Oncogene (1998) 17, 1331 ± 1342  1998 Stockton Press All rights reserved 0950 ± 9232/98 $12.00 http://www.stockton-press.co.uk/onc Cell growth control by G protein-coupled receptors: from signal transduction to signal integration

J Silvio Gutkind

Oral and Pharyngeal Cancer Branch, National Institute of Dental Research, National Institutes of Health Bethesda, Maryland 20892, USA

Keywords: G proteins; G protein-coupled receptors; G protein-coupled receptors and cell proliferation Ras; MAP kinases; Rho; cell growth The use of Pertussis toxin (Ptx), that catalyses the

ADP-ribosylation of G protein a subunits of the ai family (Ui and Katada, 1990) thereby preventing their interaction with receptors, provided the ®rst clue that a Many growth factors are known to bind and activate number of mitogens act on this family of cell surface either receptors possessing an intrinsic protein-tyrosine receptors. For example, Pouyssegur and colleagues kinase activity (Yarden et al., 1986), or those that noted that DNA-synthesis in CHO cells in response to transmit signals to the cytoplasm through the thrombin was blocked by Ptx treatment, whereas the interaction with heterotrimeric GTP-binding proteins proliferative response to PDGF was Ptx-insensitive (G proteins). The latter are collectively known as G (Chambard et al., 1987; Pouyssegur et al., 1988). A protein-coupled receptors (GPCRs) and comprise the similar approach led to the discovery that one of the largest group of cell surface receptors. With more than most potent mitogens present in serum, lysopho- 1000 members, they represent more than 1% of the sphatidic acid (LPA), a simple naturally occurring *100 000 proteins encoded by the human genome. phospholipid, was also acting on the G protein-linked The best known family of GPCRs exhibit a common class of receptors (van Corven et al., 1989). Many structural motif consisting of seven membrane-span- other Ptx sensitive mitogens were subsequently ning regions (Dohlman et al., 1987) (Figure 1). These identi®ed (see van Biesen et al., 1996 for a review). receptors can be activated by a diverse array of However, biochemical studies provided evidence that external stimuli, including growth factors, vasoactive many mitogens also acted on GPCRs, albeit their polypeptides, chemoattractants, neurotransmitters, hor- biological e€ects were mediated by Ptx-insensitive G mones, phospholipids, photons, odorants, and taste proteins and therefore were resistant to Ptx treatment. ligands. Activation of GPCRs by these agents elicits a These included thrombin acting on smooth muscle profound change in the transmembrane a helices, thus cells, bombesin on Swiss 3T3 cells and bronchial a€ecting the conformation of intracellular loops epithelium, vasopressin, bradykinin and endothelin on uncovering previously masked G protein binding sites Swiss 3T3 cells, endothelin on human mesangial cells, (Altenbach et al., 1996; Bourne, 1997; Wess, 1997). substance K on human skin ®broblasts, acetylcholine This causes the exchange of GDP for GTP bound to receptor agonists on embryonic astrocytes, angiotensin the G protein a subunit, and a conformational change on smooth muscle cells, and many others (see in three ¯exible `switch regions' of Ga, activating Ga Moolenaar, 1991). Thus, it became clear that many and causing its dissociation from the bg heterodimers ligands acting via GPCRs could elicit a mitogenic (Lambright et al., 1994, 1996; Sondek et al., 1994, response in a variety of cell types (reviewed in 1996). In turn, GTP-bound G protein a subunits or bg Rozengurt, 1986; van Biesen et al., 1996) and that complexes initiate intracellular signaling responses by these receptors can transduce proliferative signals when acting on e€ector molecules such as adenylyl cyclases, acting on Ptx sensitive or insensitive heterotrimeric G phosphodiesterases, phospholipases, or regulating the proteins. Furthermore, recent gene knock-out studies activity of ion channels, ion transporters, and a indicated that certain GPCRs are essential for cell growing number of kinases. To date, 16 distinct growth under physiological conditions (Nagata et al., mammalian G protein a subunits have been identi- 1996). ®ed, and divided into four families based upon sequence similarity: as, ai, aq, and a12 (Wilkie et al., 1992). In addition, 11-G protein g subunits and ®ve G protein b subunits have been cloned so far. Taken Oncogenic potential of G protein-coupled receptors together, it is becoming increasingly clear that GPCRs represent one of the most diverse signal transduction Constitutive activation of receptor-protein tyrosine systems in eukaryotic cells. The biochemical and kinases, either by structural alteration of the receptor biological consequences of this diversity in subunit itself or by deregulated presentation of the ligand, can composition have just begun to be appreciated. In this induce mouse ®broblasts to acquire a fully transformed review, we will describe the role of GPCRs in normal phenotype (Aaronson and Tronick, 1985; Kraus et al., and aberrant cell growth and will then focus on recent 1988; Tronick and Aaronson, 1988). Thus, it has been e€orts aimed to elucidate their downstream intracel- suggested that unrestricted activation of proliferative lular signaling pathways controlling cell proliferation. pathways contribute to the malignant state. Discovery Cell growth control by G protein coupled receptors JS Gutkind 1332

Figure 1 Divergent kinase cascades link G protein-coupled receptors to the nucleus. (see text for details)

of the mas oncogene (Young et al., 1986) provided the cally expressed in NIH3T3 cells and activated by the ®rst link between cellular transformation and GPCRs. serotonin present in the culture medium (Julius et al., The mas oncogene, which encodes a putative GPCR, 1989). Using a more de®ned experimental system, was initially cloned using standard transfection assays muscarinic m1, m3, and m5 receptors were shown to by virtue of its ability to induce tumors in mice (Young transform NIH3T3 cells in a strictly agonist-dependent et al., 1986). The natural agonist for the mas oncogene manner (Gutkind et al., 1991). As the natural agonist product is still unknown, but is likely to be a serum for these receptors, the neurotransmitter acetylcholine, component. Similarly, serotonin 1C receptors were is not present in serum, it was possible to correlate shown to harbor transforming potential when ectopi- biochemical responses with the biological e€ects Cell growth control by G protein coupled receptors JS Gutkind 1333 elicited by an exogenously added cholinergic agonist. lymphomas, in several non-human primates (Nicholas

In addition, this approach established that while Gq et al., 1992). The protein product of one of its genes, coupled receptors, m1, m3, and m5, were highly ECRF3, behaves as a functional chemokine receptor, transforming, Gi coupled receptors, m2 and m4, where with a ligand speci®city identical to that of the IL8 without focus-forming activity, under those experi- receptor (Nicholas et al., 1992). Furthermore, the mental conditions. Together, these ®ndings demon- HVS encoded GPCR responds biochemically to IL8, strated that certain GPCRs could behave as potent suggesting a role for chemokines in the molecular agonist-dependent oncogenes, and raised the possibility pathogenesis of viral infection. Interestingly, the that activating mutations in GPCRs may render them KSHV-encoded GPCR is also highly related to the transforming. Indeed, this was subsequently shown for chemokine receptor family, and has been recently the a1b G protein-linked receptor to noradrenaline shown to behave as a constitutively active Gq-coupled (Allen et al., 1991). In this case, speci®c point receptor and, as predicted, to be sucient to subvert mutations in the C-terminal juxtamembrane region of normal growth control when expressed in cultured the third intracellular loop of the a1b receptor relieved ®broblasts (Arvanitakis et al., 1997). In addition, the requirement of ligand activation for its transform- recent evidence has indicated that the KSHV-GPCR ing ability. Furthermore, the recent identi®cation of can e€ectively activate signaling pathways typical of constitutively activating mutations in TSH receptors in transforming receptors, and play an unsuspected role 30% of thyroid adenomas (Parma et al., 1993) in angiogenesis (Bais et al., in press). provided now a direct link between this class of Consistent with a role for GPCRs in normal and receptors and human cancer. Similarly, mutationally aberrant cell growth, constitutively active mutants of activated LH receptors have been identi®ed in a form Gai,Gaq,Ga0,Ga12 and Ga13 were also shown to of familial male precocious puberty that results from behave as transforming genes in a variety of cell types, hyperplastic growth of Leydig cells (Shenker et al., and mutationally activated Ga proteins were identi®ed 1993). in several disease states (reviewed in Dhanasekaran et

