPCRs and G12/13 α subunits as possible targets for cancer therapeutics.
Role of Gα12 and Gα13 in Cell Shape, Adhesion, Migration, and Invasion
Early studies of the G12/13 subfamily revealed its role in cytoskeletal regulation. In fruit flies homozygous for a mutation in the G12/13 homolog Concertina, apical constriction of cells was disrupted during the stage of ventral furrow formation that initiates gastrulation (Parks and Wieschaus 1991). The G12/13 homolog Gpa-12 in roundworms also was shown to drive cell shape changes (Yau et al. 2003). Constitutively active variants of mammalian Gα12 and Gα13 stimulated the formation of actin stress fibers in cultured cells, and genetic knockout of these α subunits disrupted cell polarity during directed migration and chemotaxis. In other studies, Gα12 and Gα13 were found to regulate actin-mediated protrusions as well as reorientation of the microtubule-organizing center toward the leading edge of cells. Activated Gα12 triggered neurite retraction and cell rounding in cultured PC12 and 1321N1 cell lines, and stimulation of the sphingosine-1-phosphate receptor drove actin stress fiber formation in a G12/13-dependent manner. Absence of Gα13 in platelets disrupted shape changes characteristic of ligand-induced activation, whereas Gα12 appeared to have no role in this pathway. The G12/13 subfamily also plays a role in the assembly of focal adhesion complexes. Expression of constitutively active Gα12 or Gα13 in human embryonic kidney cells stimulated phosphorylation of the focal adhesion proteins paxillin, p130 Crk-associated substrate, and focal adhesion kinase. However, Gα12 inhibited phosphorylation of paxillin and focal adhesion kinase in Madin-Darby canine kidney epithelial cells, suggesting cell type is an important factor in the role of G12/13 α subunits in this signaling pathway (Riobo and Manning 2005; Worzfeld et al. 2008; Juneja and Casey 2009).
The G12/13 subfamily regulates cell-cell and cell-substrate adhesion. Gα12 and Gα13 interaction with the cadherin cytoplasmic domain disrupts its association with β-catenin, a protein that participates in tethering cadherins to the actin cytoskeleton. G12/13 binding disrupted E-cadherin-mediated aggregation of human myelogenous leukemia cells and reversed the cadherin-mediated inhibition of breast cancer cell migration in wound-filling assays. Gα12 and Gα13 also modulate cell-cell adhesion through interaction with p120 catenin, a protein implicated in stabilizing cadherins at the plasma membrane. G12/13 binding increased the affinity of p120 catenin for E-cadherin and negatively regulated the ability of p120 catenin to drive a branched, dendritic phenotype in HEK cells. The cadherin-binding protein αSNAP was shown to bind Gα12 but not Gα13. Because αSNAP stabilizes VE-cadherin at the cell surface and regulates endothelial barriers, these results indicate Gα12 may employ a unique mechanism for regulating cell-cell adhesion through its selective interaction with αSNAP (Kelly et al. 2007). Cell-substrate interactions are another process in which the G12/13 subfamily may play physiological and pathological roles. Attachment of Madin-Darby canine kidney cells to the extracellular substrate laminin-5 was negatively regulated by Gα12-mediated repression of the gene encoding α6 integrin, a protein that facilitates adhesion to laminin-5. Also, Gα12 negatively regulated epithelial cell attachment and migration on collagen type I by disrupting integrin α2β1 function, via an “inside-out” signaling pathway that utilized Gα12, Rho, and Src (Kong et al. 2009). Recently, direct binding was reported between Gα13 and the β3 subunit of integrin heterodimers. This interaction occurs in platelets after attachment of integrins to an extracellular substrate, and Gα13 participates in subsequent outside-in integrin signaling that drives platelet spreading. Gα13 also interacts with talin, a protein required for integrin-mediated adhesion. Interestingly, the region of integrin β3 necessary for talin binding overlaps with a Glu-X-Glu motif of integrin β3 crucial for Gα13 interaction (Shen et al. 2013).
