THE JOURNAL OF BIOLOGICAL CHEMISTRY © 2003 by The American Society for Biochemistry and Molecular Biology, Inc.
Vol. 278, No. 42, Issue of October 17, pp. 41253–41258, 2003 Printed in U.S.A.
Direct Activation of Phospholipase C-⑀ by Rho* Received for publication, June 27, 2003, and in revised form, August 1, 2003 Published, JBC Papers in Press, August 4, 2003, DOI 10.1074/jbc.M306904200
Michele R. Wing, Jason T. Snyder, John Sondek‡, and T. Kendall Harden§ From the Department of Pharmacology, the Department of Biochemistry and Biophysics, the Lineberger Comprehensive Cancer Center, and the Curriculum in Neurobiology, University of North Carolina School of Medicine, Chapel Hill, North Carolina 27599
Unique among the phospholipase C isozymes, the recently identified phospholipase C-⑀ (PLC-⑀) contains an amino-terminal CDC25 domain capable of catalyzing nucleotide exchange on Ras family GTPases as well as a tandem array of Ras-associating (RA) domains near its carboxyl terminus that are effector binding sites for activated H-Ras and Rap. To determine whether other small GTPases activate PLC-⑀, we measured inositol phosphate accumulation in COS-7 cells expressing a broad range of GTPase-deficient mutants of Ras superfamily proteins. RhoA, RhoB, and RhoC all markedly stimulated inositol phosphate accumulation in PLC-⑀-expressing cells. This stimulation matched or exceeded phospholipase activation promoted by co-expression of PLC-⑀ with the known regulators Ras, G␣12/13, or G1␥2. In contrast, little effect was observed with the other Rho family members Rac1, Rac2, Rac3, and Cdc42. Truncation of the two carboxylterminal RA domains caused loss of responsiveness to H-Ras but not to Rho. Truncation of PLC-⑀ to remove the CDC25 and pleckstrin homology (PH) domains also did not cause loss of responsiveness to Rho, G␣12/13, or G1␥2. Comparative sequence analysis of mammalian phospholipase C isozymes revealed a unique ⬃65 amino acid insert within the catalytic core of PLC-⑀ not present in PLC-, ␥, ␦, or . A PLC-⑀ construct lacking this region was no longer activated by Rho or G␣12/13 but retained regulation by G␥ and H-Ras. GTP-dependent interaction of Rho with PLC-⑀ was illustrated in pull-down experiments with GST-Rho, and this interaction was retained in the PLC-⑀ construct lacking the unique insert within the catalytic core. These results are consistent with the conclusion that Rho family GTPases directly interact with PLC-⑀ by a mechanism independent of the CDC25 or RA domains. A unique insert within the catalytic core of PLC-⑀ imparts responsiveness to Rho, which may signal downstream of G␣12/13 in the regulation of PLC-⑀, because activation by both Rho and G␣12/13 is lost in the absence of this sequence.
Phospholipase C (PLC)1 isozymes cleave phosphatidylinositol 4,5-bisphosphate to generate the second messengers inositol * This work was supported by National Institutes of Health Grants GM29316, GM65533, and GM57391. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ‡ Recipient of support from the Pew Charitable Trusts. § To whom correspondence should be addressed: University of North Carolina School of Medicine, Dept. of Pharmacology, CB 7365, Chapel Hill, NC 27599-7365. E-mail:
[email protected]. 1 The abbreviations used are: PLC, phospholipase C; GEF, guanine nucleotide exchange factor; GST, glutathione S-transferase; HA, hemagglutinin A; PBS, phosphate-buffered saline; PH, pleckstrin homology; RA, Ras-associating. This paper is available on line at http://www.jbc.org
1,4,5-trisphosphate and diacylglycerol, which subsequently trigger Ca2⫹ release from internal stores and activation of protein kinase C, respectively (1). Historically, three different isoforms of PLC, which differ in their mechanisms of activation, were thought to underlie the formation of these two second messengers (2). The heterotrimeric G protein subunits G␣q (3– 6) and G␥ (7–9) directly activate PLC- isozymes, whereas the activity of PLC-␥ isozymes is increased via phosphorylation and translocation as a consequence of the activation of tyrosine kinase receptors (10 –12). The mechanism(s) of regulation of PLC-␦ is less clear, but it has been suggested to involve Gh proteins (13) as well as Ca2⫹ mobilization (14). A fourth member of the family of PLC enzymes, PLC-⑀, was identified recently (15–18). This ⬃250 kDa isozyme contains the structural core (EF-hands, catalytic TIM barrel, and C2 domain) common to all PLC isozymes as well as additional domains at the amino and carboxyl termini that implicate interactions with Ras family GTPases. More specifically, a CDC25 domain at the amino terminus of PLC-⑀ functions as a guanine nucleotide exchange factor for H-Ras and/or Rap1 (16, 17, 19), whereas two Ras-associating (RA) domains at the carboxyl terminus bind H-Ras and several Rap isozymes in a GTP-dependent manner (16, 18). Direct binding of H-Ras to the RA2 domain of PLC-⑀ results in