Isozyme-specific Stimulation of Phospholipase C-γ2 by Rac GTPases*♦

The regulation of the two isoforms of phospholipase C-γ, PLCγ1 and PLCγ2, by cell surface receptors involves protein tyrosine phosphorylation as well as interaction with adapter proteins and phosphatidylinositol 3,4,5-trisphosphate (PtdInsP3) generated by inositol phospholipid 3-kinases (PI3Ks). All three processes may lead to recruitment of the PLCγ isozymes to the plasma membrane and/or stimulation of their catalytic activity. Recent evidence suggests that PLCγ may also be regulated by Rho GTPases. In this study, PLCγ1 and PLCγ2 were reconstituted in intact cells and in a cell-free system with Rho GTPases to examine their influence on PLCγ activity. PLCγ2, but not PLCγ1, was markedly activated in intact cells by constitutively active Rac1G12V, Rac2G12V, and Rac3G12V but not by Cdc42G12V and RhoAG14V. The mechanism of PLCγ2 activation was apparently independent of phosphorylation of tyrosine residues known to be modified by PLCγ2-activating protein-tyrosine kinases. Activation of PLCγ2 by Rac2G12V in intact cells coincided with a translocation of PLCγ2 from the soluble to the particulate fraction. PLCγ isozyme-specific activation of PLCγ2 by Rac GTPases (Rac1 ≈ Rac2 > Rac3), but not by Cdc42 or RhoA, was also observed in a cell-free system. Herein, activation of wild-type Rac GTPases with guanosine 5′-(3-O-thio)triphosphate caused a marked stimulation of PLCγ2 but had no effect on the activity of PLCγ1. PLCγ1 and PLCγ2 have previously been shown to be indiscriminately activated by PtdInsP3 in vitro. Thus, the results suggest a novel mechanism of PLCγ2 activation by Rac GTPases involving neither protein tyrosine phosphorylation nor PI3K-mediated generation of PtdInsP3.

Inositol phospholipid-specific phospholipases C (PLCs) 3 catalyze the hydrolysis of phosphatidylinositol 4,5-bisphosphate (PtdInsP 2 ) to produce inositol 1,4,5-trisphosphate and diacylglycerol. While the products of this reaction have long been known to act as intracellular second messengers (reviewed in Refs. [1][2][3][4], it has only recently become clear that the phospholipid substrate itself may also be considered an intracellular signaling molecule in that its local concentration in the plasma membrane and possibly in other cellular membranes regulates the activities and/or subcellular distribution of a number of regulatory or structural proteins. These include enzymes, ion channels and transporters, transcription factors, scaffolding proteins, cytoskeletal proteins, as well as proteins involved in the regulation of exocytosis, endocytosis, and membrane trafficking (reviewed in Refs. [5][6][7][8]. The mammalian PLCs are divided into six subfamilies, designated ␤, ␥, ␦, ⑀, , and (reviewed in Refs. 9 and 10). The four members of the PLC␤ subfamily are activated by ␤␥ dimers of heterotrimeric G proteins and/or members of the ␣ q subfamily of G protein ␣ subunits and by certain Rho GTPases (11)(12)(13)(14). The two members of the PLC␥ family are activated by receptor and nonreceptor protein-tyrosine kinases (reviewed in Refs. 9, 10, 15, and 16). In addition, alternative mechanisms exist to regulate the activity of the two PLC␥ isoforms (reviewed in Ref. 17). The mechanisms of PLC␦ regulation are less well understood. Members of this subfamily have been reported to be under stimulatory and/or inhibitory control of a number of low molecular weight substances and proteins, including p122-RhoGAP, RhoA, the ␣ subunit of the heterodimeric GTP-binding protein G h /tissue transglutaminase 2 (reviewed in Refs. 9, 10, and 18) as well as GAP43, PLC␤ 2 , RalA, and RalB (19 -21). PLC⑀ is subject to regulation by ␤␥ dimers, ␣ 12 and ␣ 13 subunits of heterotrimeric G proteins, and several members of the Ras superfamily of small GTPases, including H-Ras, Rap1A, Rap2B, RhoA, RhoB, and RhoC (reviewed in Ref. 22). PLC is specifically expressed in mammalian sperm and is capable of inducing oscillations of intracellular Ca 2ϩ and subsequent early embryonic development when introduced into mammalian eggs upon sperm-egg fusion (23). Very recently, two members of a novel subfamily, PLC 1 and PLC 2 , have been identified and characterized (24,25). While it seems clear that PLC and PLC isozymes are activated by Ca 2ϩ (24 -26), it is currently unknown whether their activity is controlled by regulatory proteins.
