Division of Molecular Biology, Department of Molecular and Cellular Biology, Kobe University Graduate School of Medicine, 7-5-1 Kusunoki-cho, Chuo-ku, Kobe 650-0017, Japan

Correspondence to: T Kataoka, Division of Molecular Biology, Department of Molecular and Cellular Biology, Kobe University Graduate School of Medicine, 7-5-1 Kusunoki-cho, Chuo-ku, Kobe 650-0017, Japan; E-mail: kataoka@kobe-u.ac.jp

Phospholipase C is a phosphoinositide-specific phospholipase C that selectively associates with Ras and Rap small GTPases as a target. Here we explored the molecular basis of the Rap1- as well as Ras-mediated regulation of phospholipase C upon platelet-derived growth factor stimulation by using a receptor mutant deficient in its ability to phosphorylate and activate phospholipase C. Following platelet-derived growth factor treatment, this receptor induces persistent activation of ectopically expressed PLC through activation of Ras and Rap1. The rapid and initial phase of the activation is mediated by Ras, whereas Rap1 is responsible for the prolonged activation. We further demonstrate that the CDC25 homology domain, which exhibits guanine nucleotide exchange factor activity toward Rap1, but not Ras, is critical for the prolonged activation of phospholipase C. Platelet-derived growth factor prevented the hematopoietic BaF3 cells containing the mutant receptor from undergoing apoptosis, and enabled these cells to proliferate, only when phospholipase C was expressed. Therefore, the phospholipase C signal is suggested to be critical for survival and growth of BaF3 cells.

Oncogene (2002) 21, 8105-8113. doi:10.1038/sj.onc.1206003

PLC; Ras; Rap1

Introduction

Phosphoinositide-specific phospholipase C (PLC) catalyzes the hydrolysis of phosphatidylinositol 4,5-biphosphate, yielding two intracellular second messengers, diacylglycerol (DAG) and inositol 1,4,5-trisphosphate (IP₃), which direct protein kinase C activity and calcium mobilization, respectively (Singer et al., 1997). Thus, PLC plays a pivotal role in a diverse array of signal transduction pathways that regulate cell growth, differentiation and development. In mammalian cells, four classes of PLCs (PLC, PLC, PLC and PLC) have been identified, being regulated by distinct mechanisms. For instance, PLC is activated upon direct binding of the subunit (G) or the subunits (G) of heterotrimeric G proteins, whereas tyrosine phosphorylation is crucial for the activation of PLC. PLC is the most recently identified member of the PLC family. A Caenorhabditis elegans homolog of PLC, designated PLC210, was initially characterized as a novel Ras-binding protein through yeast two-hybrid screening (Shibatohge et al., 1998), followed by the isolation of human (Song et al., 2001), rat (Kelley et al., 2001) and mouse (Wu et al., unpublished data) cDNAs encoding PLC. Screening of human sequence databases by the use of amino acid sequences conserved in PLC isoforms also led to the identification of the same protein (Lopez et al., 2001).

In addition to phospholipase catalytic (X and Y) and calcium-dependent lipid-binding (C2) domains, mammalian PLC contains two Ras/Rap1-associating (RA) domains (RA1 and RA2 domains) at the C terminus and a CDC25 homology domain near the N terminus, leading to the notion that Ras family GTPases may be intimately involved in the modulation of PLC activity. Although unidentified in initial studies, a pleckstrin homology domain and EF hands common to all PLC isozymes were recently reported to be found also in PLC (Wing et al., 2001). However, their functions in the regulation of PLC remain to be clarified.

The RA2 domain of PLC associates with Ha-Ras and Rap1A (also called Krev-1) in a GTP-dependent manner, suggesting that PLC is regulated downstream of these small GTPasas (Kelley et al., 2001; Song et al., 2001). In fact, PLC was translocated to the plasma membrane and the Golgi apparatus through the binding to GTP-bound forms of Ha-Ras and Rap1A, respectively (Song et al., 2001). Ras-dependent subcellular translocation and activation of PLC were also reconstituted in vitro by the use of a liposome carrying recombinant Ras and phosphatidylinositol 4,5-biphosphate (Song et al., 2001). Additionally, PLC was significantly activated in cells when coexpressed with activated Ha-Ras (Kelley et al., 2001). Although Ras-dependent activation of PLC is observed both in vivo and in vitro as described above, it is unclear whether Rap1A as well can augment PLC activity upon the association with the RA2 domain. The RA1 domain, whose amino acid sequence is less conserved among mammalian PLC proteins, was reported to bind Ha-Ras with a much lower affinity than the RA2 domain in a GTP-independent manner (Kelley et al., 2001). Therefore, the physiological role of the RA1 domain in the regulation of PLC remains elusive, and the RA2 domain will be designed simply the 'RA domain' in this study.

The CDC25 homology domain exhibits in vitro guanine nucleotide exchange factor (GEF) activity toward Rap1A, but not toward other Ras family GTPases tested including Ha-Ras (Jin et al., 2001). Therefore, this domain may stimulate the formation of Rap1A·GTP at particular subcellular regions such as the Golgi apparatus. Indeed, the expression of PLC remarkably prolonged the time course of Rap1A·GTP formation upon epidermal growth factor (EGF) treatment (Jin et al., 2001). However, the role of the CDC25 homology domain in the regulation of signaling mediated by Rap1A and PLC remains largely unknown.