increased accumulation of inositol phosphates in COS-7 cells co-expressing constitutively active H-Ras and PLC-⑀ (18). G␣12/13, but not G␣q or other heterotrimeric G␣ subunits (17, 20), and G␥ subunits of heterotrimeric G-proteins (20) stimulate phospholipase activity of PLC-⑀; however, whether G␣12/13 and G␥ subunits activate PLC-⑀ via a direct interaction with the enzyme or through intermediate proteins remains unclear. The goals of this study were to determine whether additional members of the Ras superfamily of small GTPases regulate the phospholipase activity of PLC-⑀ and to delimit the domains of PLC-⑀ necessary for activation by G␣12/13, G␥, and Ras family GTPases. Our analyses have revealed novel regulation of PLC-⑀ by Rho family GTPases through a mechanism not involving the previously identified CDC25 and RA domains of the isozyme. Instead, we have identified an insertion within the Y region of PLC-⑀ not present in other PLC isozymes that is essential for Rho and G␣12/13 activation of PLC-⑀. In vitro binding experiments illustrate a GTP-dependent and a Y box insert-independent interaction of RhoA with PLC-⑀. EXPERIMENTAL PROCEDURES
Materials—The open reading frame of rat PLC-⑀ in pCMV-script (kind gift of Grant Kelley, State University of New York, Syracuse, NY) was subcloned into pCMV-Myc resulting in an amino-terminal Myctagged rPLC-⑀. pcDNA3 expression vectors for untagged G␣12 and G␣13 (21) were obtained from Channing Der (University of North Carolina). hG1 pcDNA3.1 (amino-terminal Myc-His) and hG␥2 pcDNA3.1 (amino-terminal double hemagglutinin A) expression vectors were as described previously (22). Hemagglutinin A-tagged wild type and GTPase deficient mutants of Rho family GTPases were obtained from the Guth-
41253
41254
Activation of PLC-⑀ by Rho
rie Institute (Sayre, PA). A rabbit PLC-⑀ antibody was generated by BioSource against a predicted exposed sequence, LKAHQTPVDILKQKAHQL, immediately carboxyl-terminal to the X box. Computational Tools—Vector NTI (Informax Inc.) and NCBI PsiBlast (www.ncbi.nlm.nih.gov/BLAST) were utilized to identify the unique insert in the Y box (TIM barrel) of PLC-⑀. Multiple sequence alignments were made with the AlignX module of Vector NTI, which employs a ClustalW alignment tool. Transfection of COS-7 Cells and Quantitation of Phospholipase C Activity—COS-7 cells were seeded in 12-well culture dishes at a density of ⬃60,000 cells per well and maintained in high glucose Dulbecco’s modified Eagle’s medium containing 10% fetal bovine serum, 100 units/ml penicillin, and 100 g/ml streptomycin at 37 °C in an atmosphere of 90% air/10% CO2. The indicated DNA vectors were transfected using FuGENE 6 (Roche Applied Science) transfection reagent (3 l of FuGENE per1 g of DNA) according to the manufacturer’s protocol. Total DNA varied from 400 to 800 ng per well and included empty vector as necessary to maintain an equal amount of DNA per well within individual experiments. Approximately twenty-four hours after transfection, culture medium was changed to inositol-free Dulbecco’s modified Eagle’s medium (ICN Biomedicals) containing 1 Ci/well myo[2-3H(N)]inositol (American Radiolabeled Chemicals), and metabolic labeling was allowed to proceed for 12–16 h. Accumulation of [3H]inositol phosphates was quantitated subsequent to the addition of 10 mM LiCl to inhibit inositol monophosphate phosphatases. The reaction was stopped after 60 min by aspiration of the medium and the addition of 50 mM formic acid followed by neutralization with 150 mM NH4OH. [3H]Inositol phosphates were quantified by Dowex chromatography as described previously (23). Construction of PLC-⑀ Fragments—PLC-⑀ fragments were constructed by polymerase chain reaction amplification of the selected region from full-length rPLC-⑀ cDNA following standard protocols and were ligated in-frame into pCMV-Myc (Clontech). Constructs were verified by sequencing, and the expression of all constructs in COS-7 cells was confirmed by Western blots with c-Myc antibody (Invitrogen). Specific constructs and corresponding amino acid ranges (see Fig 2A) include the following: CDC25-C2, 509 –1987; CDC25-RA1, 509 –2113; CDC25-Cterm, 509 –2281; EF-Cterm, 1198 –2281; and EF-Cterm⌬Y, 1198 –2281 lacking residues 1667–1728). GST-Rho Pull-down Assays—The coding sequences for human Ras, Rac1, and RhoA were amplified by PCR and subcloned into linear pGEX4T2 (Amersham Biosciences) in-frame with the GST coding sequence. BL21(DE3) cells were transformed with individual GSTGTPase plasmids, grown to mid-log phase, and protein production was induced with 1 mM isopropyl-1-thio--D-galactopyranoside for 8 h at 27 °C. Cells were harvested, resuspended in PBS, lysed with an EmulsiFlex C5 (Avestin), and clarified supernatant was loaded onto glutathione-Sepharose. GST-GTPases were eluted after several wash steps with 10 mM glutathione, and the resulting recombinant protein was dialyzed against 20 mM