We (13) and others (14) have previously reported that constitutively active Rac2 activates PLC␤ 2 in intact cDNA-transfected COS-7 and HEK293 cells and causes translocation of the enzyme from the cytosol to the plasma membrane. Both effects are mediated by the putative amino-terminal pleckstrin homology domain of PLC␤ 2 (12)(13)(14). Rac2 is specifically expressed in cells of hematopoietic origin (27,28), and mice genetically deficient in Rac2 are characterized by defects in cellular functions of hematopoietic stem cells (29,30), neutrophils (31), mast cells (32), T cells (33), and B lymphocytes (34,35). In the latter cells, the absence of Rac2 caused a reduction in the B cell antigen receptor (BCR)mediated increase in cytosolic Ca 2ϩ (34,35). The fact that PLC␥ 2 rather than PLC␤ 2 mediates this response in B cells (reviewed in Refs. 15 and 36) prompted us to examine whether PLC␥ 2 is regulated by Rac2. The results presented herein demonstrate that PLC␥ 2 , but not PLC␥ 1 , is activated by Rac GTPases in both cellular and cell-free systems by a mechanism(s) that involves translocation of the cytosolic PLC␥ 2 isozyme to cellular membranes but is apparently independent of protein tyrosine phosphorylation and inositol phospholipid 3-kinase-mediated generation of PtdInsP 3 .
Cell Culture and Transfection-COS-7 and HEK293 cells were maintained at 37°C in a humidified atmosphere of 95% air and 5% CO 2 in Dulbecco's modified Eagle's medium (Invitrogen, catalog number 41965-039) supplemented with 10% (v/v) fetal calf serum (Invitrogen, catalog number 10270-106) and 2 mM glutamine, 100 units/ml penicillin, 100 g/ml streptomycin, 1 mM sodium pyruvate, and 25 mM HEPES buffer solution (all from PAA Laboratories, Cölbe, Germany). Prior to transfection, COS-7 or HEK293 cells were seeded into 12-well plates at densities of 1 ϫ 10 5 and 2 ϫ 10 5 cells per well, respectively, and grown for 24 h in 1 ml/well of the same medium. One hour before transfection, the medium was replaced with 1 ml/well of fresh medium. For transfection of COS-7 cells, plasmid DNA (1.0 g DNA/well) was mixed with 2.0 l of Lipofectamine TM 2000 Reagent (Invitrogen) in 0.2 ml of Opti-MEM I (Invitrogen) according to the manufacturer's instructions. After the addition of the DNA-Lipofectamine TM 2000 complexes to the dishes, the cells were incubated for a further 24 h at 37°C and 5% CO 2 without changing the medium. For transfection of HEK293 cells, the CalPhos TM mammalian transfection kit (BD Biosciences, Heidelberg, Germany) was used according to the manufacturer's instructions. Plasmid DNA (1.0 g DNA/well) was mixed with 6.2 l of 2 M calcium solution (BD Biosciences), 50 l 2ϫ HEPES-buffered saline (BD Biosciences), and sterile water to obtain a total volume of 100 l. After the addition of the transfection solution to the dishes, the cells were incubated for 3.5 h at 37°C and 5% CO 2 . Subsequently, the medium was replaced with 1 ml/well of fresh medium, and the HEK293 cells were incubated for a further 20.5 h under the same conditions.
Radiolabeling of Inositol Phospholipids and Analysis of Inositol Phosphate Formation-Twenty-four hours after transfection, COS-7 or HEK293 cells were washed once with 0.5 ml/well of buffer A containing 10 mM Na 2 HPO 4 , 1.8 mM KH 2 PO 4 , 140 mM NaCl, 2.7 mM KCl, pH 7.4, and then supplied with 0.4 ml/well of Dulbecco's modified Eagle's medium containing fetal calf serum and supplements as specified above, 10 Ci/ml myo- [2-3 H]inositol (catalog number TRK317, Amersham Biosciences, Freiburg, Germany), and 10 mM LiCl. The cells were incubated in this medium for 20 h, washed once with 0.4 ml/well of buffer A, and then lysed by addition of 0.2 ml/well of 10 mM ice-cold formic acid (39). After keeping the samples on ice for 30 min, 0.3 ml/well of 10 mM NH 4 OH was added for neutralization, and the sample was centrifuged for 5 min at 15,000 ϫ g. The supernatant was loaded onto a column containing 0.25 ml of Dowex 1 ϫ 8 -200 ion exchange resin (catalog number 217425, Sigma, Deisenhofen, Germany), that had been con-verted to the formate form and equilibrated with H 2 O as described (40). The columns were washed twice with 3.5 ml each of 60 mM sodium formate and 5 mM sodium tetraborate, and inositol phosphates were eluted with 3 ml of 1 M ammonium formate and 100 mM formic acid. The eluate was supplemented with 15 ml of scintillation fluid (Ultima Gold TM , PerkinElmer Life Sciences, Rodgau-Jügesheim, Germany), and the radioactivity was quantified by liquid scintillation counting. The columns were reused after regeneration as described previously (40).
Subcellular Fractionation-HEK293 cells (2.7 ϫ 10 6 ) were grown on 100-mm dishes and transiently transfected as described above using the CalPhos TM mammalian transfection kit. Twenty-four hours after transfection, the cells were scraped into 0.15 ml of ice-cold buffer B containing 20 mM Tris/HCl, pH 7.5, 2 mM EDTA, 2 g/ml soybean trypsin inhibitor, 3 mM benzamidine, 0.1 mM phenylmethylsulfonyl fluoride, 1 M pepstatin, 1 M leupeptin, and 1 g/ml aprotinin. The cells were homogenized by forcing the suspension ten times through a 0.45 ϫ 25-mm needle attached to a disposable syringe. After removal of unbroken cells and nuclei by centrifugation at 300 ϫ g for 10 min at 4°C, particulate (P) and soluble (S) fractions were separated by centrifugation at 12,000 ϫ g for 15 min at 4°C.

