Ga_q proteins: molecular pharmacology and therapeutic potential

Abstract

Seven transmembrane G protein-coupled receptors (GPCRs) have gained much interest in recent years as it is the largest class among cell surface receptors. G proteins lie in the heart of GPCRs signalling and therefore can be therapeutically targeted to overcome complexities in GPCR responses and signalling. G proteins are classified into four families (G_i, G_s, G_12/13 and G_q); G_q is further subdivided into four classes. Among them G_αq and G_αq/11 isoforms are most crucial and ubiquitously expressed; these isoforms are almost 88% similar at their amino acid sequence but may exhibit functional divergences. However, uncertainties often arise about G_αq and G_αq/11 inhibitors, these G proteins might also have suitability to the invention of novel-specific inhibitors for each isoforms. YM-254890 and UBO-QIC are discovered as potent inhibitors of G_αq functions and also investigated in thrombin protease-activated receptor (PAR)-1 inhibitors and platelet aggregation inhibition. The most likely G protein involved in PAR-1 stimulates responses is one of the G_αq family isoforms. In this review, we highlight the molecular structures and pharmacological responses of G_αq family which may reflect the biochemical and molecular role of G_αq and G_αq/11. The advanced understanding of G_αq and G_αq/11 role in GPCR signalling may shed light on our understanding on cell biology, cellular physiology and pathophysiology and also lead to the development of novel therapeutic agents for a number of diseases.

Introduction

Seven transmembrane G protein-coupled receptors (GPCRs) are the largest and most important class of cell surface receptors numbering in the hundreds [1–4]. There has been an enormous amount of work done on GPCRs and numerous processes and pathways have been discovered. These processes include aspects of biased signalling, action of receptor-directed antibodies, allosteric modulators, GPCR-heterodimer targeted compounds and transactivation of protein tyrosine kinase and more recently protein serine/threonine kinase receptors [4–10]. However, progress in exploiting this knowledge about GPCRs has not been translated into the pipeline of new therapeutic agents that might have been envisaged [11]. Thus, new areas of investigation in the area of GPCRs and GPCR signalling are required.

The major primary and proximal effectors of GPCR signalling are the guanosine triphosphate (GTP)-binding, GTPase active, G proteins. G proteins are a point of convergence in GPCR signalling and as such provide an opportunity to overcome some of the complexities that have been identified in therapeutically exploiting upstream GPCR responses and signalling [12–15]. G proteins are classified into four families according to their α subunits: G_i, G_s, G_12/13 and G_q. The G_s and G_i families regulate adenylyl cyclase activity, while G_q activates phospholipase C (PLC) β and G_12/13 can activate small GTPase families [16]. The G_q family consists of four members—G_q, G₁₁, G₁₄ and G_15/16 [17, 18] and their respective α subunits are thus Gα_q, Gα₁₁, Gα₁₄ and Gα_15/16. Two of these isoforms, Gα_q and Gα_q/11, share 88% similarity at the amino acid sequence, and are the most prominent and ubiquitously expressed proteins in this family and are the topic of this review [19].

The Gα_q family of G proteins which includes the Gα_q and Gα_q/11 isoforms are especially significant due to the vast range of cellular responses that are mediated by these protein effectors. In their customary role, Gα_q proteins, activated by the upstream GPCRs, trigger PLC activity, which in turn hydrolyses membrane-bound phosphatidylinositol-4,5-bisphosphate (PIP2) to the generic diacylglycerol (DAG) which activates protein kinase C and inositol 1,4,5 trisphosphate (IP₃) which in turn initiates the release of calcium ions from the sarcoplasmic reticulum. Though Gα_q proteins play a critical role in GPCR-mediated signal transduction pathway, the role of heterotrimeric G proteins in GPCR signalling is discernibly the most under researched molecules in modern cell biology [20, 21].