Hepes (pH 7.5), 150 mM NaCl, and 5% glycerol. Purified GTPases were loaded with either GDP or GTP␥S by a 30-min incubation at room temperature in the presence of 2 mM EDTA and either 200 M GDP or GTP␥S followed by the addition of 10 mM MgCl2. Loaded GTPases were immobilized on glutathione-Sepharose (Amersham Biosciences) for 1 h at 4 °C and washed twice with PBS containing 5 mM MgCl2. The concentrations of GTPases were determined after SDSPAGE and Coomassie staining, and all pull-downs were carried out with equivalent amounts of each GTPase. COS-7 cells were seeded at 3 ⫻ 106 cells per 150-mm dish and transfected with FuGENE 6 (Roche Applied Science) ⬃8 h later with the indicated DNAs. Forty-eight hours post-transfection, cells were rinsed twice with 10 ml of ice-cold PBS and lysed in 1.5 ml lysis buffer (50 mM Tris pH 7.5, 150 mM NaCl, 1% Triton X-100, and 5 mM MgCl2, containing protease inhibitors). COS-7 cell lysates were rocked for 30 min at 4 °C and then centrifuged at 30,000 ⫻ g for 30 min at 4 °C. Two hundred microliters of each supernatant was incubated overnight at 4 °C with ⬃8 g GDP- or GTP␥S-loaded GST-RhoA, GST-Rac1, or GST-H-Ras. Glutathione-Sepharose resin was washed twice with lysis buffer and twice with PBS containing 5 mM MgCl2, resuspended in 25 l of 2⫻ Laemmli sample buffer, and subjected to SDS-PAGE and transfer to nitrocellulose. Western blotting was performed using anti-Myc (Invitrogen) and anti-PLC-2 (Santa Cruz Biotechnology) primary antibodies. Ten microliters of each COS-7 cell lysate was used to confirm the expression of transfected DNA.
RESULTS
Activation of PLC-⑀ by the Small GTPase Rho—Previous studies illustrated that H-Ras and Rap isozymes interact with the Ras-associating domains of PLC-⑀ to stimulate phospholipase activity (16, 18). To determine whether other small GTPases also activate PLC-⑀, COS-7 cells were transfected with DNA vectors encoding PLC-⑀ and various small GTPases of the Rho family prior to measuring the total inositol phosphate accumulation indicative of PLC-⑀ activity (Fig. 1). As observed previously, transfection of PLC-⑀ alone resulted in a small increase in [3H]inositol phosphates that was substantially augmented by cotransfection with G1␥2 or GTPase-deficient mutants of G␣12, G␣13, or H-Ras (Fig 1A). Surprisingly, constitutively active forms of RhoA, RhoB, and RhoC also markedly stimulated inositol phosphate accumulation after cotransfection with PLC-⑀ (Fig 1B, and data not shown). In fact, the constitutively active Rho isoforms consistently produced higher inositol phosphate accumulation relative to the known activators of PLC-⑀. In contrast, constitutively active Rac1, Rac2, Rac3, or Cdc42 had little or no effect on inositol phosphate formation catalyzed by cotransfected PLC-⑀ (Fig 1B, and data not shown). Enhanced phospholipase C activity was dependent on PLC-⑀, because transfection of the Rho isozymes alone did not cause marked increases in inositol phosphate accumulation (Fig 1C). Furthermore, whereas wild-type versions of the Rho isozymes augmented the phospholipase activity of PLC-⑀, this activation was routinely enhanced with constitutively active mutants of the GTPases (Fig 1C) consistent with the possibility that PLC-⑀ is a direct, downstream effector of Rho. Western blots indicated approximately equivalent levels of expression of each small GTPase under the conditions of these assays (Fig 1C, insets), and no changes in the levels of expression of PLC-⑀ occurred as a consequence of coexpression with the various GTPases (data not shown). Definition of Regions of PLC-⑀ Essential for Activation by Rho, G␣12/13, and G␥—Interaction of PLC-⑀ with H-Ras is dependent on the presence of the second Ras-associating domain of PLC-⑀, and disruption of this interaction by mutation of lysine 2150 to glutamate in rat PLC-⑀ prevents the phospholipase-stimulating activity of H-Ras (18). To assess the role of the RA domains in the regulation of PLC-⑀ by other activators, COS-7 cells were transfected with cDNA encoding either Rho, G␣12/13, or G1␥2 and one of a series of PLC-⑀ fragments (Fig 2A). Although deletion of either the last (CDC25-RA1) or both of the RA domains (CDC25-C2) from an amino-terminally truncated form of PLC-⑀ (CDC25-Cterm) eliminated activation by GTPase-deficient H-Ras, the marked stimulation of phospholipase C by constitutively active Rho (as well as by G␣12/13 or G1␥2; data not shown) was essentially unaffected (Fig 2B). Therefore, the RA domains of the isozyme are necessary for activation by H-Ras but not by other activators, including those of the Rho family of small GTPases. Results from our previous study indicated that activation of PLC-⑀ by G␥ was independent of either H-Ras or phosphatidylinositol-3 kinase (PI3K), and a pleckstrin homology (PH) domain was identified in PLC-⑀ as a potential site of interaction with G␥ (20). To test whether the PH domain of PLC-⑀ is essential for activation by G1␥2, G␣12/13, or Rho, these proteins were co-expressed with an amino-terminally truncated fragment of PLC-⑀ lacking the PH domain (EF-Cterm) and encoding sequence from the EF-hands of PLC-⑀ to its carboxyl terminus. Surprisingly, G1␥2, as well as G␣12/13, RhoA, RhoB, and H-Ras all activated this fragment of PLC-⑀, indicating that the CDC25 and PH domains are not essential for PLC-⑀ activation by any of these G proteins (Fig 2C). Thus, in combination with the previous results that eliminated the RA domains
