- Email: [email protected]
New Aspects of G-Protein-Coupled Receptor Signalling and Regulation
New Aspects of G-Protein-Coupled Receptor Signalling and Regulation Graeme Milligan
Research on the structure, regulation and signalling properties of the family of seven-transmembrane-helix, heterotrimeric guanine nucleotide-binding protein (G-protein)-coupled receptors (GPCRs) continues at a frantic pace. This reflects their central role in transmission of hormone- and neurotransmitter-encoded information across the plasma membrane of cells. The location of the ligandbinding sites on the extracellular face of the membrane has made them obvious targets for therapeutic intervention in a wide range of conditions resulting from endocrine imbalance. Furthermore, based on the identification of many novel GPCR sequences emerging from expressed sequence tags (ESTs) and other DNA sequencing programmes, it has become clear that the GPCR family is likely to be considerably larger than appreciated in even the recent past. Although neither the natural ligands nor synthetic pharmaceuticals have yet been identified for these so-called ‘orphan’ GPCRs, they offer the potential for a plethora of new therapeutic targets. Within a short review, it is impossible to cover all the current developments in this field and the topics selected represent a personal view of recent highlights of areas that provide both novel and general insights into the function and regulation of GPCRs.
•
Receptor Dimerization
It is well established that single transmembrane element growth factor receptors with intrinsic tyrosine kinase activity, such as those for epidermal growth factor and platelet-derived growth factor, have to dimerize and auto-phosphorylate in response to ligand to recruit adapter and effector enzymes and thus initiate downstream signal transduction cascades (Heldin 1995). Despite this, only recently has the suggestion that receptor dimerization might also play a central role in signalling by guanine
Graeme Milligan is at the Molecular Pharmacology Group, Division of Biochemistry and Molecular Biology, Institute of Biomedical and Life Sciences, University of Glasgow, Glasgow, UK G12 8QQ.
TEM Vol. 9, No. 1, 1998
nucleotide-binding protein (G-protein)-coupled receptors (GPCRs) gained substantial credence. Following expression in insect Sf9 cells, a proportion of the immunoreactivity of an epitope-tagged form of the human 2-adrenoceptor was shown to migrate in sodium dodecyl sulphatepolyacrylamide gel electrophoresis (SDS-PAGE) with an apparent Mr consistent with a dimeric species (Hebert et al. 1996). Artefacts of this nature might be anticipated in such a system as a result of non-specific aggregation or the detergent used for solubilization of the membranes. However, as a membrane permeant crosslinking agent increased the proportion of apparently dimeric species it was suggested that dimers were preformed in the intact cell.
Importantly, addition of a peptide corresponding to a region of transmembrane element VI of the receptor to membranes expressing the 2adrenoceptor resulted in a reduction in both the presence of the dimer and the capacity of the agonist isoprenaline to stimulate adenylyl cyclase activity. Moreover, addition of isoprenaline increased the proportion of immunologically detected dimer and protected against the effects of the peptide derived from transmembrane element VI. Very similar results, in terms of the ability to crosslink receptors to form dimers in intact cells and the effects of a peptide derived from transmembrane element VI, have been reported for the dopamine D2 receptor (Ng et al. 1996). Furthermore, for the dopamine D2 receptor, it has been suggested that differences in the binding capacity of two radiolabeled ligands are a reflection of the capacity of [3H]spiperone to bind only to the receptor monomer and [3H]nemonapride to bind to both receptor monomer and dimer (Ng et al. 1996). Further studies have provided evidence for oligomerization of histamine H2 receptors (Fukushima et al. 1997) and vasopressin V2 receptors (Hebert et al. 1996). In the case of the metabotropic glutamate receptor mGluR5, cysteine residues in the long extracellular domain (a feature of the mGluR receptor family) are involved in the production of disulphide-dependent dimers (Romano et al. 1996). Recent studies on the ␦-opioid receptor have utilized combinations of receptors that have been epitope tagged with either Flag or c-myc sequences. Following their co-expression, immunoprecipitation with one epitope tag antibody resulted in co-immunoprecipitation of the other (Cvejic and Devi 1997). These results demonstrate interactions between the differentially tagged proteins, which can be modified by treatment with certain, but not all, agonists. However, unlike the situation with the 2-adrenoceptor noted above, agonists at the ␦-opioid receptor which regulated dimerization reduced dimer levels (Cvejic and
© 1998, Elsevier Science Ltd, 1043-2760/98/$19.00. PII: S1043-2760(98)00004-6
13
Receptor–G-Protein Fusion Proteins
Many postulates of pharmacological receptor theory, such as the two-state model of receptor function and the mobile receptor hypothesis, are based on the relative stoichiometries of receptor and G protein being close to 1:1. Simple measurements of levels of expression indicate that G-protein levels are much higher than those of specific GPCRs (Milligan 1996), although such gross estimates fail to consider compartmentalization and the potential inaccessibility of much of the G-protein population to specific GPCRs (Neubig 1994). There
14
B
A
50 50
V (pmol min−1 mg−1)
•
cal linkage of the two proteins resulted in more efficient coupling. It has also been inferred that the 2-adrenoceptor–Gs␣ fusion protein is more resistant to agonist-induced desensitization than the receptor alone (Bertin et al. 1994, 1997). Building upon this premise and the capacity of sustained intracellular levels of cAMP to limit growth factorinduced activation of the extracellularly regulated mitogen-activated protein (MAP) kinases and thence cell mitogenesis, Bertin et al. (1997) have reported that expression of the 2-adrenoceptor–Gs␣ fusion protein in p21ras-transformed cancer cell lines inhibits growth in a -adrenoceptor agonist-dependent manner. A fusion protein between the ␣2A-adrenoceptor and the ␣ subunit of Gi1 has been used to measure directly the catalytic centre GTPase activity of Gi1 following addition of an ␣2-adrenoceptor agonist (Wise et al. 1997a) (Figure 1). These studies were performed on a mutant of Gi1␣ in which Cys351 was replaced by Gly. The rationale for this was that Cys351 is the acceptor site in Gi1␣
has also been considerable interest in the concept described as ‘receptor channeling’ (Kenakin 1995) whereby, if GPCRs are able to activate multiple G proteins and thence a range of signaling cascades, agonists may be found which display selectivity for activation of specific GPCR–G-protein tandems. An interesting approach to ensure that the stoichiometry of expression of a GPCR and G protein is 1:1 has been to generate fusion proteins in which the N-terminus of a Gprotein ␣ subunit is linked directly to the C-terminal tail of a GPCR. The first example was a fusion protein between the 2-adrenoceptor and its cognate G protein Gs␣ (Bertin et al. 1994). Expression of this construct in S49 cyc⫺ cells, which do not express Gs␣, resulted in agonist activation of adenylyl cyclase (Bertin et al. 1994). As S49 cells express the 2-adrenoceptor endogenously, it was of interest to note that following expression of the 2-adrenoceptor–Gs␣ fusion protein, the EC50 for isoprenaline stimulation of adenylyl cyclase was shifted more than tenfold to lower concentrations, suggesting that physi-
V (pmol min−1 mg−1 protein)
Devi 1997). As morphine was unable to reduce dimer levels and is unable to promote internalization of the receptor, it has been suggested that these two aspects may be related (Cvejic and Devi 1997). Perhaps even more interesting in this regard has been the observation that two distinct point mutations of the angiotensin AT1A receptor, which both impair ligand binding, can restore binding when co-expressed (Monnot et al. 1996). Furthermore, co-transfection of chimeric ␣2-adrenergic and M3 muscarinic acetylcholine receptors has demonstrated that functional intramolecular interactions can occur between GPCRs (Maggio et al. 1993). The rescue of function following co-expression of bindingpositive but signal transductionnegative and binding-negative but signal transduction-positive forms and chimeras of the luteinizing hormone (LH) receptor (Osuga et al. 1997) have allowed similar conclusions to be reached. It has been suggested that agents able to interfere with receptor dimerization may offer a novel approach to restricting signal transduction and, as such, might be useful therapeutic agents. However, the difficulties of targeting a compound that must disrupt protein– protein interactions to an intracellular location rather than simply occluding the ligand-binding site with antagonist drugs is likely to limit development of such a strategy.
