Peter J. Little, Discipline of Pharmacy, School of Medical Sciences, RMIT University, Melbourne, Vic. 3083, Australia. E-mail: firstname.lastname@example.org
Objectives This review discusses the latest developments in G protein coupled receptor (GPCR) signalling related to the transactivation of cell surface protein kinase receptors and the therapeutic implications.
Key findings Multiple GPCRs have been known to transactivate protein tyrosine kinase receptors for almost two decades. More recently it has been discovered that GPCRs can also transactivate protein serine/threonine kinase receptors such as that for transforming growth factor (TGF)-β. Using the model of proteoglycan synthesis and glycosaminoglycan elongation in human vascular smooth muscle cells which is a component of an in vitro model of atherosclerosis, the dual tyrosine and serine/threonine kinase receptor transactivation pathways appear to account for all of the response to the agonists, endothelin and thrombin.
Summary The broadening of the paradigm of GPCR receptor transactivation explains the broad range of activities of these receptors and also the efficacy of GPCR antagonists in cardiovascular therapeutics. Deciphering the mechanisms of transactivation with the aim of identifying a common therapeutic target remains the next challenge.
Cell surface receptors and their signalling pathways are the major regulators of cell physiology and the largest class of drug targets.[1,2] The major classes of cell surface receptors are G-protein-coupled receptors (GPCRs) and kinase receptors that phosphorylate targets on tyrosine or serine/threonine residues.[4–6] Kinase receptors have an extensive range of docking sites for accessory molecules, which elicit signalling pathways to major functions such as cell migration and proliferation. Agonists and antagonists of receptors modify cellular behaviour and represent large groups of drugs. Modulators of intracellular kinases, especially inhibitors, modify signalling pathways and, for example, inhibit oncogenic pathways. Initial studies have described linear pathways from receptor through intracellular signalling pathways but further research has quickly revealed that signalling is much more complex. Pathways intersect to inhibit or enhance cellular responses providing exquisite complexity but also providing multiple opportunities to develop drugs for specific pathologies.
Seven transmembrane GPCRs are the largest group of cell surface receptors in mammalian systems, comprising approximately 1000 members.[9,10] These receptors have four extracellular domains and four intracellular domains, with seven membrane-spanning domains. As these receptors lack intrinsic enzymatic activity, they couple intracellularly to heterotrimeric G proteins, which are so named because they bind guanosine triphosphate (GTP). These proteins function as intracellular messengers, stimulating or inhibiting an intracellular signalling cascade. G proteins consist of three subunits, designated α, β and γ. The α subunit binds guanosine diphosphate (GDP) and the G protein is inactive. Stimulation of the GPCR causes the α subunit to exchange the bound GDP for GTP, leading to the dissociation of α subunit from the βγ subunits and both are then able to modulate effector enzymes or ion channels. G proteins are classified according to their effector molecules. Currently there are 20 α, six β and 11 γ subunits that have been cloned. The interaction between GPCRs and the large number of G-protein subunits can be highly diverse, as one GPCR is capable of interacting with more than one G protein, resulting in multifunctional signalling. GPCRs may also interact with molecules other than G proteins to turn a signal on or off. G-protein-receptor kinases (GRKs) phosphorylate activated GPCRs to which arrestin molecules can bind, causing desensitization.[11,15] Arrestin binding can also promote internalization of the receptor via a clathrin-mediated process.[16,17] The receptor may be either degraded or recycled to the cell surface via this process. Recently it has been shown by Lefkowitz and others that arrestins can also promote activation of signalling pathways by acting as ‘scaffolding’ to which other proteins may bind. Other binding proteins include mitogen activated protein kinase (MAPK) cascade components such as Raf-1 and Erk1/2, the non-receptor tyrosine kinases c-Src and Yes, and phosphodiesterases.[11,15] Agonists of GPCRs range from radiation (light) to cytokines to large polypeptide structures. The involvement of GPCR in cellular functions is very extensive and although many existing drugs target GPCRs and their linear signalling pathways the potential to exploit the more sophisticated responses (transactivation of multiple diverse cell surface protein kinases) elicited by these receptors for therapeutic purposes has only just begun.