Activation of PLC-⑀ by Rho
FIG. 1. Rho GTPases, but not Rac or Cdc42, activate PLC-⑀. A, constitutively active forms of G␣12, G␣13, and H-Ras as well as G1␥2 were expressed in COS-7 cells in the absence and presence of full-length PLC-⑀, and phospholipase C activity was quantified as described under “Experimental Procedures.” Inositol phosphate accumulation promoted by each G protein subunit in the absence of exogenous PLC-⑀ was subtracted from corresponding samples in which the G proteins were co-expressed with PLC-⑀. The dotted line indicates the activity observed with expression of PLC-⑀ alone. The results are representative of those obtained in three or more experiments. B, constitutively active forms of the Rho GTPases RhoA, RhoB, Rac1, and Cdc42 were expressed in the absence and presence of PLC-⑀ in COS-7 cells, and phospholipase activity was quantified. As in panel A, inositol phosphate accumulation in the absence of PLC-⑀ was subtracted from corresponding PLC-⑀-expressing samples. The dotted line indicates activity observed with expression of PLC-⑀ alone. The results are representative of those obtained in three or more experiments. C, increasing amounts of wild-type or constitutively active RhoA, RhoB, and RhoC were expressed in the absence or presence of PLC-⑀. The dotted line indicates the phospholipase activity of cells expressing PLC-⑀ in the absence of exogenous G protein under the conditions of the assay. Insets confirm expression of HA-tagged wild-type and GTPase-deficient Rho subunits from an equivalent well. For all experiments, DNA of the indicated G protein was transfected with 50 ng of empty vector or 50 ng of PLC-⑀, and data are the mean ⫾ S.D. for duplicate samples in one experiment. The results are representative of those obtained in three experiments.
as important determinants for PLC-⑀ activation by Rho, G␣12/13, or G␥, these modulators require only the core catalytic domain of PLC-⑀ (EF-hands to C2 domain) to enhance phospholipase activity. Identification of an Insertion in the Y Box of PLC-⑀ —As illustrated in the experiments above, PLC-⑀ is unique among
41255
FIG. 2. The known small GTPase-interacting domains of PLC-⑀ are not required for Rho activation. A, domain architecture of PLC-⑀. From the amino terminus, PLC-⑀ consists of a cysteine-rich domain (Cys) not previously reported, a CDC25 domain (CDC25), a PH domain, a series of EF-hands (EF), two regions of high sequence conservation among all PLCs (X and Y), which correspond to the catalytic TIM-barrel, a C2 domain, and two RA domains. Domain boundaries are drawn to scale, and the corresponding amino acid ranges are for the rat PLC-⑀. Note that this representation differs significantly from previous reports on several points. For example, the CDC25 domain has been extended at its amino terminus to include a conserved region (Src0) necessary for the structural integrity of the catalytic portion (Src1–3) of the CDC25 region of Sos1 (38). Similarly, the region containing EFhands is amino-terminally extended to encompass additional EF-hands 1 and 2 and several insertions not previously identified in our earlier study (20). Also, the Y box of PLC-⑀ is extended to accommodate a ⬃65 amino acid insertion not previously recognized. Finally, we have identified a cysteine-rich domain amino-terminal to the CDC25 domain with significant homology to LDLa modules typically found within the extracellular domains of low density lipoprotein receptors. Fragments used in these studies are represented below the schematic of PLC-⑀, and the amino-terminal boundaries in these constructs were defined prior to the discovery of the Src0 and additional EF-hands. B, the indicated PLC-⑀ fragments or empty vector (100 ng) was co-expressed with GTPase-deficient RhoA or H-Ras (50 ng), and PLC activity was quantified. The inositol phosphate accumulation stimulated by each G protein in the absence of exogenous PLC-⑀ was subtracted from PLC-⑀expressing samples. C, the CDC25 and PH domains of PLC-⑀ are dispensable for phospholipase activation by all known activators. The indicated, constitutively active GTPase or G1␥2 (10 –300 ng to yield maximum stimulation as seen in Fig. 1) was co-transfected with empty vector or PLC-⑀ EF-Cterm (300 ng). Data shown are the mean ⫾ S.D. for triplicate samples in one experiment. Results are representative of three or more experiments in each case.