40
30
20
30
20
10
10
0
40
0 0
0.5
1.0
1.5
[GTP] (µM)
2.0
2.5
0
20
40
60
80
100 120 140 160
180
V (pmol min−1 mg−1)/S (µM)
Figure 1. UK14304 stimulates high affinity GTPase activity of an ␣2A-adrenoceptor– Cys351Gly Gi1␣ fusion protein. P2 particulate membrane fractions from pertussis toxintreated COS-7 cells transfected to express the ␣2A-adrenoceptor–Cys351Gly Gi1␣ fusion protein were used to measure high affinity GTPase activity. (A) High affinity GTPase activity was measured over a range of concentrations of GTP in the absence (filled symbols) or presence (open symbols) of UK14304 (10 M). (B) The data are presented as an Eadie–Hofstee transformation. In the example displayed, the increase in Vmax produced by UK14304 was 18.8 pmol min⫺1 mg⫺1 membrane protein. As these membranes expressed the fusion protein at 6.2 pmol mg⫺1 protein, UK14304 stimulation of turnover of GTP by the fusion proteinassociated G protein could be calculated as 3.0 min⫺1. From Wise et al. 1997a, with permission.
TEM Vol. 9, No. 1, 1998
for ADP ribosylation catalyzed by pertussis toxin. As this modification prevents productive coupling between Gi1 and GPCRs, cells expressing the ␣2A- a d r e n o c e p t o r – C y s 351G l y G i1 fusion protein could be treated with pertussis toxin to eliminate potential contacts between the receptor and endogenously expressed forms of Gi and thus allow agonist activation of the fusion protein to be measured in isolation. Many GPCRs and the ␣ subunits of all widely expressed G proteins become post-translationally palmitoylated (Milligan et al. 1995). A palmitoylation-resistant mutant (Cys3Ser) of Gi1␣ cannot be activated by a co-expressed ␣2A-adrenoceptor (Wise et al. 1997b). To assess whether this reflected an absolute requirement for acylation in the process of agonist-induced information transfer or related instead to the role of acylation in appropriate targeting of proteins, fusion proteins were generated between the ␣2Aadrenoceptor and acylation-resistant forms of Gi1␣. In all cases, agonist activation of the GTPase activity of the mutant G protein was observed (Wise and Milligan 1997), demonstrating that the proximity provided by linking the GPCR and G protein together was both necessary and sufficient to initiate G-protein signalling rather than the inherent acylation status of the proteins. Such fusion proteins can also be used to examine the basis for agonist efficacy (Wise et al. 1997c). By measuring agonist-induced Vmax for the GTPase activity of the ␣2A-adrenoceptor–Cys351GlyGi1 fusion protein, a range of efficacies were recorded for different agonists that followed the same rank order as that seen with independent co-expression of the receptor and Cys351GlyGi1. •
Regulators of G-Protein Signalling
As with the heterotrimeric G proteins, small Mr GTP-binding proteins, such as p21ras, display intrinsic GTPase activity that controls the activation/ deactivation status of the proteins. However, the observed GTPase activTEM Vol. 9, No. 1, 1998
ity of the purified proteins is extremely low and inconsistent with estimates of the rates of GTP exchange and hydrolysis predicted from in vivo analyses. This discrepancy led to searches for accessory proteins that would stimulate GTP hydrolysis and resulted in the identification of Ras-GAP (GTPase activating protein). Subsequently, other molecules that regulate the time frame of the presence of GTP in the nucleotide-binding pocket have been discovered. These include guanine nucleotide dissociation-stimulator (GDS) and guanine nucleotide dissociationinhibitor (GDI) proteins for a wide range of monomeric G proteins. In addition, the structural basis for these functions has been elucidated [see, for example (Rittinger et al. 1997)]. Initially, these observations were considered to have little relevance to signal transduction through heterotrimeric G proteins as, when purified, these G proteins display markedly higher GTPase activity than the small Mr G proteins. However, the measured catalytic rates (in the region of 2–5 min⫺1) were still inconsistent with the known kinetics of signal transduction, and simply could not be incorporated into an understanding of the visual process in which rapid on/off responses of the visual G protein (transducin) are required to prevent retinal bleaching. The demonstration that a subunit of the cGMP phosphodiesterase effector target of transducin could function as a GAP for transducin (Arshavsky and Bownds 1992) appeared to obviate this concern and the demonstration that phospholipase C1 could also function as a GAP for Gq␣ (Biddlecome et al. 1996) suggested that it might simply be a matter of time before all effector enzymes were shown to function as GAPs for their cognate G proteins. This has yet to be realized. Ongoing genetic studies of the pheromone response pathway in yeast, which shares high levels of similarity with mammalian G-proteincoupled signal transduction, resulted
in identification of sst (supersensitivity to pheromone) mutants. One locus of these mutants is the SST2 gene. This protein was subsequently shown to act as a GAP for the yeast G-protein ␣ subunit GPA1. Parallel studies on egg laying (EGL) in the nematode Caenorhabditis elegans identified the product of the EGL-10 gene as a homologous protein able to reduce the activity of the Go1 gene of this nematode. The use of yeast two-hybrid screens and subsequent cloning by homology has resulted in the identification of a considerable family of mammalian proteins highly related to the SST2 and EGL-10 gene products (Druey et al. 1996, Hunt et al. 1996, Watson et al. 1996, Dohlman and Thorner 1997). These proteins bind to a variety of mammalian G-protein ␣ subunits and appear to stabilize the transition state required for their hydrolysis of GTP (Berman et al. 1996, Tesmer et al. 1997). This family of proteins [collectively named regulators of G-protein signalling (RGSs)] continues to expand and, although a number of them are widely expressed, many others show marked specificity of distribution. Furthermore, although initial studies had suggested that they might be limited in function to modulation of the function of Gi-like G proteins, this has now been expanded to include the phospholipase-regulating Gq family G proteins (Huang et al. 1997, Helper et al. 1997, Yan et al. 1997). Recent studies have begun to concentrate on the capacity of RGS proteins to act as attenuators of signal transduction cascades. As such, there is the possibility that positive and negative regulators of RGS proteins could be used to regulate the effectiveness of G-protein-coupled signalling processes. This concept is promoted by observations of the regulation of expression of certain RGS proteins by ligands at GPCRs (Druey et al. 1996) and by cellular differentiation (Ogier-Denis et al. 1997). Furthermore, it has been demonstrated that they can play a role in regulating GPCR-mediated 15
signal transduction (Doupnik et al. 1997, Ogier-Denis et al. 1997). •
Localization of GPCRs
It is an inherent necessity that GPCRs reside, for at least a significant part of the time, at the plasma membrane. Recent studies have begun to unravel aspects of how they are targeted there, how they may move within subdomains of the plasma membrane in response to agonist ligands, and their intracellular trafficking pathways. A series of fascinating studies on the delivery of GPCRs to the plasma membrane has been performed on polarized canine kidney cells. Within the three highly homologous ␣2-adrenoceptors, both the ␣2A-adrenoceptor and the ␣2Badrenoceptor are targeted to the basolateral membrane (Wozniak and Limbird 1996). However, although the ␣2A-adrenoceptor is delivered directly to this surface, initially the ␣2B-adrenoceptor appears to be inserted at random into the apical and basolateral surfaces but subsequently is selectively retained by the basolateral surface (Wozniak and Limbird 1996). By contrast, although part of the steady state ␣2C-adrenoceptor population has a basolateral plasma membrane location, to which it is delivered directly, at least in these cells, a proportion of the cellular levels of this receptor has an intracellular location (Wozniak and Limbird 1996). In contrast to the ␣2-adrenoceptors, in the same cells, the A1 adenosine receptor is selectively enriched in the apical membrane (Saunders et al. 1996). Furthermore, disruption of microtubules interfered with the targeting of the A1 adenosine receptor to the apical surface but not the initial apical component of ␣2B-adrenoceptor distribution (Saunders and Limbird 1997). The possibility that specialized regions of the plasma membrane might concentrate signalling components has been raised both by observations of a non-uniform distribution of fluorescent agonists and antagonists at GPCRs in both fixed tissue and confocal microscopy studies [for a re-
16
Figure 2. Detection of the distribution of a thyrotropin-releasing hormone (TRH) receptor– green fluorescent protein fusion protein in HEK293 cells. A rat TRH receptor–green fluorescent protein fusion construct was expressed in HEK293 cells and imaged after fixation in a confocal microscope. Two adjacent cells are observed in the plane of the microscope, as is the edge of a third cell above them. The construct was targeted predominantly to the plasma membrane of the cells.