The G12/13 subfamily drives metastatic invasion of cancer cells. Expression of constitutively active Gα12 or Gα13 stimulated several breast cancer and prostate cancer cell lines to penetrate an artificial extracellular matrix. In contrast to early results in cultured fibroblasts, activated Gα12 and Gα13 failed to stimulate proliferation of breast cancer cells, suggesting the function of the G12/13 subfamily in cancer progression, i.e., stimulating aberrant proliferation versus metastasis, is cell type-specific (Juneja and Casey 2009). Several mechanistic details of G12/13-mediated cell invasion have been revealed in recent years. LPA receptor-driven migration and focal adhesion kinase autophosphorylation were blocked by peptides mimicking the Gα12 and Gα13 C-termini. Also, depletion or inhibition of JNK diminished the ability of Gα12 to induce breast cancer cell invasion through an artificial extracellular matrix. Secretion of matrix metalloproteinases (MMP) by some tumor cells is mediated by the G12/13 subfamily. MMP-1 facilitated cellular invasion through three-dimensional collagen, and depletion of Gα13 by RNA interference hindered this effect in several cancer cell lines and upregulated cell-cell adhesion mediated by E-cadherin. Also, activation of Gα12 increased the expression of MMP-2 and the interleukins IL-6 and IL-8 in breast cancer cells (Juneja et al. 2011; Chia et al. 2014; Chow et al. 2016). Furthermore, Gα13 bound the cortactin-binding protein Hax-1 as part of a complex that mediates LPA receptor signaling, driving migration of cancer cells (Gomathinayagam et al. 2014). Other, non-GPCR pathways may modulate the ability of G12/13 to induce cell migration. For example, lamellipodia formation and migration driven by several receptor tyrosine kinases was blunted in embryonic cells cultured from Gα13 knockout mice (Shan et al. 2006).
Many cytoskeletal effects mediated by the G12/13 subfamily require Rho activation. Early reports of Gα12 and Gα13 driving cytoskeletal rearrangements described these events as sensitive to Clostridium botulinum C3 exoenzyme, an inhibitor of Rho activity (Buhl et al. 1995). GTP-bound Rho stimulates its signaling effectors Rho-associated kinase (ROCK) and mDia, and the resulting inactivation of myosin light chain phosphatase allows myosin II to become activated, driving contractile responses involving filamentous actin (Hall 2012). Stimulation of cell migration through the LPA receptor was blocked by inhibition of Rho or ROCK. Signaling from Rho drives contraction of the cell’s trailing edge, a critical step in migration. G12/13 stimulation of Rho was found to mediate polarity of neutrophils and contraction at the back and sides of cells during migration (Xu et al. 2003). In cultured neonatal cardiomyocytes, inhibition of Rho or ectopic expression of a peptide mimicking the C-terminus of Gα12 blocked α-adrenergic receptor-induced cytoskeletal rearrangements. Despite the importance of Rho in a variety of G12/13-mediated cytoskeletal responses, other targets of Gα12 and/or Gα13 drive pathways independent of the RhoGEF-Rho axis. Forced expression of E-cadherin in cultured breast cancer cells inhibited migration in wound-filling assays, and coexpression of constitutively active Gα12 restored migration of these cells. This effect of Gα12 was unperturbed by a specific inhibitor of ROCK, suggesting Gα12 regulation of cadherin function occurs through a Rho-independent mechanism. Gα12 and Gα13 also upregulated the activity of glycogen synthase kinase-3, a protein involved in neurite retraction induced by LPA. Interestingly, only Gα13 was sensitive to C3 exoenzyme, suggesting a Rho-independent mechanism specific to Gα12 in this cytoskeletal response. Furthermore, in breast cancer cells implanted in the mammary fat pad of athymic mice, expression of dominant-negative RhoGEF hindered metastatic invasion of these cells, adding to evidence implicating the G12/13 subfamily as a signaling component in metastasis (Juneja and Casey 2009).