There are some areas of science in which for various valid, but sometimes pragmatic or even investigator-determined reasons, terminology confuses non-specialist and even sometimes specialist readers. For example, in the intensely studied area of transforming growth factor-β (TGF-β) signalling, the immediate downstream target of the TGF-β type 1 receptor (TGFβR1) is the Smad transcription factor and its carboxy terminal phosphorylation (phosphoSmadC); however, the phenomena of indirect kinase-mediated phosphorylation of Smads in the linker region (phosphoSmadL) acquired the name, obviously confusingly, of “non-Smad signalling” or “Smad-independent signalling” [22–25]. For the Gα_q family of G proteins the members are Gα_q and Gα_q/11—precisely the former member would be Gα_q/q but this is truncated to simply Gα_q. In some cases, the protein Gα_q/11 is referred to as Gα₁₁, a term which does not identify the protein as a Gα_q family member [26]. The terminology, thus, creates serious opacity and compromises understanding in this area. This issue is further exemplified by the use of the phrase or the definition of a compound having activity as a selective Gα_q inhibitor when that does not allow for the existence of the isoforms, Gα_q and Gα_q/11, where we speculate in this review that these two proteins might have different functions and might also be amenable to the discovery of inhibitors which are selective for each isoform [27].

A major area of confusion, which flows on from the lack of clarity in defining the two isoforms, is the role of inhibitors. We have previously reflected on the short-comings and confusion surrounding the pharmacological tools in this area [25]. Two natural products, YM-254890 from the bacteria Chromobacterium sp. QS3666 and UBO-QIC also known as FR300359 from the leaves of the plant, Ardisia crenata sims are very potent inhibitors of Gα_q functions and have been investigated in the therapeutic context of inhibitors of the thrombin protease-activated receptor (PAR)-1 and the inhibition of platelet aggregation [21, 25, 28–30]. The availability of these compounds has been severely restricted; however, various groups [27], including our own [31], are proceeding along the pathway to the synthesis of these compounds which will almost certainly provide a base for experiments. These compounds will define more broadly the roles of the individual Gα_q isoforms in GPCR signalling and will help us to understand the cell biology of these two isoforms.

In our recent studies of the role of PAR-1 in the stimulation of the expression of the genes which encode the enzymes that mediate the elongation of chondroitin/dermatan sulphate glycosaminoglycan (GAG) chains on proteoglycans, we have found that all of the signalling goes via the transactivation of protein tyrosine kinases (which are responsible for the generation of phosphoErk) and protein serine/threonine kinase receptors (which generate phosphoSmad2C) [32–34]. The most likely G protein involved in these responses is one of the Gα_q family isoforms; however, there is little evidence for the involvement of calcium ions or protein kinase C in these responses which would reflect the traditional PLC mediated pathway [35]. We have not yet defined the role of G proteins in these responses but it is interesting to note that the response does not appear to arise from typical Gα_q signalling raising the possibility that there may be novel pathways involving and downstream of Gα_q.

This review sets out the current knowledge of the Gα_q family of G proteins including the molecular structures and pharmacological responses on the basis that new experimental data will shortly become available. We assume that these data will be helpful to define the biochemical and/or molecular role of Gα_q and Gα_q/11 and perhaps will help us to answer some of the very interesting questions such as the role of G protein flipping in GPCR biased signalling and GPCR transactivation-dependent cell signalling [9].

