Molecular targeting of Gα and Gβγ subunits: a potential approach for cancer therapeutics

Abstract

G protein-coupled receptors (GPCRs) signal through G protein α and βγ subunit families to regulate a wide range of physiological and pathophysiological processes. As such, GPCRs are major targets for therapeutic drugs. Downstream targets of GPCRs have also gained interest as a therapeutic approach to complex pathologies involving multiple GPCRs. One such approach involves targeting of the G proteins themselves. Several small molecule Gα and Gβγ modulators have been developed and been tested in various animal models of disease. Here we will discuss the requirements for targeting Gα and Gβγ subunits, the mechanisms of action of currently identified inhibitors, and focus on the potential utility of Gα and Gβγ inhibitors in the treatment of various cancers.

The G Protein-Coupled Receptor system as a therapeutic target

GPCRs comprise a large family of receptors for physiologically relevant ligands such as adrenalin or serotonin, which makes them attractive drug targets [1]. For each GPCR ligand there are often multiple subtypes involved in specific cellular functions. For example, there are 13 different GPCR family members that bind serotonin, but they are differentially expressed and couple to distinct signal transduction mechanisms [2]. Targeting of these specific GPCR subtypes holds the promise of being able to almost surgically manipulate the biology controlled by these receptors, and thus greatly limiting potential side effects of pharmacological therapy. A potential downside to targeting GPCR subtypes is that some pathologies, such as cancer and heart disease, are the result of dysregulation of a number of GPCR signaling pathways and circulating factors. Therefore, targeting one receptor may not be sufficient to be an effective treatment.

As an alternate approach to treatment of diseases involving multiple ligand-dependent signaling inputs, an interest has developed in more broad-based pharmacological targeting of key steps in common pathways downstream of multiple receptors that are directly involved in mediating a cellular disease pathway[3, 4]. This approach sacrifices specificity for the sake of increased efficacy, but for complex and deadly diseases like cancer, efficacy is of paramount concern. In the GPCR signaling system there are multiple second messenger cascades activated by G proteins downstream of GPCRs that have been considered as targets [3, 5]. In the classic G protein signaling system, GPCRs couple to heterotrimeric G proteins that, when activated by exchange of GDP for GTP on the G protein α subunit, are conformationally competent to interact with high affinity with second messenger generating enzymes or ion channels[1, 6]. Superimposed on this basic system are regulators of the G proteins (RGS proteins) that stimulate GTP hydrolysis on Gα subunits[7], and regulators of the GPCRs such as G protein coupled-receptor kinases (GRKs), arrestin, and associated pathways regulated by arrestin [8]. Many of these systems have been investigated as potential therapeutic avenues either with genetic or small molecule based approaches [9–13]. In this review we will discuss the potential for pharmacological targeting the G protein subunits directly with a particular focus on their utility in treating cancer.

Structural features of G proteins amenable to small molecule binding

Although peptide-based inhibitors of G protein α subunits and G protein βγ subunits have been developed [10, 14–16], we will focus on small molecule inhibitors because of their potential therapeutic utility, with the exception of one peptidic inhibitor that has potent actions in cellular systems. Gα and Gβγ represent distinct molecular problems with respect to small molecule binding. Gα subunits have a catalytic site and numerous clefts that have the potential to bind to small molecules [17, 18] and inhibit Gα nucleotide exchange and GTP hydrolysis activity. Gβγ does not have a catalytic pocket directly amenable to small molecule targeting but does have a concave surface at a protein-protein interaction “hot spot” that has proven amenable to small molecule binding [17, 19–21].