Role of Gα12 and Gα13 at the Organismal Level: Development and Morphogenesis
Although many investigations of Gα12 and Gα13 have focused on their role at the level of individual cells (e.g., signal transduction mechanisms, interaction with effector proteins), other studies have examined the function of these α subunits in a multicellular or organismal context, including physiological events such as morphogenesis and embryonic development. Early studies revealed the Drosophila G12/13 homolog Concertina as crucial in formation of the posterior midgut and ventral furrows during gastrulation (Parks and Wieschaus 1991). In another model organism, Caenorhabditis elegans, depletion of the homolog Gpa-12 by RNA interference caused egg-laying defects, was embryonic lethal in most offspring, and caused locomotion and sensory defects in the few survivors. Disruption of Ce-RhoGEF, a homolog of mammalian RhoGEFs responsive to G12/13, phenocopied the Gpa-12 depleted roundworms to a lesser degree, suggesting G12/13 stimulation of RhoGEF-Rho pathways is an ancient signaling axis. The greater severity of phenotype in Gpa-12 knockdowns compared to Ce-RhoGEF knockdowns, along with evidence of Gpa-12 expression in multiple cell types lacking Ce-RhoGEF, suggests Gpa-12 drives other signaling pathways in this organism (Suzuki et al. 2009). G12/13 signaling is also involved in vertebrate gastrulation. Studies in zebrafish show G12/13 α subunits driving Rho-mediated and Rho-independent pathways during epiboly, a stage of gastrulation in which the blastoderm spreads over the yolk cell through coordinated shape changes and movements of several cell populations. An important event in this developmental process is downregulation of cell-cell adhesion by G12/13 binding to the cytoplasmic tail of E-cadherin. Suppression of Gα12 and Gα13 expression also disrupted a separate process of convergence and extension, a stage of gastrulation that requires a RhoGEF-Rho signaling axis (Lin et al. 2009).
Genetic knockouts revealed differences between Gα12 and Gα13 in embryonic development. Mice lacking Gα13 died at embryonic day 9.5 with defects in angiogenesis, whereas Gα12 knockout mice developed in normal fashion. However, Gα12 function is not superfluous during embryogenesis. In mice haploinsufficient for Gα13, a condition that itself does not cause embryonic lethality, absence of Gα12 resulted in death by day 9.5 of embryogenesis. Also, mice lacking both Gα12 and Gα13 died at least one day earlier in embryogenesis than Gα13(−/−) mice. Other experimental approaches have revealed novel developmental functions of the G12/13 subfamily. Mice in which Gα13 is ablated solely in endothelial cells showed an embryonic lethal phenotype similar to the full Gα13 knockout. An interesting finding came from Gα13 knockout mice in which Gα13 expression was restored specifically in endothelial cells; these embryos lived several days beyond day 9.5 and unmasked a phenotype of intracranial bleeding and aberrant brain growth outside the skull (Worzfeld et al. 2008). Tissue-specific knockout of both Gα12 and Gα13 in neurons and glial cells resulted in overmigration of cortical neurons, suggesting a negative regulatory role for the G12/13 subfamily at this developmental stage. Also, G12/13 α subunits were implicated in signaling initiated by the adhesion receptor Gpr56, a pathway important in development of oligodendrocytes and migration of neural progenitor cells (Ackerman et al. 2015). Neutrophils expressing dominant-negative G12/13 α subunits showed an aberrant increase in random migration that disrupted the ability of B lymphocytes to settle properly in the marginal zone of the spleen. In addition to roles in the nervous and immune systems, the G12/13 subfamily regulates vascular smooth muscle contraction. Studies of mice conditionally lacking Gα12 and Gα13 in smooth muscle cells revealed a requirement for these α subunits in the increased vascular tone that underlies salt-induced hypertension. The G12/13 subfamily also regulates differentiation of smooth muscle cells, through a Rho-dependent pathway that drives SRF-mediated expression of smooth muscle α-actin and other key genes (Worzfeld et al. 2008; Mack 2011).