Even more prevalent than activating mutations, al., 1995). For example, mutationally activated Gas, paracrine and autocrine stimulation of multiple referred to as the gsp oncogene (Landis et al., 1989), GPCRs for neuropeptides and prostaglandins has results in hyperplasia of endocrine cells, and are been implicated in a number of human neoplasias. present in human thyroid and pituitary tumors Particularly, bombesin, gastrin-releasing peptide (reviewed in Dhanasekaran et al., 1995), and in the (GRP), neuromedin B, bradykinin, cholecystokinin McCune-Albright syndrome, which exhibits autono- (CCK), galanin, neurotensin and vasopressin have mous hyperproliferation in multiple endocrine glands been implicated in human small cell lung cancer (Weinstein et al., 1991). Activating mutations have also

(SCLC) (Cuttitta et al., 1985; Moody and Cuttitta, been identi®ed for Gai2, referred to as the gip2 1993; Schuller, 1991a,b; Sethi et al., 1992). Many of oncogene, in a subset of ovarian sex cord stromal these neuropeptides are secreted by tumor-derived cells tumors and adrenal cortical tumors (Lyons et al., or by the solid tumor, suggesting an autocrine or 1990). In addition, Ga12, referred as the gep oncogene paracrine mechanism of action (Sethi and Rozengurt, (Xu et al., 1993, 1994), was isolated as a transforming 1991). Similarly, neuropeptide receptors and their gene from a soft tissue sarcoma-derived cell (Chan et agonists have been implicated in colon adenomas and al., 1993), although its role in tumorigenesis is still carcinomas (Hoosein et al., 1988), and gastric unclear. No naturally occurring activated mutations in hyperplasia and cancer (Tahara, 1990). Interestingly, members of the Gaq family have been described to antibodies against these peptides or receptor antago- date. Collectively, these observations implicate G nists have been shown to be e€ective in in vitro and in protein a subunits and their linked cell surface vivo models for SCLC, further supporting a potential receptors in hyperproliferative diseases, including causative role for GPCRs in this disease, thus cancer. providing an opportunity for novel treatment mod- alities (Langdon et al., 1992; Sethi et al., 1992; Thomas et al., 1992a). Sequences encoding functional GPCRs have also Mitogenic signaling through G protein-coupled receptors been found in the genome of a number of DNA viruses, including human cytomegalovirus (HCMV) Recent work has revealed that multiple intracellular (Bankier et al., 1991; Chee et al., 1990), herpes virus signaling pathways mediate the proliferative e€ects of saimiri (HVS) (Nicholas et al., 1992) and Kaposi's GPCRs. Initial studies addressing this issue had sarcoma associated herpesvirus (KSHV) (Arvanitakis focused primarily on second messenger generating et al., 1997). The HCMV, which infects leukocytes, systems, such as adenylyl cyclases and phospolipases. ®broblasts and epithelial cells, contains in its genome However, although certain growth promoting agents three predicted open reading frames, UL27, UL28, such as LPA and thrombin activate G proteins of the and UL33, encoding typical GPCRs, which are Gi family and thus inhibiting adenylyl cyclase, no believed to represent functional chemokine receptors direct correlation has been observed between induction (Ahuja and Murphy, 1993; Neote et al., 1993). of DNA-synthesis and decreased intracellular levels of Although their function in the viral life cycle is still cAMP. By contrast, many studies have implicated unclear, they are believed to help elude natural host phospholipase C-b activation and enhanced phospha- defenses through molecular mimicry of proteins tidylinositol bisphosphate (PIP2) hydrolysis in mitogen- normally involved in host defense mechanisms esis (Rozengurt, 1986). However, experiments utilizing (Ahuja and Murphy, 1993). The HVS causes fatal mutant tyrosine kinase receptors provided evidence lymphoproliferative diseases, including leukemias and that PIP2 hydrolysis may be neither necessary nor Cell growth control by G protein coupled receptors JS Gutkind 1334 sucient for mitogenesis (Coughlin et al., 1989; these receptors were known to activate G proteins of

Mohammadi et al., 1992). Consistent with those the Gi family, and were therefore sensitive to Ptx. In observations, a number of agonists acting on GPCRs, contrast, other receptors were known to couple to

can e€ectively induce PIP2-hydrolysis, but fail to phospholipase-C PIP2 hydrolysis by a Ptx-insensitive G

stimulate growth when added alone to quiescent cells protein, likely of the Gq family (Moolenaar, 1991). (Moolenaar, 1991). Although we cannot rule out that Thus, GPCRs could signal MAPK activation utilizing

in certain cellular settings the persistent activation of either Gi or Gq-dependent pathways. How these

the phospholipase C-PIP2 hydrolysis pathway might heterotrimeric G proteins then communicate to

elicit a mitogenic response and even lead to neoplastic MAPK was unclear. For Gq-coupled receptors, PKC- conversion, the available body of information strongly dependent and independent pathways were reported,

suggests that additional e€ector pathways participate but expression of an activated mutant of Gaq did not in the proliferative signaling by G protein-linked enhance MAPK activity when stably expressed in receptors. In this regard, recent work has revealed NIH3T3 cells (our unpublished results) or transiently

that the family of extracellular signal-regulated kinases, expressed in some cellular settings. For Gi-coupled (ERKs) or MAP kinases, is a central component of the receptors, activation of MAPK was shown to be intracellular signaling pathway(s) controlling cell largely PKC-independent (Hawes et al., 1995). In the

proliferation (see Marshall, 1995). For example, the latter case, activated forms of Gai2, the gip2 oncogene, enzymatic activity of ERK1 and ERK2, p44MAPK and was shown to induce MAPK activation when expressed p42MAPK, respectively, referred herein as MAPKs, in rodent ®broblasts (Gupta et al., 1992). However, increases in response to mitogenic stimulation, and MAPK stimulation by gip2 was found to be Ptx- impeding their function prevents cell proliferation in insensitive, suggesting an indirect mechanism of response to a number of growth stimulating agents activation likely to result from secondary events (Pages et al., 1993). Moreover, constitutive activation involved in the acquisition of the transformed of molecules upstream of the MAPK pathway is phenotype (Gupta et al., 1992). Thus, the mechanism

sucient for tumorigenesis (Mansour et al., 1994; of activation of MAPK by receptors linked to Gq or

Schlessinger, 1993). Gi-related G proteins was largely unknown until Growth factor receptors of the tyrosine kinase class recently. activate MAPKs in a multistep process. For example, binding of EGF to its cognate receptors leads to the tyrosine phosphorylation, in the EGF receptor, of a A role for bg subunits of heterotrimeric G proteins docking site for the adaptor protein GRB2/SEM-5 linking G protein-coupled receptors to the Ras-MAPK (reviewed in Schlessinger, 1993). This causes the pathway recruitment of a guanine nucleotide exchange protein for Ras, SOS, to the plasma membrane, and the To investigate the mechanism of MAPK activation by subsequent exchange of GDP for GTP-bound to Ras. GPCRs, a number of laboratories have used the GTP-bound Ras then proceeds to activate a cascade of transient coexpression of an epitope-tagged form of protein kinases de®ned sequentially as MAP kinase MAPK together with GPCRs in readily transfectable kinase kinase, and represented by A-Raf, B-Raf, and cell lines, such as COS-7 cells. An advantage of this Raf-1, and MAP kinase kinase such as MEK1 and system is that several proteins can be simultaneously MEK2. MEKs ultimately phosphorylate p44MAPK and expressed at very high levels, without the in¯uence of p42MAPK, on both threonine and tyrosine residues, biological changes that might be manifested during thereby increasing their enzymatic activity. In turn, prolonged culturing of cells. Upon transfection into MAPKs phosphorylate and regulate the activity of key COS-7 cells, the epitope tagged MAPK could then be enzymes and nuclear proteins, which ultimately immunoprecipitated with an anti-epitope monoclonal regulate the expression of genes essential for prolifera- antibody, and its enzymatic activity assayed using tion (reviewed in Davis, 1993). Although the mechan- standard in vitro kinase assays (Crespo et al., 1994b; ism of activation of MAPKs by GPCRs is still poorly Pages et al., 1993). In this cellular setting, it was understood, recent e€orts from a number of labora- observed that MAPK was potently activated upon

tories is helping to elucidate how this family of cell ligand addition by either Gq-coupled or Gi-coupled surface receptors signal to MAPKs. receptors, respectively, in a Ptx-insensitive and -sensitive fashion (Crespo et al., 1994b; Faure et al., 1994; Koch et al., 1994). However, under identical