view, see (McGrath et al. 1996)] and from the use of antibodies raised either against peptide sequences derived from GPCRs or following expression of ‘epitope-tagged’ GPCRs (von Zastrow and Kobilka 1992, Molino et al. 1997). Such studies begin to hint at the likely complexities of GPCR distribution. GPCR clustering and sequestration, which is likely to contribute to the processes of desensitization and resensitization of GPCRs (Barak et al. 1994, Pippig et al. 1995), appears to involve
both clathrin-coated and non-clathrincoated vesicle pathways (Zhang et al. 1996). In the case of non-clathrincoated vesicles, although its significance remains a matter of some debate, a series of studies has indicated a selective enrichment of signaling polypeptides, including heterotrimeric G proteins, to glycosphingolipid-rich regions named caveolae, which are highly enriched for the presence therein of scaffolding proteins of the caveolin family (Song et al. 1996, Li et al. 1996). Intriguingly, two recent TEM Vol. 9, No. 1, 1998
Figure 3. Cellular redistribution of G i1␣ by agonist activation of the thyrotropin-releasing hormone (TRH) receptor. E2M11 HEK293 cells (Svoboda et al. 1996) express high levels of both the rat TRH receptor and murine Gi1␣. These cells were grown on cover slips and were either (A) untreated or (B) exposed to TRH (10 M, 16 h) and then prepared for immunofluorescence confocal microscopy to detect Gi1␣ (green) and -tubulin (brown). Before treatment with TRH, Gi1␣ was evenly distributed around the plasma membrane of the cells but, following sustained treatment with agonist, the remaining Gi1␣ now displayed a predominantly punctate perinuclear localization. The reduced overall immunostaining of Gi1␣ in (B) reflects the substantial degree of downregulation of Gi1␣ produced by TRH treatment for this period of time (Svoboda et al. 1996).
TEM Vol. 9, No. 1, 1998
reports have indicated that the targeting of GPCRs to caveolae may require agonist activation. In cardiac myocytes, addition of a muscarinic acetylcholine receptor agonist resulted in movement of a proportion of the m2 muscarinic acetylcholine receptor population to a caveolar location and the subsequent interaction of the receptor with caveolin3, a muscle-specific form of caveolin (Feron et al. 1997). A similar occurrence has been reported for the bradykinin B2 receptor (de Weerd and Leeb-Lundberg 1997 ). As caveolin appears to bind tightly to inactive forms of G proteins and not to mutationally activated forms of G-protein ␣ subunits, it could be suggested either that agonist-mediated transfer of a GPCR to the caveolae would act to compete with caveolin for the G protein and thus allow it to be activated by the GPCR to initiate signaling or, alternatively, that movement of the GPCR to the caveolae is part of the desensitization response for the GPCR. Many GPCRs appear to be internalized via clathrin-coated vesicles. Rapid, recent progress in this area has developed from the capacity to interfere selectively in this process. Although relatively crude strategies which, for example, have utilized exposure to hypertonic sucrose have been used as an indicator of a role for clathrin-coated vesicles, recent strategies have been to examine the ability of dynamin and, more specifically, GTP-binding mutants of dynamin to interfere with GPCR internalization. Dynamin is a GTPase that plays a key role in the pinching off of endocytic vesicles from the plasma membrane. Dominant negative mutants of dynamin have been shown to interfere with agonistinduced sequestration of the 2adrenoceptor when expressed in HEK293 cells, but not with internalization of the angiotensin II type 1A receptor in the same cells (Zhang et al. 1996), arguing both for a specific role for clathrin-coated vesicles in the regulation of the 2-adrenoceptor and that this is 17
not a pathway used universally by GPCRs. Such approaches are likely to contribute significantly to the analysis of the role of GPCR internalization in desensitization. Opioid receptors have been particularly well studied in this regard as only subsets of agonist ligands are able to induce receptor internalization (Sternini et al. 1996) and this does not seem to be purely related to ligand efficacy. Furthermore, desensitization of the -opioid receptor may not reflect uncoupling from its cognate G proteins (Pak et al. 1996), unlike the situation for the intensely studied 2-adrenoceptor (Lohse 1993). G proteins can also be redistributed and internalized in response to agonist (Svoboda et al. 1996) (Figure 3) and this may contribute to long-term and maintained aspects of desensitization. In general, however, effects at the G-protein level are substantially slower than those observed for the cognate GPCR. The capacity to label a GPCR fluorescently offers the potential to examine its distribution and agonistinduced redistribution in intact cells and in real time. One such approach to this has involved tagging of GPCRs at their C-terminal tail with green fluorescent protein (GFP) (Barak et al. 1997, Tarasova et al. 1997) (Figure 2). Such studies have demonstrated that both a 2-adrenoceptor– GFP fusion protein and an equivalent construct of the cholecystokinin-A receptor have the capacity to activate adenylyl cyclase and to be redistributed from the plasma membrane to internal structures in response to agonist. It is likely that both GFP-tagged GPCRs and the interactions between fluorescently labelled GPCRs and other signaling polypeptides will be utilized in the near future to explore the details of cellular localization and protein–protein interactions. Moreover, highly fluorescent peptides that have similar pharmacology to the native peptides are now available and should be of great use for examination of cellular redistribution of GPCRs with peptide ligands.
18
•
Conclusions
Rapid progress is being made in the understanding of the cellular distribution, regulation and function of GPCRs. The expanding range of ‘knock-out’ mice in which expression of specific GPCRs has been eliminated has also provided clear and, in certain cases, unexpected information on the functions of individual GPCRs, which has either validated theories or suggested new avenues towards targeted strategies for therapeutic intervention in disease processes. •
Acknowledgements
Work in the author’s laboratory is supported by the Medical Research Council, the Biotechnology and Biosciences Research Council (UK) and the Wellcome Trust. References Arshavsky VY, Bownds MD: 1992. Regulation of deactivation of photoreceptor G protein by its target enzyme and cGMP. Nature 357:416–417. Barak LS, Tiberi M, Freedman NJ, Kwatra MM, Lefkowitz RJ, Caron MG: 1994. A highly conserved tyrosine residue in G protein-coupled receptors is required for agonist-mediated 2-adrenergic receptor sequestration. J Biol Chem 269:2790–2795. Barak LS, Ferguson SSG, Zhang J, Martenson C, Meyer T, Caron MG: 1997. Internal trafficking and surface mobility of a functionally intact 2-adrenergic receptor–green fluorescent protein conjugate. Mol Pharmacol 51:177–184. Berman DM, Kozasa T, Gilman AG: 1996. The GTPase-activating protein RGS4 stabilizes the transition state for nucleotide hydrolysis. J Biol Chem 272:27209–27212. Bertin B, Friessmuth M, Jockers R, Strosberg AD, Marullo S: 1994. Cellular signaling by an agonist-activated receptor/Gs␣ fusion protein. Proc Natl Acad Sci USA 91:8827–8831.
Steady-state balance of receptor-mediated activation and GTPase-activating proteinpromoted deactivation. J Biol Chem 271:7999–8007. Cvejic S, Devi LA: 1997. Dimerization of the ␦ opioid receptor: implication for a role in receptor internalization. J Biol Chem 272:26959–26964. de Weerd WFC, Leeb-Lundberg LMF: 1997. Bradykinin sequesters B2 bradykinin receptors and the receptor-coupled G␣ subunits G␣q and G␣i in caveolae in DDT1 MF2 smooth muscle cells. J Biol Chem 272:17858–17866. Dohlman HG, Thorner J: 1997. RGS proteins and signaling by heterotrimeric G proteins. J Biol Chem 272:3871–3874. Doupnik CA, Davidson N, Lester HA, Kofuji P: 1997. RGS proteins reconstitute the rapid gating kinetics of G␥-activated inwardly rectifying K+ channels. Proc Natl Acad Sci USA 94:10461–10466. Druey KM, Blumer KJ, Kang VH, Kehrl JH: 1996. Inhibition of G-protein-mediated MAP kinase activation by a new mammalian gene family. Nature 379:742–746. Feron O, Smith TW, Michel T, Kelly RA: 1997. Dynamic targeting of the agoniststimulated m2 muscarinic acetylcholine receptor to caveolae in cardiac myocytes. J Biol Chem 272:17744–17748. Fukushima Y, Asano T, Saitoh T, et al.:1997. Oligomer formation of histamine H2 receptors expressed in Sf9 and COS7 cells. FEBS Lett 409:283–286. Hebert TE, Moffett S, Morello JP, et al.: 1996. A peptide derived from a 2adrenergic receptor transmembrane domain inhibits both receptor dimerization and activation. J Biol Chem 271:16384–16392. Heldin C-H: 1995. Dimerization of cell surface receptors in signal transduction. Cell 80:213–223. Helper JR, Berman DM, Gilman AG, Kozasa T: 1997. RGS4 and GAIP are GTPaseactivating proteins for Gq␣ and block activation of phospholipase C by ␥-thio-GTP–Gq␣. Proc Natl Acad Sci USA 94:428–432. Huang C, Helper JR, Gilman AG, Mumby SM: 1997. Attenuation of Gi-and Gq-mediated signaling by expression of RGS4 or GIAP in mammalian cells. Proc Natl Acad Sci USA 94:6159–6163.