Although evidence for the involvement of Ras and Rap small GTPases in the regulation of PLC has been accumulating, receptor-mediated signal transduction pathways that activate PLC in response to extracellular stimuli remain obscure. Schmidt and colleagues (Schmidt et al., 2001b; Evellin et al., 2002) recently documented that PLC was activated downstream of cAMP-dependent signaling triggered by stimulation of a subset of G protein-coupled serpentine receptors, such as ₂-adrenergic, prostaglandin E₁ and M₃ muscarinic acetylcholine receptors. In these signaling pathways, Rap2B is reported to be specifically involved in the regulation of PLC although the molecular mechanisms underlying the activation of PLC by Rap2B remain to be solved (Schmidt et al., 2001b; Evellin et al., 2002).

Herein, we show that Rap1 as well as Ras plays an important role in the regulation of PLC following stimulation of the platelet-derived growth factor (PDGF) receptor. Ras mediates the rapid and transient signal for the activation of PLC, whereas Rap1 is responsible for the sustained activation. Furthermore, we describe that the CDC25 homology domain, which exhibits a Rap1-specific GEF activity, is critical for the prolongation of PLC activation in Rap1-dependent signaling.

Results

Activation of PLC by coexpression of Ras or Rap in COS-7 cells

PLC binds to Ras and Rap in a GTP-dependent manner through its C-terminal RA domain. Although the activation of PLC upon binding to the GTP-bound form of Ras has been demonstrated both in vivo (Kelley et al., 2001) and in vitro (Song et al., 2001), the role of Rap in the regulation of PLC remains elusive. As an initial step to clarify this point, the activation of PLC by coexpression of constitutively active forms of Ras subfamily members was examined in COS-7 cells (Figure 1). As previously described (Kelley et al., 2001), coexpression of a constitutively active form of Ha-Ras, Ha-Ras^G12V, resulted in an increase in the level of inositol phosphates in PLC-expressing COS-7 cells, reflecting the activation of PLC (Figure 1a). An effector region mutant of Ha-Ras, Ha-Ras^G12V,Y32F, which did not bind to PLC (Song et al., 2001), exerted no effect on inositol phosphates production, supporting a notion that the observed activation of PLC is mediated through direct binding (Figure 1a). Similarly, an activated form of Rap1A (Rap1A^G12V), but not its effector region mutant (Rap1A^G12V,D38A) that did not bind to PLC (data not shown), increased the level of inositol phosphates (Figure 1b). The effect of Rap1A was significant but modest compared to that of Ha-Ras, which may be ascribed to difference in the size of phosphatidylinositol 4,5-biphosphate pools in different subcellular regions where Ras or Rap1A exists. Like Rap1A, both Rap2A and Rap2B associated with PLC in a GTP-dependent manner as revealed by an in vitro binding assay (Figure 2a), whereas other Ras family members Rin and M-Ras did not bind to PLC (data not shown). Consistent with the binding activity to PLC, Rap2B, but not Rin and M-Ras, in their active forms, caused modest PLC activation in COS-7 cells (Figure 1c).

Specificity of the RA and CDC25 homology domains of PLC

We recently reported that the RA domain of PLC associates with both Ha-Ras and Rap1A in a GTP-dependent manner (Song et al., 2001). On the other hand, the CDC25 homology domain of PLC exhibits GEF activity toward Rap1A, but not Ha-Ras (Jin et al., 2001). On the basis of these biochemical properties, we assumed that the CDC25 homology domain might be important for amplification and prolongation of Rap1A-mediated signaling (Jin et al., 2001). However, it remains unclear whether the Rap2 signal as well is modulated by the CDC25 homology domain, and therefore we further examined the interaction of PLC with Rap2A and Rap2B. Like Rap1A, GTP-bound forms, but not GDP-bound forms, of both Rap2A and Rap2B associate with the PLC RA domain (Figure 2a). In contrast, GDP release from Rap1A was enhanced by PLC, whereas that from Rap2A or Rap2B remained unaffected (Figure 2b).

Isolation of cell lines containing PLC and a PDGF receptor mutant that is incapable of activating PLC

Previously, we addressed translocation of PLC in response to EGF stimulation to the plasma membrane and the Golgi apparatus through binding to Ras and Rap1, respectively (Song et al., 2001). However, we failed to detect a significant increase in the catalytic activity of PLC in response to EGF treatment because EGF also stimulated another subtype of PLC, PLC, through tyrosine phosphorylation. In an attempt to elucidate the role of Ras and Rap in the regulation of PLC in response to extracellular stimuli, we isolated cell lines harboring PLC and a PDGF receptor mutant that is incapable of stimulating PLC. Phosphorylation of two tyrosine residues (Y977 and Y989) at the C-terminal intracellular portion of the PDGF receptor is essential for PDGF-dependent PLC activation, and therefore, phenylalanine substitution of these tyrosines renders the PDGF receptor incapable of stimulating PLC without affecting Ras activation (Fantl et al., 1993; Satoh et al., 1993). The cDNA encoding full-length PLC or PLC lacking the N-terminal portion including the CDC25 homology domain (PLCN) was introduced into a hematopoietic BaF3 cell line-derived clone carrying the PDGF receptor mutant PDGFR(Y977F/Y989F) (designated BaF3-PDGFR(Y977F/Y989F)) (Satoh et al., 1993). The parental BaF3 cell line does not have a detectable amount of endogenous PLC. Resulting stable transfectant clones (designated BaF3-PDGFR (Y977F/Y989F)/PLC and BaF3-PDGFR(Y977F/Y989F)/PLCN, respectively) were subjected to further analyses.