G protein-coupled receptors signal to the MAPK experimental conditions, activated forms of Gai2,Gaq,

pathway Gas or Ga12 did not induce any detectable increase in MAPK activity (Crespo et al., 1994b). Activation of MAPKs in response to many agonists Why the expression of activated Ga subunits failed to acting on GPCRs is well documented. For example, mimic receptor-mediated stimulation of MAPK is ligands as diverse as bombesin, endothelin-1, somatos- unclear. However, as discussed above, activated tatin, thromboxane A2, prostaglandin F2a, interleukin- GPCRs act as guanine-nucleotide exchange factors for 8, platelet-activating factor (PAF), LHRH, TRH, heterotrimeric G proteins, inducing the release of the formyl-methionyl peptide (fMLP), C5a peptide, sphin- Ga-bound-GDP and its replacement for GTP and the gosine-1-phosphate, oxytocin, and angiotensin II, have consequent dissociation of GTP-bound-Ga from bg been shown to e€ectively activate MAPKs upon complexes (Clapham and Neer, 1997). Although it was stimulation of their cognate endogenously expressed initially thought that Ga subunits were the sole receptors, in a variety of cell types (see van Biesen et molecules responsible for signal transmission, it is now al., 1996 for a recent review). Interestingly, many of clear that both activated Ga and free bg subunits can Cell growth control by G protein coupled receptors JS Gutkind 1335 stimulate a variety of e€ector pathways (Clapham and GPCR stimulation (Daub et al., 1996; Linseman et al., Neer, 1997). That observation and the failure of 1995), and to participate in MAPK activation by mutationally activated Ga subunits to activate MAPK GPCRs. We can conclude that a number of receptor prompted the exploration of a role for bg heterodimers and non-receptor tyrosine kinases might link GPCRs in signaling to the MAPK pathway. This led to the to the Ras-MAPK pathway. The relative contribution observation that membrane-bound forms of bg-subunits of each of these, or others yet to be discovered tyrosine can directly activate biochemical routes leading to kinases, in GPCR signaling to MAPK is likely to be MAPK activation (Crespo et al., 1994b), and stimu- cell type-speci®c, and is clearly a subject that warrants lated the search for molecules acting downstream of Gbg further investigation. in this signaling pathway. Using a variety of experi- Additional signaling molecules have been recently mental approaches, it was shown that stimulation of shown to connect GPCRs and Gbg to the Ras-MAPK MAPK activity by coexpressed bg dimers required pathway. These include the protein tyrosine phospha- neither PLC-b nor PKC activation, but was blocked by tase SH-PTP1 (Gaits et al., 1996), Ras-GRF (Mat- molecules inhibiting Ras function (Crespo et al., 1994b; tingly and Macara, 1996) and PI3Kg (Lopez-Ilasaca et Koch et al., 1994). Furthermore, bg subunits were later al., 1997). Ras-GRF is a distinct Ras guanine- shown to be sucient to induce the accumulation of nucleotide exchange factor expressed in neuronal Ras in the GTP-bound, active form (Koch et al., 1994). cells, and its activity was found to be enhanced in Taken together, these ®ndings indicate that signaling response to GPCR stimulation or upon coexpression of from GPCRs to MAPK involves bg heterodimers acting Gbg by a still unclear mechanism (Mattingly and on a Ras-dependent pathway. This provided the ®rst Macara, 1996). In the case of PI3K, the observation indication that bg subunits link heterotrimeric G that wortmannin, a phosphatidylinositol-3-kinase proteins to small G proteins of the Ras superfamily, (PI3K) inhibitor, can diminish MAPK activation by and also established that the GPCR signaling pathway GPCRs (see Hawes et al., 1996), provided the ®rst converges with that emerging from receptors of the indication of a role for this lipid kinase in GPCR tyrosine kinase class at the level of Ras. signaling to MAPK. Recently, a novel PI3K isoform that does not bind the p85 PI3K non-catalytic subunit was identi®ed and termed PI3Kg (Stoyanov et al.,

Molecular mechanisms linking GPCRs and Gbg to Ras 1995). This novel PI3K is not stimulated by tyrosine phosphorylation, but instead is activated upon direct

The elucidation of the molecular events linking GPCRs physical interaction with Gbg complexes (Stoyanov et and Gbg to the Ras-MAPK pathway has received al., 1995). This PI3Kg was found to act downstream extraordinary attention in the last few years. Perhaps from Gbg and upstream of Src-like kinases and Shc, the ®rst indication that tyrosine kinases might mediate Grb2, Sos and Ras, thus suggesting a potential the activation of MAPKs by GPCRs came from the mechanism whereby heterotrimeric G proteins can observation that genistein, a relatively non-speci®c regulate non-receptor tyrosine kinases and, in turn, tyrosine kinase blocker, diminishes MAPK stimulation control the MAPK pathway (Lopez-Ilasaca et al., by LPA (Hordijk et al., 1994). More recent studies 1997). demonstrated that activation of GPCRs in several MAPK activation by GPCRs in a Ras-independent cellular systems leads to the rapid phosphorylation of manner has also been reported based upon the inability the adaptor protein Shc on tyrosine residues and the to observe accumulation of Ras in the GTP-bound formation of Shc-GRB2 complexes (Chen et al., 1996; form in response to GPCR stimulation (Pace et al., van Biesen et al., 1995). Furthermore, Lutrell et al. 1995; Takahashi et al., 1997). However, dominant (1996) have recently provided evidence that Src, or a interfering mutants for Ras prevented MAPK activa- Src-like kinase, mediates the bg induced phosphoryla- tion even in systems where GTP-bound Ras was not tion of Shc, the recruitment of GRB2 and SOS, and readily demonstrable (Pace et al., 1995). Thus, it is the subsequent stimulation of the Ras-MAPK path- possible that additional pathways might cooperate with way. Extending these observations, recent studies have limited amounts of GTP-bound Ras to stimulate implicated other non-receptor tyrosine kinases in MAPK, or that in certain cellular settings, GPCRs signaling from GPCR to MAPKs. These include may stimulate MAPKs by yet to be identi®ed several Src-like kinases such as Fyn, Lyn, Yes, and biochemical routes bypassing the requirement for Ras. the more distantly related Syk (Ptasznik et al., 1995; One such example may involve PKCs, as direct Wan et al., 1996), and a novel Ca2+ and PKC- activation of PKCs by phorbol esters has been shown dependent protein tyrosine kinase, Pyk2 (Della Rocca to induce MAPK by a mechanism that is still unclear. et al., 1997; Dikic et al., 1996; Lev et al., 1995). Although PKCs phosphorylate Raf (Heidecker et al., Interestingly, the latter is closely related to FAK (Focal 1992; Kolch et al., 1993), this does not result in Adhesion Kinase) which is involved in integrin increased Raf phosphorylating activity on its natural signaling, and participates in the formation of focal substrate, MEK (Macdonald et al., 1993). In addition, complexes containing a number of molecules, including PKC activation appears not to induce Ras-GTP Src, paxillin, dynamin and Grb2. FAK was earlier accumulation in the vast majority of cell types (Bos, shown to be activated by GPCRs (Gutkind and 1995), however, it activates MAPK in a Ras-dependent Robbins, 1992; Rankin et al., 1994), and may manner in several cellular settings (Burgering and Bos, represent an additional candidate mediating GPCR 1995; Hawes et al., 1995; Thomas et al., 1992b). As signaling to MAPK. Receptors of the tyrosine kinase Ras alone is not sucient to activate Raf fully in vitro class have been also shown to play a role in GPCR (Macdonald et al., 1993) or when coexpressed in Sf9 signaling. For example, PDGF and EGF receptors cells (Williams and Roberts, 1994), it is possible that were found to become tyrosine-phosphorylated upon PKCs might act on the MAPK pathway by facilitating Cell growth control by G protein coupled receptors JS Gutkind 1336 the full activation of Raf upon binding to Ras (see These ®ndings suggested that MAPKs but not Burgering and Bos, 1995). Thus, PKC activation would JNKs were located downstream from Ras, as PDGF then be expected to play a critical role on MAPK and other agonists known to activate the Ras pathway stimulation under conditions of submaximal activation failed to stimulate JNK activity in a number of of Ras by cell surface receptors. This might help cellular settings (Coso et al., 1995a; Kyriakis et al.,

explain some con¯icting observations regarding Gq- 1994), and raised the possibility that proteins other coupled receptors, as these receptors can activate than Ras may directly regulate biochemical pathways MAPK in a PKC-dependent (Hawes et al., 1995), leading to the activation of JNK. In this regard, the fully PKC-independent (Charlesworth and Rozengurt, Ras superfamily of GTPases comprises more than 50 1997), or partially PKC-dependent (Crespo et al., members, which have been divided in six families 1994a) manner. Thus, depending on the cellular based upon sequence similarity: Rab, Arf, Sar, Ran,

system and the level of receptor expression, Gq- Rho and Ras (see Hall, 1994). The Rab, Arf and Sar coupled receptors would induce MAPK through a groups have been shown to participate in the

biochemical route including Gaq acting on phospholi- transport of proteins and vesicles among di€erent

pase C and PKC, or initiated by Gqbg acting on a Ras- intracellular compartments (Boguski and McCormick, dependent mechanism. At low receptor level or low 1993; Hall, 1994). The Rho-family of GTP-binding concentration of agonist, the PKC component may proteins consists of the Rac, Rho and Cdc42

play a critical role, as the limited availability of Gqbg subfamilies, and have been shown to regulate several might not be sucient to cause substantial Ras aspects of cytoskeleton functioning (Hall, 1994). For activation. example, the Rac subfamily includes Rac1 and Rac2, We can conclude that although additional molecules the former involved in membrane ru‚ing (Hall, 1994) may participate in MAPK activation by GPCRs and and the latter in NADPH oxidase-catalyzed super-