Role of G proteins and Gα_q proteins specifically in GPCR signalling

The proximal effectors of GPCR signalling are the G proteins, so-called because of their ability to bind GTP and guanosine diphosphate (GDP). There are monomeric G proteins such as Ras (downstream of GPCRs) but the G proteins which are involved in GPCR proximal signalling are part of heterotrimeric complexes involving a poly functional α subunit and a supporting, sometimes functional, βγ dimer [36–39]. The α and γ subunits have membrane binding domains which serve as lipid anchors in the cell membrane (Fig. 1). When the GPCR is not activated, in the absence of a cognate ligand, the G protein α subunit is bound to GDP and the whole complex of GPCR and GDP bound G α subunit, is quiescent. Upon ligand engagement with the GPCR, the receptor changes conformation leading to changes in the G protein α subunit and exchange of the bound GDP for GTP. The heterotrimeric G protein complex simultaneously dissociates into the GTP-bound α subunit and the independent βγ subunit. The GTP-bound Gα subunit remains anchored to the cell membrane but it is now capable of two-dimensional lateral movement in the membrane, away for the GPCR, where it can play multiple functional roles and continue the signal transduction. Thus, it is this change in conformation, modified and activated the GTP-bound and membrane-anchored G protein which is the effector of integrating and promulgating the original GPCR signal (Fig. 1). In the present context activated GTP-bound Gα_q family proteins bind and activate PLC with the well-known distal cascade arising from the hydrolysis of PIP2 in the membrane, leading to activation of two signalling pathways and ultimately changes in gene expression and, to the extent that the gene expression profile defines phenotype, modifications in cellular phenotype. Whereas, in contrast to kinase receptors [40], GPCRs do not possess enzymatic activity, Gα subunits possess GTPase enzymatic activity (see domain analysis in Fig. 2). The GTPase activity of G proteins catalyses the hydrolysis of the GTP to GDP and reverses the activation state of the Gα subunit which can then restore its relationship with an inactive GPCR (Fig. 1). Although a considerable amount of work has led to the description of this elegant biochemical process, there is much which is unknown such as the range of partners that can bind to activated Gα_q subunit, the pathways that can be initiated and whether or not there are distinct pathways associated with Gα_q and Gα_q/11 isoforms. All of this new information will be revealed from studies with YM-254890 and UBO-QIC and their derivatives or surrogates and it is expected that these studies will shed some major light in the area of GPCR signalling. For noting, compound YM-254890 was intermittently available a decade or so ago from Yamaguichi Pharmaceuticals which, following a corporate takeover, subsequently became the property of Astellas Pharmaceuticals. YM-254890 has only recently become available commercially from a generic biotech company. UBO-QIC was made commercially available for a short time a couple of years ago; however, at the present time, it is not available commercially and its only supply is from the publishing source [21].

Fig. 1.

G protein coupled receptor-mediated activation of signal transduction pathway: Membrane-bound GPCR is associated with Gα, Gβ and Gγ proteins where Gα protein is bound with GDP which is functionally inactive form. In the presence of extracellular ligand, GDP is replaced by GTP and both Gα and Gβγ protein heterodimer relocate along the membrane to associate with the specific ion channels and activate intracellular signalling pathways. In the presence of modulator, the Gα activates the GTPase activity and hydrolyses the GTP and the pocket is occupied by the GDP which is coupled with the release of ligand from GPCR

Fig. 2.

Gα_q (GNAQ) and Gα_q/11(GNA11) share very high sequence homology (upper panel) and functionally active domain structures (lower panel). Human Gα_q and Gα_q/11 proteins were compared to find sequence homology. The dark shed region represents the highest homology. Sequences with absolutely no similarity are not shaded and AA belongs to the same group are marked with light shade. The GTP-binding domain (Site 1), adenylyl cyclase interaction site (Site 2) and transmembrane receptor binding site (Site 3) are very conserved. AA sequence of both proteins were analysed by NCBI CDD (conserved domain database) to find out functionally homologous domain. Diagram of both proteins were generated by arbitrarily marking the possible functionally active sites (lower panel). Both proteins have GTP/Mg⁺² ion binding site, GTPase activity domain and transmembrane domain at the identical location