Direct GPCR signalling pathways
Major signalling pathways downstream of GPCRs include the phospholipase Cβ/inositol phosphate system, the adenylate cyclase (AC) and cyclic adenosine monophosphate (cAMP) pathway, and Rho kinase and phosphoinositide 3-kinase (PI3-K) mediated pathways. These pathways have all been very well characterized and reviewed and are mentioned briefly here.
In the phospholipase Cβ/inositol phosphate pathway, GPCR activation causes dissociation of the Gα and Gβγ subunits, which are able to activate the membrane-bound enzyme phospholipase Cβ, leading to the formation of two second messengers, diacylglycerol (DAG), which activates protein kinase C (PKC), and inositol-(1,4,5)-triphosphate (IP3), which diffuses into the cytoplasm and causes a massive intracellular calcium release. In the AC/cAMP system the pathway is activated by the α subunit of the Gs protein or inactivated by the α subunit of the Gi protein. The membrane-bound inactive enzyme AC, once activated by Gαs, converts ATP to cyclic adenosine-3′,5′-monophosphate (cAMP). Increased levels of cAMP activate protein kinase A (PKA). This signalling cascade is terminated by the breakdown of cAMP by a family of phosphodiesterases that hydrolyse cAMP to AMP. Inhibitors of phosphodiesterases are a large class of drugs that enhance effects through this pathway. In the Rho kinase pathway, low molecular weight GTPases activate their own signalling pathways. Members of this class include Rac, Ras and Rho. Unlike classical G proteins, activation of the members of the Rho family is not through interaction with agonist-bound GPCRs. The exchange of GDP for GTP on these proteins is controlled through guanine nucleotide exchange factors (GEFs), which catalyse the exchange of GDP for GTP. The Rho kinase signalling pathway is used by GPCR agonists to activate a range of cellular responses. In the final direct GPCR signalling pathway, PI3-K phosphorylates inositol-containing lipids, such as phosphatidylinositol, and some phosphoinositides, on the 3′-OH position of the inositol ring. Targets of PI3-K include phosphatidylinositol, phosphatidylinositol-4-phosphate and phosphatidylinositol-4,5-diphosphate. Products of PI3-K are further converted by kinases and phosphatases, and are used in intracellular signalling pathways. Downstream mediators of the PI3-K lipid product PtdIns(3,4,5) triphosphate include Tec kinases such as Bruton's tyrosine kinase and induced T-cell kinase, GEFs and Pleckstrin homology domain kinase 1 (PDK-1). PDK-1 activation causes phosphorylation of Akt or protein kinase B, leading to inhibition of pro-apoptotic factors and inactivation of GSK-3, which allows glycogen synthase to function. Akt activates the mammalian target of rapamycin (mTOR), leading to an increase in protein synthesis and cell growth. Activation of PI3-K/Akt signalling mediates vascular smooth muscle cell (VSMC) proliferation, cell survival and migration, stimulated by insulin-like growth factor. Other growth factors such as PDGF and angiotensin II (Ang II) use this pathway to maintain the differentiated phenotype of VSMC and induce VSMC hypertrophy, respectively.