Emerging Roles of the G12/13 Subfamily
Early characterization of Gα12 and Gα13 pointed to their role in cancer progression, but more recent studies have implicated this G protein subfamily in other diseases. In many cases of autosomal dominant polycystic kidney disease (ADPKD), the integral plasma membrane protein polycystin-1 harbors mutations in regions that include its C-terminus, which binds Gα12 and downregulates its apoptotic signaling function. The role of Gα12-polycystin-1 interaction in kidney pathologies is not well understood, but interesting clues have emerged from tissue-specific expression of Gα12 variants. In mice engineered to express mutationally activated Gα12 in podocytes, the kidney disease states of glomerulosclerosis and protein secretion into the urine showed accelerated progression. Also, Gα12 and polycystin-1 exhibit positive and negative regulation, respectively, of the proteolytic removal of extracellular E-cadherin regions. Because altered cell-cell adhesion in kidney epithelial cells is a hallmark of ADPKD, these regulatory mechanisms involving multiple binding partners of Gα12 may provide targets for therapeutic intervention in this disease (Boucher et al. 2012). Gα12 signaling also is implicated in pathologies of cardiac cells, in which cardiomyocytes respond to injury by aberrantly differentiating to matrix-secreting myofibroblasts. Stimulation of angiotensin II type 1 receptors drives a pathway that utilizes Gα12-AKAP-Lbc coupling to activate Rho, leading to collagen secretion and other pro-fibrotic responses. Gα12 also couples to AKAP-Lbc to drive α1-adrenergic signaling that leads to hypertrophy of cardiomyocytes (Cavin et al. 2014). Gα12 and Gα13 may also play a key role in blood pressure disorders, as evidenced by a requirement for G12/13-mediated signaling in the development of salt-induced hypertension in mice (Worzfeld et al. 2008). Advances in understanding G12/13 signaling networks in kidney epithelia, cardiac myocytes, smooth muscle cells, and other tissues may reveal new therapeutic strategies for several disease states.
A new role for the G12/13 subfamily was discovered in signaling pathways regulating the transcriptional coactivator Yes-associated protein (YAP). The Hippo-YAP pathway regulates organ size in mammals through coordinated control of cell proliferation and apoptosis, and dysregulation of this pathway is associated with tumor progression. G12/13 α subunits provide conduits for the receptor PAR1 to drive inhibition of Lats kinases, and the resulting dephosphorylated state of the Lats target YAP allows its import to the nucleus and subsequent activity. G12/13 signaling to YAP has been implicated in proliferation of ovarian cancer cells and also a pathway in which extracellular Wnt proteins initiate noncanonical signaling through frizzled receptors (Mo et al. 2012). Other signaling pathways recently discovered as G12/13-mediated may provide clues to the mechanism in which these α subunits promote cancer progression and metastatic invasion. Gα12 was found recently to downregulate key miRNAs that govern levels of MDM2 and Zeb transcripts, leading to aberrant expression of these proteins important in epithelial-mesenchymal transition. Also, recent evidence shows Gα12-mediated regulation of miRNAs leading to diminished p53 production, resulting in transition to a mesenchymal state and tumor growth (Yang et al. 2015). Other studies have yielded unexpected links between the G12/13 subfamily and signaling pathways previously characterized as responsive to other G proteins. For example, studies of bone marrow-derived macrophages lacking Gα13 revealed its requirement in a pathway driven by sphingosine-1-phosphate that increases activity of an adenylyl cyclase subtype, AC VII. Overexpression of Gα12 rescued signaling to AC VII in these cells, suggesting both G12/13 subfamily members can participate in this pathway (Jiang et al. 2013).
This chapter describes signaling molecules, cellular changes, and physiological and pathological responses regulated by the G12/13 subfamily of G protein α subunits. Gα12 and Gα13 are important components in an expanding list of biological events that include cell growth, polarity, migration, cell fate determination, cytoskeletal rearrangements, cell-cell and cell-substrate adhesion, and embryonic development. Compared to other G protein subfamilies, G12/13 α subunits interact with an unusually large and diverse array of target proteins, and this diversity of effectors has created a complex web of possible signaling mechanisms and regulatory relationships between different target proteins. The best studied signaling pathway downstream of Gα12 and Gα13 is the RhoGEF-Rho axis, which drives multiple cytoskeletal responses and is essential in G12/13-mediated tumorigenic growth and invasion. However, it is increasingly clear that other downstream targets of Gα12 and/or Gα13 serve as signaling effectors in physiology and disease. A comprehensive understanding of G12/13-mediated signaling will require defining these pathways and connecting individual binding partners to specific G12/13-mediated events. Also, it will be important to elucidate the structural features of Gα12 and Gα13 that mediate specific interactions and develop molecular tools to selectively disrupt these events. Thus far, X-ray crystallographic studies have examined Gα12 and Gα13 as isolated proteins, as well as Gα13 in complex with purified domains of RhoGEFs. Structures of Gα13 and Gα12 bound to various effector proteins should provide crucial new details of signaling mechanisms in these non-RhoGEF pathways, and may reveal competition between certain effector proteins for α subunit binding, or other regulatory mechanisms.