Gbg, many of the likely candidates may have already oxide formation in neutrophils. The Rho subfamily been identi®ed. Interestingly, several of the proteins has at least 6 members, RhoA, RhoB, RhoC, Rho6, described above exhibit a very restricted tissue Rho7, and Rho8. RhoA has been shown to distribution (Lev et al., 1995; Mattingly and Macara, participate in the formation of actin-stress ®bers, as 1996; Stoyanov et al., 1995). Thus, the nature of the well as in mediating redistribution of cytoskeletal

molecules linking GPCRs and Gbg to MAPK stimula- components (Hall, 1994). The Cdc42 group consists of tion is expected to depend on the repertoire of Cdc42Hs (referred here as Cdc42), G25K and RhoG signaling molecules available in each particular (Hall, 1994), and its function in mammalian cells is cellular system. still largely unknown. Of interest, it had been reported that GTP-bound forms of the Rho-related proteins Rac1 and Cdc42 can speci®cally associate and activate Novel signaling pathways communicate GPCRs to the a novel serine-threonine kinase, PAK (Manser et al., nucleus 1994). This situation was highly analogous to that of the Ras-Raf interaction, thus suggesting that the Rho GPCRs and tyrosine kinase receptors converge at the family of GTP binding proteins might also initiate level of Ras to initiate the activity of a kinase cascade activity of a kinase cascade (Manser et al., 1994). This leading to MAPK activation and transcriptional observation prompted several laboratories to examine regulation. Thus, stimulation of either type of receptor whether the Rho family of GTPases participates in would be expected to elicit a similar response at the signaling to the JNK pathway. level of nuclear transcription factors. However, Recently, it was found that the small GTP-binding activation of GPCRs in NIH3T3 cells were found to proteins Rac1 and Cdc42 initiate an independent induce a very distinct pattern of expression of kinase cascade regulating JNK activity (Coso et al., immediate early genes of the jun and fos family 1995b), and that Rac and Cdc42 are an integral part of (Coso et al., 1995a). In particular, activation of m1 the signaling route linking many cell surface receptors, muscarinic receptors but not of PDGF receptors led to including GPCRs, to JNK (Coso et al., 1996; Minden a remarkable expression of c-jun (Coso et al., 1995a). et al., 1995). Many novel components of this pathway Surprisingly, this response did not correlate with have been recently identi®ed, and they include, MAPK activation (Coso et al., 1995a), thus suggesting sequentially, Pak and other members of the Ste20 that GPCRs control a distinct biochemical route family of serine-threonine kinase, members of the Ste11 regulating gene expression. In this regard, recent work family of kinases, including MEKK, and two members demonstrated that the activity of the transcription of the Ste7 class of kinases, including Sek/JNKK1/ factor c-jun is controlled by a novel family of enzymes MKK4 and MKK7 (Fanger et al., 1997; Tournier et structurally related but clearly distinct from MAPKs. al., 1997; Whitmarsh et al., 1997). Detailed examina- These enzymes, named jun kinases (JNKs) (Derijard et tion of the pathway linking GPCRs to JNK provided al., 1994) or stress activated-protein kinases (SAPKs) evidence that free bg dimers (Coso et al., 1996) and, in

(Kyriakis et al., 1994), selectively phosphorylate the N- some cellular systems, Ga12 (Prasad et al., 1995) convey terminal transactivating domain of the c-jun protein signals from this class of receptors to JNK. thereby increasing its transcriptional activity. Interest- Furthermore, we and others have recently shown ingly, when JNK activity was investigated, it was that JNK is potently activated by several observed that GPCRs, but not PDGF receptors, naturally occurring human oncogenes (Coso et al., potently activated JNK in NIH3T3 cells (Coso et al., 1995b; Crespo et al., 1997; Minden et al., 1995; 1995a), thus establishing that the GPCR signaling Teramoto et al., 1997; Whitehead et al., 1997). pathways diverge at the level of JNK from those However, whereas a function for JNK in apoptosis utilized by certain tyrosine kinase receptors. has been recently established (Goillot et al., 1997; Cell growth control by G protein coupled receptors JS Gutkind 1337 Johnson et al., 1996; Xia et al., 1995), the role for JNK signal transduction in yeast, we can postulate that still in cellular transformation is still unclear. unknown sca€olding proteins may also participate in the mammalian pathways connecting heterotrimeric G proteins to MAPK cascades.

Unsuspected role for Gbg in signal transduction in mammalian cells ± The search for sca€olding proteins Additional pathways connect G proteins to the nucleus: Whereas the role of Ga subunits in signal transduction Rho links GPCRs to the serum response element and to has been extensively studied, a newly available body of cellular transformation evidence suggests that in mammalian cells, bg subunits of heterotrimeric G proteins communicate GPCRs with As described above, activation of GPCRs and tyrosine the MAPK and JNK pathways acting, respectively, on kinase receptors induces a distinct pattern of expres- a Ras and Rac1/Cdc42-dependent biochemical route. sion of early response genes of the c-jun and c-fos Thus, the emerging picture is that bg heterodimers family. However, these two classes of receptors elicited provide a molecular bridge between heterotrimeric G a similar response on c-fos expression (Coso et al., proteins and small GTP-binding proteins. Although 1995a). The induction of c-fos transcription is mediated this may represent a novel concept in signal transduc- by several promoter elements (Hipskind and Nord- tion, its similarity with the pathway linking G protein- heim, 1991). Among these, the Serum Response coupled pheromone receptors to MAPK-like proteins Element (SRE) is believed to play a central regulatory in the budding yeast S. cereviseae is striking, and (Treisman, 1994, 1995), as this sequence has been suggests that the Gbg connection to small GTPases shown to be necessary and sucient for the rapid represents a biologically relevant example of a pathway induction of c-fos by most growth-promoting stimuli extraordinarily conserved throughout evolution. In S. (Greenberg et al., 1987; Johansen and Prywes, 1994). A cereviseae, extracellular ligands (a or a factors) activate number of proteins that bind the c-fos SRE can cell surface pheromone receptors which, in turn, induce mediate SRE-dependent transcription (Treisman, the dissociation of a heterotrimeric G protein into a 1994, 1995), including a transcription factor of (GPA1) and bg (Ste4, Ste18) subunits (see Herskowitz, 67 kD, termed Serum Response Factor (SRF), that 1995). Free bg dimers then activate a serine-threonine binds the SRE in vivo and in vitro as a dimer kinase, Ste20, stimulating the activity of a linear (Treisman, 1994, 1995). Another protein forms a cascade of kinases including sequentially Ste11 and ternary complex with SRE and the SRF dimer and Ste7, which phosphorylate and activate the yeast has been therefore termed Ternary Complex Factor or MAPK homolog Fus3 and Kss1. Extensive search for p62TCF (Treisman, 1995). TCF can be phosphorylated molecules linking yeast bg complexes to Ste20 has led by both MAPK and JNK (Cavigelli et al., 1995; Gille to the recent discovery that Cdc42 participates in Ste20 et al., 1995; Whitmarsh et al., 1997), thus providing a activation (Simon et al., 1995) and that Ste4 might mechanism whereby SRE activity can be regulated in directly bind and activate Cdc24, a nucleotide exchange response to the activation of the Ras-MAPK and factor for Cdc42 (Zhao et al., 1995). Thus, in yeast, the Rac1/Cdc42 JNK signaling pathways. Interestingly, G protein b subunit can initiate activity of a MAPK SRE can also be regulated in a TCF-independent cascade by binding an exchange factor for the Cdc42 manner, and certain TCF-independent signaling path- GTPase, and then this GTP-binding protein physically ways acting on SRE have been shown to be controlled interacts with the most upstream kinase, Ste20, causing by the small GTP-binding protein Rho (Hill et al., its activation (Herskowitz, 1995). Available data 1995). Regarding the latter, evidence presented by Hill suggest that additional molecules might be also et al. (1995) suggests that Rho signals to the SRE involved, including a protein designated Ste5, which through a novel pathway independent of any MAPKs binds yeast bg and plays a role as a platform or described to date. Furthermore, recent work estab- sca€old recruiting Ste11, Ste7, Kss1/Fus3 (Akada et lished that GPCRs can signal through this novel TCF- al., 1996; Whiteway et al., 1995). independent pathway (Hill et al., 1995), and that RhoA In mammalian cells a more complex array of or other Rho-related proteins act as integral compo- signaling molecules appears to connect GPCRs and nents of this biochemical route connecting GPCRs,