GPCR signalling via cell surface receptor transactivation
As well as the direct pathways described briefly above it has been known for a decade and a half that GPCRs can mediate the activation and downstream signalling of protein tyrosine kinase (PTK) receptors in a process known as receptor ‘transactivation’.[27,28] In 1996 Axel Ulrich and colleagues discovered that a G-protein-receptor coupled agonist such as Ang II could lead to the ‘transactivation’ of a protein tyrosine coupled receptor, specifically the epidermal growth factor (EGF) receptor, leading to phosphorylation of the EGF receptor and the generation of products such as phospho ERK1/2 (pErk1/2) that are downstream of the EGF receptor.[27,28] This is very important because GPCR agonists are not generally capable of generating a cell growth signal but the ability to transactivate a fully competent growth factor receptor such as the EGF receptor gives the GPCR the ability to generate a cell growth response. Since this original demonstration there have been many examples in which GPCR agonists, including Ang II, thrombin and endothelin, have been shown to be able to transactivate multiple PTK receptors, including PDGF, FGF and IGF-1.[29–34]
Protein tyrosine kinase receptor transactivation via GPCR activation is a pathway that greatly expands the repertoire of responses to GPCR agonists. Traditionally, activation of GPCR through ligand binding was thought to directly activate two G-protein-mediated pathways, the phospholipase C (PLC) pathway and the AC pathway, which were separate, compartmentalized and did not interact with other tyrosine kinase mediated pathways. However, recent evidence has demonstrated that G-protein-mediated pathways do cross over and interact with tyrosine kinase pathways, and that agonists of GPCR are able to stimulate signal transduction events that are associated with receptor tyrosine kinase mediated pathways, such as the activation of the JAK/STAT pathway and MAPK/Erk1/2 pathways. GPCRs hijack tyrosine kinase signalling machinery by cross-talking with tyrosine kinase receptors, such as the EGFR and the PDGF receptors, via several intracellular mechanisms. One of these is via a Ca2+-dependent pathway. Ang II is able to stimulate and increase intracellular Ca2+ via the PLC/IP3 pathway and Ang II-induced EGFR transactivation requires an increase of intracellular calcium. However, there are other studies which demonstrate that EGFR transactivation stimulated by Ang II is not dependent on intracellular calcium concentration, but is inhibited by N-acetylcysteine – a reactive oxygen species (ROS) scavenger. ROS are known to act as second messengers and activate a wide variety of serine–threonine and tyrosine kinases and may activate the EGFR by targeting the cysteine regions of the active sites of tyrosine phosphatases, which in turn activate tyrosine kinases. ROS can also stimulate the production of a heparin-binding EGF-like growth factor (HB-EGF) via its cleavage by a metalloproteinase. HB-EGF is expressed as a 208 amino acid transmembrane precursor that is proteolytically processed to give rise to mature soluble forms, which are able to activate the EGFR. Evidence for the role of HB-EGF in EGFR transactivation is supported by the observation that neutralizing antibodies to HB-EGF block EGFR activation. An integral process in GPCR to protein tyrosine kinase receptor transactivation appears to be the involvement of caveolae.[40,41] Multiple receptor components are present in lipid rafts and disruption of these rafts by cholesterol depletion, for example with β-cyclodextrin, prevents the generation of pErk1/2 from GPCR agonists such as Ang II but does not prevent the generation of pErk1/2 following treatment of cells with EGF. This data indicates that caveolin and cholesterol-rich micro-domains are involved in the GPCR-mediated transactivation of EGFR. Much has been learnt but much more needs to be determined about the process of GPCR transactivation of tyrosine kinase receptors.
GPCRs can transactivate cell surface serine/threonine kinase receptors
GPCR transactivation of protein tyrosine kinase is well established and accepted. There have, however, been several examples of GPCR transactivation of serine/threonine kinase receptors, notably the TGF-β type I receptor (TβRI) in which the authors have not adopted the use of the term transactivation; this is a receptor-to-receptor process for which we have recently offered a definition. We proposed the mechanistic definition for receptor–receptor and, in this context, GPCR–kinase transactivation as ‘the agonist occupancy of its cognate GPCR complex which leads in a relatively short time and in the absence of “de novo” protein synthesis to the activation of and cytosolic generation of the immediate downstream product(s) of a second cell surface protein kinase receptor’. The definition requires a conceptual component that takes into account the temporal characteristics of the response because long-term receptor occupancy can lead to a multitude of secondary and tertiary responses. The proposed definition includes the concept of the pathway being independent of de novo protein synthesis.
We recently observed that two separate GPCR agonists can rapidly generate carboxy terminal polyphosphorylated Smad 2 (pSmad2C), the immediate downstream product of activation of TβRI.[43,44] The data for the dose–response relationship of the effect of endothelin-1 on the level of pSmad2C in human vascular smooth muscle cells is reproduced in Figure 1. As this seemed analogous to the GPCR generation of pErk1/2 from protein tyrosine kinase receptors, for which the term transactivation is used, we have suggested that the use of transactivation is also appropriate for the GPCR activation of serine/threonine kinase receptors. Our definition allows for and includes both the GPCR generation of pErk1/2 from protein tyrosine kinases and pSmad2C from serine/threonine kinase receptors to be classified as transactivation.