Structural features of the α subunit

The G protein α subunit consists of two distinct domains, a GTPase or Ras-like domain and an α helical domain connected by two flexible random coil linkers[22, 23] (Figure 1A). The guanine nucleotide binding site is comprised of amino acids contributed by both domains at the interface between these two domains. The mechanism for how alteration in GPCR conformation upon activation by ligand catalyzes the nucleotide exchange reaction on G proteins has been the subject of intense interest for many years [24, 25]. A recent breakthrough is the elucidation of the three dimensional structure of a complex of the agonist-bound β-adrenergic receptor and Gα_sβ₁γ₂, by X-ray crystallography [26]. This structure reveals an agonist-dependent opening of the intracellular surface of the receptor allowing for extensive interaction between various intracellular receptor surfaces and various domains of the Gα subunit, leading to alterations in the Gα subunit GDP binding pocket. These GPCR-G protein interfaces could represent small molecule targets that have not yet been exploited, although short Gα-derived peptides can selectively disrupt this interface[27]. The detailed mechanism for how receptors transmit conformational information through the Gα subunit will not be discussed here, but an unexpected result, supported by complementary electron microscopy and dynamics studies, was that the α helical domain moves apart from the GTPase domain in the GPCR-activated transition state [26, 28–30]. This suggests that the release of GDP either requires these inter-domain movements, or that the movements are a consequence of GDP dissociation. Interestingly, peptides, such as GPR/goloco peptides[10], and some small molecules (discussed below) that inhibit GDP release seem to prevent the inter-domain movements of the G protein α subunit, suggesting that the inter-domain movements are required, rather than a consequence of GDP release.

Figure 1.

Structural representations of Gα_q and Gβγ with potential sites for small molecule interactions. (a) Gα_q bound to YM-254890. The α helical domain is in blue, the Ras-like domain is in red and the linkers connecting the two domains are in green. GDP is in CPK and YM-254890 is in space fill yellow. Rendered from PDB 3AH8. (b) Surface representation of Gβ. In red is the contact surface for the SIGK peptide used in a completion screen for small molecule discovery. Rendered from PDB 1XHM using MolSoft ICM.

Structural features of the Gβγ subunit

As discussed above, targeting Gβγ is more complicated because it involves inhibiting protein-protein interactions rather than a catalytic function. Protein-protein interactions are generally thought to be more difficult to block because of the often two-dimensional nature of the binding surfaces [31]. Unlike catalytic binding pockets that bind small molecule substrates with amino acid coordination in three dimensions, small molecule interaction with two dimensional protein-protein interaction surfaces may not derive sufficient binding energy from limited potential bonding interactions. The Gβ subunit is the prototype for the WD40 repeat, 7 bladed β-propeller protein family characterized by 7 blades, each consisting of 4 antiparallel β-sheets, arranged circularly with the last of the 7 blades packing against the first [5, 17, 19]. This arrangement results in formation of a toroid shape with a hole in the middle and a major protein binding surface on the top of the torus over the hole [5, 32] (Figure 1B). Thus, this protein interaction surface on Gβ is not flat but rather is formed by the ridge of the “donut” with a concave center projecting into the hole. This 3D surface may be more amenable to small molecule binding than a typical flat protein-protein interaction interface[5].

Mechanisms of small molecule actions on Gα subunits

Several small molecule G protein α subunit inhibitors have been developed that all have an apparently common mechanism of action. They bind to G protein α subunits to prevent intrinsic and receptor-stimulated GDP release.

Suramin

Developed by Bayer in 1916, suramin is a symmetric polysulphonated napthylamine-benzamide derivative that has been used to treat African Sleeping Disease and river blindness (Figure 2A). In 1996, Freissmuth and colleagues discovered that suramin is a relatively specific inhibitor of GDP release on the Gs family G protein α subunits (IC₅₀ ~250 nM), and inhibited nucleotide exchange on Gα_i and Gα_o, at 20 and 10 fold higher concentrations, respectively[33]. Because these assays were conducted with purified G protein α subunits, suramin must be binding directly to G protein α subunits. Suramin and its analogues inhibit GPCR-stimulated nucleotide exchange on Gα_s, and Gα_s-dependent stabilization of high affinity agonist binding to adrenergic receptors, presumably by preventing GDP release thereby preventing formation of the high affinity Gα-receptor, nucleotide-free transition state. Derivatization of suramin led to identification of analogues that are highly selective antagonists for Gs relative to Gi or Gq [34].