Bertin B, Strosberg AD, Marullo S: 1997. Human 2-adrenergic receptor/Gs␣ fusion protein, expressed in 2 ras-dependent murine carcinoma cell lines, prevents tumor growth in syngenic mice. Int J Cancer 71:1029–1034.
Hunt TW, Fields TA, Casey PJ, Peralta EG: 1996. RGS10 is a selective activator of G␣i GTPase activity. Nature 383:175–177.
Biddlecome GH, Berstein G, Ross EM: 1996. Regulation of phospholipase C-1 by Gq and m1 muscarinic cholinergic receptor.
Kenakin T: 1995. Agonist-receptor efficacy II: agonist trafficking of receptor signals. Trends Pharmacol Sci 16:232–238.
TEM Vol. 9, No. 1, 1998
Li S, Couet J, Lisanti MP: 1996. Src tyrosine kinases, G␣ subunits and H-Ras share a common membrane-anchored scaffolding protein, caveolin. J Biol Chem 271:29182–29190. Lohse MJ: 1993. Molecular mechanisms of membrane receptor desensitization. Biochim Biophys Acta 1179:171–188. Maggio R, Vogel Z, Wess J: 1993. Coexpression studies with mutant muscarinic/adrenergic receptors provide evidence for intermolecular ‘cross-talk’ between G-protein-linked receptors. Proc Natl Acad Sci USA 90:3103–3107.
tization of the -opioid receptor is mediated by loss of membrane receptors rather than uncoupling from G protein. Mol Pharmacol 50:1214–1222. Pippig S, Andexinger S, Lohse MJ: 1995. Sequestration and recycling of 2-adrenergic receptors permit receptor resensitization. Mol Pharmacol 47:666–676. Rittinger K, Walker PA, Eccleston JF, Smerdon SJ, Gamblin SJ: 1997. Structure at 1.65Å of RhoA and its GTPase-activating protein in complex with a transitionstate analogue. Nature 389:758–762.
McGrath JC, Arribas S, Daly CJ: 1996. Fluorescent ligands for the study of receptors. Trend Pharmacol Sci 17:393–399.
Romano C, Yang W-L, O’Malley KL: 1996. Metabotropic glutamate receptor 5 is a disulfide-linked dimer. J Biol Chem 271:28612–28616.
Milligan G: 1996. The stoichiometry of expression of protein components of the stimulatory adenylyl cyclase cascade and the regulation of information transfer. Cell Signal 8:87–96.
Saunders C, Limbird LE: 1997. Disruption of microtubules reveals two independent apical targeting mechanisms for G-proteincoupled receptors in polarized renal epithelial cells. J Biol Chem 272:19035–19045.
Milligan G, Parenti M, Magee AI: 1995. The dynamic role of palmitoylation in signal transduction. Trends Biochem Sci 20:181–186.
Saunders C, Keefer JR, Kennedy AP, Wells JN, Limbird LE: 1996. Receptors coupled to pertussis toxin-sensitive G-proteins traffic to opposite surfaces in Madin– Darby canine kidney cells. A1 adenosine receptors achieve apical and ␣2A-adrenergic receptors achieve basolateral localization. J Biol Chem 271:995–1002.
Molino M, Bainton DF, Hoxie JA, Coughlin SR, Brass LF: 1997. Thrombin receptors on human platelets. Initial localization and subsequent redistribution during platelet activation. J Biol Chem 272:6011–6017. Monnot C, Bihoreau C, Conchon S, Curnow KM, Corvol P, Clauser E: 1996. Polar residues in the transmembrane domains of the type 1 angiotensin II receptor are required for binding and coupling. Reconstitution of the binding site by coexpression of two deficient mutants. J Biol Chem 271:1507–1513. Neubig RR: 1994. Membrane organization in G protein mechanisms. FASEB J 8:939–946. Ng GYK, O’Dowd BF, Lee SP, et al.: 1996. Dopamine D2 receptor dimers and receptor-blocking peptides. Biochem Biophys Res Commun 227:200–204. Ogier-Denis E, Petiot A, Bauvy C, Codogno P: 1997. Control of the expression and activity of the G␣-interacting protein (GAIP) in human intestinal cells. J Biol Chem 272:24599–24603.
Song KS, Li S, Okamoto T, Quilliam LA, Sargiacomo M, Lisanti MP: 1996. Copurification and direct interaction of ras with caveolin, an integral membrane protein of caveolae microdomains. Detergentfree purification of caveolae membranes. J Biol Chem 271:9690–9697. Sternini C, Spann M, Anton B, et al.: 1996. Agonist-selective endocytosis of opioid receptor by neurons in vivo. Proc Natl Acad Sci USA 93:9241–9246 Svoboda P, Kim GD, Grassie MA, Eidne KA, Milligan G: 1996. Thyrotropin-releasing hormone-induced subcellular redistribution and down-regulation of Gi1␣: analysis of agonist regulation of co-expressed Gi1␣ species variants. Mol Pharmacol 49:646–655. Tarasova NI, Stauber RH, Choi JK, et al.: 1997. Visualization of G protein-coupled receptor trafficking with the aid of the green fluorescent protein. Endocytosis and recycling of cholecystokinin receptor type A. J Biol Chem 272:14817–14824.
Osuga Y, Hayashi M, Kudo M, Conti M, Kobilka B, Hsueh AJW: 1997. Co-expression of defective luteinizing hormone receptor fragments reconstitutes ligandinduced signal generation. J Biol Chem 272:25006–25012.
Tesmer JJG, Berman DM, Gilman AG, Sprang SR: 1997. Structure of RGS4 bound to AlF4⫺-activated Gi1␣: stabilization of the transition state for GTP hydrolysis. Cell 89:251–261.
Pak Y, Kouvelas A, Scheideler MA, Rasmussen J, O’Dowd BF, George SR: 1996. Agonist-induced functional desensi-
von Zastrow M, Kobilka BK: 1992. Ligandregulated internalization and recycling of human beta 2-adrenergic receptors
TEM Vol. 9, No. 1, 1998
between the plasma membrane and endosomes containing transferrin receptors. J Biol Chem 267:3530–3538. Watson N, Linder ME, Druey KM, Kehrl JH, Blumer KJ: 1996. RGS family members: GTPase-activating proteins for heterotrimeric G-protein ␣-subunits. Nature 383:172–175. Wise A, Milligan G: 1997. Rescue of functional interactions between the ␣2Aadrenoreceptor and acylation-resistant forms of Gi1␣ by expressing the proteins from chimeric open reading frames. J Biol Chem 272:24673–24678. Wise A, Carr IC, Milligan G: 1997a. Measurement of agonist-induced guanine nucleotide turnover by the G protein Gi1␣ when constrained within an ␣2A-adrenoceptor-Gi1␣ fusion protein. Biochem J 325:17–21. Wise A, Grassie MA, Parenti M, Lee M, Rees S, Milligan G: 1997b. A cysteine-3 to serine mutation of the G protein Gi1␣ abrogates functional activation by the ␣2A-adrenoceptor but not interactions with the ␥ complex. Biochemistry 36:10620–10629. Wise A, Carr IC, Groarke DA, Milligan G: 1997c. Measurement of agonist efficacy using an ␣2A-adrenoceptor–Gi1␣ fusion protein. FEBS Lett 419:141–146. Wozniak M, Limbird LE: 1996. The three ␣2adrenergic receptor subtypes achieve basolateral localization in Madin–Darby canine kidney II cells via different targeting mechanisms. J Biol Chem 271:5017–5024. Yan Y, Chi PP, Bourne HR: 1997. RGS4 inhibits Gq-mediated activation of mitogenactivated protein kinase and phosphoinositide synthesis. J Biol Chem 272:11924–11927. Zhang J, Ferguson SSG, Barak LS, Menard L, Caron MG: 1996. Dynamin and arrestin reveal distinct mechanisms for G protein-coupled receptor internalization. J Biol Chem 271:18302–18305.
Do you want to reproduce material from Trends in Endocrinology and Metabolism? This publication and the individual contributions contained in it are protected by the copyright of Elsevier Science. Except as outlined in the terms and conditions (see p. IV), no part of Trends in Endocrinology and Metabolism may be reproduced, either in print or in electronic form, without written permission from Elsevier Science. Please send any permission requests to: Elsevier Science Ltd, PO Box 800, Oxford, UK OX5 1DX.