Activation of Ras and Rap1 following PDGF stimulation in BaF3-derived transfectants

GTP-bound forms of Ras and Rap1 in BaF3-derived transfectants were detected by the pull-down assays (Figure 3). In BaF3-PDGFR(Y977F/Y989F)/PLC cells, PDGF caused a rapid and transient increase in Ha-Ras·GTP with a peak level at 1 to 5 min (Figure 3a). In contrast, the Rap1·GTP level increased more slowly and remained elevated for at least 20 min (Figure 3b). In BaF3-PDGFR(Y977F/Y989F)/PLCN cells, PDGF-dependent Rap1·GTP formation became transient as shown in Figure 3c, suggesting that the CDC25 homology domain is required for the sustained Rap1 activation observed in the cells expressing full-length PLC. The time course of Rap1 activation in BaF3-PDGFR(Y977F/Y989F) cells was similar to that of BaF3-PDGFR(Y977F/Y989F)/PLCN cells (data not shown).

IP₃ production following PDGF stimulation in BaF3-derived transfectants

Intracellular IP₃ levels were quantitated following PDGF treatment of BaF3-derived cells (Figure 4). We also stimulated these cells with interleukin-3 (IL-3) as a control and detected an increase in the level of intracellular IP₃. PDGF did not induce elevation of the IP₃ level in the absence of PLC because the PDGF receptor mutant could not activate PLC (Figure 4a). In marked contrast, PDGF caused a remarkable increase in the IP₃ level when full-length PLC was ectopically expressed (Figure 4b). The increase in the IP₃ level is presumably ascribed to the activation of PLC mediated by Ras and Rap1 because both small GTPases are activated in response to PDGF as described (Figure 3). To further clarify the involvement of Ras and Rap in the PLC-dependent accumulation of IP₃, a dominant-negative mutant of Ras (Ha-Ras^S17N) or a Rap GTPase-activating protein (SPA-1) was expressed in BaF3-PDGFR(Y977F/Y989F)/PLC cells by the use of a retroviral expression system (Figure 5). Ha-Ras^S17N potently inhibited the PDGF-triggered IP₃ production, but a slight increase in the IP₃ level was observed with delayed time kinetics at 10 to 20 min. In contrast, in SPA-1-expressing cells, the rapid response remained unaffected whereas the prolonged increase in the IP₃ level was significantly abrogated. Therefore, it is feasible that the PLC-dependent IP₃ production, particularly at the initial phase, is mediated mainly by Ras, while Rap may be responsible for the sustained activation of PLC. The activity of PLC may also be modulated by the CDC25 homology domain because this domain exhibits Rap1-specific GEF activity (Figure 2), and is responsible for the prolongation of the Rap1 signal (Figure 3). Indeed, prolonged IP₃ production was impaired in BaF3-PDGFR(Y977F/Y989F)/PLCN cells, and consequently the time course resembled that observed in the SPA-1-expressing cells (Figures 4c and 5). Considering that the CDC25 homology domain is specific to Rap1, but not Rap2, Rap1 rather than Rap2 is likely to be implicated in the regulation of PLC.

Proliferation of PLC-expressing BaF3 cells

Lastly, we examined whether the PLC pathway stimulates survival and growth of BaF3 cells, which undergo apoptosis upon IL-3 deprivation (Johnson, 1998). Cell number increase of two BaF3-PDGFR(Y977F/Y989F)/PLC cell clones (# 8 and # 10) in the presence or absence of PDGF is illustrated in Figure 6. PDGF did not stimulate growth of BaF3-PDGFR(Y977F/Y989F) and BaF3-PDGFR(Y977F/Y989F)/vector cells. In contrast, both PLC-expressing clones proliferated in the presence of PDGF. Additionally, growth rates of the two clones in the presence of IL-3 were higher compared to that of BaF3-PDGFR(Y977F/Y989F) cells. Apoptosis of the two clones was also prevented when stimulated by PDGF although BaF3-PDGFR(Y977F/Y989F) cells underwent apoptosis in the presence of PDGF. Therefore, it is likely that PLC, when activated by PDGF, suppresses apoptotic cell death, and stimulates cell proliferation. BaF3-PDGFR(Y977F/Y989F)/PLC (clone # 8) cells did not undergo apoptosis under growth factor-deprived conditions (Figure 6b), which may be due to basal unstimulated activity of PLC at a high expression level. Inhibition of Ras- and Rap-mediated activation of PLC by retroviral expression of Ha-Ras^S17N and SPA-1, respectively, almost completely abolished cell growth, suggesting that PLC activities in distinct subcellular regions, such as the plasma membrane and the Golgi apparatus, are indispensable for cell growth (Figure 6d).