Figure 2.

Structures of compounds that bind to Gα or Gβγ subunits.

In the absence of mutagenic mapping or structural data, the detailed molecular mechanism for suramin-dependent inhibition of GDP release is not understood. Suramin is large for a small molecule (1297.29 Da) and has the potential to occupy a large surface of Gα_s. Interestingly, purified adenylyl cyclase reduced the ability of suramin to inhibit nucleotide exchange on purified Gα_s, indicating that binding of adenylyl cyclase and suramin to Gα_s are mutually exclusive[33]. This suggests that suramin exerts its effect by binding to the effector binding sites on Gα subunits, but other interpretations involving alterations in Gα conformation by adenylyl cyclase are possible. Suramin is highly sulfonated with a strong negative charge that limits its utility as a reagent or a drug because it cannot cross cell membranes. Nevertheless, it is the first example of an organic molecule-based approach to inhibition of G protein-dependent signaling, and it has an overall mechanism of action that is similar to other G protein α subunit small molecule inhibitors.

Imidazopirazines

The imidazo-pyrazine derivative BIM-46174 (Figure 2B) was identified in a cell-based differential screen in an MCF7 breast cancer cell line for compounds that would inhibit cholera-toxin-induced, but not forskolin-induced, cAMP accumulation[35]. This screening strategy was designed to find compounds that inhibit Gs but not receptors or adenylyl cyclase. BIM-46174 and a more stable derivative BIM-46187 inhibit activation of multiple G proteins by GPCRs and bind directly to α subunits to inhibit nucleotide exchange [35, 36]. BIM-46174 inhibition of nucleotide exchange on the free Gα subunits suggests that it inhibits the conformational changes associated with G protein α subunit GDP release. Again, the lack of structural or mutagenic data limits our understanding of the mechanistic details at structural level of how this compound acts [36].

BIM-46174 contains a free sulfhydryl group and BIM-46187 is a disulfide bridged dimer of BIM-46174 [36]. The developers argue that because both the free compound and the dimer inhibit G protein activity that the –SH group is not involved in the activity of the compound, but it is not clear how carefully the disulfide linkage was controlled in these experiments to ensure that no free BIM-46714 is present in the BIM-46I87 samples. Unlike suramin, BIM-46174 is effective in cells, allowing the cell biology of the compound to be assessed. The disulfide is likely to be in a reduced state in the cellular environment. In cells, these compounds block all signaling pathways initiated by GPCRs through heterotrimeric G protein families, and thus, although not directly tested on all the individual Gα subunit family members in vitro, this class of compounds appears to bind and inhibit all G protein α subunit families equally.

YM-254890, a selective Gα_q inhibitor

A cyclic depsipeptide, YM-254890 (Figure 2C) was derived from Chromobacterium and has potent antithrombotic effects. Recently it has been shown to specifically inhibit Gq signaling[37, 38]. Although not technically a small organic molecule, but because it has actions in cells and animals, it has potential as a therapeutic or lead molecule. Similar to suramin and BIM-46174, YM-254890 inhibits Gα_q by inhibiting the nucleotide exchange reaction by binding directly to the Gα subunit and preventing either spontaneous or receptor-stimulated GDP release[37]. The compound inhibits Gq signaling downstream of GPCR activation but does not inhibit activation of signaling stimulated by transfected Gq209L, a mutation that stabilizes the GTP bound form of the Gα subunit[37]. This suggests that compounds that inhibit Gα subunits by inhibiting GDP release will be not be effective at treating diseases resulting from constitutively activating G protein subunit mutations that block GTP hydrolysis but may be effective at treating cancers or inflammatory processes related to GPCR-dependent activation of G protein pathways.