Gbg to parallel MAPK cascades. These include lipid likely through Gbg and Ga12, to both transcriptional kinases, non-receptor and receptor tyrosine kinases, activation of the SRE and neoplastic transformation SH2 and SH3-containing adapter molecules, protein- (Fromm et al., 1997). phosphatases, PKCs, and Ras-GRF. However, we can not exclude the possibility that G proteins may directly regulate the activity of unidenti®ed guanine-nucleotide Network of GTPases exchange factors, as observed in yeast. Although no mammalian homolog for Ste5 has been so far Another interesting area for investigation involves the described, a recent report indicates that a PDZ- interrelationship among small GTPases. Microinjection containing protein termed InaD acts as a sca€old studies suggest that Cdc42, Rac and RhoA can be linking Gaq to several molecular components of the considered, sequentially, as part of a `small GTP- visual system of the fruit ¯y (Tsunoda et al., 1997). A binding protein cascade' (Nobes and Hall, 1995a,b). mammalian InaD homolog has been recently described However, the mechanism by which one small G protein (Philipp and Flockerzi, 1997), and several putative activates the subsequent one is yet unknown. Another PDZ-containing adaptor can be found in the interesting ®nding is that although activation of JNK GeneBank (unpublished observation). Based upon by G protein-linked and tyrosine kinase receptors is these recent ®ndings and the critical role of Ste5 in initiated by Rac1/Cdc42, the dominant negative Cell growth control by G protein coupled receptors JS Gutkind 1338 mutant of Ras can also block JNK-stimulation by Receptor and pathway cross-talk these receptors (Coso et al., 1995b), further supporting a functional relationship among these GTPases. In this Recent studies have revealed that activation of regard, evidence for a regulatory role for a Ras-like angiotensin receptors in smooth muscle cells causes protein on the activation of another, Cdc42, has the activation of tyrosine kinase receptors for PDGF already been obtained in the yeast pheromone path- (Diverse-Pierluissi et al., 1997; Du et al., 1996), and way. In this case, the Ras homolog Rsr1 appears to stimulation of LPA receptors leads to the tyrosine play an important regulatory role by binding Cdc24 phosphorylation of EGF receptors in Cos-7 (Daub et and, probably, by positioning this guanine nucleotide al., 1996) and HEK 293 cells (Luttrell et al., 1997), the exchange factor within the cell, thus allowing its latter resulting from the activation of Src-like kinases interaction with Cdc42 and its targets (Zheng et al., by GPCRs (Luttrell et al., 1997). While the biological 1995). These observations represent the ®rst example of signi®cance of cross-talk between GPCRs and tyrosine convergence between Ras-like (Rsr1) and Rho-like kinase receptors is still unclear, at the biochemical level (Cdc42) biochemical routes, and might help explain tyrosine phosphorylated receptors might provide a the regulatory e€ect of Ras on Rac/Cdc42 in docking site for the formation of multiprotein signal mammalian cells. Conversely, Rho-like proteins might transducing complexes. Similarly, GPCR stimulation also regulate Ras exchange factors, Sos and Ras-GRF, lead to the activation of FAK (Gutkind and Robbins, upon interaction with their DH domains, a possibility 1992), a component of the integrin-signaling pathway that has not yet been fully explored. (Hanks and Polte, 1997; Schlaepfer and Hunter, 1996). Thus, biological events triggered by GPCRs might involve the lateral dissemination of the signal, thus Regulation of kinase cascades by second-messenger ge- engaging the functioning of other cell surface receptors. nerating systems Another emerging theme is the ability of one kinase cascade to in¯uence the biochemical or biological Novel signaling e€ects for elevated intracellular [Ca2+] outcome of another. For example, recent work has have been recently reported. These include the ability established that the balance between JNK and MAPK to enhance the guanine-nucleotide exchange activity of activation can determine whether cells will undergo Ras-GRF through a conserved domain termed `IQ' programmed cell death or survive (Cuvillier et al., (Buchsbaum et al., 1996; Farnsworth et al., 1995; 1996; Xia et al., 1995). In addition, Cobb and co- Freshney et al., 1997); to activate a novel tyrosine workers have recently shown that activation of Rac1- kinase, Pyk2, implicated in GPCR signaling to JNK Pak can synergize the submaximal stimulation of the and MAPK in certain cells (Della Rocca et al., 1997; MAPK pathway, probably by direct phosphorylation Dikic et al., 1996; Lev et al., 1995; Tokiwa et al., 1996); of MEK, resulting in enhanced ability to be activated to activate a novel calcium-calmodulin-dependent by Raf (Frost et al., 1997). The implication of this P13K activity (Joyal et al., 1997); and to activate the study is that proteins at the level of MEK (MEK1, dephosphorylation of nuclear transcription factors MEK2, MKK3, MKK4/SEK/JNKK1, MEK5, through the calmodulin-dependent phosphatase, calci- MKK6, MKK7) may be susceptible to regulation by neurin (Chow et al., 1997; Sugimoto et al., 1997). Thus, components of other kinase cascades, an exciting it is becoming increasingly clear that limited changes in prediction that warrants further investigation. intracellular [Ca2+], caused by many G protein-linked and tyrosine kinase receptors, might exert a profound e€ect on many aspects of signal transmission. Signal integration in the nucleus Another classical second messenger that can in¯uence signal transmission is cAMP. In many cell Perhaps the most obvious and yet poorly understood types, increased levels of cAMP will lead to a PKA- site for signal integration lies in the nucleus. dependent inhibition of MAPK stimulation (Burgering Regulation of transcriptional activity by phosphoryla- et al., 1993; Chen and Iyengar, 1994, 1995). The precise tion of transcription factors appears now to be the target for PKA is still unclear, but current evidence norm rather than the exception. These events can cause suggests that PKA phosphorylates Raf thus preventing increased activity of transcription factors by enhancing its activation by Ras (see Burgering and Bos, 1995), their binding to co-activators, increasing their protein and that PKA activates certain inhibitory molecules, stability, a€ecting their ability to dimerize with other such as the small GTP-binding protein Rap1 transcription factors, or causing their dissociation from (Altschuler et al., 1995). Surprisingly, however, in transcriptional repressors (Hill and Treisman, 1995; some cells, including PC12 and Swiss 3T3 cells, cAMP Karin and Hunter, 1995; Su and Karin, 1996; elevating agents induce MAPK activation (Faure and Treisman, 1996). However, phosphorylation of tran- Bourne, 1995; Faure et al., 1994; Vossler et al., 1997), scription factors can also cause a dramatic decrease in and a recent report has suggested that the latter the ability to bind DNA, hence inhibiting their involves Rap1b acting on B-Raf (Vossler et al., 1997). functional activity (Hill and Treisman, 1995). In the Interestingly, several adenylyl cyclases have been last few years we have witnessed an explosion in the