Functional analysis of Gα_q and Gα_q/11

Insofar as the functions of Gα_q and Gα_q/11 are concerned, having identical numbers of amino acids and a very high amino acid (AA) sequence homology would suggest that the proteins have similar if not identical functions. However, consideration of AA sequence and analysis of the functional domain homology indicated that, besides their known role as activators of PLC, they might actually have additional and distinct biochemical functions. The primary structures of Gα_q and Gα_q/11 are comprised of several α-helices and β-strands. Conserved domain database analysis located a GTP/Mg₂ ⁺ binding site, around the first 50 AA, close to the N-terminal end. Besides this GTP/Mg²⁺ binding site, both proteins have a conserved adenylyl cyclase interaction site, which is located almost at the central part of the protein and a C-terminal end transmembrane receptor binding site (Fig. 2, upper panel). The adenylyl cyclase site on Gα_q is located close to 180AA from the N-terminal end of both proteins and has around 60% sequence homology with the cognate site in the much larger Gα_s protein (1024AA) which is traditionally associated with adenylate cyclase signalling [41]. Structural analysis also indicated a switch region, which is located almost at the central part of the proteins and this region is responsible for changing the conformation of these proteins depending upon their binding to GTP or GDP (Fig. 2, lower panel). The inherent GTPase activity of Gα_q and Gα_q/11 has been established almost a decade ago; however, the domain, which is responsible for this activity, is not very well defined. Our conserved functional domain search indicated that the GTPase activity domain of Gα_q is similar to both the RAS as well as ARF-like superfamily of small GTPases, which includes Ras, Rho, Rab, and Sar1/Arf families of proteins. The domain is located between 178-282AA where several β-strands are surrounded by an α-helix which is the predicted structure of the enzymatic activity. On the other hand, in the case of Gα_q/11 the predicted site, associated with the GTPase activity, is located between 184-278AA and it is more similar to the ARF-like but not the RAS-like superfamily family of small GTPases. Therefore, it can be speculated that these two proteins are perhaps linked to different signal transduction pathways to activate specific intracellular signalling cascades. Analysis of a human gene database also showed that both proteins are linked to the distinct signal transduction pathways. For example, Gα_q is associated with thyroid stimulating hormone and serotonin receptor 2 and ELK-SRF/GATA4 signalling pathway where as Gα_q/11 is not.

We used pathway analysis (an online network analysis programme) to identify the biological network they are involved. In this analysis, Gα_q is well connected to the transcription factor SP1, SP1-SMAD2 and SP1-Activate Protein C driven signalling pathways. Surprisingly, our analysis did not identify any network where Gα_q/11 is involved (Fig. 3). As mentioned above, the limited availability to specific tools (pharmacological inhibitors) to address the possibility that Gα_q and Gα_q/11 having specific functions it is only speculation that Gα_q and Gα_q/11 might have different functions but there is sufficient information to make this a proposal worthy of further investigation.

Fig. 3.

Pathway analysis for Gα_q and Gα_q/11. Both Gα_q and Gα_q/11 proteins were placed separately in PathwayNet analysis programme (an online network analysis programme developed by Princeton University) to identify the molecular network they belong to based on the existing information. No network association was identified so far where Gα_q/11 is involved. These figures are reproduced from our analysis in PathwayNet

The speculation, that Gα_q and Gα_q/11 might have distinct functional roles, is also supported by an examination of the ribbon structures of the two proteins (Fig. 4). Ribbon diagrams are three-dimensional representations of protein structure as described above in terms of α-helices shown as coiled ribbons or β-strands shown as thick tubes. Ribbon diagrams show a visualisation of the twists, folds and unfolded areas of proteins and in the current context, they are used to show that the two isoforms of Gα_q proteins, Gα_q and Gα_q/11 actually have discernibly different three-dimensional structures. For example, when we compared Fig. 4a of both proteins, we found a group of β-sheet is surrounded by two α-helixes in Gα_q. On the other hand, those β-sheets are closer to one of the α-helixes in Gα_q/11. Thus, the derived ribbon structures show a visual depiction of the structural features described above and clearly show differences empirically capable of yielding functional differences between Gα_q and Gα_q/11.

Fig. 4.

3D protein structure and functional domain analysis of Gα_q (GNAQ) and Gα_q/11(GNA11). The 3D ribbon structures of both proteins were derived based on ModBase, a database of comparative protein structure model analysis. A few of the all possible models with similar alignment were picked up to compare the 3D structure. The protein structures are reproduced from ModBase [69]