Discussion

Both Ha-Ras and Rap1A in their GTP-bound forms bind to the RA domain of PLC, suggesting that PLC may function as a direct target of not only Ha-Ras but also Rap1A (Song et al., 2001). This was further supported by our observation that PLC colocalized with Ha-Ras and Rap1A in the plasma membrane and the Golgi apparatus, respectively, after treatment with EGF (Song et al., 2001). However, it remained unclear whether phosphoinositide-hydrolyzing activity of PLC is regulated by Rap1A, although Ha-Ras-dependent activation of PLC was shown both in vivo and in vitro (Kelley et al., 2001; Song et al., 2001). In this study, we further explored a role for Rap1A in the regulation of PLC activity. When coexpressed in COS-7 cells, activated Rap1A, like Ha-Ras, significantly stimulated PLC activity presumably through direct interaction (Figure 1). Therefore, it is feasible that Rap1A directly binds to, and thereby stimulates PLC in the Golgi apparatus in response to extracellular stimuli such as EGF treatment. Considering that protein kinase C exists in the Golgi apparatus (de Matteis et al., 1993), and can act as a downstream target of DAG yielded by PLC, the Rap1A/PLC signaling pathway may provide a link between growth factor receptors and Golgi-localized protein kinase C.

It is important to examine whether the activity of PLC is actually stimulated in response to extracellular signals in a Ras- or Rap1-dependent manner. However, we were unable to detect EGF-dependent increase in PLC activity specifically because other subtypes of PLCs were also activated following EGF treatment. To accomplish this, we isolated cell lines harboring PLC and a PDGF receptor mutant that activates Ras and Rap1, but not PLC. Indeed, the activity of PLC, as determined by intracellular IP₃ production, was stimulated upon PDGF treatment, and remained elevated at least for 30 min after PDGF was challenged (Figure 4).

Disruption of the Rap pathway by infection of a SPA-1-expressing virus resulted in a transient increase in the intracellular IP₃ level after PDGF treatment (Figure 5), suggesting that Rap-mediated signaling is required for the sustained activation of PLC. In contrast, Ras mediates the rapid and transient activation of PLC as demonstrated by the effect of Ha-Ras^S17N (Figure 5). Although biological functions of PLC downstream of a variety of cell surface receptors remain obscure, Ras-mediated transient and Rap-mediated sustained activation of PLC may have different roles in cell response. In this respect, it is noteworthy that, in PC12 pheochromocytoma cells, Rap1 is involved in nerve growth factor-dependent sustained activation of extracellular signal-regulated kinases, which is prerequisite for neuronal differentiation, whereas Ras is responsible for the rapid activation (York et al., 1998). Moreover, Ras- and Rap-dependent PLC activation may possibly occur in distinct subcellular compartments not only in COS-7 cells (Song et al., 2001), but also in other cell types. Therefore, distinct sets of protein kinase C isozymes and calcium channels may selectively function in Ras/PLC and Rap/PLC signaling pathways, exerting different biological functions. Notably, PDGF-dependent growth of PLC-expressing BaF3-derived cells was totally sensitive to the inhibitory action of Ha-Ras^S17N or SPA-1 (Figure 6). Thus, both Ras- and Rap1-dependent functions of PLC in the plasma membrane and the Golgi apparatus, respectively, are indispensable for proliferation of this cell line.

The CDC25 homology domain of PLC exhibits GEF activity toward Rap1A in vitro, leading to the hypothesis that this domain may be required for prolonged activation of PLC through a positive feedback mechanism (Jin et al., 2001). In support of this, the CDC25 homology domain was essential for prolonged formation of Rap1·GTP accompanied by prolonged translocation of PLC to the Golgi apparatus following EGF stimulation (Jin et al., 2001). As illustrated in Figure 4, the increase in the intracellular level of IP₃ upon PDGF treatment in BaF3-derived cells that express a PLC mutant lacking the CDC25 homology domain was transient in contrast to the sustained increase in full-length PLC-expressing cells. Collectively, it is conceivable that the CDC25 homology domain functions as a module responsible for prolongation of Rap1/PLC signaling.

It remains unclear which subtype of the Rap family is involved in the activation of PLC in BaF3-derived transfectants because both Rap1 and Rap2 bind to the RA domain in a GTP-dependent manner, thereby stimulating the activity of PLC (Figures 1 and 2). We also cannot distinguish roles of Rap1 and Rap2 based on the inhibitory action of SPA-1 because SPA-1 serves as a GTPase-activating protein for both Rap1A and Rap2A (Kurachi et al., 1997). Intriguingly, the CDC25 homology domain of PLC shows a GEF activity toward Rap1A, but neither Rap2A nor Rap2B, in vitro (Figure 2) and is needed for the PDGF-triggered Rap-dependent sustained activation of PLC (Figure 4). Hence, Rap1A rather than Rap2A and Rap2B is likely to be responsible for PLC signaling. On the other hand, Rap2 was reported to be specifically involved in the regulation of PLC following stimulation of Gs-coupled receptors (Schmidt et al., 2001b; Evellin et al., 2002), where the cAMP-dependent Rap GEF Epac1 serves as a regulator. However, PDGF stimulation of PLC presumably does not involve cAMP production. Additionally, more than 50% of Rap2A exists in the GTP-bound form in cells, leading to the notion that regulatory mechanisms for Rap2A are different from those for Rap1A (Ohba et al., 2000). Thus, a GEF specific to Rap1 may play a crucial role for triggering Rap1 activation in our system.