YM-254890 binding to Gα_q is the only example of a small molecule-G protein complex for which structural information is available, thus allowing for an atomic level understanding of its mechanism of action[39]. The compound interacts directly with a pocket of the α1 helix and β2 strands of the Ras-like GTPase domain and the αA helix of the α helical domain (Figure 1A). In addition, the compound interacts with the inter-domain linkers including switch 1. This binding mode might be expected to prevent inter-domain movements and provides a plausible mechanism for inhibiting GDP release. Many of the direct contacts for YM-254890 with Gα_q are conserved amongst members of the Gα_q family (Gαq, 11, 14, 15 and 16) but vary significantly in the other G protein α subunit family members. Thus, there is an opportunity for either rational design or high throughput screening (HTS) to identify α subunit subtype-selective inhibitors that capitalize on the amino acid diversity in this pocket.

Mechanisms of small molecule actions on Gβγ subunits

M119/gallein

Multiple classes of small molecule inhibitors have been identified that bind to G protein βγ subunits and modulate interactions with select groups of effectors downstream of Gβγ [40] (Box 1). We developed the concept that different classes of small molecules interact with Gβγ in unique binding modes that lead to differential modulation of pathways downstream of Gβγ. A major class of antagonist molecules identified in this screen is the M119 class which includes M119, gallein and M119K amongst others (Figure 2D) [40, 41]. In part because of the ready availability of gallein and M119, this class of compounds has been the most extensively validated, however no atomic level information detailing the interactions between G βγ and these molecules has been published. The nature of the screen used to identify these molecules (Box 1) indicates that the molecules bind to a protein interaction “hot spot” on the top surface of the β subunit WD40 repeat that is normally involved in interactions with of the switch II helix of Gα and various effectors[20, 40].

Box 1. Steps in screeningcreen for identification and characterization of Gβγ-binding small molecules.

1. Test libraries of compounds for their ability to compete for binding of SIGK peptide to Gβγ. Compounds that bind to Gβγ can be identified by screening for completion for the binding of SIGK-peptide to Gβγ. The structure of the SIGK-Gβγ complex has been solved and SIGK is known to bind to a site that overlaps with many Gβγ targets. Given the relative inflexibility of Gβγ it is likely that compounds that inhibit SIGK binding do so by directly competing for SIGK binding, thus some idea of the compound binding site is given a priori. For example, compounds of the M119 family (Figure 2D) were identified in a small scale screen of the NCI diversity set for competition for SIGK binding to Gβγ in an ELISA format.

2. Secondary assays for disruption of βγ-dependent effector regulation. Compounds identified in the primary screen with low μM IC50’s in the primary assay are chosen and tested in vitro for effects on Gβγ-dependent target regulation. For example, compounds can be tested for their ability to inhibit purified Gβγ-dependent activation of purified PLCβ enzymatic activity. If compounds inhibit the Gβγ-dependent component of these types of assays, without inhibiting intrinsic basal PLC (or other effector) activity it is good evidence that the compounds bind directly to Gβγ. Specificity for different Gβγ targets can be examined by comparison of compound potency and efficacy in a variety of in vitro assays or in cell based assays of Gβγ signaling as discussed below. For example, some key effectors that have been tested in vitro include PLCβ, adenylyl cyclase, phosphoinositide 3-kinase and GRK2.

3. Assays for effectiveness in cells. For small molecules that inhibit Gβγ-dependent PLCβ regulation a standard cell based assay for inhibition of this pathway is formyl-Met-Leu-Phe (fMLP)-dependent regulation of Ca²⁺ release in differentiated HL60-neutrophil like cells. fMLP signaling is representative of chemoattractant and chemokine receptor signaling that relies largely on Gβγ released from Gi heterotrimers to regulate multiple signaling pathways in neutrophils including PLC, PI3K and pREX activation. For high throughput analysis of compounds, cell based assays could be conducted prior to in vitro assays since a single cell based experiment can give a readout of multiple signaling pathways and identifies molecules that are cell permeable.