Production of Recombinant Rho GTPases and Recombinant PLC␥ Isozymes in Baculovirus-infected Insect
Cells-For production of recombinant Rho GTPases, Sf9 cells (Invitrogen) were grown at 27°C in suspension culture in TNM-FH medium containing 10% (v/v) fetal calf serum (catalog number P04 -83500, PAN Biotech, Aidenbach, Germany) supplemented with 0.2% (w/v) Pluronic F-68 (Invitrogen), 50 g/ml gentamicin (PAA Laboratories), and 2.5 g/ml amphotericin B (Fungizone, Invitrogen) in a 1800-ml Fernbach culture flask. Cells (1.2 ϫ 10 9 ) were incubated at 27°C with recombinant baculovirus in 400 ml of medium at 80 rpm on a rotary shaker with an amplitude of 25 mm. Three days after infection, the cells were harvested at room temperature by centrifugation at 300 ϫ g for 5 min followed by two washes with 100 ml each of buffer A. To obtain detergent-solubilized Rho GTPases, the cells were resuspended in 15 ml per 10 9 intact cells at the time of cell harvesting of ice-cold buffer C containing 20 mM Tris/HCl, pH 8.0, 1 mM EDTA, 1 mM dithiothreitol, 100 mM NaCl, 3.75 mM MgCl 2 , 0.1 mM phenylmethylsulfonyl fluoride, 1 g/ml leupeptin, 1 g/ml aprotinin, and 3 M GDP and homogenized using a precooled 5-ml Teflon-glass homogenizer. Nuclei and unbroken cells were removed by centrifugation at 300 ϫ g for 10 min at 4°C. The membrane fraction was collected from the resulting supernatant by centrifugation at 12,000 ϫ g for 15 min at 4°C. Rho GTPases were solubilized by resuspending the membranes in 2 ml per 10 9 intact cells at the time of cell harvesting of ice-cold buffer C containing 23 mM sodium cholate and incubating this mixture for 90 min at 4°C with vigorous vortexing every 10 min. Insoluble material was removed from this suspension by centrifugation at 12,000 ϫ g for 15 min at 4°C. The resulting detergent extract was aliquoted, snap-frozen in liquid N 2 , and stored at Ϫ80°C. For production of recombinant PLC␥ isozymes, Sf9 cells were grown at 27°C in adherent culture in TNM-FH medium (catalog number T3285, Sigma) supplemented with 10% (v/v) fetal calf serum (catalog number F7524, Sigma) and 50 g/ml gentamicin. Cells (120 ϫ 10 6 ) were incubated at 27°C in 96 ml of medium. Three days after infection, the cells were detached from the plastic surface, harvested by centrifugation at 300 ϫ g for 5 min at room temperature, washed twice at room temperature with 32 ml each of buffer A, and then resuspended in 1.2 ml of ice-cold buffer B. The cells were homogenized by forcing the suspension six times through a 0.45 ϫ 25-mm needle attached to a disposable syringe. The homogenate was centrifuged at 100,000 ϫ g for 1 h at 4°C and the resulting supernatant aliquoted, snap-frozen in liquid N 2 , and stored at Ϫ80°C.
Phospholipase C Assay-Phospholipase C activity was determined as described (41), with minor modifications. In brief, 5 l of detergent extract prepared from membranes of Rho-GTPase-baculovirus-infected insect cells were reconstituted with 10 l of the soluble fraction of PLC␥-baculovirus-infected insect cells appropriately diluted in buffer D containing 60 mM HEPES/NaOH, pH 7.2, 84 mM KCl, 3.6 mM EGTA, 2.4 mM dithiothreitol, and incubated in a volume of 60 l containing 50 mM HEPES/NaOH, pH 7.2, 70 mM KCl, 3 mM EGTA, 2 mM dithiothreitol, 536 M phosphatidylethanolamine, 33.4 M [ 3 H]PtdInsP 2 (185 GBq/mol), and the concentrations of sodium deoxycholate and free Ca 2ϩ specified in the legends to Figs. 5-8. The concentration of CaCl 2 required to adjust the concentration of free Ca 2ϩ to the desired value was calculated using the program EqCal for Windows (Biosoft, Ferguson, MO). The reaction was terminated, and the samples were analyzed for inositol phosphates as described (41).