identi®ed and shown to be regulated by Gas,Gai, number of potential transcription-factor kinases 2+ Gbg,Ca or PKC in a subtype-speci®c manner (Davis, 1995; Goedert et al., 1997), and we are only (Sunahara et al., 1996; Taussig and Gilman, 1995). just beginning to understand how the interplay among These ®ndings suggest that many cell surface receptors these kinases and positive and negative modulatory will activate adenylyl cyclases thus initiating a complex events control transcriptional activity. series of regulatory events on MAPK cascades, an area Another area that is currently being revisited is the that is still largely unexplored. complexity of regulatory elements found in the Cell growth control by G protein coupled receptors JS Gutkind 1339 promoter region of growth-regulating genes. For From signal transduction to signal integration example, the idea that JNK activation by cell surface receptors leads to c-jun expression through an AP1 site The simplest model of how GPCRs and other cell in the c-jun promoter has been recently challenged surface receptors communicate to the nucleus and (Coso et al., 1997). Unexpectedly, it was observed that control gene expression, involves the activation of a critical regulatory element in the c-jun promoter small GTP-binding proteins of the Ras superfamily binds members of the MEF2 family of transcription which, in turn, control the activity of parallel kinase factors (Coso et al., 1997), and one such MEF2 cascades resulting in the phosphorylation of critical protein, MEF2C, was recently found to be regulated nuclear transcription factors. However, activation of by the p38 MAPK (Han et al., 1997). Thus, GPCRs leads to a number of signaling events that, information currently available suggests that expres- although dissectable for their examination under sion of c-jun may be under the control of p38-related controlled experimental conditions, do not occur in pathways. Consistent with this hypothesis, stimulation isolation in their natural settings. As outlined above, of cells derived from JNK3 knock out animals, did not upon GPCR activation, both Ga and Gbg stimulate result in either detectable activation of JNK nor biochemical routes leading to the concomitant activa- phosphorylation of c-Jun (Yang et al., 1997). Never- tion of several small GTP-binding proteins of the Ras- theless, enhanced expression of c-jun messages in superfamily, a variety of second messenger generating response to excitatory stimuli was similar to that of systems, and even other cell surface receptors, such as control cells (Yang et al., 1997). Thus, the c-jun PDGF and EGF receptors. These, in turn, stimulate a promoter may be the site of convergence of several number of highly interconnected cytoplasmic signaling kinase cascades. Similarly, as described above, the c-fos pathways that lead to a temporally distinct pattern of promoter contains an SRE that can be regulated by activation of members of the MAPK superfamily, both MAPK and JNK acting on the TCF, and by a yet including many recently identi®ed (Goedert et al., 1997; to be identi®ed pathway, initiated by Rho, a€ecting Lawler et al., 1997). Ultimately, these kinases will SRF function (Hill et al., 1995). However, the c-fos reach the nucleus and thus control gene expression by promoter also contains CRE, SIE, Spl and other sites changing the status of phosphorylation of nuclear adjacent to the SRE (Hipskind and Nordheim, 1991; proteins, including transcription factors, and thereby Johansen and Prywes, 1994; Runkel et al., 1991), a€ecting an intrincated balance of nuclear regulatory suggesting that the regulation of the c-fos promoter molecules. Therefore, the ®nal biological outcome, may depend on the multiple positive and negative including cell proliferation, most likely results from a interactions likely to occur among their corresponding network of responses, rather than from a single series binding proteins. Thus, whereas the regulation of of sequential events. Recent advances in our under- transcription from isolated SRE sequences has standing of basic mechanisms that regulate signal received renewed attention, how coincident incoming transduction, will now a€ord the opportunity to signals are integrated at the level of the c-fos promoter unravel the intricacies of how these signals are to regulate c-fos expression is still largely unknown. In integrated to elicit biological responses, and how summary, the emerging picture is that transcriptional perturbation of this signaling network can result in activation or repression most likely results from the cancer. Furthermore, it will also provide unique activity of a number of interlinked regulatory opportunities to identify previously unsuspected molecules and pathways, rather than from a linear molecular targets for pharmacological intervention in series of signaling events. this disease.