Gβγ-binding compounds have been identified that either inhibit or potentiate Gβγ signaling. The M119 class inhibits interactions with downstream targets. Conceptually, the nature of inhibitory interactions can be readily explained as direct competition for binding interactions. There is no evidence for allosteric interactions mediating the effects of the compounds. Gβγ is relatively inflexible, but NMR spectroscopy analysis of Gβγ in the presence and absence of protein or peptide binding partners reveals both small and large allosteric changes [42]. Analysis of small molecule binding using this same NMR-based approach revealed no allosteric alterations with binding of the M119 or gallein. A selective reduction in NMR peak intensity for a single position was found upon binding of one compound from another class (M201), indicating a change in local dynamics, but again no conformational alterations were detected. Thus it is likely that compounds that inhibit interactions with downstream targets do so by directly interfering with crucial contacts between Gβγ subunits and effectors.

How compounds potentiate Gβγ signaling is conceptually more complex. Several scenarios can be imagined. For example, compounds may interfere with Gα-βγ subunit interactions and release free βγ that can signal downstream without nucleotide exchange. Such a scenario would require that the compound interactions with Gβγ that drive subunit dissociation do not interfere with Gβγ effector interactions. This mechanism was identified for the peptide SIRK, which binds to Gβγ at the α/βγ interface and drives ERK activation in cells [43, 44]. Such compounds would not necessarily be expected to enhance Gβγ-dependent effector activation in vitro, and would not be expected to inhibit the particular target of Gβγ whose activation is enhanced in cells. Alternatively, compounds binding at the Gβγ-effector interface compounds could increase the affinity of Gβγ for the target molecule by simply acting as a “molecular glue”. Such a scenario is testable in vitro by monitoring either the apparent affinity or direct binding of targets to Gβγ. Finally, compounds could alter the dynamics or local conformation of the βγ-effector binding site in a way that would allow for more efficient effector activation.

Receptor specific biased Gβγ inhibition

An exciting recent study has identified a novel mechanism for selective inhibition of Gβγ signaling downstream of specific receptors [45]. In a screen for antagonists for the receptor for 5-oxo-6E,8Z,11Z,14Z-eicosatetraenoic acid (5-oxo-ETE) receptor (OXE-R), a compound was identified that inhibited Gβγ signaling without influencing Gα subunit signaling. Surprisingly this compound inhibits OXE-R dependent signaling pathways without affecting Gβγ signaling by other Gi coupled receptors, indicating that the compound binds to the OXE-R to selectively block Gβγ signaling. This is difficult to explain based on current models of Gβγ activation, where Gα activation and dissociation is thought to be the mechanism for Gβγ activation, and implies a more active role of specific receptor-Gβγ interactions in regulation of activation of Gβγ pathways. The existence of such a mechanism implies that other specific receptor ligands can be developed that bias Gα vs. Gβγ pathways.

Small molecule G protein inhibitors in the treatment of cancer

There is considerable evidence for the utility of small molecule G protein α and βγ subunit inhibitors as potential therapeutic agents in a number of scenarios including heart failure, inflammation and pain (Table I) that have been discussed in other reviews[5, 46]. An interesting potential application of G protein inhibitors is in the treatment of various cancers that are under the control of a complex array of GPCR ligands regulating multiple steps in the development of primary tumors and metastasis[47, 48]. Because there are several reviews on the role of GPCRs and G proteins in cancer, I will briefly discuss key conceptual points that are relevant to considering G proteins as therapeutic targets in cancer, and consider the empirical data supporting the utility of small molecule G protein inhibitors in cancer treatment.

GPCRs in cancer

G protein signaling has an underappreciated role in contributing to cancer processes. Many GPCRs are overexpressed in various cancer cell types where they respond to autocrine and paracrine signals and regulate tumor growth and metastasis[47]. For example, many inflammatory mediators released in the tumor microenvironment such as IL-8 and prostaglandin E2 (PGE2) bind to overexpressed GPCRs on tumor cells to regulate tumor cell growth and survival [49, 50]. Thrombin-dependent activation of protease-activated receptors (PARs) promotes growth and survival as well as release of matrix metalloproteases involved in tumor escape and cell migration[51]. PGE2 and lysophosphatidic acid (LPA) contribute to angiogenesis important for oxygen delivery to growing tumor[52]. Chemokines secreted by stromal cells and metastatic target tissues are crucial for migration and invasion processes associated with metastatic cancer[53].