PLC isozymes in requiring only the catalytic core for activation by Rho, G␣12, and G␣13. To identify unique regions within the catalytic core of PLC-⑀ that potentially could impart this responsiveness, pair-wise and multiple sequence alignments were performed comparing PLC-⑀ isoforms with the other
41256
Activation of PLC-⑀ by Rho
FIG. 3. PLC-⑀ possesses a unique insertion within the Y box highly conserved among PLC isozymes. A, multiple sequence alignment of the major PLC subfamilies highlighting the ⬃65 amino acid insertion within the amino-terminal portion of the Y box between -strands T5 and T6 as defined (39) and unique to PLC-⑀ isozymes. Species of origin are denoted with a two-letter suffix as follows: rn, Rattus norvegicus; hs, Homo sapiens; and ce, Caenorhabditis elegans. B, crystal structure (39) of the catalytic TIM-barrel of rPLC-␦1 (Protein Data Bank accession number 1DJX) highlighting the surface-exposed loop (red), which corresponds to residues 508 –513 of PLC-␦ (SSPGTS) and denotes the corresponding location of the Y-insert within PLC-⑀. A ball-and-stick representation of inositol 1,4,5-trisphosphate (Ins(1,4,5)P3) is located at the active site formed by the X and Y boxes highlighted in yellow and green.
classes of PLC. This analysis revealed a previously unidentified ⬃65 residue insertion (residues 1669 –1730 in PLC-⑀) within the Y box of all forms of PLC-⑀, including the distantly related PLC-⑀ from Caenorhabditis elegans, that is not present in other PLC isozymes (Fig 3A). Examination of the PLC-␦ structure indicates that this insert maps to a small surface-exposed loop within the catalytic core distinct from the highly disordered and poorly conserved X-Y linker characteristic of all PLC isozymes (Fig 3B). The unique placement of this insertion within PLC-⑀ suggests that it would be available to interact with other proteins that specifically regulate PLC-⑀ compared with other PLC isozymes. Given the clear demarcation of this insertion between regions of high sequence conservation corresponding to known secondary structure elements within PLC␦1, we anticipated that deletion of this region would not destroy
proper folding of PLC-⑀. As such, the unique insert was deleted from the amino-terminally truncated form of PLC-⑀ (EFCterm), and the resultant protein, EF-Cterm⌬Y, was shown to retain phospholipase C activity when over-expressed in COS-7 cells (Fig 4A). Furthermore, EF-Cterm⌬Y retained capacity for activation by H-Ras and G1␥2 (Fig 4A). In marked contrast, EF-Cterm⌬Y was completely unresponsive to stimulation of phospholipase activity by either G␣12/13 or Rho (Fig 4A), directly implicating this unique ⬃65 amino acids insertion in the regulation of PLC-⑀ by both G␣12/13 and Rho. This differential responsiveness was further illustrated in an experiment in which amounts of transfected DNA were varied over several orders of magnitude, and phospholipase stimulation was normalized to account for differences in intrinsic activity of the phospholipase variants (Fig 4B). Constitutively active RhoA
Activation of PLC-⑀ by Rho
41257
FIG. 5. RhoA binds directly to PLC-⑀ in a GTP-dependent manner. GST-RhoA, GST-Rac1, and GST-H-Ras were loaded with GDP or GTP␥S, immobilized on glutathione-Sepharose, and incubated with COS-7 cell lysates expressing Myc-PLC-⑀ EF-C, Myc-PLC-⑀ EF-C⌬Y, PLC-2, or Myc-ROCK. Western blots generated after simultaneous incubation with primary antibodies for both Myc and PLC-2 are shown. The lysate panel represents 1/20 of total lysates used for each pull-down.