S]GTP␥S Binding-Binding of [ 35 S]GTP␥S
to Rho GTPases was assayed as described previously (11) with minor modifications. Briefly, the detergent extracts prepared from membranes of Rho-GTPasebaculovirus-infected insect cells were diluted 15-fold in buffer C without GDP containing 23 mM sodium cholate and then incubated at 30°C in an incubation mixture (40 l) containing 25 mM HEPES/NaOH, pH 8.0, 1 mM EDTA, 1 mM dithiothreitol, 20 mM MgCl 2 , 100 mM NaCl, 0.1% (v/v) GENAPOL C-100 (Calbiochem, high performance liquid chromatography grade), and 1 M [ 35 S]GTP␥S (46.25 GBq/mmol). The incubation was terminated after 4.5 h by the addition of 2 ml of ice-cold buffer E containing 50 mM Tris/HCl, pH 8.0, 100 mM NaCl, and 5 mM MgCl 2 and rapid filtration through 0.45-m pore size nitrocellulose filters, followed by four washes with 2 ml each of ice-cold buffer E. The filters were dried and the retained radioactivity was determined by liquid scintillation counting. Nonspecific binding was defined as the binding not competed for by 1 mM unlabeled GTP␥S. Under these conditions, GTP␥S binding reaches a plateau after 3 h of incubation (results not shown). Given the fact that the dissociation constant (K D ) of guanine nucleoside di-and triphosphate binding to monomeric Rho GTPases is typically Յ1 nM (Ref.

42), 4 this procedure can be used to estimate the amount of Rho GTPases in detergent extracts prepared from baculovirus-infected insect cells.
Miscellaneous-Recombinant baculoviruses were produced as described (43). A baculovirus-encoding Escherichia coli ␤-galactosidase (44) was a gift from Dr. Michael Ruffing. Protein concentrations were determined according to Bradford (45) using bovine IgG as standard. SDS-PAGE and immunoblotting were performed according to standard protocols (46), except that immunoreactive proteins were visualized using the ECL Western blotting detection system (Amersham Biosciences). The sources of all other reagents are specified in Refs. 11-13 and 41. All experiments were performed at least three times. Similar results and identical trends were obtained each time. Data from representative experiments are shown as means Ϯ standard deviation of triplicate determinations.

RESULTS
We have previously shown that a constitutively active mutant of Rac2, Rac2 G12V , causes a marked stimulation of a fusion protein of PLC␤ 2 and enhanced green fluorescent protein, PLC␤ 2 -enhanced green fluorescent protein in intact cDNA-transfected COS-7 and HEK293 cells (13). To examine the PLC isozyme specificity of this effect, the formation of inositol phosphates was measured in COS-7 cells that had been cotransfected with vector encoding wild-type Rac2 or Rac2 G12V together with vector encoding PLC␤ 1 , PLC␤ 2 , PLC␤ 3 , PLC␥ 1 , PLC␥ 2 , or PLC␦ 1 . Fig. 1 shows that expression of the PLC isozymes caused a variable increase in inositol phosphate formation in the absence of Rac2 and Rac2 G12V , which was maximal (ϳ3.9-fold) for PLC␤ 2 , followed by PLC␥ 1 , PLC␤ 1 , PLC␥ 2 , PLC␦ 1 , and PLC␤ 3 . Expression of wild-type Rac2 had no effect on inositol phosphate formation both in the absence and in the presence of exogenous PLC isozymes. Expression of Rac2 G12V did not affect inositol phosphate formation in the absence of exogenous PLC isozymes and in the presence of PLC␤ 1 , PLC␤ 3 , PLC␥ 1 , or PLC␦ 1 but caused a marked stimulation of inositol phosphate formation in cells 4 M. R. Ahmadian, personal communication. expressing PLC␤ 2 . Interestingly, expression of Rac2 G12V also caused a marked stimulation of inositol phosphate formation in cells expressing PLC␥ 2 . Taking into account that only about 74 and 29% of basal inositol phosphate formation in cells expressing PLC␤ 2 and PLC␥ 2 , respectively, was due to the activity of the exogenous PLC isozyme and that Rac2 G12V caused no change in inositol phosphate formation in the absence of exogeneous PLC isozymes, Rac2 G12V caused an about 4.9-and 24.0-fold stimulation of PLC␤ 2 and PLC␥ 2 , respectively, in this experiment. In additional experiments, we found that PLC␥ 2 was stimulated by Rac G12V to a similar extent in cotransfected HEK293 cells (results not shown). The inset of Fig. 1 shows that PLC␥ 1 was in fact expressed in COS-7 cells transfected with the corresponding cDNA and that the levels of expression of PLC␤ 2 , PLC␥ 1 , and PLC␥ 2 were similar in the absence or presence of wild-type or constitutively active Rac2. Similar results were obtained for the expression of PLC␤ 1 , PLC␤ 3 , and PLC␦ 1 in cotransfected COS-7 cells (data not shown). Thus, the stimulation of inositol phosphate formation by Rac2 G12V in cells expressing PLC␤ 2 or PLC␥ 2 was not explained by an increase in PLC expression, and the inability of Rac2 G12V to stimulate inositol phosphate formation in cells that had been cotransfected with the cDNAs of PLC␤ 1 , PLC␤ 3 , PLC␥ 1 , and PLC␦ 1 was not due to the lack of expression of these PLC isozymes. We occasionally observed increased expression of exogenous proteins in COS-7 cells coexpressing these proteins with high levels of Rac2 G12V (data not shown). The reason(s) for these changes are currently unknown. Fig. 2 shows that the ability to stimulate the activity of PLC␤ 2 and PLC␥ 2 in cotransfected COS-7 cells was not restricted to Rac2 G12V but was also observed for the constitutively active mutants of Rac1 and Rac3, Rac1 G12V and Rac3 G12V , but not for the corresponding