Tissue distribution of Gα_q and Gα_q/11

Tissue-specific gene (and protein) expression studies are very important to understand the biology of any target peptide or protein, and they are also useful for drug development and biomarker discovery. Several new databases were developed recently generated to correlate gene expression and a variety of human diseases [42, 43]. We accessed one of such databases (Gene Cards^® Human Gene Database) to investigate the tissue-specific distributions of Gα_q and Gα_q/11. Both Gα_q and Gα_q/11 proteins are expressed in the majority of the human tissues and organs such as heart, smooth muscle, neurons, brain, pancreas, thyroid, colon and uterus (Fig. 5). Interestingly, expression was not observed in a few organs such as breast, stomach and spleen indicating that these two proteins may have common as well as other functions specific to particular tissues. The observed similar organ and tissue-specific expression patterns also suggest that expression of both genes are regulated by a common upstream regulatory mechanism, which is, absent or inactivated in tissues or organs where their expression is restricted. However, in the absence of functional data it is not possible to speculate further on the role of these two isoforms in all the tissues.

Fig. 5.

Gα_q family proteins—tissue distribution. Tissue distribution of Gα_q and Gα_q/11 is very similar. The tissue distribution profiles of both proteins across the major human tissues were adopted from the public data base Gene Cards [70]

Possible Gα_q pathways other than through PLC

We have recently discovered that the GPCR, the thrombin receptor, PAR-1, which was known to transactivate the protein tyrosine kinase receptor, epidermal growth factor receptor (EGFR) can also transactivate the protein serine/threonine kinase TGFβR1 [32, 33, 44]. Detailed investigations of the distal components of these signalling pathways and comparison with those reported for the GPCR transactivation of EGFR show that the distal mechanisms are completely distinct. For example, the PAR-1 transactivation of EGFR involves matrix metalloproteinases (MMPs) and the cell surface cleavage of a cognate ligand whereas the transactivation of TGFβR1 does not involve MMPs [44].

In the cardiovascular studies in our laboratory, the functional read-out is the process of thrombin-stimulated elongation of the chondroitin/dermatan sulphate GAG chains on the proteoglycan biglycan secreted by vascular smooth muscle cells (VSMC), as it relates to enhanced lipid binding which is associated with the early step in atherosclerosis [45–47]. GAG hyperelongation has been investigated as changes in the apparent size of the GAG chains and more recently we studied the regulation of gene expression for the enzymes which are rate limiting for the hyperelongation of GAG chains on biglycan [33, 44]. Our data indicate that PLC and its downstream pathways are not responsible for GAG hyperelongation. Thus, mimicking the increase in intracellular calcium with the calcium ionophore, ionomycin, does not stimulate GAG elongation [35]. Calcium channel blockers partially block growth factor-stimulated GAG hyperelongation but the mechanism of inhibition does not depend solely upon the blocking of calcium channels [35]. TGF-β, which is a very efficacious stimulator of GAG hyperelongation has no effect on cellular IP₃ levels [48]. When VSMCs were transfected with angiotensin (AII) Type 1 receptors there was a huge increase in AII stimulated IP₃ production but no effect on proteoglycan synthesis (Little, Thomas and Ivey, unpublished observations). In accordance with these unpublished findings, Yang et al. [49] studying Gα_q coupled glutamate receptor (mGluR) responses in cultured neurones and found that the stimulation of the phosphorylation of the mitogen-activated protein (MAP) kinase JNK, occurred via transactivation of the EGFR and did not involve the traditional PLC, PKC and IP3 pathways. These results suggested that the PLC pathway and the pathway to the transactivation of the EGFR are occurring in parallel and therefore implying that there is an unknown mechanism which links the role of the Gα_q to the transactivation of EGFR [49].

As previously mentioned, our conserved functional domain search indicated that the GTPase activity domain of Gα_q is similar to both the RAS as well as ARF like superfamily of small GTPases, which includes Ras, Rho, Rab, and Sar1/Arf families of proteins whereas for Gα_q/11 the GTP/GDP binding site is more similar to the ARF-like but not the RAS-like superfamily family of small GTPases. This information provides support for the existence of Gα_q/Gα_q/11 signalling pathways distinct from the PLC pathways although these pathways have been very little explored. New roles are merging for the β-arrestin molecule other than its action of heterotrimeric G protein desensitisation and the internalisation of GPCRs. Data reveal that β-arrestin molecules can act as signal transducers of Erk activation independent of G proteins [50, 51]. The increase in expression of β-arrestin1 allows for β-arrestin2 to regulate Erk activity in HEK cells expressing angiotensin type II receptor in [52, 53]. However, overexpression of β-arrestin1 leads to activation of small GTPase, RhoA [52]. Thus β-arrestin signalling can activate Gα_q mediated pathways independent of PLC.