G12/13 α subunits are unique among G proteins in their ability to drive oncogenic transformation of cells as overexpressed rather than constitutively active forms. Studies of breast, prostate, small cell lung carcinoma, oral squamous carcinoma, pancreatic, and cervical cancer cell lines, along with large-scale sequencing of multiple human tumors, suggest Gα12 and Gα13 contribute to cancerous growth and metastatic invasion as wildtype proteins. This differs from other G protein subfamilies in which some cancers show a high frequency of GTPase-deficient mutants (O’Hayre et al. 2013). A major area of future work will be to unravel the mechanisms that allow aberrantly high levels of G12/13 α subunits to signal for tumorigenic responses in the absence of activating mutations. Studies of mechanisms governing Gα12 and Gα13 gene expression, protein turnover, miRNA-mediated downregulation of mRNAs, and other posttranscriptional and posttranslational regulatory events will be vital to our understanding of G12/13 signaling in cancer. Also, because many cancers show overexpression of GPCRs that couple to the G12/13 subfamily, it will be important to investigate whether overexpressed Gα12 or Gα13 augment the ability of specific receptors to drive unregulated growth, migration, or metastatic invasion.
Gα12 and Gα13 are closely related in amino acid sequence and share numerous functions, but each α subunit has evolved unique signaling roles and binding partners. Many experimental systems have studied and manipulated the signaling function of these proteins as a group, using reagents such as dominant-negative downstream effectors that do not differentiate between the two α subunits. However, other studies have focused on the individual properties of Gα12 and Gα13, using purified components in vitro or other systems employing cells or organisms. Investigations of invertebrate G12/13 signaling, in which a single protein of this subfamily is encoded, have provided useful information toward distinguishing ancient, “core” signaling functions of this subfamily from more specialized functions that evolved in Gα12 or Gα13 after gene duplication and divergence. In this chapter, we have attempted to summarize the upstream receptors, downstream effectors, and cellular roles of Gα12 and Gα13 with an eye toward categorizing molecules and events as Gα12-specific, Gα13-specific, or responsive to both proteins. It is important to note that many studies of signaling through an individual α subunit did not apply the same scrutiny to the other G12/13 subfamily member. An important future goal will be to determine the respective signaling functions of Gα12 and Gα13, particularly in diseases linked to the G12/13 subfamily, and to understand the evolutionary advantages afforded by the gene duplication and divergence that yielded these two α subunits. The increasing list of downstream effectors that show preferential binding to Gα12 or Gα13 may allow development of dominant-negative proteins for querying the role of individual G12/13 α subunits in specific pathways.
- Ackerman SD, Garcia C, Piao X, Gutmann DH, Monk KR. The adhesion GPCR Gpr56 regulates oligodendrocyte development via interactions with Gα12/13 and RhoA. Nat Commun. 2015;6–6122.Google Scholar
- Montgomery ER, Temple BRS, Peters KA, Tolbert CE, Booker BK, Martin JW, Hamilton TP, Tagliatela AC, Smolski WC, Rogers SL, Jones AM, Meigs TE. Gα12 structural determinants of Hsp90 interaction are necessary for serum response element-mediated transcriptional activation. Mol Pharmacol. 2014;85:586–97.PubMedCrossRefPubMedCentralGoogle Scholar
- Yang YM, Lee WH, Lee CG, An J, Kim ES, Kim SH, Lee SK, Lee CH, Dhanasekaran DN, Moon A, Hwang S, Lee SJ, Park JW, Kim KM, Kim SG. Gα12 gep oncogene deregulation of p53-responsive microRNAs promotes epithelial-mesenchymal transition of hepatocellular carcinoma. Oncogene. 2015;34:2910–21.PubMedCrossRefGoogle Scholar