mutant of RhoA, RhoA G14V . The stimulation of both PLC␤ 2 and PLC␥ 2 was maximal in cells that coexpressed Rac3 G12V , followed by cells coexpressing Rac2 G12V and cells coexpressing Rac1 G12V . Note that this rank order does not necessarily reflect the rank order of sensitivity of the two PLC isozymes to the three Rac GTPases, since the relative abundance of the latter proteins in the cotransfected cells analyzed in Fig. 2 remained unknown. Consistent with earlier in vitro findings (11,12), expression of Cdc42 G12V caused an about 2.3-fold increase of inositol phosphate formation in cells expressing PLC␤ 2 . In contrast, Cdc42 G12V did not affect this activity in cells expressing PLC␥ 2 . Thus, although the sensitivity of PLC␤ 2 and PLC␥ 2 to stimulation by activated Rac1, Rac2, and Rac3, respectively, appears to be similar, the sensitivity to stimulation by activated Cdc42 distinguishes PLC␤ 2 from PLC␥ 2 . Only marginal, if any, changes of inositol phosphate formation were observed in cells coexpressing the constitutively active Rho GTPases with PLC␥ 1 . In addi-tional experiments (results not shown), we found that coexpression of RhoA G14V , but not of wild-type RhoA, with human PLC⑀ 1b in COS-7 cells caused an about 13.5-fold stimulation of this PLC isozyme, indicating that RhoA G14V was in fact constitutively active in this system. Activation of PLC␥ 2 by cell surface receptors has previously been shown to involve protein phosphorylation at one or several of four tyrosine residues present at positions 753, 759, 1197, and 1217 of PLC␥ 2 (47)(48)(49)(50). To examine whether phosphorylation of any one of these residues was involved in stimulation of inositol phosphate formation by activated Rac GTPases in COS-7 cells expressing PLC␥ 2 , mutants of PLC␥ 2 carrying substitutions of one, two, or all four tyrosine residues by phenylalanine residues were coexpressed with either wild-type Rac2 or Rac2 G12V . Fig. 3A shows that expression of Rac2 G12V , but not of wildtype Rac2, caused a marked increase in inositol phosphate formation in cells expressing PLC␥ 2 , regardless whether Rac2 G12V was coexpressed with wild-type or with mutant PLC␥ 2 . Although the formation of inositol phosphates was lower in cells expressing Rac2 G12V together with the mutant PLC␥ 2 polypeptides, basal inositol phosphate formation was enhanced ϳ4.1-fold by Rac2 G12V even in cells expressing the PLC␥ 2 mutant lacking all four tyrosine residues concurrently, PLC␥ 2 FFFF . Fig.  3B shows that the expression of wild-type and mutant PLC␥ 2 was slightly variable in this experiment, which may contribute, at least to some extent, to the differences in Rac2 G12V -stimulated inositol phosphate formation observed in Fig. 3A. In additional experiments (results not shown), we found that PLC␥ 2 FFFF was consistently expressed at lower levels than wild-type PLC␥ 2 in cotransfected COS-7 cells. These caveats notwithstanding, the results shown in Fig. 3 clearly demonstrate that phosphorylation of the tyrosine residues present in positions 753, 759, 1197, and 1217 is not required for Rac2 G12V to cause stimulation of inositol phosphate formation by PLC␥ 2 .
To investigate the mechanisms by which Rac2 G12V mediates stimulation of inositol phosphate formation in COS-7 cells and HEK293 cells expressing PLC␥ 2 , the influence of wild-type Rac2 and constitutively active Rac2 G12V on the subcellular distribution of PLC␥ 2 was examined in HEK293 cotransfected with vector encoding PLC␥ 2 and either empty vector or vector encoding wild-type Rac2 or Rac2 G12V . Homogenates of transfected cells were fractionated, and aliquots of the postnuclear particulate fraction containing plasma membranes (P) and the soluble fraction (S) were analyzed by immunoblotting. Fig. 4, top panel, shows that PLC␥ 2 was mostly, if not entirely, soluble in the absence of exogenous Rac GTPases and in the presence of wild-type Rac2. Very interestingly, expression of Rac2 G12V caused the appearance of a considerable amount of immunoreactive PLC␥ 2 in the particulate fraction (Fig. 4, top panel). Note that both wild-type Rac2 and Rac2 G12V were present in the partic-FIGURE 2. Specificity of the stimulatory effect of Rho GTPases on the activity of PLC␤ 2 and PLC␥ 2 in intact cells. COS-7 cells were cotransfected as indicated at the abscissa with 750 ng each per well of either empty vector (Mock) or vector encoding either wild-type Rac1, Rac3, Cdc42, or RhoA, or constitutively active Rac1 G12V , Rac2 G12V , Rac3 G12V , Cdc42 G12V , or RhoA G14V , together with 250 ng each per well of either empty vector (Control) or vector encoding PLC␤ 2 (PLC␤ 2 ), PLC␥ 1 (PLC␥ 1 ), or PLC␥ 2 (PLC␥ 2 ). The cDNAs of Rac2 and Rac2 G12V were used at 50 ng each per well, because transfection of COS-7 cells with higher quantities of vector encoding Rac2 G12V caused the cells to round up and to detach from the plastic surface of the culture plate. The total amount of DNA was maintained constant at 1.0 g in each transfection by adding empty vector. The formation of inositol phosphates was measured as described in the legend to Fig. 1. ulate fraction and were expressed at similar levels in cotransfected cells (Fig. 4, upper center panel). These results indicate that expression of constitutively active Rac2 G12V in HEK293 cells causes a considerable portion of exogenous PLC␥ 2 to translocate from the soluble to the particulate fraction.