Although understanding of the mechanism has not been fully elucidated, GPCR can also transactivate serine/threonine kinase cell surface receptors such as that for transforming growth factor (TGF)-β. The kinase signalling is mediated by the type I receptor known as TβRI.[45,46] The immediate downstream product of the activation of TβRI (also known as activin-like kinase or ALK-V) is pSmad2C.[45,47,48] pSmad2C complexes with Smad4 and in some circumstances to inhibitory Smad7 and translocates to the nucleus, where it elicits the response to the agonist. We have recently found that the GPCR agonists thrombin and endothelin-1 can elicit the rapid formation of pSmad2C in vascular smooth muscle cells. The original data for the response to endothelin-1 shows that there is a two-fold increase in pSmad2C levels at 1 h (Figure 1). The thrombin-mediated increase in pSmad2C is insensitive to the protein synthesis inhibitor cycloheximide, implying a receptor-to-receptor transactivation independent of gene activation and de novo protein synthesis (Little and Burch, unpublished observation). The stimulation of pSmad2C by the GPCR agonists thrombin and endothelin-1 is blocked by the respective receptor antagonists JN5177094 and bosentan;[44,50] these responses to GPCR agonists are blocked by the ALK-V antagonist SB431542, which confirms the involvement of TβRI (ALK-V). The responses to the GPCR agonists are not due to the agonist-mediated release and autocrine action of TGF-β. Using neutralizing antibodies to TGF-β we previously showed that the response to PDGF is partially blocked, indicating that the effect of PDGF is mediated by the release of TGF-β. However, the generation of pSmad2C in vascular smooth muscle cells treated with thrombin is not blocked in the presence of the pan TGF-β antibody. Thrombin acts as a protease to activate its receptor, which occurs by binding of the tethered ligand to the active site of protease-activated receptor 1. The small peptide, thrombin receptor activating peptide (TRAP), which acts as a mimetic, also stimulates an increase in pSmad2C. These data provide compelling evidence for the transactivation of a serine/threonine kinase receptor, TβRI, by at least two GPCR agonists, thrombin and endothelin-1. Detailed studies are underway in our laboratory to further unravel the mechanism of the transactivation and to compare it to the mechanism of transactivation of the tyrosine kinase EGFR, which also leads to stimulation of proteoglycan synthesis in these cells.
The pSmad2C responses arising from stimulation by thrombin and endothelin-1 are functionally important because both lead to the stimulation of proteoglycan synthesis.[44,50] Thus GPCRs can transactivate both protein tyrosine kinase and protein serine/threonine kinase receptors. The question therefore is what is the relative importance of the multiple signalling pathways described above and what is their combined contribution to GPCR signalling? Using thrombin as a model GPCR agonist in vascular smooth muscle cells and proteoglycan synthesis as a model response, we have observed that 36% of the stimulation of proteoglycan synthesis is blocked by AG1478 and is therefore attributable to transactivation of the EGF receptor. Furthermore, 44% of the stimulation of proteoglycan synthesis is blocked by SB431542 and therefore attributable to transactivation of TβRI. Although these two components sum to 80% we have not yet determined if these pathways are independent or redundant to assess the total components attributable to both transactivation pathways acting concomitantly. Taken with other evidence that calcium ions are not involved in the GPCR stimulation of proteoglycan synthesis, these data strongly imply that the transactivation pathways are the predominant signalling pathways for the stimulation of proteoglycan synthesis in these cells.
Other investigators prior to our work identified GPCR to serine/threonine kinase receptor transactivation. Sheppard and colleagues have shown that in mouse lung epithelial cells thrombin and LPA stimulate phosphorylation of Smad2 and the formation of pSmad2 within 1 h, with the response rising to a peak at 4 h.[55,56] The response in these cells is dependent on the RhoA/ROCK pathway and also integrin signalling. The basal levels of pSmad2C were increased in cells transfected with constitutively active RhoA and abolished with a dominant negative RhoA. In cells stimulated with thrombin and LPA, pharmacological inhibition of ROCK prevents the increases in pSmad2. The GPCR agonist mediated increase in pSmad2 levels in mouse lung epithelial cells is also dependent on αVβ6 integrins; immunoneutralizing antibodies to αVβ6 prevent thrombin- and LPA-induced Smad2 phosphorylation.[55,56]
Another study in mouse pulmonary artery smooth muscle cells provided evidence for transactivation of serine/threonine kinase BMP receptors by the GPCR agonist serotonin. Smads 2 and 3 are the natural targets of TβRI but the targets of serine/threonine kinase phosphorylation by other TGF-β super family receptor members are Smads 1, 5 and 8. Serotonin, via its GPCR, 5-HTR 1B/1D, transactivates BMPR1A, leading to the rapid generation of pSmad1, 5 and 8; as above this response is reported to be sensitive to pharmacological inhibitors of ROCK activity.