either the GDP- or GTP␥S-bound state. As expected, Rac1, which activates PLC-2 (24⫺26) but not PLC-⑀ (Fig 1B), precipitated in a GTP␥S-dependent manner with PLC-2. Taken together, these results illustrate interaction of PLC-⑀ with RhoA and suggest that, although the unique Y box insert is required for Rho activation of PLC-⑀, it does not solely account for Rho interaction. DISCUSSION
FIG. 4. Rho and G␣12/13 do not activate PLC-⑀ lacking the unique insertion within the Y box. A, COS-7 cells were transfected with the indicated GTPase-deficient G protein subunits (maximally activating amounts as in Fig. 1, A and B) in the absence of exogenous PLC or the presence of either amino-terminally truncated PLC-⑀, (EFCterm, 300 ng) or the identical construct with the Y box insert deleted (EF-Cterm⌬Y, 300 ng). Total inositol phosphate accumulation was quantified, and values in the absence of exogenous PLC were subtracted from the corresponding values in the presence of PLC-⑀ fragments. B, COS-7 cells were transfected with varying amounts of DNA encoding constitutively active RhoA or H-Ras and either EF-Cterm or EF-Cterm⌬Y (300 ng), and inositol phosphate accumulation was quantified. To account for the difference in basal activity of the PLC-⑀ fragments observed in panel A, values were normalized to represent the fold increase in inositol phosphate accumulation.
elevated the phospholipase activity of EF-Cterm by ⬎10-fold, whereas the mutant form of PLC-⑀ lacking the insertion was essentially unresponsive to the GTPase mutant of RhoA over a broad range of amounts of transfected DNA. Again, H-Ras activation of PLC-⑀ was independent of the Y-box insertion irrespective of the amount of H-Ras DNA that was transfected. GTP-dependent Interaction of Rho with PLC-⑀ —To determine whether PLC-⑀ interacts directly with Rho and whether the Y box insert is uniquely responsible for this interaction, COS-7 cells were transfected with DNA encoding one of the following: Myc-EF-Cterm; Myc-EF-Cterm ⌬Y; Myc-Rho kinase (ROCK); or PLC-2 (Fig. 5). Cell lysates obtained after 24 h of culture under each condition were incubated with GST-RhoA, GST-Rac1, or GST-H-Ras, each prepared in either the GDP- or GTP␥S-bound form. Both PLC-⑀ EF-Cterm and EF-Cterm ⌬Y precipitated with GST-RhoA and GST-H-Ras in the GTP␥Sbound but not the GDP-bound form. In contrast, no interaction was observed between the PLC-⑀ isozymes and GST-Rac1 in
The recent identification of PLC-⑀ as a Ras- and G␣12/13activated enzyme that also exhibits Ras GEF activity has added complexity to inositol lipid signaling. PLC-␥ was previously thought to underlie inositol phosphate signaling in response to tyrosine kinase receptors, and, likewise, G proteincoupled, receptor-stimulated inositol phosphate production was thought to occur through PLC- isozymes and G␣q or G␥. Activation of PLC-⑀ by G␣12/13, G␥, and Ras reveals a convergence of tyrosine kinase/Ras GTPase and heterotrimeric G protein signaling through a common effector. Results reported here illustrate that Rho family GTPases regulate PLC-⑀ and do so through a unique region, further amplifying the complexity of regulation both of PLC-⑀ and of inositol lipid signaling in general. The initial identification of the rat and human PLC-⑀ genes revealed the presence of tandem RA domains at the carboxyl terminus (16 –18), which suggested interaction of this novel isozyme with Ras family G proteins. GTP-dependent binding of H-Ras with the second RA domain was directly confirmed and shown to be essential for Ras stimulation of phospholipase activity of PLC-⑀ (16, 18). Whether increased lipase activity is due to direct activation of PLC-⑀ or translocation of the enzyme to the membrane, as is the case with other Ras effectors, remains unclear. The current data indicate that, in addition to Ras, the Rho family GTPases Rho A, B, and C, but not the highly homologous Rac1, 2, 3, or Cdc42, activate PLC-⑀. Indeed, the extent of activation observed with the Rho GTPases was consistently greater than that observed with H-Ras. Moreover, truncation/deletion analyses revealed that Rho-dependent activation of PLC-⑀ is independent of the CDC25, PH, and RA domains, which distinctly contrasts with the requirement of an intact RA2 domain for Ras activation of PLC-⑀. Lomasney and co-workers reported that PLC-⑀ is not simply a Ras effector but also acts as a Ras activator, suggesting an intriguing positive feedback loop (17) with PLC-⑀ activating Ras via the GEF activity of a CDC25 domain in the amino terminus. However, Kataoka and co-workers, working with a