The next experiments were designed to investigate the regulation of PLC␥ isozymes by Rho GTPases in a cell-free system. To this end, the PLC␥ isozymes and Rho GTPases were separately produced in baculovirus-infected insect cells and functionally reconstituted to study the effect of Rho GTPase activation by the poorly hydrolyzable GTP analogue GTP␥S on the ability of the PLC␥ isozymes to hydrolyze PtdInsP 2 . Fig. 5 shows that both PLC␥ 1 and PLC␥ 2 were produced in baculovirusinfected insect cells and were present in large excess of endogenous phospholipases C in the soluble fraction of infected cells. The enzymatic activity of the two recombinant PLC␥ isozymes was dependent on and markedly stimulated by Ca 2ϩ (Fig. 5, right panel). The concentration dependence on free Ca 2ϩ was similar for the two PLC␥ isozymes with half-maximal and maximal effects at ϳ175 nM and 10 M free Ca 2ϩ , respectively. In the experiment shown in Fig. 6, samples containing either PLC␥ 1 or PLC␥ 2 and displaying similar phospholipase C activities at 1 M free Ca 2ϩ (Fig. 6, left panel) were reconstituted with a detergent extract prepared from membranes of baculovirus-infected insect cells expressing Rac2 and then assayed for phospholipase C activity in the presence of 30 nM free Ca 2ϩ and 1 mM sodium deoxycholate (Fig. 6, right  panel). Under these conditions, addition of GTP␥S caused a marked (ϳ5.2-fold) increase in inositol phosphate formation by the preparation containing PLC␥ 2 but did not affect inositol phosphate formation by the preparation containing PLC␥ 1 . There was no effect of GTP␥S on phospholipase C activity when the two PLC␥ preparations were reconstituted with a detergent extract prepared from membranes of ␤-galactosidase-expressing insect cells or when soluble preparations of the latter cells were reconstituted with the detergent extract containing Rac2 (data not shown). The ability of GTP␥S to cause stimulation of PLC␥ 2 in the presence of Rac2 was concentration-dependent and specific (Fig. 7).
In the presence of 0.65 M GDP, half-maximal and maximal effects of GTP␥S were observed at ϳ1.6 and 100 M, respectively (Fig. 7, left The latter proteins were used as markers for the soluble fraction (S) and the particulate fraction containing plasma membranes (P), respectively, of cotransfected HEK293 cells.  (PLC␥ 2 FF ), or PLC␥ 2 Y753F,Y759F,Y1197F,Y1217F (PLC␥ 2 FFFF ) together with 50 ng each of either empty vector (Control) or vector encoding wild-type Rac2 (Rac2) or constitutively active Rac2 G12V (Rac2 G12V ). The total amount of DNA was maintained constant at 1.0 g in each transfection by adding empty vector. The formation of inositol phosphates was measured as described in the legend to Fig. 1. B, COS-7 cells were cotransfected as described above with either empty vector (Mock) or vector encoding wild-type or mutant PLC␥ 2 together with vector encoding wild-type Rac2 or constitutively active Rac2 G12V . Cells from one well were lysed in 100 l of SDS-PAGE sample preparation buffer. The lysate was subjected to SDS-PAGE, and immunoblotting was performed using antibodies reactive against PLC␥ 2 . NOVEMBER 25, 2005 • VOLUME 280 • NUMBER 47 panel). Thus, maximal activation of Rac2 with GTP␥S caused an ϳ15fold stimulation of inositol phosphate formation by PLC␥ 2 in this experiment. Among the purine nucleoside di-and triphosphates tested, only the poorly hydrolyzable guanine nucleoside triphosphate analogues GTP␥S and GppNHp caused a robust stimulation of inositol phosphate formation in the presence of Rac2 and PLC␥ 2 , ADP, GDP, and GDP␤S   caused no change in this activity. Only minimal, if any, effects were observed upon addition of ATP, ATP␥S, and GTP (Fig. 7, right panel).