Taken together, these data suggest that multiple pathways including the process of growth factor-stimulated GAG hyperelongation do not involve PLC and its downstream pathways. As responses through GPCRs must, by definition, involve G proteins, if PAR-1-mediated GAG hyperelongation includes a G protein, presumably a Gα_q family G protein, but that does not involve activation of PLC, then this would possibly be an alternative pathway(s) of PAR-1 signalling. Thus, as most signal transduction pathways involve protein:protein interactions, it would be interesting to investigate the potential signal transduction binding partners for the conformationally modified, activated, GTP bound and membrane anchored G protein which arises after the stimulation of PAR-1 in VSMCs.

Pharmacology of Gα_q inhibitors

UBO-QIC/FR300359

The plant derived compound, FR300359, also known as UBO-QIC, is like the better known as YM-254890, a cyclic depsipeptide [21, 25]. UBO-QIC is isolated from the Ardisia crenata sims plant [54], a plant of the primrose family with multiple common names including Christmas berry and Australian holly and it is native to East Asia but an invasive species in the south-eastern United States and elsewhere.

UBO-QIC inhibits platelet aggregation in rabbits and causes dose-related hypotension in anaesthetised normotensive rodents which is consistent with the observed effect on blood pressure in Gα_q knock down mice [55, 56]. The blood pressure lowering effect was attributed to the ability of UBO-QIC to partially mediate nitric oxide release from endothelial cells and inhibit calcium migration caused by voltage-dependent and receptor operated channels [57]. In HEK cells transfected with Transient receptor potential cation channel subfamily V member 4 (TRPV4), PAR-2 or G protein-coupled receptor 11, mediated intracellular calcium release was abolished by UBO-QIC when compared to control; however, extracellular calcium influx through the TRPV4 ion channel was unaffected thus showing that PAR-2 coupling to TRPV4 is not mediated by Gα_q signalling [30]. Neither of the above studies assessed the level of expression of Gα_q and Gα_q/11 nor were there other studies that addressed the specificity of action of UBO-QIC towards these isoforms.

Kostenis and her colleagues [21] have recently reported a major study on the characterisation of UBO-QIC. To understand G protein specificity of UBO-QIC, they used an experimental approach based on bioluminescence resonance energy transfer (BRET) that relies on co-transfection of GPCRs with a panel of Gα subunit sensors. These sensor based studies were proven to be a valuable tool to monitor the activation associated conformational changes within G proteins upon stimulation of GPCR. Upon pre-treatment with UBO-QIC (FR3 00359), it was observed that agonist BRET responses of Gα_q- and Gα₁₁-containing heterotrimers were completely prevented but those of all other Gα isoforms were not affected [21, 58–61]. Inhibition of Gα_q, Gα₁₁ and Gα₁₄ signalling by UBO-QIC was observed at micromolar concentrations whereas, signalling via other Gα_q family member, Gα₁₆ was completely unaffected in the presence of excess UBO-QIC. Thus, their results clearly showed that UBO-QIC is selective for the inhibition of Gα_q, Gα₁₁ and Gα₁₄. By using CRISPR-Cas9-mediated genome editing and rescue experiments, these authors [21] showed unambiguously that UBO-QIC selectively blocks the activity of Gα_q and Gα_q/11. This work was taken further in exploring the therapeutic potential of targeting G proteins in a model of oncogenesis involving the malignancy of melanoma cells. UBO-QIC blocked melanoma cell proliferation and migration and triggered differentiation [21] underpinning the need for further proof of concept studies on the targeting of G proteins, particularly, Gα_q family of proteins in disease models.