In addition to signaling pathways involving Ras or Rap, PLC has been implicated in pathways immediately downstream of heterotrimeric G proteins. A constitutively activated form of G12, but not other G subunits, such as Gs and Gi, stimulated PLC activity although the mechanism whereby the interaction between G12 and PLC occurs remains totally unknown (Lopez et al., 2001). Moreover, G enhanced PLC activity when coexpressed in COS-7 cells presumably through the association of G with a pleckstrin homology domain newly identified in PLC (Wing et al., 2001). The effect of G seems to be independent of the Ras pathways, implying that multiple regulatory signals converge at PLC. Future studies will further reveal the molecular mechanisms underlying the complicated regulation of PLC.

It is important to clarify the physiological function of PLC activated downstream of Rap1. To date, Rap1 has been implicated in an array of intracellular signaling pathways, which regulate proliferation, differentiation, lymphocyte aggregation, T-cell anergy or platelet activation (Bos, 1998; Bos et al., 2001). In addition, Rap1 may exert its function in a subcellular region-specific manner because Rap1 is present both in the Golgi apparatus and in the plasma membrane (Beranger et al., 1991; Wienecke et al., 1996; Matsubara et al., 1999; Ohba et al., 2000; York et al., 2000; Gao et al., 2001). Several lines of evidence have supported the notion that Rap1 plays an important role in integrin-mediated leukocyte adhesion. An active form of Rap1 potently induces the activation of integrins and subsequent cell aggregation (Katagiri et al., 2000; Reedquist et al., 2000), whereas a dominant-negative form of Rap1 and GTPase-activating proteins for Rap1 including SPA-1 inhibited cell adhesion triggered by ligation of the T-cell receptor or CD31 (Tsukamoto et al., 1999; Katagiri et al., 2000; Reedquist et al., 2000). Likewise, lipopolysaccharide-induced activation of integrins in macrophages requires Rap1 (Schmidt et al., 2001a). Currently, however, the involvement of PLC in these Rap1-dependent signaling pathways is totally unknown. Recently, DAG has been implicated in the recruitment of protein kinase D to the Golgi apparatus, which is important for protein transport to the cell surface (Baron and Malhotra, 2002). PLC may contribute to the generation of DAG in the Golgi apparatus and may be involved in protein kinase D regulation although other lipid metabolic pathways play a major role in controlling the DAG level. Further studies will be required for the elucidation of biological functions of the Rap1/PLC pathway particularly in the Golgi apparatus.

Materials and methods

Plasmids

cDNAs for full-length human PLC and its deletion mutant PLCN (amino acids 1088-2302) lacking the CDC25 homology domain were subcloned into the mammalian expression vector pFLAG-CMV2 (Sigma), generating pFLAG-CMV2-PLC and pFLAG-CMV2-PLCN, respectively (Jin et al., 2001). Hemagglutinin (HA)-tagged small GTPases (Ha-Ras, Ha-Ras^G12V, Ha-Ras^G12V,Y32F, Rap1A^G12V, Rap1A^G12V,D38A, Rap2B^G12V, Rin^G29V and M-Ras^Q71L) were subcloned into the mammalian expression vector pEF-BOS (Mizushima and Nagata, 1990). The Ha-Ras^S17N cDNA was subcloned into the retroviral vector pLPCX containing the puromycin-resistant gene (Clontech), generating pLPCX-Ha-Ras^S17N. pLXSN-FLAG-SPA-1 (for retroviral expression of FLAG-tagged SPA-1), containing the neomycin-resistant gene, was kindly provided by Dr Nagahiro Minato (Kyoto University, Kyoto, Japan).

Cell lines

COS-7 cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS). BaF3-PDGFR(Y977F/Y989F) cells (Satoh et al., 1993) were cultured in RPMI 1640 supplemented with 10% FBS, IL-3 (approximately 1 nM) and G418 (0.3 mg/ml). pFLAG-CMV2-PLC, pFLAG-CMV2-PLCN or pFLAG-CMV2 (vector) was introduced into BaF3-PDGFR(Y977F/Y989F) cells with a hygromycin B-resistant marker plasmid by electroporation essentially as described (Satoh et al., 1993). Stable transfectants (designated BaF3-PDGFR(Y977F/Y989F)/PLC, BaF3-PDGFR(Y977F/Y989F)/PLCN and BaF3-PDGFR(Y977F/Y989F)/vector, respectively) were selected and cultured in RPMI 1640 supplemented with 10% FBS, IL-3 (approximately 1 nM), G418 (0.3 mg/ml) and hygromycin B (0.5 mg/ml). Viable cells were counted after Trypan blue staining.