Stimulation of Phospholipase C-␥ 2 by Rac GTPases
To examine the specificity of PLC␥ 2 stimulation by Rho GTPase family members, the recombinant Rho GTPases were produced in baculovirus-infected insect cells, extracted from the membrane of infected cells with detergent-containing buffer, and reconstituted with samples containing PLC␥ 2 . The amounts of GTP␥S-binding proteins present in the detergent extracts were assayed by [ 35 S]GTP␥S binding. Fig. 8, left panel, shows that the Rho GTPases Cdc42, RhoA, Rac1, Rac2, and Rac3 were present in the detergent extracts used in this experiment at different concentrations (Cdc42 Ͼ RhoA Ͼ Rac3 Ͼ Rac1 Ϸ Rac2). When equal volumes of these detergent extracts were reconstituted with PLC␥ 2 , addition of GTP␥S instead of GDP caused an ϳ6.5-fold stimulation of inositol phosphate formation in the presence of either Rac1 and Rac2 and an ϳ4.0-fold stimulation in the presence of Rac3. In marked contrast, there was no effect of GTP␥S in the presence of detergent extracts prepared from insect cells expressing ␤-galactosidase, Cdc42, or RhoA (Fig. 8, right panel). In additional experiments (data not shown), we found that GTP␥S caused an ϳ7.3and 3.7-fold activation of inositol phosphate formation upon reconstitution of the latter two extracts with samples containing recombinant PLC␤ 2⌬ (10) and PLC⑀ (22), respectively, confirming that Cdc42 and RhoA were present in these extracts as functional proteins. There was no effect of GTP␥S on inositol phosphate formation when the detergent extracts used in Fig. 8 were reconstituted with soluble fractions of insect cells expressing ␤-galactosidase or PLC␥ 1 (not shown). Given the fact that Cdc42, RhoA, and Rac3 were present in the detergent extracts at ϳ8.0-, 4.2-, and 2.3-fold excess, respectively, of Rac1 and Rac2 (Fig. 8, left panel), these results collectively suggest that PLC␥ 2 is specifically activated by Rac GTPases and that the rank order of Rac GTPases to mediate this activation is Rac1 Ϸ Rac2 Ͼ Rac3.

DISCUSSION
The observation that inactivation of the genes encoding the Rho guanine nucleotide exchange factors (Rho GEFs) Vav1 and Vav2 (51,52) (57) has been taken to suggest that Rho GEFs and Rho GTPases may activate PLC␥ isozymes in intact cells indirectly through enhanced formation of PtdInsP 3 , followed by PH domain-and/or SH2 domainmediated translocation of the PLC␥ isozymes to the plasma membrane or to specific plasma membrane microdomains referred to as lipid rafts or glycolipid-enriched microdomains, GEMs (36). Very recently, RhoA has been shown to be activated, again indirectly and downstream of PI3K(s), in response to BCR stimulation and to promote activation of PLC␥ 2 , Ca 2ϩ mobilization, and B cell proliferation by augmenting the synthesis of the PLC substrate PtdInsP 2 (58).
The results of this study suggest that Rac GTPases may also control the enzymatic activity and/or the subcellular localization of the PLC␥ 2 isozyme independently of enhanced formation of PtdInsP 3 and PtdInsP 2 . Thus, both PLC␥ 1 and PLC␥ 2 have previously been shown to be activated by PtdInsP 3 with very similar efficacies and potencies (57), yet only PLC␥ 2 was activated in this study by Rac GTPases both in intact cells and in a cell-free system (cf. Figs. 1, 2, and 6). Likewise, if Rac GTPases were able to stimulate PLC␥ 2 indirectly by promoting the formation of PtdInsP 2 in intact cells or stabilizing the inositol phospholipids in the cell-free system, all of the PLCs tested should be activated rather than specific representatives of the PLC isozyme family. Since the concentration of free Ca 2ϩ is tightly controlled in the cell-free system, PLC␥ isozyme-specific stimulation of PLC␥ 2 by Rac GTPases in this system also argues against the possibility that the stimulatory effect is due to an increase in the concentration of free Ca 2ϩ , which has previously been shown to cause a C2-domain-mediated PLC␥ 2 translocation (59).
The functional defects observed in mice deficient in PLC␥ 2 are not restricted to B cells but are also observed in platelets, natural killer cells, monocytes/macrophages, and mast cells (60,61 Rho and Rac GTPases as wells as Cdc42, and with Clostridium sordelii lethal toxin, which inactivates Rac GTPases, possibly Cdc42, but not Rho GTPases, caused a complete abrogation of degranulation and an inhibition of intracellular Ca 2ϩ mobilization that normally follows Fc⑀RI ligation. In contrast, treatment of cells with Clostridium botulinum exoenzyme C3 or the fusion toxin C2IN-C3, which selectively inactivate Rho GTPases, was without effect (63,64). Along the same lines, constitutively active mutants of Cdc42 and Rac1 have previously been shown to stimulate exocytosis of secretory granules in RBL-2H3 cells by stimulating InsP 3 formation and Ca 2ϩ mobilization upon antigen stimulation of Fc⑀RI (65). Furthermore, constitutively active Cdc42 and Rac1 reconstituted Fc⑀RI-mediated Ca 2ϩ mobilization and degranulation in B6A4C1 mutant RBL cells that are defective in antigen-stimulated PLC␥ isozyme activity. Overexpression of PLC␥ 1 together with either activated Cdc42 or Rac1 synergistically stimulated degranulation (66). Interestingly, bacterially expressed glutathione S-transferase fusion proteins of constitutively active Cdc42 Q61L and, to a lesser extent, wild-type Cdc42 interacted in vitro with a protein immunoreactive with antibodies raised against amino acids 1249 -1262 of bovine PLC␥ 1 .