19
Research on the structure, regulation and signalling properties of the family of seven-transmembrane-helix, heterotrimeric guanine nucleotide-binding protein (G-protein)-coupled receptors (GPCRs) continues at a frantic pace. This reflects their central role in transmission of hormone- and neurotransmitter-encoded information across the plasma membrane of cells. The location of the ligandbinding sites on the extracellular face of the membrane has made them obvious targets for therapeutic intervention in a wide range of conditions resulting from endocrine imbalance. Furthermore, based on the identification of many novel GPCR sequences emerging from expressed sequence tags (ESTs) and other DNA sequencing programmes, it has become clear that the GPCR family is likely to be considerably larger than appreciated in even the recent past. Although neither the natural ligands nor synthetic pharmaceuticals have yet been identified for these so-called ‘orphan’ GPCRs, they offer the potential for a plethora of new therapeutic targets. Within a short review, it is impossible to cover all the current developments in this field and the topics selected represent a personal view of recent highlights of areas that provide both novel and general insights into the function and regulation of GPCRs.
•
Receptor Dimerization
It is well established that single transmembrane element growth factor receptors with intrinsic tyrosine kinase activity, such as those for epidermal growth factor and platelet-derived growth factor, have to dimerize and auto-phosphorylate in response to ligand to recruit adapter and effector enzymes and thus initiate downstream signal transduction cascades (Heldin 1995). Despite this, only recently has the suggestion that receptor dimerization might also play a central role in signalling by guanine
Graeme Milligan is at the Molecular Pharmacology Group, Division of Biochemistry and Molecular Biology, Institute of Biomedical and Life Sciences, University of Glasgow, Glasgow, UK G12 8QQ.
TEM Vol. 9, No. 1, 1998
nucleotide-binding protein (G-protein)-coupled receptors (GPCRs) gained substantial credence. Following expression in insect Sf9 cells, a proportion of the immunoreactivity of an epitope-tagged form of the human 2-adrenoceptor was shown to migrate in sodium dodecyl sulphatepolyacrylamide gel electrophoresis (SDS-PAGE) with an apparent Mr consistent with a dimeric species (Hebert et al. 1996). Artefacts of this nature might be anticipated in such a system as a result of non-specific aggregation or the detergent used for solubilization of the membranes. However, as a membrane permeant crosslinking agent increased the proportion of apparently dimeric species it was suggested that dimers were preformed in the intact cell.
Importantly, addition of a peptide corresponding to a region of transmembrane element VI of the receptor to membranes expressing the 2adrenoceptor resulted in a reduction in both the presence of the dimer and the capacity of the agonist isoprenaline to stimulate adenylyl cyclase activity. Moreover, addition of isoprenaline increased the proportion of immunologically detected dimer and protected against the effects of the peptide derived from transmembrane element VI. Very similar results, in terms of the ability to crosslink receptors to form dimers in intact cells and the effects of a peptide derived from transmembrane element VI, have been reported for the dopamine D2 receptor (Ng et al. 1996). Furthermore, for the dopamine D2 receptor, it has been suggested that differences in the binding capacity of two radiolabeled ligands are a reflection of the capacity of [3H]spiperone to bind only to the receptor monomer and [3H]nemonapride to bind to both receptor monomer and dimer (Ng et al. 1996). Further studies have provided evidence for oligomerization of histamine H2 receptors (Fukushima et al. 1997) and vasopressin V2 receptors (Hebert et al. 1996). In the case of the metabotropic glutamate receptor mGluR5, cysteine residues in the long extracellular domain (a feature of the mGluR receptor family) are involved in the production of disulphide-dependent dimers (Romano et al. 1996). Recent studies on the ␦-opioid receptor have utilized combinations of receptors that have been epitope tagged with either Flag or c-myc sequences. Following their co-expression, immunoprecipitation with one epitope tag antibody resulted in co-immunoprecipitation of the other (Cvejic and Devi 1997). These results demonstrate interactions between the differentially tagged proteins, which can be modified by treatment with certain, but not all, agonists. However, unlike the situation with the 2-adrenoceptor noted above, agonists at the ␦-opioid receptor which regulated dimerization reduced dimer levels (Cvejic and
© 1998, Elsevier Science Ltd, 1043-2760/98/$19.00. PII: S1043-2760(98)00004-6
13
Receptor–G-Protein Fusion Proteins
Many postulates of pharmacological receptor theory, such as the two-state model of receptor function and the mobile receptor hypothesis, are based on the relative stoichiometries of receptor and G protein being close to 1:1. Simple measurements of levels of expression indicate that G-protein levels are much higher than those of specific GPCRs (Milligan 1996), although such gross estimates fail to consider compartmentalization and the potential inaccessibility of much of the G-protein population to specific GPCRs (Neubig 1994). There
14
B
A
50 50
V (pmol min−1 mg−1)
•
cal linkage of the two proteins resulted in more efficient coupling. It has also been inferred that the 2-adrenoceptor–Gs␣ fusion protein is more resistant to agonist-induced desensitization than the receptor alone (Bertin et al. 1994, 1997). Building upon this premise and the capacity of sustained intracellular levels of cAMP to limit growth factorinduced activation of the extracellularly regulated mitogen-activated protein (MAP) kinases and thence cell mitogenesis, Bertin et al. (1997) have reported that expression of the 2-adrenoceptor–Gs␣ fusion protein in p21ras-transformed cancer cell lines inhibits growth in a -adrenoceptor agonist-dependent manner. A fusion protein between the ␣2A-adrenoceptor and the ␣ subunit of Gi1 has been used to measure directly the catalytic centre GTPase activity of Gi1 following addition of an ␣2-adrenoceptor agonist (Wise et al. 1997a) (Figure 1). These studies were performed on a mutant of Gi1␣ in which Cys351 was replaced by Gly. The rationale for this was that Cys351 is the acceptor site in Gi1␣
has also been considerable interest in the concept described as ‘receptor channeling’ (Kenakin 1995) whereby, if GPCRs are able to activate multiple G proteins and thence a range of signaling cascades, agonists may be found which display selectivity for activation of specific GPCR–G-protein tandems. An interesting approach to ensure that the stoichiometry of expression of a GPCR and G protein is 1:1 has been to generate fusion proteins in which the N-terminus of a Gprotein ␣ subunit is linked directly to the C-terminal tail of a GPCR. The first example was a fusion protein between the 2-adrenoceptor and its cognate G protein Gs␣ (Bertin et al. 1994). Expression of this construct in S49 cyc⫺ cells, which do not express Gs␣, resulted in agonist activation of adenylyl cyclase (Bertin et al. 1994). As S49 cells express the 2-adrenoceptor endogenously, it was of interest to note that following expression of the 2-adrenoceptor–Gs␣ fusion protein, the EC50 for isoprenaline stimulation of adenylyl cyclase was shifted more than tenfold to lower concentrations, suggesting that physi-
V (pmol min−1 mg−1 protein)
Devi 1997). As morphine was unable to reduce dimer levels and is unable to promote internalization of the receptor, it has been suggested that these two aspects may be related (Cvejic and Devi 1997). Perhaps even more interesting in this regard has been the observation that two distinct point mutations of the angiotensin AT1A receptor, which both impair ligand binding, can restore binding when co-expressed (Monnot et al. 1996). Furthermore, co-transfection of chimeric ␣2-adrenergic and M3 muscarinic acetylcholine receptors has demonstrated that functional intramolecular interactions can occur between GPCRs (Maggio et al. 1993). The rescue of function following co-expression of bindingpositive but signal transductionnegative and binding-negative but signal transduction-positive forms and chimeras of the luteinizing hormone (LH) receptor (Osuga et al. 1997) have allowed similar conclusions to be reached. It has been suggested that agents able to interfere with receptor dimerization may offer a novel approach to restricting signal transduction and, as such, might be useful therapeutic agents. However, the difficulties of targeting a compound that must disrupt protein– protein interactions to an intracellular location rather than simply occluding the ligand-binding site with antagonist drugs is likely to limit development of such a strategy.