Retroviral gene transfer

The packaging cell line Phoenix Eco (American Type Culture Collection No; SD3444) was transfected with pLPCX-Ha-Ras^S17N and pLXSN-FLAG-SPA-1 and cultured in the presence of puromycin (1 g/ml, for pLPCX-Ha-Ras^S17N) or G418 (0.3 mg/ml, for pLXSN-FLAG-SPA-1) for 2 weeks. Recombinant retroviruses were obtained according to the instructions by Dr Garry Nolan (Stanford University, CA, USA). BaF3-PDGFR(Y977F/Y989F)/PLC cells infected with the Ha-Ras^S17N virus were selected by puromycin (1 g/ml). Expression of Ha-Ras^S17N and SPA-1 was confirmed by immunoblotting.

Quantitation of total inositol phosphates in COS-7 cells

Total inositol phosphates in COS-7 cells were quantitated essentially as described (Kelley et al., 2001). Briefly, COS-7 cells cultured in 12-well culture plates were transfected with expression plasmids using LipofectAMINE 2000 (Gibco-BRL) after reaching 90-95% confluency. Twenty-four hours after transfection, medium was changed to inositol-free DMEM supplemented with 4 Ci per well of myo-[³H]inositol (84 Ci/mmol, NEN). After 12 h incubation, 10 mM LiCl was added to the culture medium, and cells were incubated for additional 1 h. The reaction was stopped by adding ice-cold percholoric acid to the final concentration of 4.5% (v/v), and [³H]inositol phosphates were separated by column chromatography using Dowex AG1-X8 (200-400 mesh, formate form). Radioactivities of total inositol phosphates were quantitated by liquid scintillation counting.

In vitro binding assay

Maltose-binding protein (MBP)-PLCRA(2094-2302) was previously described (Song et al., 2001). Glutathione S-transferase (GST)-Rap1A, GST-Rap2A and GST-Rap2B were expressed in Escherichia coli by using the expression vector pGEX-5X (Amersham), and purified with glutathione-Sepharose (Amersham). In vitro binding of GST-tagged Rap GTPases (10 pmol) preloaded with GTPS or GDP to MBP-PLCRA(2094-2302) (50 pmol) was assayed in 100 l of the reaction mixture as described (Song et al., 2001). GST-tagged Rap GTPases were detected by immunoblotting using an anti-GST antibody (Shima et al., 1997).

In vitro GEF assay

FLAG-tagged PLC was expressed in Spodoptera frugiperda Sf9 cells, and affinity-purified by the use of anti-FLAG M2 resin (Sigma) as described previously (Song et al., 2001). GST-tagged Rap GTPases (4 pmol) preloaded with [³H]GDP (3000 c.p.m./pmol, NEN) were incubated with PLC (1 pmol) in 50 l of the reaction mixture, and in vitro GEF activities were assayed as described (Jin et al., 2001).

In vivo GEF (pull-down) assay

BaF3-PDGFR(Y977F/Y989F)/PLC cells transfected with pEF-BOS-HA-Ha-Ras by electroporation (for Ha-Ras) or BaF3-PDGFR(Y977F/Y989F)/PLC cells (for Rap1) were serum-starved in RPMI 1640 supplemented with bovine serum albumin (1 mg/ml) for 2 h. Thereafter, cells were stimulated with PDGF (50 ng/ml), and GTP-bound forms of Ha-Ras and Rap1 were detected by pull-down assays using GST-Raf-1-Ras-binding domain (RBD) (for Ha-Ras) or GST-Ral guanine nucleotide dissociation stimulator (RalGDS)-Rap1-interacting domain (RID) (for Rap1) as previously described (Gao et al., 2001). Ha-Ras and Rap1 were visualized by immunoblotting using anti-HA (12CA5, Roche) and anti-Rap1 (sc-65, Santa Cruz) antibodies, respectively.

Measurement of IP₃ in BaF3-derived transfectants

BaF3-derived transfectants were serum-starved in RPMI 1640 supplemented with bovine serum albumin (1 mg/ml) for 2 h, and stimulated with PDGF (50 ng/ml). Intracellular IP₃ concentrations were determined by using the IP₃ [³H] radioreceptor assay kit (NEN) according to the manufacturer's instructions.

Acknowledgements

We thank Dr Nagahiro Minato for the SPA-1 cDNA and Dr Garry Nolan for helpful advice. This investigation was supported by Grants-in-aid for Scientific Research in Priority Areas and for Scientific Research (B) and (C) from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

Baron CL, Malhotra V. (2002). Science, 295: 325-328. Article MEDLINE

Beranger F, Goud B, Tavitian A, de Gunzburg J. (1991). Proc. Natl. Acad. Sci. USA, 88: 1606-1610. MEDLINE

Bos JL. (1998). EMBO J., 17: 6776-6782. Article MEDLINE

Bos JL, de Rooij J, Reedquist KA. (2001). Nat. Rev. Mol. Cell Biol., 2: 369-377. Article MEDLINE

De Matteis MA, Santini G, Kahn RA, Tullio GD, Luini A. (1993). Nature, 364: 818-821. MEDLINE

Evellin S, Nolte J, Tysack K, vom Dorp F, Thiel M, Oude Weernink PA, Jakobs KH, Webb E, Lomasney JW, Schmidt M. (2002). J. Biol. Chem., 277: 16805-16813.