Very recently, Jacobson et al. [62] reported the effect of UBO-QIC on other subunits of G proteins, such as βγ subunits. Their studies showed that at lower concentration, UBO-QIC inhibited multiple examples of known Gα_q signalling pathways as well as other non-Gα_q-events such as A₁ adenosine and M₂ muscarinic acetylcholine production, indicating that its effect is not limited to Gαq. UBO-QIC completely inhibited signalling responses activated by Gq-coupled P2Y1 receptor and Gi-coupled A1A or P2Y12 receptor. UBO-QIC completely blocked Gq mediated Akt activity and only partially inhibited Gi mediated activity. This shows that the full selectively spectrum of UBO-QIC is still unknown and thus the need for a specific Gαq antagonist.

YM-254890 and related cyclic depsipeptides

The compound known as YM-254890, a cyclic depsipeptide isolated from the bacteria Chromobacterium sp. QS3666, has been used as a Gα_q family inhibitor. YM-254890 inhibits the signal transduction of Gα_q by inhibiting the exchange of GDP for GTP preventing the activation of the G protein, rather than GPCR-Gα_q interactions [63]. When bound to GDP, the non-polar side chains of YM-254890 form hydrogen bonds with the Switch I region of Gα_q, however, this is a conformation that cannot be maintained when bound with GTP [64]. At a molecular level YM derivatives appear to inhibit Gα_q by binding to a pocket in the Gα_q that stabilises the protein in an inactive conformation which prevents the activation of the Gα_q by a cognate GPCR [65].

YM-254890 inhibits platelet aggregation induced by ADP, a response that is mediated via GPCRs, P2Y₁ and P2Y12 receptors [29]. YM-254890 has no effect on a specific P2Y12 signal transduction pathway which is conventionally associated with the G_i signalling pathways, indicating that the compound has some specificity for Gα_q. It was also shown that YM-254890 inhibited Gα_q-coupled GPCR signalling by inhibiting calcium mobilisation in P2Y1-expressing C6-15 cells but not cAMP accumulation which is coupled to the P2Y12 receptor-mediated signal transduction pathway [63]. Aside from antiplatelet activity, by electrically inducing carotid artery thrombosis in rodents, YM-254890 was also shown to have antithrombotic and thrombolytic effects [28].

YM-254890 was discovered and developed by Yamaguichi Pharmaceuticals Japan; Yamaguichi was subsequently acquired by Astellas Pharmaceuticals, Japan. This compound was made available to researchers 10 years ago for a short period of time and a small number of interesting studies were published. The initial results indicated that YM-254890 is a useful tool for investigating Gα_q/11-coupled receptor signalling and the physiological roles of Gα_q/11. Gα_q knockout mice have blood pressure lower than the control [56]. Therefore, it is possible that high blood pressure in patients can be controlled by modulating the activity of Gα_q, This indicates the potential of YM-254890 as a Gα_q inhibitor to be an anti-hypertensive agent. There are very limited data available on the selectivity of YM-254890 on other G protein subunits; thus, more studies are needed to profile this inhibitor.

The peptide antagonist GP-2A

In 2004, Tanski et al. [66] discovered a competitive Gα_q inhibitor, G Protein antagonist- 2A, also known as GP-2A. GP-2A is a peptide that selectively inhibits the action of Gα_q by M1 muscarinic cholinergic receptors. GP-2A, as presented here, should be distinguished from the small chemical entity known as GP-2A which is a pyrazole derivative and a selective cannabinoid receptor CB2 agonist [67]. The effect of GP-2A on the signalling pathway of Gα_q and its role in cell proliferation with rat pulmonary artery smooth muscle cells have been investigated. Angiotensin II-mediated proliferation, PLCβ activation and Erk1/2 phosphorylation were inhibited by more than 50% in the presence of GP-2A [66]. The EGFR can be activated by EGF to generate an intracellular signalling pathway leading to the phosphorylation of several downstream effector proteins such as Erk1/2 [68]. Tanski et al. have evaluated angiotensin II, as a specific Gα_q agonist, to effectively reduce Erk1/2 activation mediated by PLCβ via Gα_q in the presence of GP-2A indicating its association with the phosphorylation of Erk1/2 in rat pulmonary artery smooth muscle cells [66]. GP-2A is commercially available; so further studies on its peptide and potentially other derivatives might provide more information on the role of Gα_q family members in diverse biological processes and diseases.