In addition to roles in responding to the local growth environment, a recent high throughput analysis for somatic mutations in human cancers revealed a significant prevalence of mutations in GPCRs including adhesion GPCRs, chemokine receptors and metabotropic glutamate receptors [54]. As examples, the adhesion family receptor brain specific angiogenesis inhibitor 1 (BAI1), and type 1 and 8 metabotropic glutamate receptors were found frequently mutated in lung squamous and adenocarcinomas. Many of these mutations are in the coding regions of these genes suggesting they may have functional effects leading to tumor initiation.

In each cancer type a different complex array of GPCRs and GPCR ligands are required at multiple levels including: tumor initiation, tumor cell growth, survival, escape migration and invasion. A strategy that targets individual GPCRs in some cases may inhibit cancer progression, but a broad-based strategy that targets common factors downstream of GPCRs may be more effective at multiple stages of cancer development.

G proteins in cancer

Mutations in G protein α subunits drive particular cancers. Gα_s mutations have been found to be associated with endocrine adenomas [55]. The first report described an activating mutation of Gα_s at R201 to either C or H[56]. Both of these mutations inhibit the GTP hydrolytic activity of the G protein α_s subunit resulting a constitutively GTP bound G protein subunit and constitutive cAMP production. More recently, the same high throughput analysis cited above for GPCRs, found Gα_s R201 mutations to be prevalent in pancreatic cancers (12%), and Gα_s upregulated in 12% of ovarian cancers and a significant percentage of breast cancers [54]. Additionally, a functional mutation in Gα_o at R243H was found at high prevelance found in breast cancers (5–15%) [54]. The Gα_o at R243H mutation was analyzed functionally and shown to be constitutively active due to a high basal ability to exchange GDP for GTP [57]. Compounds such as suramin or BIM-46174 would be expected to inhibit cancers driven by exchange activated Gα mutants such as Gα_o R243H but not GTPase deficient mutants such as Gα_s R201.

In another recent study, somatic mutations in Gα_q or Gα₁₁ were found to be prevalent in certain types of melanomas [58, 59]. Ocular uveal melanomas in particular have a high prevelance (83%) of either Gα_q or Gα₁₁ mutations. Both Q209L and R183C mutations were identified in either Gα_q or Gα₁₁ with the Q209L mutation being more prevalent. As discussed above, YM-254890 inhibits signaling by R183C but not Q209L, indicating that it would be useful for treatment for tumors containing the R183C mutation.

RGS proteins

One could easily envision that if GPCRs can turn on G proteins to regulate various aspects of cancer, then other proteins that regulate G protein activity could also be involved in cancer. Indeed, alterations in levels of RGS proteins have been shown to be associated with various cancer states[60]. RGS2, for example is downregulated in androgen-resistant prostate cancer[61] and in acute myelogenous leukemia[62]. Downregulation of RGS2 would be expected to enhance signaling through Gq dependent GPCR signaling pathways. RGS4 is downregulated in human breast cancer tissue, and overexpression of RGS4 inhibits chemokine receptor-dependent cell migration and invasion[63]. RGS4 downregulation would be expected to enhance signaling through Gα_q, Gα_i and Gβγ-dependent signaling pathways. RGS protein single nucleotide polymorphisms (SNPs) are associated with lung and bladder cancers [60, 64].

In many cancers, dysregulation of GPCR signaling occurs with upregulation of multiple GPCRs and GPCR ligands, which may be combined with alterations in RGS protein levels as well. Additionally, as discussed above, GPCRs signaling is involved at multiple levels of the cancer process including proliferation, cell survival, tumor escape, cell migration, invasion and tissue homing. In this complex scenario, molecules that target common G proteins, downstream of multiple GPCRs/RGS proteins, would likely be more efficacious than targeting single GPCRs. Cancers caused by mutations in the G protein α subunits are rarer, but again, targeting the mutant G protein itself could be efficacious.