Efficacy of GPCR antagonism as a therapeutic strategy
There are a number of drugs that target GPCRs and are highly efficacious in therapeutic outcomes. In the current context we have described results for endothelin-1 and thrombin transactivation of tyrosine and serine/threonine kinase receptors. Based on the extensive data for the GPCR transactivation of tyrosine kinase receptors we speculate that this response, including transactivation of serine/threonine kinase receptors, can be generalized to most if not all GPCRs; in considering the therapeutic implications of this new paradigm we wish to focus on the most prominent GPCR agonist in cardiovascular medicine – Ang II. Angiotensin converting enzyme inhibitors (ACEI) block the generation of Ang II and effectively block all Ang II-dependent responses in the body. ACEI are amongst the most efficacious drugs in cardiovascular medicine, with many clinical trials having demonstrated their effectiveness in the prevention of cardiovascular events.[60,61] ACEI have recently been joined, after a delay of many years, by Ang II receptor antagonists (ARBs) and the latter are also highly efficacious in the prevention of cardiovascular disease.[62,63] The primary target for ACEI and ARBs, and the rationale for their development, was hypertension, and they are highly efficacious in reducing blood pressure.[62,64] However, the beneficial actions in the prevention of cardiovascular disease will exceed the blood pressure lowering effects and the outcome for these drugs at a mechanistic and clinical level shows that they have far greater efficacy than other drugs that lower blood pressure to an equivalent extent. These latter actions are known as ‘pleiotropic’ and include a long list of actions such as renoprotective, anti-atherogenic, antioxidant and antiplatelet effects. For some receptor agonist drugs, pleiotropic actions need to be divided into two classes, where some are directly related to the primary action of a drug and therefore potentially predictable while others are genuinely off target, such as the plethora of drug actions that are often termed anti-inflammatory. ACEI and ARBs are clearly associated with both types of pleiotropic actions in the cardiovascular system.
Our recent data, following on from that of others, indicates that GPCRs can signal via transactivation of both protein tyrosine kinase and protein serine/threonine kinase receptors.[42,44,50] Our readout has been proteoglycan synthesis but there is good reason to think that this may be one of a series of signalling outputs that are activated by transactivation. Whereas most rapid initial responses to GPCR activation, such as vasoconstriction in response to Ang II, are probably physiological, it is possible, even likely, that the prolonged responses that result from transactivation of important cell surface receptor kinases may represent pathological responses. In the context of Ang II signalling described above we would speculate that a major factor that underpins the highly efficacious actions of anti-Ang II strategies is that the inhibition of the Ang II signalling in preventing the activation of multiple cell surface kinase pathways may be the prime cause of the cardiovascular pathology. The inhibition of signalling by Ang II receptors either by direct inhibition by ARBs or reduced AngII levels following ACEI treatment may be preventing a plethora of events that are derived from protein tyrosine kinase and protein serine/threonine kinase receptors and not established pathways downstream of Ang II receptors.
The mechanistic implications of the contribution of cell surface receptor kinase transactivation to the effects of GPCRs are profound but somewhat predictable. It appears very likely that multiple GPCR agonists can transactivate multiple kinase receptors (Figure 2). In our model of proteoglycan synthesis[68,69] both endothelin-1 and thrombin, acting through their respective GPCRs, can transactivate both EGFR and TβRI.[44,50] At face value this appears to render the role of these pathways as a therapeutic target somewhat problematical. However, what is required is an understanding of the mechanism(s) of transactivation such that these signalling mechanisms and not the receptors themselves may be the novel therapeutic targets. For example, GPCR transactivation of EGFR involves, amongst multiple mechanisms, a triple membrane bypass mechanism in which receptor occupancy leads to activation of matrix metalloproteinases (MMPs), which cleave HB-EGF from the cell surface. The released HB-EGF acts on the EGFR to elicit a full response. Multiple GPCR agonists such as thrombin, lysophosphatidic acid and endothelin-1 use this pathway.[16,70] This process is inhibited by broad-spectrum MMP inhibitors such as GM6001 and also inhibition of individual MMPs using pharmacological and molecular biological approaches. Thus the inhibition of MMPs leads to the blocking of multiple pathways through which GPCR agonists activate tyrosine kinase EGFRs.