References

Aaronson SA and Tronick SR. (1985). Carcinog. Compr. Buchsbaum R, Telliez JB, Goonesekera S and Feig LA. Surv., 10, 35 ± 49. (1996). Mol. Cell. Biol., 16, 4888 ± 4896. Ahuja SK and Murphy PM. (1993). J. Biol. Chem., 268, Burgering BM and Bos JL. (1995). Trends Biochem. Sci., 20, 20691 ± 20694. 18 ± 22. AkadaR,KallalL,JohnsonDIandKurjanJ.(1996). BurgeringBM,PronkGJ,vanWeerenPC,ChardinPand Genetics, 143, 103 ± 117. Bos JL. (1993). EMBO J., 12, 4211 ± 4220. Allen LF, Lefkowitz RJ, Caron MG and Cotecchia S. (1991). Cavigelli M, Dol® F, Claret FX and Karin M. (1995). EMBO Proc. Natl. Acad. Sci. USA, 88, 11354 ± 11358. J., 14, 5957 ± 5964. Altenbach C, Yang K, Farrens DL, Farahbakhsh ZT, Chambard JC, Paris S, G LA and Pouyssegur J. (1987). Khorana HG and Hubbell WL. (1996). Biochemistry, 35, Nature, 326, 800 ± 803. 12470 ± 12478. ChanAM,FlemingTP,McGovernES,ChedidM,MikiT Altschuler DL, Peterson SN, Ostrowski MC and Lapetina and Aaronson SA. (1993). Mol. Cell. Biol., 13, 762 ± 768. EG. (1995). J. Biol. Chem., 270, 10373 ± 10376. Charlesworth A and Rozengurt E. (1997). Oncogene, 14, Arvanitakis L, Geras-Raaka E, Varma A, Gershengorn MC 2323 ± 2329. and Cesarman E. (1997). Nature, 385, 347 ± 350. Chee MS, Satchwell SC, Preddie E, Weston KM and Barrell Bankier AT, Beck S, Bohni R, Brown CM, Cerny R, Chee BG. (1990). Nature, 344, 774 ± 777. MS, Hutchison CAD, Kouzarides T, Martignetti JA, Chen J and Iyengar R. (1994). Science, 263, 1278 ± 1281. Preddie E and et al. (1991). DNA Seq., 2, 1 ± 12. Chen J and Iyengar R. (1995). Biochem Soc. Trans., 23, 129 ± Boguski MS and McCormick F. (1993). Nature, 366, 643 ± 133. 654. ChenY,GrallD,SalciniAE,PelicciPG,PouyssegurJand Bos JL. (1995). Eur. J. Cancer, 31A, 1051 ± 1054. Van Obberghen-Schilling E. (1996). EMBO J., 15, 1037 ± Bourne HR. (1997). Curr. Opin. Cell. Biol., 9, 134 ± 142. 1044. Cell growth control by G protein coupled receptors JS Gutkind 1340 Chow CW, Rincon M, Cavanagh J, Dickens M and Davis Gupta SK, Gallego C, Johnson GL and Heasley LE. (1992). RJ. (1997). Science, 278, 1638 ± 1641. J. Biol. Chem., 267, 7987 ± 7990. Clapham DE and Neer EJ. (1997). Annu. Rev. Pharmaco. Gutkind JS, Novotny EA, Brann MR and Robbins KC. Toxicol., 37, 167 ± 203. (1991). Proc. Natl. Acad. Sci. USA, 88, 4703 ± 4707. Coso OA, Chiariello M, Kalinec G, Kyriakis JM, Woodgett Gutkind JS and Robbins KC. (1992). Biochem. Biophys. Res. J and Gutkind JS. (1995a). J. Biol. Chem., 270, 5620 ± Commun., 188, 155 ± 161. 5624. Hall A. (1994). Annu. Rev. Cell. Biol., 10, 31 ± 54. Coso OA, Chiariello M, Yu JC, Teramoto H, Crespo P, Xu Han J, Jiang Y, Li Z, Kravchenko VV and Ulevitch RJ. N, Miki T and Gutkind JS. (1995b). Cell, 81, 1137 ± 1146. (1997). Nature, 386, 296 ± 299. Coso OA, Montaner S, Fromm C, Lacal JC, Prywes R, Hanks SK and Polte TR. (1997). Bioessays, 19, 137 ± 145. Teramoto H and Gutkind JS. (1997). J. Biol. Chem., 272, Hawes BE, Luttrell LM, van Biesen T and Lefkowitz RJ. 20691 ± 20697. (1996). J. Biol. Chem., 271, 12133 ± 12136. Coso OA, Teramoto H, Simonds WF and Gutkind JS. Hawes BE, van Biesen T, Koch WJ, Luttrell LM and (1996). J. Biol. Chem., 271, 3963 ± 3966. Lefkowitz RJ. (1995). J. Biol. Chem., 270, 17148 ± 17153. Coughlin SR, Escobedo JA and Williams LT. (1989). Heidecker G, Kolch W, Morrison DK and Rapp UR. (1992). Science, 243, 1191 ± 4. Adv. Cancer. Res., 58, 53 ± 73. Crespo P, Schuebel KE, Ostrom AA, Gutkind JS and Herskowitz I. (1995). Cell, 80, 187 ± 197. Bustelo XR. (1997). Nature, 385, 169 ± 172. Hill CS and Treisman R. (1995). Cell, 80, 199 ± 211. Crespo P, Xu N, Daniotti JL, Troppmair J, Rapp UR and Hill CS, Wynne J and Treisman R. (1995). Cell, 81, 1159 ± Gutkind JS. (1994a). J. Biol. Chem., 269, 21103 ± 21109. 1170. Crespo P, Xu N, Simonds WF and Gutkind JS. (1994b). Hipskind RA and Nordheim A. (1991). J. Biol. Chem., 266, Nature, 369, 418 ± 420. 19583 ± 19592. Cuttitta F, Carney DN, Mulshine J, Moody TW, Fedorko J, Hoosein NM, Kiener PA, Curry RC, Rovati LC, McGilbra Fischler A and Minna JD. (1985). Nature, 316, 823 ± 826. DK and Brattain MG. (1988). Cancer Res., 48, 7179 ± Cuvillier O, Pirianov G, Kleuser B, Vanek PG, Coso OA, 7183. Gutkind S and Spiegel S. (1996). Nature, 381, 800 ± 803. Hordijk PL, Verlaan I, van Corven EJ and Moolenaar WH. Daub H, Weiss FU, Wallasch C and Ullrich A. (1996). (1994). J. Biol. Chem., 269, 645 ± 651. Nature, 379, 557 ± 560. Johansen FE and Prywes R. (1994). Mol. Cell. Biol., 14, Davis RJ. (1993). J. Biol. Chem., 268, 14553 ± 14556. 5920 ± 5928. Davis RJ. (1995). Mol. Reprod. Dev., 42, 459 ± 467. Johnson NL, Gardner AM, Diener KM, Lange-Carter CA, Della Rocca GJ, van Biesen T, Daaka Y, Luttrell DK, Gleavy J, Jarpe MB, Minden A, Karin M, Zon LI and Luttrell LM and Lefkowitz RJ. (1997). J. Biol. Chem., 272, Johnson GL. (1996). J. Biol. Chem., 271, 3229 ± 337. 19125 ± 19132. Joyal JL, Burks DJ, Pons S, Matter WF, Vlahos CJ, White Derijard B, Hibi M, Wu IH, Barrett T, Su B, Deng T, Karin MF and Sacks DB. (1997). J. Biol. Chem., 272, 28183 ± M and Davis RJ. (1994). Cell, 76, 1025 ± 1037. 28186. Dhanasekaran N, Heasley LE and Johnson GL. (1995). Julius D, Livelli TJ, Jessell TM and Axel R. (1989). Science, Endocr. Rev., 16, 259 ± 270. 244, 1057 ± 1062. Dikic I, Tokiwa G, Lev S, Courtneidge SA and Schlessinger Karin M and Hunter T. (1995). Curr. Biol., 5, 747 ± 757. J. (1996). Nature, 383, 547 ± 50. Koch WJ, Hawes BE, Allen LF and Lefkowitz RJ. (1994). Diverse-Pierluissi M, Remmers AE, Neubig RR and Dunlap Proc. Natl. Acad. Sci. USA, 91, 12706 ± 12710. K. (1997). Proc. Natl. Acad. Sci. USA, 94, 5417 ± 5421. KolchW,HeideckerG,KochsG,HummelR,VahidiH, Dohlman HG, Caron MG and Lefkowitz RJ. (1987). Mischak H, Finkenzeller G, Marme D and Rapp UR. Biochemistry, 26, 2657 ± 2664. (1993). Nature, 364, 249 ± 252. Du J, Sperling LS, Marrero MB, Phillips L and Delafontaine Kraus MH, Pierce JH, Fleming TP, Robbins KC, Di Fiore P. (1996). Biochem. Biophys. Res. Commun., 218, 934 ± PP and Aaronson SA. (1988). Ann. NY. Acad. Sci., 551, 939. 320 ± 335; discussion 336. Fanger GR, Gerwins P, Widmann C, Jarpe MB and Johnson Kyriakis JM, Banerjee P, Nikolakaki E, Dai T, Rubie EA, GL. (1997). Curr.Opin.Genet.Dev.,7, 67 ± 74. Ahmad MF, Avruch J and Woodgett JR. (1994). Nature, Farnsworth CL, Freshney NW, Rosen LB, Ghosh A, 369, 156 ± 60. Greenberg ME and Feig LA. (1995). Nature, 376, 524 ± Lambright DG, Noel JP, Hamm HE and Sigler PB. (1994). 527. Nature, 369, 621 ± 628. Faure M and Bourne HR. (1995). Mol. Biol. Cell., 6, 1025 ± Lambright DG, Sondek J, Bohm A, Skiba NP, Hamm HE 1035. and Sigler PB. (1996). Nature, 379, 311 ± 319. Faure M, Voyno-Yasenetskaya TA and Bourne HR. (1994). Landis CA, Masters SB, Spada A, Pace AM, Bourne HR and J. Biol. Chem., 269, 7851 ± 7854. Vallar L. (1989). Nature, 340, 692 ± 696. Freshney NW, Goonesekera SD and Feig LA. (1997). FEBS Langdon S, Sethi T, Ritchie A, Muir M, Smyth J and Lett., 407, 111 ± 115. Rozengurt E. (1992). Cancer Res., 52, 4554 ± 4557. Fromm C, Coso OA, Montaner S, Xu N and Gutkind JS. Lawler S, Cuenda A, Goedert M and Cohen P. (1997). FEBS (1997). Proc. Natl. Acad. Sci. USA, 94, 10098 ± 10103. Lett., 414, 153 ± 158. Frost JA, Steen H, Shapiro P, Lewis T, Ahn N, Shaw PE and LevS,MorenoH,MartinezR,CanollP,PelesE,Musacchio Cobb MH. (1997). EMBO J., 16, 6426 ± 6438. JM, Plowman GD, Rudy B and Schlessinger J. (1995). Gaits F, Li RY, Bigay J, Ragab A, Ragab-Thomas MF and Nature, 376, 737 ± 745. Chap H. (1996). J. Biol. Chem., 271, 20151 ± 20155. Linseman DA, Benjamin CW and Jones DA. (1995). J. Biol. Gille H, Strahl T and Shaw PE. (1995). Curr. Biol., 5, 1191 ± Chem., 270, 12563 ± 12568. 