Fantl WJ, Johnson DE, Williams LT. (1993). Annu. Rev. Biochem., 62: 453-481. MEDLINE

Gao X, Satoh T, Liao Y, Song C, Hu C-D, Kariya K, Kataoka T. (2001). J. Biol. Chem., 276: 42219-42225.

Jin TG, Satoh T, Liao Y, Song C, Gao X, Kariya K, Hu C-D, Kataoka T. (2001). J. Biol. Chem., 276: 30301-30307. Article MEDLINE

Johnson DE. (1998). Front. Biosci., 3: d313-d324. MEDLINE

Katagiri K, Hattori M, Minato N, Irie S, Takatsu K, Kinashi T. (2000). Mol. Cell. Biol., 20: 1956-1969. Article MEDLINE

Kelley GG, Reks SE, Ondrako JM, Smrcka AV. (2001). EMBO J., 20: 743-754. Article MEDLINE

Kurachi H, Wada Y, Tsukamoto N, Maeda M, Kubota H, Hattori M, Iwai K, Minato N. (1997). J. Biol. Chem., 272: 28081-28088. Article MEDLINE

Lopez I, Mak EC, Ding J, Hamm HE, Lomasney JW. (2001). J. Biol. Chem., 276: 2758-2765. Article MEDLINE

Matsubara K, Kishida S, Matsuura Y, Kitayama H, Noda M, Kikuchi A. (1999). Oncogene, 18: 1303-1312. Article MEDLINE

Mizushima S, Nagata S. (1990). Nucleic Acids Res., 18: 5322. MEDLINE

Ohba Y, Mochizuki N, Matsuo K, Yamashita S, Nakaya M, Hashimoto Y, Hamaguchi M, Kurata T, Nagashima K, Matsuda M. (2000). Mol. Cell. Biol., 20: 6074-6083. Article MEDLINE

Reedquist KA, Ross E, Koop EA, Wolthuis RM, Zwartkruis FJ, van Kooyk Y, Salmon M, Buckley CD, Bos JL. (2000). J. Cell Biol., 148: 1151-1158. Article MEDLINE

Satoh T, Fantl WJ, Escobedo JA, Williams LT, Kaziro Y. (1993). Mol. Cell. Biol., 13: 3706-3713. MEDLINE

Schmidt A, Caron E, Hall A. (2001a). Mol. Cell. Biol., 21: 438-448. Article MEDLINE

Schmidt M, Evellin S, Oude Weernink PA, vom Dorp F, Rehmann H, Lomasney JW, Jakobs KH. (2001b). Nat. Cell Biol., 3: 1020-1024. Article MEDLINE

Shibatohge M, Kariya K, Liao Y, Hu C-D, Watari Y, Goshima M, Shima F, Kataoka T. (1998). J. Biol. Chem., 273: 6218-6222. MEDLINE

Shima F, Yamawaki-Kataoka Y, Yanagihara C, Tamada M, Okada T, Kariya K, Kataoka T. (1997). Mol. Cell. Biol., 17: 1057-1064. MEDLINE

Singer WD, Brown HA, Sternweis PC. (1997). Annu. Rev. Biochem., 66: 475-509. Article MEDLINE

Song C, Hu C-D, Masago M, Kariya K, Yamawaki-Kataoka Y, Shibatohge M, Wu D, Satoh T, Kataoka T. (2001). J. Biol. Chem., 276: 2752-2757. Article MEDLINE

Tsukamoto N, Hattori M, Yang H, Bos JL, Minato N. (1999). J. Biol. Chem., 274: 18463-18469. Article MEDLINE

Wienecke R, Maize JC, Shoarinejad F, Vass WC, Reed J, Bonifacino JS, Resau JH, de Gunzburg J, Yeung RS, DeClue JE. (1996). Oncogene, 13: 913-923. MEDLINE

Wing MR, Houston D, Kelley GG, Der CJ, Siderovski DP, Harden TK. (2001). J. Biol. Chem., 276: 48257-48261. MEDLINE

York RD, Molliver DC, Grewal SS, Stenberg PE, McCleskey EW, Stork PJ. (2000). Mol. Cell. Biol., 20: 8069-8083. Article MEDLINE

York RD, Yao H, Dillon T, Ellig CL, Eckert SP, McCleskey EW, Stork PJ. (1998). Nature, 392: 622-626. Article MEDLINE

Figure 1 Activation of PLC by coexpression of Ras or Rap. (a) Activation of PLC by constitutively active Ha-Ras, but not its effector region mutant. Ha-Ras^G12V or its effector region mutant Ha-Ras^G12V,Y32F was expressed with or without PLC in COS-7 cells, and inositol phosphates were quantitated. Results are shown as the means±s.d. (n=3). *P<0.01 compared with mock-transfected cells. (b) Activation of PLC by constitutively active Rap1A, but not its effector region mutant. Rap1A^G12V or its effector region mutant Rap1A^G12V,D38A was expressed with or without PLC in COS-7 cells, and inositol phosphates were quantitated. Results are shown as the means±s.d. (n=3). **P<0.05 compared with mock-transfected cells. (c) Activation of PLC by constitutively active forms of Ras family GTPases. Inositol phosphates in COS-7 cells ectopically expressing PLC and indicated small GTPases were quantitated. Fold increase in the level of inositol phosphates relative to that in cells expressing PLC alone is shown as the means±s.d. (n=3)