There was no effect, however, of activated Cdc42 on PLC␥ 1 activity upon reconstitution of the two purified proteins in a cell-free system (65). At first glance, these findings are not easily consistent with the findings reported here. However, PLC␥ 2 was apparently not examined in the latter studies, and the results presented here do not exclude the possibility that activated Cdc42 and Rac1 interact with PLC␥ 1 and that Cdc42 interacts with PLC␥ 2 without enhancing the catalytic activities of the PLC␥ isozymes. Moreover, activated Cdc42 appeared to interact with a species of PLC␥ 1 that migrated slightly faster on SDS-polyacrylamide gels than the majority of recombinant PLC␥ 1 , suggesting that it may be a modified form of the enzyme that binds most effectively to activated Cdc42 (65). The site(s) of interaction between activated Rho GTPases and PLC␤ 2 has been mapped to the amino-terminal 138 or 144 amino acids of PLC␤ 2 containing the putative PH domain (aa 11-135) of the enzyme (12)(13)(14). Activated Rho GTPases appear to directly and specifically interact with this portion of PLC␤ 2 with affinities (K D ) in the micromolar range (14). The mode of interaction of activated Rac GTPases with PLC␥ 2 and the structural requirements of this interaction are currently unknown. Importantly, the PLC␥ isoforms carry two putative PH domains, one at an amino-terminal position corresponding to the position of the putative PH of PLC␤ 2 (PLC␥ 1 , aa 18 -144; PLC␥ 2 , aa 11-133) and a second, split PH domain located between the two catalytic subdomains X and Y (PLC␥ 1 , PH-n, aa 482-527, PH-c, aa 872-937; PLC␥ 2 , PH-n, aa 468 -513, PH-c, aa 849 -914) (reviewed in Ref. 67). The split PH domain of PLC␥ 1 has been shown to be involved in protein-protein interactions (68,69). It is thus tempting to speculate that activated Rac GTPases interact with PLC␥ 2 either via the amino-terminal or via one or both components of the split PH domain. Interestingly, both PLC␥ isoforms have been shown to mediate agonist-induced Ca 2ϩ entry, at least in part by a phospholipase C activity-independent manner (59,70). Very recent evidence suggests that PH-c of PLC␥ 1 controls cell surface expression of the canonical transient receptor potential channel 3 (TRPC3) by interacting with a complementary partial PH-like domain present within its amino-terminal portion (71). Whether activated Rac GTPases mediate translocation of TRP channels to the plasma membrane via PLC␥ 2 remains an intriguing question to be clarified by future experimentation, in particular since there is precedence for an enhancement of rapid vesicular translocation and insertion of TRP channels by activated Rac1 (72).
Although the changes of B cell development and signaling exhibited by mice deficient in either Rac2 (34,35) or PLC␥ 2 (60) are complex, certain features were observed in both cases, suggesting that the functions of Rac GTPases and PLC␥ 2 in B cells may overlap. These include a paucity of mature B cells, such as B1 B cells, in the periphery, reduced serum concentration of IgM, and a decreased ability to respond to T cell-independent antigens in vivo (34,35,60). While it seems clear that activation of Rac1 and Rac2 is not required for BCR-mediated activation at high levels of BCR cross-linking (34,35), it is likely that Rac GTPases are important under more limiting conditions, i.e. when the antigen is present at low abundance or interacts with the BCR with low affinity, i.e. only for short periods of time (34). Under these circumstances, efficient mechanisms are required to amplify the extracellular signal at the level of the plasma membrane. Interestingly, B cells have been shown to acquire antigen from antigen-presenting cells after formation of synapses, which are characterized by a central supramolecular activation cluster-containing antigen-bound BCR and PLC␥ 2 at high densities. The cSMACs are surrounded by peripheral rings, referred to pSMACs, containing high concentrations of integrins such as LFA-1 (73,74). The interaction of LFA-1 with its main ligand, ICAM-1, which is present on leukocytes, follicular dendritic cells, dendritic cells, and vascular endothelial cells (74), has been shown to lower the threshold of B cell activation (75). Although the relationship between lipid rafts and B cell synapses is still tenuous (76), the observation that Rac GTPases may be targeted to the plasma membrane by integrins via lipid rafts and represent major mediators of integrin signaling (reviewed in Refs. 77 and 78) raises the intriguing possibility that Rac GTPases coordinate, both temporally and spatially, BCR-mediated PLC␥ 2 activation with integrin engagement and thus allow B cells to recognize antigens even at low concentrations in a context-dependent way.