Evidence for efficacy of small molecule G protein-targeting in cancer

The pan Gα GDP-release inhibitor BIM-46174, in the low to mid μM range, inhibits the growth of multiple types of human cancer cell lines and drug-resistant cancer cell lines in culture[35]. This effect seems to be through a compound-dependent enhancement of apoptosis rather than through altering cell cycle and mitosis, indicating that GPCR ligands in the culture medium are enhancing survival. BIM-46174 also inhibits invasion of cancer cells in response to Wnt3a and neurotensin. In vivo intraperitoneal (IP) injection of BIM-46174 partially inhibited growth of human lung and pancreatic tumor xenografts in nude mice, and was strongly synergistic with the common chemotherapy drug cisplatin.

Gα_q, Gα₁₂ and Gα₁₃ have all been shown to promote tumorigenesis, however; BIM-45174 does not inhibit migration or invasion of cancer cells driven by overexpression of constitutively active Gα₁₂, Gα₁₃ or free Gβγ. This supports the notion that the compound acts by inhibiting GPCR-dependent nucleotide exchange on Gα subunits, not by directly inhibiting downstream signaling by Gα. Inhibition of nucleotide exchange on Gα subunits blocks Gβγ signaling by preventing Gβγ release from the Gαβγ heterotrimer. Thus, inhibition of Gβγ signaling could by a mechanism by which Gα inhibitors block some aspects of cancer cell function.

Considerable evidence indicates that Gβγ-signaling plays a role in cancer processes. The pan βγ inhibitor, GRK2ct (βARKct) inhibits growth of prostate cancer cells by virtue of its ability to inhibit Gβγ-dependent activation of ERK[65]. Small molecule Gβγ-binding could, in theory, act similarly but βγ binding compounds identified thus far do not influence GPCR- and Gβγ-dependent ERK activation in cells[21]. However, as new molecules are discovered and characterized, it seems likely that inhibitors of Gβγ-dependent ERK activation will be identified.

Gβγ release from heterotrimeric G proteins initiates key signaling events downstream of chemokine receptors, which are Gi-coupled receptors involved in multiple aspects of cancer. These receptors regulate migration and invasion processes underlying metastasis [53], but are also involved in tumor cell survival and proliferation[47]. One study examined effects of the M119 class of Gβγ inhibitors on migration and invasion of an MDA-MB-231 breast cancer cell line in response to a mixture of ligands in NIH3T3 cell culture medium[66]. M119K (Figure 2D) (at low μM concentrations) strongly inhibited cell migration and invasion. Close examination of the cells showed that M119K inhibited Rac-dependent lamellipodia formation, consistent with the ability of the M119 class of compounds to inhibit Gβγ-dependent activation of the Rac exchange factor pREX [41, 67]. The compound also strongly inhibited migration and lamellipodia formation in response to the CXCR4 chemokine receptor ligand CXCL12 (SDF1α), but only weakly inhibited epidermal growth factor (EGF)-stimulated migration, demonstrating specificity for GPCR- dependent signaling pathways. A separate study examining the role of Gβγ in breast cancer cell growth showed that M119 significantly inhibited the growth rate of MDA-MB-231 cells [68]. This study went on to show that inhibition of Gβγ signaling by expression of Gβγ sequestering proteins inhibited tumor growth and lung metastasis in mice, but the effects of Gβγ inhibitory compounds were not examined in this in vivo setting. Although multiple studies have examined the efficacy of Gβγ inhibitory compounds in other therapeutic models in vivo (Table 1), there are as yet no studies examining effects of the M119 class or other βγ inhibitors in in vivo cancer models.

Table 1. In addition to cancer Gβγ is a potential target for therapeutics in multiple conditions.