41258
Activation of PLC-⑀ by Rho
different amino-terminal splice variant of human PLC-⑀, have suggested that the GEF domain is selective for Rap over Ras (19, 27). Although the CDC25 domain of PLC-⑀ could yet prove to interact with Rho GTPases as downstream signaling molecules, this domain clearly is not involved in the activating effects observed with Rho GTPases in our study. Previous studies have illustrated that Ras and G␣12/13 subunits of heterotrimeric G proteins activate PLC-⑀ through separate mechanisms (17, 18, 20). However, the regions of the enzyme essential for G␣12/13 regulation have remained undefined. We illustrate here that the amino terminus, CDC25 domain, PH domain, and RA domains are not necessary for G␣12/13 regulation of PLC-⑀. Identification of a unique sequence within the Y domain of PLC-⑀ suggested that this region might be responsible for Rho, G␣12/13, and/or G␥ activation of the isozyme. This sequence is predicted to be an exposed region on the surface of the enzyme based on the crystal structure of PLC-␦. The functional importance of this sequence is supported by our observations that its removal eliminated Rho- and G␣12/13- (but not Ras- or G␥-) promoted activation of PLC-⑀. GTP-dependent interaction of Rho with PLC-⑀ was illustrated in pull-down experiments with GST-Rho. Whereas the unique sequence in the Y domain of PLC-⑀ was necessary to observe Rho-dependent stimulation of phospholipase C activity, binding of GTP-activated Rho was retained in a mutant PLC-⑀ lacking the insert. This sequence may constitute part, but not all of a binding site, in essence being necessary, but not sufficient, to bind G protein alone. Therefore, the structural basis for the interaction of Rho with PLC-⑀ and its concomitant activation remain to be defined. In a previous study, we identified a PH domain and EF-hand region in PLC-⑀ and illustrated that G␥ subunits stimulate phospholipase activity of PLC-⑀ (20). Because direct interactions of G␥ with the PH domains of PLC- isozymes (28, 29) and other proteins (30, 31) have been well characterized, we hypothesized that the PH domain of PLC-⑀ might mediate the stimulating actions of G␥. The experiments reported here illustrate that the PH domain is not necessary for G␥ activation of PLC-⑀ when expressed in COS-7 cells. Indeed, mutant PLC-⑀ fragments missing the amino-terminal region, CDC25 domain, and PH domain apparently retained full regulation by G␥, and the region of G␥ regulation of PLC-⑀ was delimited to the catalytic core of the enzyme (EF-hand, X, Y, and C2 domains). An ␣-helix within the Y box of PLC-2 appears to mediate its interaction with G␥ (32). A similar region may be involved in G␥ regulation of PLC-⑀, although the Tyr-578 residue reported to be essential for G␥ activation of PLC-2 is not present in PLC-⑀. The addition of Rho to the known regulators of PLC-⑀ introduces a previously unsuspected signaling pathway into an effector that generates inositol 1,4,5-trisphosphate and diacylglycerol. As such, PLC-⑀ is implicated as a potential effector in Rho-dependent processes, including cytoskeletal rearrangement, vascular smooth muscle contraction, gene expression, and cell growth. Many of these processes are regulated by G protein-coupled receptors (i.e. for thrombin, lysophosphatidic acid, endothelin, etc.), perhaps via the G␣12 family of G proteins (33). The recent identification of Rho GEFs that are activated by G␣12/13 (i.e. p115RhoGEF and LARG) bridges the gap between heterotrimeric and Rho GTPases (34 –37). Because both G␣12/13 and Rho activation of PLC-⑀ are dependent on a relatively short sequence within the Y box, we speculate,