40
30
20
30
20
10
10
0
40
0 0
0.5
1.0
1.5
[GTP] (µM)
2.0
2.5
0
20
40
60
80
100 120 140 160
180
V (pmol min−1 mg−1)/S (µM)
Figure 1. UK14304 stimulates high affinity GTPase activity of an ␣2A-adrenoceptor– Cys351Gly Gi1␣ fusion protein. P2 particulate membrane fractions from pertussis toxintreated COS-7 cells transfected to express the ␣2A-adrenoceptor–Cys351Gly Gi1␣ fusion protein were used to measure high affinity GTPase activity. (A) High affinity GTPase activity was measured over a range of concentrations of GTP in the absence (filled symbols) or presence (open symbols) of UK14304 (10 M). (B) The data are presented as an Eadie–Hofstee transformation. In the example displayed, the increase in Vmax produced by UK14304 was 18.8 pmol min⫺1 mg⫺1 membrane protein. As these membranes expressed the fusion protein at 6.2 pmol mg⫺1 protein, UK14304 stimulation of turnover of GTP by the fusion proteinassociated G protein could be calculated as 3.0 min⫺1. From Wise et al. 1997a, with permission.
TEM Vol. 9, No. 1, 1998
for ADP ribosylation catalyzed by pertussis toxin. As this modification prevents productive coupling between Gi1 and GPCRs, cells expressing the ␣2A- a d r e n o c e p t o r – C y s 351G l y G i1 fusion protein could be treated with pertussis toxin to eliminate potential contacts between the receptor and endogenously expressed forms of Gi and thus allow agonist activation of the fusion protein to be measured in isolation. Many GPCRs and the ␣ subunits of all widely expressed G proteins become post-translationally palmitoylated (Milligan et al. 1995). A palmitoylation-resistant mutant (Cys3Ser) of Gi1␣ cannot be activated by a co-expressed ␣2A-adrenoceptor (Wise et al. 1997b). To assess whether this reflected an absolute requirement for acylation in the process of agonist-induced information transfer or related instead to the role of acylation in appropriate targeting of proteins, fusion proteins were generated between the ␣2Aadrenoceptor and acylation-resistant forms of Gi1␣. In all cases, agonist activation of the GTPase activity of the mutant G protein was observed (Wise and Milligan 1997), demonstrating that the proximity provided by linking the GPCR and G protein together was both necessary and sufficient to initiate G-protein signalling rather than the inherent acylation status of the proteins. Such fusion proteins can also be used to examine the basis for agonist efficacy (Wise et al. 1997c). By measuring agonist-induced Vmax for the GTPase activity of the ␣2A-adrenoceptor–Cys351GlyGi1 fusion protein, a range of efficacies were recorded for different agonists that followed the same rank order as that seen with independent co-expression of the receptor and Cys351GlyGi1. •
Regulators of G-Protein Signalling
As with the heterotrimeric G proteins, small Mr GTP-binding proteins, such as p21ras, display intrinsic GTPase activity that controls the activation/ deactivation status of the proteins. However, the observed GTPase activTEM Vol. 9, No. 1, 1998
ity of the purified proteins is extremely low and inconsistent with estimates of the rates of GTP exchange and hydrolysis predicted from in vivo analyses. This discrepancy led to searches for accessory proteins that would stimulate GTP hydrolysis and resulted in the identification of Ras-GAP (GTPase activating protein). Subsequently, other molecules that regulate the time frame of the presence of GTP in the nucleotide-binding pocket have been discovered. These include guanine nucleotide dissociation-stimulator (GDS) and guanine nucleotide dissociationinhibitor (GDI) proteins for a wide range of monomeric G proteins. In addition, the structural basis for these functions has been elucidated [see, for example (Rittinger et al. 1997)]. Initially, these observations were considered to have little relevance to signal transduction through heterotrimeric G proteins as, when purified, these G proteins display markedly higher GTPase activity than the small Mr G proteins. However, the measured catalytic rates (in the region of 2–5 min⫺1) were still inconsistent with the known kinetics of signal transduction, and simply could not be incorporated into an understanding of the visual process in which rapid on/off responses of the visual G protein (transducin) are required to prevent retinal bleaching. The demonstration that a subunit of the cGMP phosphodiesterase effector target of transducin could function as a GAP for transducin (Arshavsky and Bownds 1992) appeared to obviate this concern and the demonstration that phospholipase C1 could also function as a GAP for Gq␣ (Biddlecome et al. 1996) suggested that it might simply be a matter of time before all effector enzymes were shown to function as GAPs for their cognate G proteins. This has yet to be realized. Ongoing genetic studies of the pheromone response pathway in yeast, which shares high levels of similarity with mammalian G-proteincoupled signal transduction, resulted
in identification of sst (supersensitivity to pheromone) mutants. One locus of these mutants is the SST2 gene. This protein was subsequently shown to act as a GAP for the yeast G-protein ␣ subunit GPA1. Parallel studies on egg laying (EGL) in the nematode Caenorhabditis elegans identified the product of the EGL-10 gene as a homologous protein able to reduce the activity of the Go1 gene of this nematode. The use of yeast two-hybrid screens and subsequent cloning by homology has resulted in the identification of a considerable family of mammalian proteins highly related to the SST2 and EGL-10 gene products (Druey et al. 1996, Hunt et al. 1996, Watson et al. 1996, Dohlman and Thorner 1997). These proteins bind to a variety of mammalian G-protein ␣ subunits and appear to stabilize the transition state required for their hydrolysis of GTP (Berman et al. 1996, Tesmer et al. 1997). This family of proteins [collectively named regulators of G-protein signalling (RGSs)] continues to expand and, although a number of them are widely expressed, many others show marked specificity of distribution. Furthermore, although initial studies had suggested that they might be limited in function to modulation of the function of Gi-like G proteins, this has now been expanded to include the phospholipase-regulating Gq family G proteins (Huang et al. 1997, Helper et al. 1997, Yan et al. 1997). Recent studies have begun to concentrate on the capacity of RGS proteins to act as attenuators of signal transduction cascades. As such, there is the possibility that positive and negative regulators of RGS proteins could be used to regulate the effectiveness of G-protein-coupled signalling processes. This concept is promoted by observations of the regulation of expression of certain RGS proteins by ligands at GPCRs (Druey et al. 1996) and by cellular differentiation (Ogier-Denis et al. 1997). Furthermore, it has been demonstrated that they can play a role in regulating GPCR-mediated 15
signal transduction (Doupnik et al. 1997, Ogier-Denis et al. 1997). •
Localization of GPCRs
It is an inherent necessity that GPCRs reside, for at least a significant part of the time, at the plasma membrane. Recent studies have begun to unravel aspects of how they are targeted there, how they may move within subdomains of the plasma membrane in response to agonist ligands, and their intracellular trafficking pathways. A series of fascinating studies on the delivery of GPCRs to the plasma membrane has been performed on polarized canine kidney cells. Within the three highly homologous ␣2-adrenoceptors, both the ␣2A-adrenoceptor and the ␣2Badrenoceptor are targeted to the basolateral membrane (Wozniak and Limbird 1996). However, although the ␣2A-adrenoceptor is delivered directly to this surface, initially the ␣2B-adrenoceptor appears to be inserted at random into the apical and basolateral surfaces but subsequently is selectively retained by the basolateral surface (Wozniak and Limbird 1996). By contrast, although part of the steady state ␣2C-adrenoceptor population has a basolateral plasma membrane location, to which it is delivered directly, at least in these cells, a proportion of the cellular levels of this receptor has an intracellular location (Wozniak and Limbird 1996). In contrast to the ␣2-adrenoceptors, in the same cells, the A1 adenosine receptor is selectively enriched in the apical membrane (Saunders et al. 1996). Furthermore, disruption of microtubules interfered with the targeting of the A1 adenosine receptor to the apical surface but not the initial apical component of ␣2B-adrenoceptor distribution (Saunders and Limbird 1997). The possibility that specialized regions of the plasma membrane might concentrate signalling components has been raised both by observations of a non-uniform distribution of fluorescent agonists and antagonists at GPCRs in both fixed tissue and confocal microscopy studies [for a re-
16
Figure 2. Detection of the distribution of a thyrotropin-releasing hormone (TRH) receptor– green fluorescent protein fusion protein in HEK293 cells. A rat TRH receptor–green fluorescent protein fusion construct was expressed in HEK293 cells and imaged after fixation in a confocal microscope. Two adjacent cells are observed in the plane of the microscope, as is the edge of a third cell above them. The construct was targeted predominantly to the plasma membrane of the cells.