The knowledge of the mechanisms through which GPCR agonists activate serine/threonine kinase receptors such as TβRI are much less well advanced. In our human vascular smooth muscle cell model, we have confirmed that MMPs are involved in the transactivation of the EGFR, leading to inhibition of the pErk1/2 and of proteoglycan synthesis stimulated by thrombin. However, the thrombin-mediated transactivation of TβRI, leading to the generation of pSmad2C and stimulation of proteoglycan core protein synthesis and elongation of chondroitin sulphate/dermatan sulphate glycosaminoglycan (GAG) chains on biglycan is not inhibited by GM6001 (Burch and Little, unpublished observations). Further studies are required to identify the mechanism of transactivation of TβRI, with the aim of identifying processes that may be common to multiple vasoactive GPCR agonists. Our data has excluded a role for calcium and conventional GPCR signalling pathways in transactivation, so this allows for the discovery of novel but common pathways for the GCPR transactivation of the serine/threonine kinase TβRI and the family of protein tyrosine kinase receptors.
One interesting final point concerns the breadth and type of cell surface protein kinase receptors that may be the subject of transactivation from GPCRs. For example, the TGF-β receptors TβRI, which are predominantly serine/threonine kinase, also possesses some tyrosine kinase activity. Derynck and colleagues have demonstrated that TGF-β stimulation acting through the TβRI directly phosphorylates ShcA proteins on both tyrosine and serine residues. Accordingly, with respect to being targets for transactivation by GPCRs, there does not appear to be a fundamental biochemical difference between tyrosine and serine/threonine kinase cell-surface receptors and therefore no apparent reason why the currently accepted phenomenon of transactivation of cell-surface protein tyrosine kinase receptors should not apply to cell-surface serine/threonine kinase receptors. It is an interesting point as to whether or not the tyrosine kinase activity of TβRI is subject to transactivation from GPCRs and this point is presently being considered in our laboratory.
The discovery of the importance of cell surface receptor kinase transactivation in the signalling pathways for GPCRs opens up a whole area of drug discovery. There has been a focus on the discovery of inhibitors of GPCRs and the programmes to discover and develop inhibitors of Ang II receptors and protease activated receptors for thrombin have been amongst the most active in cardiovascular therapeutic drug discovery.[73,74] These have lead to the discovery of extremely efficacious agents for the prevention of cardiovascular disease. Now that transactivation pathways have been identified as playing such a potentially important role in the pathophysiological responses to GPCRs, developing an understanding of the mechanism(s) of transactivation provides a new layer of possibilities for inhibiting the actions of GPCR agonists and selectively preventing their role in a range of pathologies from atherosclerosis to lung fibrosis. The therapeutic implications are that either it is possible to discover common or shared mechanisms of transactivation such that an inhibitor of one target can block both GPCRs to tyrosine kinase and serine/threonine kinase receptor transactivation or alternatively if such a common mechanism does not exist or emerge then it will be necessary to provide multiple concomitant therapeutic strategies to block the profligate downstream pathways resulting from the transactivation of multiple kinase pathways by GPCRs that are responsible for complex pathologies.
Conflict of interest
The Authors declare that they have no conflicts of interest to disclose.
Work described in this report was supported by a National Health and Medical Research NHMRC Project Grant (#1022800) (PJL and NO) and a National Heart Foundation of Australia grant-in-aid (PJL) and Diabetes Australia Research Trust grants (PJL and NO). The PhD Program of MLB generously received support through a National Heart Foundation of Australia post-graduate scholarship and a post-graduate award from GlaxoSmithKline Australia. WZ acknowledges support from the National Natural Science Fund of China (No. 30670652; no. 30711120565; no. 30970935).