although we have no direct evidence, that Rho mediates the G␣12/13 activation of PLC-⑀. Acknowledgements—The authors are indebted to Svetlana Gershburg for excellent technical assistance and to Krister Wennerberg and Keith Burridge for helpful discussions. REFERENCES 1. Berridge, M. J., and Irvine, R. F. (1987) Annu. Rev. Biochem. 56, 159 –193 2. Rhee, S. G., and Bae, Y. S. (1997) J. Biol. Chem. 272, 15045–15048 3. Smrcka, A. V., Hepler, J. R., Brown, K. O., and Sternweis, P. C. (1991) Science 251, 804 – 807 4. Taylor, S. J., Chae, H. Z., Rhee, S. G., and Exton, J. H. (1991) Nature 350, 516 –518 5. Waldo, G. L., Boyer, J. L., Morris, A. J., and Harden, T. K. (1991) J. Biol. Chem. 266, 14217–14225 6. Wu, D. Q., Lee, C. H., Rhee, S. G., and Simon, M. I. (1992) J. Biol. Chem. 267, 1811–1817 7. Camps, M., Carozzi, A., Schnabel, P., Scheer, A., Parker, P. J., and Gierschik, P. (1992) Nature 360, 684 – 686 8. Boyer, J. L., Waldo, G. L., and Harden, T. K. (1992) J. Biol. Chem. 267, 25451–25456 9. Blank, J. L., Brittain, K. A., and Exton, J. H. (1992) J. Biol. Chem. 267, 23069 –23075 10. Wahl, M. I., Nishibe, S., Suh, P.-G., Rhee, S. G., and Carpenter, G. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, 1568 –1572 11. Meisenhelder, J., Suh, P.-G., Rhee, S. G., and Hunter, T. (1989) Cell 57, 1099 –1122 12. Rhee, S. G., and Choi, K. D. (1992) J. Biol. Chem. 267, 12393–12396 13. Baek, K. J., Kang, S., Damron, D., and Im, M. (2001) J. Biol. Chem. 276, 5591–5597 14. Kim, Y. H., Park, T. J., Lee, Y. H., Baek, K. J., Suh, P. G., Ryu, S. H., and Kim, K. T. (1999) J. Biol. Chem. 274, 26127–26134 15. Shibatohge, M., Kariya, K., Liao, Y., Hu, C. D., Watari, Y., Goshima, M., Shima, F., and Kataoka, T. (1998) J. Biol. Chem. 273, 6218 – 6222 16. Song, C., Hu, C. D., Masago, M., Kariyai, K., Yamawaki-Kataoka, Y., Shibatohge, M., Wu, D., Satoh, T., and Kataoka, T. (2001) J. Biol. Chem. 276, 2752–2757 17. Lopez, I., Mak, E. C., Ding, J., Hamm, H. E., and Lomasney, J. W. (2001) J. Biol. Chem. 276, 2758 –2765 18. Kelley, G. G., Reks, S. E., Ondrako, J. M., and Smrcka, A. V. (2001) EMBO J. 20, 743–754 19. Jin, T. G., Satoh, T., Liao, Y., Song, C., Gao, X., Kariya, K., Hu, C. D., and Kataoka, T. (2001) J. Biol. Chem. 276, 30301–30307 20. Wing, M. R., Houston, D., Kelley, G. G., Der, C. J., Siderovski, D. P., and Harden, T. K. (2001) J. Biol. Chem. 276, 48257– 48261 21. Zohn, I. E., Klinger, M., Karp, X., Kirk, H., Symons, M., Chrzanowska-Wodnicka, M., Der, C. J., and Kay, R. J. (2000) Oncogene 19, 3866 –3877 22. Blake, B. L., Wing, M. R., Zhou, J. Y., Lei, Q., Hillmann, J. R., Behe, C. I., Morris, R. A., Harden, T. K., Bayliss, D. A., Miller, R. J., and Siderovski, D. P. (2001) J. Biol. Chem. 276, 49267– 49274 23. Brown, H. A., Lazarowski, E. R., Boucher, R. C., and Harden, T. K. (1991) Mol. Pharmacol. 40, 648 – 655 24. Illenberger, D., Schwald, F., Pimmer, D., Binder, W., Maier, G., Dietrich, A., and Gierschik, P. (1998) EMBO J. 17, 6241– 6249 25. Illenberger, D., Walliser, C., Nurnberg, B., Diaz Lorente, M., and Gierschik, P. (2003) J. Biol. Chem. 278, 3006 –3014 26. Snyder, J. T., Singer, A. U., Wing, M. R., Harden, T. K., and Sondek, J. (2003) J. Biol. Chem. 278, 21099 –21104 27. Song, C., Satoh, T., Edamatsu, H., Wu, D., Tadano, M., Gao, X., and Kataoka, T. (2002) Oncogene 21, 8105– 8113 28. Wang, T., Dowal, L., El Maghrabi, M. R., Rebecchi, M., and Scarlata, S. (2000) J. Biol. Chem. 275, 7466 –7469 29. Barr, A. J., Ali, H., Haribabu, B., Snyderman, R., and Smrcka, A. V. (2000) Biochemistry 39, 1800 –1806 30. Touhara, K., Inglese, J., Pitcher, J. A., Shaw, G., and Lefkowitz, R. J. (1994) J. Biol. Chem. 269, 10217–10220 31. Mahadevan, D., Thanki, N., Singh, J., McPhie, P., Zangrilli, D., Wang, L. M., Guerrero, C., LeVine, H., III, Humblet, C., Saldanha, J., Gutkind, J. S., and Najmabadi-Haske, T. (1995) Biochemistry 34, 9111–9117 32. Sankaran, B., Osterhout, J., Wu, D., and Smrcka, A. V. (1998) J. Biol. Chem. 273, 7148 –7154 33. Seasholtz, T. M., Majumdar, M., and Brown, J. H. (1999) Mol. Pharmacol. 55, 949 –956 34. Kozasa, T., Jiang, X., Hart, M. J., Sternweis, P. M., Singer, W. D., Gilman, A. G., Bollag, G., and Sternweis, P. C. (1998) Science 280, 2109 –2111 35. Hart, M. J., Jiang, X., Kozasa, T., Roscoe, W., Singer, W. D., Gilman, A. G., Sternweis, P. C., and Bollag, G. (1998) Science 280, 2112–2114 36. Kuner, R., Swiercz, J. M., Zywietz, A., Tappe, A., and Offermanns, S. (2002) Eur. J. Neurosci. 16, 2333–2341 37. Fukuhara, S., Chikumi, H., and Gutkind, J. S. (2000) FEBS Lett. 485, 183–188 38. Boriak-Sjodin, P. A., Margarit, S. M., Bar-Sagi, D., and Kuriyan, J. (1998) Nature 394, 337–343 39. Essen, L. O., Perisic, O., Cheung, R., Katan, M., and Williams, R. L. (1996) Nature 380, 595– 602