view, see (McGrath et al. 1996)] and from the use of antibodies raised either against peptide sequences derived from GPCRs or following expression of ‘epitope-tagged’ GPCRs (von Zastrow and Kobilka 1992, Molino et al. 1997). Such studies begin to hint at the likely complexities of GPCR distribution. GPCR clustering and sequestration, which is likely to contribute to the processes of desensitization and resensitization of GPCRs (Barak et al. 1994, Pippig et al. 1995), appears to involve
both clathrin-coated and non-clathrincoated vesicle pathways (Zhang et al. 1996). In the case of non-clathrincoated vesicles, although its significance remains a matter of some debate, a series of studies has indicated a selective enrichment of signaling polypeptides, including heterotrimeric G proteins, to glycosphingolipid-rich regions named caveolae, which are highly enriched for the presence therein of scaffolding proteins of the caveolin family (Song et al. 1996, Li et al. 1996). Intriguingly, two recent TEM Vol. 9, No. 1, 1998
Figure 3. Cellular redistribution of G i1␣ by agonist activation of the thyrotropin-releasing hormone (TRH) receptor. E2M11 HEK293 cells (Svoboda et al. 1996) express high levels of both the rat TRH receptor and murine Gi1␣. These cells were grown on cover slips and were either (A) untreated or (B) exposed to TRH (10 M, 16 h) and then prepared for immunofluorescence confocal microscopy to detect Gi1␣ (green) and -tubulin (brown). Before treatment with TRH, Gi1␣ was evenly distributed around the plasma membrane of the cells but, following sustained treatment with agonist, the remaining Gi1␣ now displayed a predominantly punctate perinuclear localization. The reduced overall immunostaining of Gi1␣ in (B) reflects the substantial degree of downregulation of Gi1␣ produced by TRH treatment for this period of time (Svoboda et al. 1996).
TEM Vol. 9, No. 1, 1998
reports have indicated that the targeting of GPCRs to caveolae may require agonist activation. In cardiac myocytes, addition of a muscarinic acetylcholine receptor agonist resulted in movement of a proportion of the m2 muscarinic acetylcholine receptor population to a caveolar location and the subsequent interaction of the receptor with caveolin3, a muscle-specific form of caveolin (Feron et al. 1997). A similar occurrence has been reported for the bradykinin B2 receptor (de Weerd and Leeb-Lundberg 1997 ). As caveolin appears to bind tightly to inactive forms of G proteins and not to mutationally activated forms of G-protein ␣ subunits, it could be suggested either that agonist-mediated transfer of a GPCR to the caveolae would act to compete with caveolin for the G protein and thus allow it to be activated by the GPCR to initiate signaling or, alternatively, that movement of the GPCR to the caveolae is part of the desensitization response for the GPCR. Many GPCRs appear to be internalized via clathrin-coated vesicles. Rapid, recent progress in this area has developed from the capacity to interfere selectively in this process. Although relatively crude strategies which, for example, have utilized exposure to hypertonic sucrose have been used as an indicator of a role for clathrin-coated vesicles, recent strategies have been to examine the ability of dynamin and, more specifically, GTP-binding mutants of dynamin to interfere with GPCR internalization. Dynamin is a GTPase that plays a key role in the pinching off of endocytic vesicles from the plasma membrane. Dominant negative mutants of dynamin have been shown to interfere with agonistinduced sequestration of the 2adrenoceptor when expressed in HEK293 cells, but not with internalization of the angiotensin II type 1A receptor in the same cells (Zhang et al. 1996), arguing both for a specific role for clathrin-coated vesicles in the regulation of the 2-adrenoceptor and that this is 17
not a pathway used universally by GPCRs. Such approaches are likely to contribute significantly to the analysis of the role of GPCR internalization in desensitization. Opioid receptors have been particularly well studied in this regard as only subsets of agonist ligands are able to induce receptor internalization (Sternini et al. 1996) and this does not seem to be purely related to ligand efficacy. Furthermore, desensitization of the -opioid receptor may not reflect uncoupling from its cognate G proteins (Pak et al. 1996), unlike the situation for the intensely studied 2-adrenoceptor (Lohse 1993). G proteins can also be redistributed and internalized in response to agonist (Svoboda et al. 1996) (Figure 3) and this may contribute to long-term and maintained aspects of desensitization. In general, however, effects at the G-protein level are substantially slower than those observed for the cognate GPCR. The capacity to label a GPCR fluorescently offers the potential to examine its distribution and agonistinduced redistribution in intact cells and in real time. One such approach to this has involved tagging of GPCRs at their C-terminal tail with green fluorescent protein (GFP) (Barak et al. 1997, Tarasova et al. 1997) (Figure 2). Such studies have demonstrated that both a 2-adrenoceptor– GFP fusion protein and an equivalent construct of the cholecystokinin-A receptor have the capacity to activate adenylyl cyclase and to be redistributed from the plasma membrane to internal structures in response to agonist. It is likely that both GFP-tagged GPCRs and the interactions between fluorescently labelled GPCRs and other signaling polypeptides will be utilized in the near future to explore the details of cellular localization and protein–protein interactions. Moreover, highly fluorescent peptides that have similar pharmacology to the native peptides are now available and should be of great use for examination of cellular redistribution of GPCRs with peptide ligands.
18
•
Conclusions
Rapid progress is being made in the understanding of the cellular distribution, regulation and function of GPCRs. The expanding range of ‘knock-out’ mice in which expression of specific GPCRs has been eliminated has also provided clear and, in certain cases, unexpected information on the functions of individual GPCRs, which has either validated theories or suggested new avenues towards targeted strategies for therapeutic intervention in disease processes. •
Acknowledgements
Work in the author’s laboratory is supported by the Medical Research Council, the Biotechnology and Biosciences Research Council (UK) and the Wellcome Trust. References Arshavsky VY, Bownds MD: 1992. Regulation of deactivation of photoreceptor G protein by its target enzyme and cGMP. Nature 357:416–417. Barak LS, Tiberi M, Freedman NJ, Kwatra MM, Lefkowitz RJ, Caron MG: 1994. A highly conserved tyrosine residue in G protein-coupled receptors is required for agonist-mediated 2-adrenergic receptor sequestration. J Biol Chem 269:2790–2795. Barak LS, Ferguson SSG, Zhang J, Martenson C, Meyer T, Caron MG: 1997. Internal trafficking and surface mobility of a functionally intact 2-adrenergic receptor–green fluorescent protein conjugate. Mol Pharmacol 51:177–184. Berman DM, Kozasa T, Gilman AG: 1996. The GTPase-activating protein RGS4 stabilizes the transition state for nucleotide hydrolysis. J Biol Chem 272:27209–27212. Bertin B, Friessmuth M, Jockers R, Strosberg AD, Marullo S: 1994. Cellular signaling by an agonist-activated receptor/Gs␣ fusion protein. Proc Natl Acad Sci USA 91:8827–8831.
Steady-state balance of receptor-mediated activation and GTPase-activating proteinpromoted deactivation. J Biol Chem 271:7999–8007. Cvejic S, Devi LA: 1997. Dimerization of the ␦ opioid receptor: implication for a role in receptor internalization. J Biol Chem 272:26959–26964. de Weerd WFC, Leeb-Lundberg LMF: 1997. Bradykinin sequesters B2 bradykinin receptors and the receptor-coupled G␣ subunits G␣q and G␣i in caveolae in DDT1 MF2 smooth muscle cells. J Biol Chem 272:17858–17866. Dohlman HG, Thorner J: 1997. RGS proteins and signaling by heterotrimeric G proteins. J Biol Chem 272:3871–3874. Doupnik CA, Davidson N, Lester HA, Kofuji P: 1997. RGS proteins reconstitute the rapid gating kinetics of G␥-activated inwardly rectifying K+ channels. Proc Natl Acad Sci USA 94:10461–10466. Druey KM, Blumer KJ, Kang VH, Kehrl JH: 1996. Inhibition of G-protein-mediated MAP kinase activation by a new mammalian gene family. Nature 379:742–746. Feron O, Smith TW, Michel T, Kelly RA: 1997. Dynamic targeting of the agoniststimulated m2 muscarinic acetylcholine receptor to caveolae in cardiac myocytes. J Biol Chem 272:17744–17748. Fukushima Y, Asano T, Saitoh T, et al.:1997. Oligomer formation of histamine H2 receptors expressed in Sf9 and COS7 cells. FEBS Lett 409:283–286. Hebert TE, Moffett S, Morello JP, et al.: 1996. A peptide derived from a 2adrenergic receptor transmembrane domain inhibits both receptor dimerization and activation. J Biol Chem 271:16384–16392. Heldin C-H: 1995. Dimerization of cell surface receptors in signal transduction. Cell 80:213–223. Helper JR, Berman DM, Gilman AG, Kozasa T: 1997. RGS4 and GAIP are GTPaseactivating proteins for Gq␣ and block activation of phospholipase C by ␥-thio-GTP–Gq␣. Proc Natl Acad Sci USA 94:428–432. Huang C, Helper JR, Gilman AG, Mumby SM: 1997. Attenuation of Gi-and Gq-mediated signaling by expression of RGS4 or GIAP in mammalian cells. Proc Natl Acad Sci USA 94:6159–6163.