1200. Lopez-Ilasaca M, Crespo P, Pellici PG, Gutkind JS and Goedert M, Cuenda A, Craxton M, Jakes R and Cohen P. Wetzker R. (1997). Science, 275, 394 ± 397. (1997). EMBO J., 16, 3563 ± 3571. Luttrell LM, Della Rocca GJ, van Biesen T, Luttrell DK and Goillot E, Raingeaud J, Ranger A, Tepper RI, Davis RJ, Lefkowitz RJ. (1997). J. Biol. Chem., 272, 4637 ± 4644. Harlow E and Sanchez I. (1997). Proc. Natl. Acad. Sci. Luttrell LM, Hawes BE, van Biesen T, Luttrell DK, Lansing USA, 94, 3302 ± 3307. TJ and Lefkowitz RJ. (1996). J. Biol. Chem., 271, 19443 ± Greenberg ME, Siegfried Z and Zi€ EB. (1987). Mol. Cell 19450. Biol., 7, 1217 ± 1225. Cell growth control by G protein coupled receptors JS Gutkind 1341 LyonsJ,LandisCA,HarshG,VallarL,GrunewaldK, Stoyanov B, Volinia S, Hanck T, Rubio I, Loubtchenkov M, Feichtinger H, Duh QY, Clark OH, Kawasaki E, Bourne Malek D, Stoyanova S, Vanhaesebroeck B, Dhand R, HR et al. (1990). Science, 249, 655 ± 659. Nurnberg B et al. (1994). Science, 269, 690 ± 693. Macdonald SG, Crews CM, Wu L, Driller J, Clark R, Su B and Karin M. (1996). Curr. Opin. Immunol., 8, 402 ± Erikson RL and McCormick F. (1993). Mol. Cell. Biol., 411. 13, 6615 ± 6620. Sugimoto T, Stewart S and Guan K-L. (1997). J. Biol. Chem., Manser E, Leung T, Salihuddin H, Zhao ZS and Lim L. 272, 29415 ± 29418. (1994). Nature, 367, 40 ± 46. Sunahara RK, Dessauer CW and Gilman AG. (1996). Annu. Mansour SJ, Matten WT, Hermann AS, Candia JM, Rong S, Rev. Pharmacol. Toxicol., 36, 461 ± 480. Fukasawa K, Vande Woude GF and Ahn NG. (1994). Tahara E. (1990). J. Cancer Res. Clin. Oncol., 116, 121 ± 131. Science, 265, 966 ± 970. Takahashi T, Kawahara Y, Okuda M, Ueno H, Takeshita A Marshall CJ. (1995). Cell, 80, 179 ± 185. and Yokoyama M. (1997). J. Biol. Chem., 272, 16018 ± Mattingly RR and Macara IG. (1996). Nature, 382, 268 ± 16022. 272. Taussig R and Gilman AG. (1995). J. Biol. Chem., 270, 1±4. Minden A, Lin A, Claret FX, Abo A and Karin M. (1995). Teramoto H, Salem P, Robbins KC, Bustelo XR and Cell, 81, 1147 ± 1157. Gutkind JS. (1997). J. Biol. Chem., 272, 10751 ± 10755. Mohammadi M, Dionne CA, Li W, Li N, Spivak T, Thomas F, Arvelo F, Antoine E, Jacrot M and Poupon MF. Honegger AM, Jaye M and Schlessinger J. (1992). (1992a). Cancer Res., 52, 4872 ± 4877. Nature, 358, 681 ± 684. Thomas SM, DeMarco M, D'Arcangelo G, Halegoua S and Moody TW and Cuttitta F. (1993). Life Sci., 52, 1161 ± 1173. Brugge JS. (1992b). Cell, 68, 1031 ± 1040. Moolenaar WH. (1991). Cell Growth Di€er., 2, 359 ± 364. Tokiwa G, Dikic I, Lev S and Schlessinger J. (1996). Science, Nagata A, Ito M, Iwata N, Kuno J, Takano H, Minowa O, 273, 792 ± 794. Chihara K, Matsui T and Noda T. (1996). Proc. Natl. Tournier C, Whitmarsh AJ, Cavanagh J, Barrett T and Davis Acad. Sci. USA, 93, 11825 ± 11830. RJ. (1997). Proc. Natl. Acad. Sci. USA, 94, 7337 ± 7342. Neote K, DiGregorio D, Mak JY, Horuk R and Schall TJ. Treisman R. (1994). Curr. Opin. Genet. Dev., 4, 96 ± 101. (1993). Cell, 72, 415 ± 425. Treisman R. (1995). EMBO J., 14, 4905 ± 4913. Nicholas J, Cameron KR and Honess RW. (1992). Nature, Treisman R. (1996). Curr. Opin. Cell. Biol., 8, 205 ± 215. 355, 362 ± 365. Tronick SR and Aaronson SA. (1988). Adv. Second Nobes CD and Hall A. (1995a). Biochem. Soc. Trans., 23, Messenger Phosphoprotein Res., 21, 201 ± 214. 456 ± 459. Tsunoda S, Sierralta J, Sun Y, Bodner R, Suzuki E, Becker Nobes CD and Hall A. (1995b). Cell, 81, 53 ± 62. A, Socolich M and Zuker CS. (1997). Nature, 388, 243 ± Pace AM, Faure M and Bourne HR. (1995). Mol. Biol. Cell., 249. 6, 1685 ± 1695. Ui M and Katada T. (1990). Adv. Second Messenger Pages G, Lenormand P, L'Allemain G, Chambard JC, Phosphoprotein Res., 24, 63 ± 69. Meloche S and Pouyssegur J. (1993). Proc. Natl. Acad. vanBiesenT,HawesBE,LuttrellDK,KruegerKM, Sci. USA, 90, 8319 ± 8323. TouharaK,Por®riE,SakaueM,LuttrellLMand Parma J, Duprez L, Van Sande J, Cochaux P, Gervy C, Lefkowitz RJ. (1995). Nature, 376, 781 ± 784. Mockel J, Dumont J and Vassart G. (1993). Nature, 365, van Biesen T, Luttrell LM, Hawes BE and Lefkowitz RJ. 649 ± 651. (1996). Endocr. Rev., 17, 698 ± 714. Philipp S and Flockerzi V. (1997). FEBS Lett., 413, 243 ± van Corven EJ, Groenink A, Jalink K, Eichholtz T and 248. Moolenaar WH. (1989). Cell, 59, 45 ± 54. Pouyssegur J, Chambard JC, G LA, Magnaldo I and Seuwen VosslerMR,YaoH,YorkRD,PanMG,RimCSandStork K. (1988). Philo.Trans.R.Soc.Lond.BBiol.Sci.,320, PJ. (1997). Cell, 89, 73 ± 82. 427 ± 436. Wan Y, Kurosaki T and Huang XY. (1996). Nature, 380, Prasad MV, Dermott JM, Heasley LE, Johnson GL and 541 ± 544. Dhanasekaran N. (1995). J. Biol. Chem., 270, 18655 ± Weinstein LS, Shenker A, Gejman PV, Merino MJ, Fried- 18659. man E and Spiegel AM. (1991). N.Engl.J.Med.,325, Ptasznik A, Traynor-Kaplan A and Bokoch GM. (1995). J. 1688 ± 1695. Biol. Chem., 270, 19969 ± 19973. Wess J. (1997). FASEB J., 11, 346 ± 354. Rankin S, Morii N, Narumiya S and Rozengurt E. (1994). Whitehead IP, Campbell S, Rossman KL and Der CJ. (1997). FEBS Lett., 354, 315 ± 319. Biochim. Biophys. Acta., 1332, F1 ± 23. Rozengurt E. (1986). Science, 234, 161 ± 166. Whiteway MS, Wu C, Leeuw T, Clark K, Fourest-Lieuvin A, Runkel L, Shaw PE, Herrera RE, Hipskind RA and Thomas DY and Leberer E. (1995). Science, 269, 1572 ± Nordheim A. (1991). Mol. Cell. Biol., 11, 1270 ± 1280. 1575. Schlaepfer DD and Hunter T. (1996). Cell Struct. Funct., 21, Whitmarsh AJ, Yang SH, Su MS, Sharrocks AD and Davis 445 ± 450. RJ. (1997). Mol. Cell. Biol., 17, 2360 ± 2371. Schlessinger J. (1993). Trends Biochem. Sci., 18, 273 ± 275. Wilkie TM, Gilbert DJ, Olsen AS, Chen XN, Amatruda TT, Schuller HM. (1991a). Exp. Lung Res., 17, 837 ± 852. Korenberg JR, Trask BJ, de Jong P, Reed RR, Simon MI Schuller HM. (1991b). Cancer Cells, 3, 496 ± 503. et al. (1992). Nat. Genet., 1, 85 ± 91. Sethi T, Langdon S, Smyth J and Rozengurt E. (1992). Williams NG and Roberts TM. (1994). Cancer Metastasis Cancer Res., 52, 2737s ± 2742s. Rev., 13, 105 ± 116. Sethi T and Rozengurt E. (1991). Cancer Res., 51, 3621 ± Xia Z, Dickens M, Raingeaud J, Davis RJ and Greenberg 3623. ME. (1995). Science, 270, 1326 ± 1331. ShenkerA,LaueL,KosugiS,MerendinoJr,JJ,MinegishiT Xu N, Bradley L, Ambdukar I and Gutkind JS. (1993). Proc. and Cutler Jr, GB. (1993). Nature, 365, 652 ± 654. Natl. Acad. Sci. USA, 90, 6741 ± 6745. Simon MN, De Virgilio C, Souza B, Pringle JR, Abo A and Xu N, Voyno-Yasenetskaya T and Gutkind JS. (1994). Reed SI. (1995). Nature, 376, 702 ± 705. Biochem. Biophys. Res. Commun., 201, 603 ± 609. Sondek J, Bohm A, Lambright DG, Hamm HE and Sigler Yang DD, Kuan CY, Whitmash AJ, Rincon M, Zheng TS, PB. (1996). Nature, 379, 369 ± 374. Davis RJ, Rakic P and Flavell RA. (1997). Nature, 389, Sondek J, Lambright DG, Noel JP, Hamm HE and Sigler 865 ± 870. PB. (1994). Nature, 372, 276 ± 279. Cell growth control by G protein coupled receptors JS Gutkind 1342 Yarden Y, Escobedo JA, Kuang WJ, Yang-Feng TL, Daniel Zhao ZS, Leung T, Manser E and Lim L. (1995). Mol. Cell. TO,TremblePM,ChenEY,AndoME,HarkinsRN, Biol., 15, 5246 ± 5257. Francke U et al. (1997). Nature, 323, 226 ± 232. Zheng Y, Bender A and Cerione RA. (1995). J. Biol. Chem., Young D, Waitches G, Birchmeier C, Fasano O and Wigler 270, 626 ± 630. M. (1986). Cell, 45, 711 ± 719.