Figure 2 Specificity of the RA and CDC25 homology domains. (a) In vitro binding of the RA domain to Rap GTPases. GST-tagged Rap GTPases were preloaded with GTPS (T) or GDP (D) and incubated with MBP-PLCRA (2094-2302) immobilized on amylose resin. GTPases bound to MBP-PLCRA(2094-2302) (Bound Rap) and one-tenth aliquots of GTPases in the reaction mixture (Input of Rap) were detected by immunoblotting using an anti-GST antibody. Representative results of three independent experiments are shown. (b) GEF activity of the CDC25 homology domain. Rap GTPases preloaded with [³H]GDP were incubated at 30°C for 20 min with excess amounts of unlabeled GTP in the presence (black bars) or absence (gray bars) of PLC. [³H]GDP remaining bound to the protein was quantitated and expressed as the percentage of the values at the zero time point. Results are shown as the means±s.d. (n=3)

Figure 3 Activation of Ras and Rap following PDGF stimulation. (a) BaF3-PDGFR(Y977F/Y989F)/PLC cells were transiently transfected with the HA-tagged Ha-Ras cDNA. The GTP-bound form of HA-Ha-Ras in these cells was detected by the pull-down assay following PDGF stimulation for indicated times. Representative results of three independent experiments are shown. (b) The GTP-bound form of endogenous Rap1 in BaF3-PDGFR(Y977F/Y989F)/PLC cells was detected by the pull-down assay following PDGF stimulation for indicated times. Representative results of three independent experiments are shown. (c) The GTP-bound form of endogenous Rap1 in BaF3-PDGFR(Y977F/Y989F)/PLCN cells was detected by the pull-down assay following PDGF stimulation for indicated times. Representative results of three independent experiments are shown

Figure 4 Time courses of IP₃ production following IL-3 or PDGF stimulation. (a) BaF3-PDGFR(Y977F/Y989F) cells were stimulated with IL-3 (circles) or PDGF (triangles), and intracellular IP₃ levels were quantitated. Representative results of three independent experiments performed in duplicate are shown. (b) BaF3-PDGFR(Y977F/Y989F)/PLC cells were stimulated with IL-3 (circles) or PDGF (triangles), and intracellular IP₃ levels were quantitated. Representative results of three independent experiments performed in duplicate are shown. (c) BaF3-PDGFR(Y977F/Y989F)/PLCN cells were stimulated with IL-3 (circles) or PDGF (triangles), and intracellular IP₃ levels were quantitated. Representative results of three independent experiments performed in duplicate are shown

Figure 5 Effect of dominant-negative Ras and the Rap GTPase-activating protein SPA-1 on PDGF-induced production of IP₃ in BaF3-PDGFR(Y977F/Y989F)/PLC cells. BaF3-PDGFR(Y977F/Y989F)/PLC cells were mock-infected (circles) or infected with a retrovirus expressing Ha-Ras^S17N (triangles) or SPA-1 (squares). Following stimulation with PDGF, intracellular IP₃ levels were quantitated. Representative results of three independent experiments performed in duplicate are shown

Figure 6 Effect of PLC on proliferation of BaF3-derived cells. (a) Proliferation of BaF3-PDGFR(Y977F/Y989F) cells. BaF3-PDGFR(Y977F/Y989F) cells (1.0´10⁵ cells) were cultivated in the presence of FBS (circles), FBS plus PDGF (triangles) or FBS plus IL-3 (squares), and cell numbers were counted. Mean values of three or four independent experiments performed in duplicate are shown. (b) Proliferation of BaF3-PDGFR(Y977F/Y989F)/vector cells. BaF3-PDGFR(Y977F/Y989F)/vector cells were cultivated as in (a). Mean values of three or four independent experiments performed in duplicate are shown. (c) Proliferation of BaF3-PDGFR(Y977F/Y989F)/PLC (clone # 8) cells. BaF3-PDGFR(Y977F/Y989F)/PLC (clone # 8) cells were cultivated as in (a). Mean values of three or four independent experiments performed in duplicate are shown. (d) Proliferation of BaF3-PDGFR(Y977F/Y989F)/PLC (clone # 10) cells. BaF3-PDGFR(Y977F/Y989F)/PLC (clone # 10) cells were cultivated as in (a). Mean values of three or four independent experiments performed in duplicate are shown. (e) Effect of Ha-Ras^S17N and SPA-1. BaF3-PDGFR(Y977F/Y989F)/PLC (clone # 8) cells were infected with retroviruses expressing indicated proteins. Infected cells (5.0´10⁵ cells) were cultivated in the presence of FBS plus PDGF (black bars) or FBS plus IL-3 (gray bars) for 3 days, and cell numbers were counted. Mean values of three or four independent experiments performed in duplicate are shown