Bertin B, Strosberg AD, Marullo S: 1997. Human 2-adrenergic receptor/Gs␣ fusion protein, expressed in 2 ras-dependent murine carcinoma cell lines, prevents tumor growth in syngenic mice. Int J Cancer 71:1029–1034.
Hunt TW, Fields TA, Casey PJ, Peralta EG: 1996. RGS10 is a selective activator of G␣i GTPase activity. Nature 383:175–177.
Biddlecome GH, Berstein G, Ross EM: 1996. Regulation of phospholipase C-1 by Gq and m1 muscarinic cholinergic receptor.
Kenakin T: 1995. Agonist-receptor efficacy II: agonist trafficking of receptor signals. Trends Pharmacol Sci 16:232–238.
TEM Vol. 9, No. 1, 1998
Li S, Couet J, Lisanti MP: 1996. Src tyrosine kinases, G␣ subunits and H-Ras share a common membrane-anchored scaffolding protein, caveolin. J Biol Chem 271:29182–29190. Lohse MJ: 1993. Molecular mechanisms of membrane receptor desensitization. Biochim Biophys Acta 1179:171–188. Maggio R, Vogel Z, Wess J: 1993. Coexpression studies with mutant muscarinic/adrenergic receptors provide evidence for intermolecular ‘cross-talk’ between G-protein-linked receptors. Proc Natl Acad Sci USA 90:3103–3107.
tization of the -opioid receptor is mediated by loss of membrane receptors rather than uncoupling from G protein. Mol Pharmacol 50:1214–1222. Pippig S, Andexinger S, Lohse MJ: 1995. Sequestration and recycling of 2-adrenergic receptors permit receptor resensitization. Mol Pharmacol 47:666–676. Rittinger K, Walker PA, Eccleston JF, Smerdon SJ, Gamblin SJ: 1997. Structure at 1.65Å of RhoA and its GTPase-activating protein in complex with a transitionstate analogue. Nature 389:758–762.
McGrath JC, Arribas S, Daly CJ: 1996. Fluorescent ligands for the study of receptors. Trend Pharmacol Sci 17:393–399.
Romano C, Yang W-L, O’Malley KL: 1996. Metabotropic glutamate receptor 5 is a disulfide-linked dimer. J Biol Chem 271:28612–28616.
Milligan G: 1996. The stoichiometry of expression of protein components of the stimulatory adenylyl cyclase cascade and the regulation of information transfer. Cell Signal 8:87–96.
Saunders C, Limbird LE: 1997. Disruption of microtubules reveals two independent apical targeting mechanisms for G-proteincoupled receptors in polarized renal epithelial cells. J Biol Chem 272:19035–19045.
Milligan G, Parenti M, Magee AI: 1995. The dynamic role of palmitoylation in signal transduction. Trends Biochem Sci 20:181–186.
Saunders C, Keefer JR, Kennedy AP, Wells JN, Limbird LE: 1996. Receptors coupled to pertussis toxin-sensitive G-proteins traffic to opposite surfaces in Madin– Darby canine kidney cells. A1 adenosine receptors achieve apical and ␣2A-adrenergic receptors achieve basolateral localization. J Biol Chem 271:995–1002.
Molino M, Bainton DF, Hoxie JA, Coughlin SR, Brass LF: 1997. Thrombin receptors on human platelets. Initial localization and subsequent redistribution during platelet activation. J Biol Chem 272:6011–6017. Monnot C, Bihoreau C, Conchon S, Curnow KM, Corvol P, Clauser E: 1996. Polar residues in the transmembrane domains of the type 1 angiotensin II receptor are required for binding and coupling. Reconstitution of the binding site by coexpression of two deficient mutants. J Biol Chem 271:1507–1513. Neubig RR: 1994. Membrane organization in G protein mechanisms. FASEB J 8:939–946. Ng GYK, O’Dowd BF, Lee SP, et al.: 1996. Dopamine D2 receptor dimers and receptor-blocking peptides. Biochem Biophys Res Commun 227:200–204. Ogier-Denis E, Petiot A, Bauvy C, Codogno P: 1997. Control of the expression and activity of the G␣-interacting protein (GAIP) in human intestinal cells. J Biol Chem 272:24599–24603.
Song KS, Li S, Okamoto T, Quilliam LA, Sargiacomo M, Lisanti MP: 1996. Copurification and direct interaction of ras with caveolin, an integral membrane protein of caveolae microdomains. Detergentfree purification of caveolae membranes. J Biol Chem 271:9690–9697. Sternini C, Spann M, Anton B, et al.: 1996. Agonist-selective endocytosis of opioid receptor by neurons in vivo. Proc Natl Acad Sci USA 93:9241–9246 Svoboda P, Kim GD, Grassie MA, Eidne KA, Milligan G: 1996. Thyrotropin-releasing hormone-induced subcellular redistribution and down-regulation of Gi1␣: analysis of agonist regulation of co-expressed Gi1␣ species variants. Mol Pharmacol 49:646–655. Tarasova NI, Stauber RH, Choi JK, et al.: 1997. Visualization of G protein-coupled receptor trafficking with the aid of the green fluorescent protein. Endocytosis and recycling of cholecystokinin receptor type A. J Biol Chem 272:14817–14824.
Osuga Y, Hayashi M, Kudo M, Conti M, Kobilka B, Hsueh AJW: 1997. Co-expression of defective luteinizing hormone receptor fragments reconstitutes ligandinduced signal generation. J Biol Chem 272:25006–25012.
Tesmer JJG, Berman DM, Gilman AG, Sprang SR: 1997. Structure of RGS4 bound to AlF4⫺-activated Gi1␣: stabilization of the transition state for GTP hydrolysis. Cell 89:251–261.
Pak Y, Kouvelas A, Scheideler MA, Rasmussen J, O’Dowd BF, George SR: 1996. Agonist-induced functional desensi-
von Zastrow M, Kobilka BK: 1992. Ligandregulated internalization and recycling of human beta 2-adrenergic receptors
TEM Vol. 9, No. 1, 1998
between the plasma membrane and endosomes containing transferrin receptors. J Biol Chem 267:3530–3538. Watson N, Linder ME, Druey KM, Kehrl JH, Blumer KJ: 1996. RGS family members: GTPase-activating proteins for heterotrimeric G-protein ␣-subunits. Nature 383:172–175. Wise A, Milligan G: 1997. Rescue of functional interactions between the ␣2Aadrenoreceptor and acylation-resistant forms of Gi1␣ by expressing the proteins from chimeric open reading frames. J Biol Chem 272:24673–24678. Wise A, Carr IC, Milligan G: 1997a. Measurement of agonist-induced guanine nucleotide turnover by the G protein Gi1␣ when constrained within an ␣2A-adrenoceptor-Gi1␣ fusion protein. Biochem J 325:17–21. Wise A, Grassie MA, Parenti M, Lee M, Rees S, Milligan G: 1997b. A cysteine-3 to serine mutation of the G protein Gi1␣ abrogates functional activation by the ␣2A-adrenoceptor but not interactions with the ␥ complex. Biochemistry 36:10620–10629. Wise A, Carr IC, Groarke DA, Milligan G: 1997c. Measurement of agonist efficacy using an ␣2A-adrenoceptor–Gi1␣ fusion protein. FEBS Lett 419:141–146. Wozniak M, Limbird LE: 1996. The three ␣2adrenergic receptor subtypes achieve basolateral localization in Madin–Darby canine kidney II cells via different targeting mechanisms. J Biol Chem 271:5017–5024. Yan Y, Chi PP, Bourne HR: 1997. RGS4 inhibits Gq-mediated activation of mitogenactivated protein kinase and phosphoinositide synthesis. J Biol Chem 272:11924–11927. Zhang J, Ferguson SSG, Barak LS, Menard L, Caron MG: 1996. Dynamin and arrestin reveal distinct mechanisms for G protein-coupled receptor internalization. J Biol Chem 271:18302–18305.
Do you want to reproduce material from Trends in Endocrinology and Metabolism? This publication and the individual contributions contained in it are protected by the copyright of Elsevier Science. Except as outlined in the terms and conditions (see p. IV), no part of Trends in Endocrinology and Metabolism may be reproduced, either in print or in electronic form, without written permission from Elsevier Science. Please send any permission requests to: Elsevier Science Ltd, PO Box 800, Oxford, UK OX5 1DX.
19