Experiments with knockouts of Gβγ targets, protein based inhibitors of Gβγ including the c-terminus of GPCR kinase 2 (GRK2), Gβγ-binding peptides, and small molecules implicate Gβγ as a therapeutic target in multiple conditions.

Therapeutic Condition	Knockout, GRK2ct or peptide data	M119/Gallein
Inflammation	Gβγ plays a critical role in leukocyte migration in response to chemoattractants [70]	Gallein blocks acute inflammatory responses [41]
Pain	Gβγ-dependent regulation of PLC may inhibit opioid signaling. Blockade of this pathway could increase the potency of μOR agonists [71]	M119 and gallein potentiate the analgesic actions of morphine and limit tolerance and dependence [21, 80]
Heart Failure	GRK2ct prevents overload hypertrophy [72–74]	M119 and gallein protect mice from heart failure in a chronic adrenergic receptor stimulation model and a calsequestrin model [81]
Hypertension	Vascular GRK2 overexpression causes HTN, GRK2ct reverses [75]	Not tested
Drug Addiction	GRK2ct and peptide based Gβγ inhibitors inhibit synergistic cooperativity of addictive drugs [76, 77]	Not tested
Arterial restenosis	Adenoviral delivery of GRK2ct prevents restenosis in animal models [79].	Not tested

Specificity issues with G protein inhibitors and cancer

Current treatments for cancer are notoriously non-specific to the point of being toxic. Newer treatments such as Gleevec have relatively high specificity but only work on a subset of cancers. If single GPCRs could be found that are the major driving factors in certain cancers, they would be ideal high-specificity anticancer targets to exploit in an emerging era of personalized medicine. In this review we have discussed how various cancers may be controlled by a wide array of factors such that a more efficacious approach is one that has broad specificity. In the case of targeting Gα subunits there is of course the danger that inhibiting individual Gα subunit pathways will have a range of effects beyond the cancer cells that are targeted but this consideration would clearly be outweighed by the benefits of potential controlling or eliminating cancers.

Similar considerations exist for Gβγ subunit signaling where Gβγ is central to the functions of all GPCRs. Several lines of evidence indicate that the Gβγ inhibition strategy is far more specific than might be presumed. One key to the specificity of this approach is that the small molecule Gβγ inhibitors that have been characterized do not block all Gβγ functions. For example, Gβγ inhibitors do not block activation of Gα subunits by GPCRs despite a requirement for Gβγ in this process. Another key to specificity is that most GPCRs do not signal downstream via Gβγ, rather GαGTP is a primary mediator in most cases. Finally, Gβγ inhibitors only block a subset of downstream targets, so for the relatively limited subset of responses mediated by Gβγ in select tissues, only some of those responses are inhibited by the Gβγ inhibitors identified thus far.

Concluding remarks

In this review I discussed the progress in small molecule G protein inhibitor development and the potential for G protein subunit inhibitors in the treatment of complex diseases such as cancer. Pan-Gα inhibitors may be well suited for cancer treatment, but for other therapies, selective inhibition of Gα pathways may be required. The identification of a selective Gq class of inhibitor indicates that Gα subunit subtype inhibitor development is possible, and the structural data indicate a way forward for development of novel Gα subunit subtype selective inhibitors. Gβγ inhibitors have more selectivity than pan Gα subunit inhibitors because Gβγ signaling is most often downstream of Gi-coupled receptors, and the Gβγ inhibitors discovered thus far only inhibit some signals downstream of Gβγ [5, 69]. Nevertheless, Gβγ subtype-selective inhibitors would be highly desirable. The structural basis for development of Gβγ subtype-selective inhibitors is not clear but such molecules might be discovered as novel Gβγ inhibitors emerge.

Overall, these types of data provide support for the emerging idea that GPCR signaling can be a major driver of tumor growth and metastases. Targeting specific GPCRs may represent novel avenues for treatment of cancer but targeting pathways downstream of receptors may also be effective and in some cases more effective in treating certain types of cancer.