Future medicinal chemistry and pharmacological developments

Gα_q family of proteins has not been sufficiently studied to understand fully their role in cell signalling and cell biology. Numerous groups have been involved in the synthesis of the depsipeptides YM-254890 and UBO-QIC/FR300359, however limited data is available on the actions of these peptides. Our effort to synthesise YM-254890 produced structurally most of the molecule but this compound did not inhibit thrombin-stimulated increase of intracellular calcium, which is a response most likely to be mediated by Gα_q family proteins (Little and Brimble, unpublished data). The remaining residues between the molecules we reported and YM-254890 must provide the inhibitory interaction with the Gα_q protein. Other groups have taken alternative pathways to synthesise simplified form of YM-254890 derivatives [27] but the resultant molecules are at least tenfold less potent than YM-254890.

We have recently been able to successfully perform siRNA-(GE Dharmacon M-010960-02-0005 and M-008562-00-0005) based selective knock down of Gα_q and Gα_q/11 VSMCs (Little, Osman and Kamato, unpublished observations). This result will allow us to look for Gα_q- and Gα_q/11-specific responses. We are currently speculating, based on our preliminary data, that in some systems or some cells, UBO-QIC/FR300359 and YM-254890 might behave as Gα_q and Gα_q/11 selective inhibitors, respectively, or that YM-254890 is a selective Gα_q/11 inhibitor, but these speculations obviously require further experimental data. These speculations also relate to the medicinal chemistry in that the substituents absent from our recently reported molecule might not only provide the potent inhibitory activity towards G_q family proteins but also be a selective inhibitor of Gα_q or Gα_q/11.

Finally, due to the very high homology in AA sequences, the molecular approaches are undoubtedly required to define the role of Gα_q and Gα_q/11 isoforms in particular to the cellular responses. However, molecular approaches in this area, for example G protein knock down, can lead to rebound increases in other G proteins with unexpected results [13]. Ultimately the combined approaches of molecular interventions with pharmacological tools will be required to define the role of Gα_q family of proteins in cell biology and this approach would potentially identify a pathophysiological role and their specific inhibitors.

Conclusions

The two Gα_q family proteins, Gα_q and Gα_q/11, warrant intense investigation for their ability to provide extremely useful information on the understanding of the functioning of GPCRs and also as potential therapeutic targets in their own right. Although such as all-encompassing target may intuitively be too broad to provide the desired specificity in a therapeutic target and be limited by off-target effects, the outcome is simply not clear until the target is evaluated in several diseases. Furthermore, in a conceptual sense, targets may often be activated or hyperactivated in a particular disease state and it is this state that elevates the target to a therapeutic target over and above the role in normal physiology.

G proteins are most likely at the centre of the phenomena of biased signalling which has enormous potential in explaining and exploiting the therapeutic potential of many GPCRs. We consider it is likely that biased signalling arises, at least in one form, from the ability of different GPCR biased ligands to “flip” the proximal G protein partner from one G protein to another, for example from Gα_q to Gα_s and thus elicit a qualitatively different response. However, how far does such a paradigm extend? What of Gα_q and Gα_q/11—can a biased agonist flip the proximal signalling partner from one Gα_q isoform to another with different effects on phenotype or is this the limit of the paradigm of biased signalling? These questions will best be answered with the broad availability of the pharmacological tools specially YM-254890 and UBO-QIC and their derivatives. When more confirmatory information is available on the potency and the (Gα_q versus Gα_q/11) specificity of these important agents, then it can be established with the two depsipeptides coupled with siRNA studies whether or not the two isoforms are involved in biased signalling and indeed if there are independent signalling pathways arising from each of these two G proteins.

In summary, an advanced understanding of the role of Gα_q and Gα_q/11 in GPCR signalling will provide an enormous addition to our understanding of cell biology and potentially of cellular physiology and pathophysiology, and may lead to the discovery and development of new therapeutic agents capable of modulating diseases process in years that have not yet been envisaged.