Rac-mediated Stimulation of Phospholipase Cγ2 Amplifies B Cell Receptor-induced Calcium Signaling*♦

Background: Phospholipase Cγ2 (PLCγ2) is stimulated by Rac GTPases through direct protein-protein interaction. Results: The Rac-PLCγ2 interaction markedly enhances B cell-receptor-mediated Ca2+ mobilization and nuclear translocation of the Ca2+-regulated transcription factor NFAT in B cells. Conclusion: Rac-mediated stimulation of PLCγ2 activity amplifies B cell receptor-induced Ca2+ signaling. Significance: A specific Rac-resistant PLCγ2 variant is used to determine the physiological cell signaling relevance of a functional Rac-PLCγ2 interaction in an appropriate cellular context. The Rho GTPase Rac is crucially involved in controlling multiple B cell functions, including those regulated by the B cell receptor (BCR) through increased cytosolic Ca2+. The underlying molecular mechanisms and their relevance to the functions of intact B cells have thus far remained unknown. We have previously shown that the activity of phospholipase Cγ2 (PLCγ2), a key constituent of the BCR signalosome, is stimulated by activated Rac through direct protein-protein interaction. Here, we use a Rac-resistant mutant of PLCγ2 to functionally reconstitute cultured PLCγ2-deficient DT40 B cells and to examine the effects of the Rac-PLCγ2 interaction on BCR-mediated changes of intracellular Ca2+ and regulation of Ca2+-regulated and nuclear-factor-of-activated-T-cell-regulated gene transcription at the level of single, intact B cells. The results show that the functional Rac-PLCγ2 interaction causes marked increases in the following: (i) sensitivity of B cells to BCR ligation; (ii) BCR-mediated Ca2+ release from intracellular stores; (iii) Ca2+ entry from the extracellular compartment; and (iv) nuclear translocation of the Ca2+-regulated nuclear factor of activated T cells. Hence, Rac-mediated stimulation of PLCγ2 activity serves to amplify B cell receptor-induced Ca2+ signaling.


The Rho GTPase Rac is crucially involved in controlling multiple B cell functions, including those regulated by the B cell receptor (BCR) through increased cytosolic Ca 2؉ . The underlying molecular mechanisms and their relevance to the functions of intact B cells have thus far remained unknown.
We have previously shown that the activity of phospholipase C␥ 2 (PLC␥ 2 ), a key constituent of the BCR signalosome, is stimulated by activated Rac through direct protein-protein interaction. Here, we use a Rac-resistant mutant of PLC␥ 2 to functionally reconstitute cultured PLC␥ 2

-deficient DT40 B cells and to examine the effects of the Rac-PLC␥ 2 interaction on BCR-mediated changes of intracellular Ca 2؉ and regulation of Ca 2؉ -regulated and nuclear-factor-of-activated-T-cellregulated gene transcription at the level of single, intact B cells. The results show that the functional Rac-PLC␥ 2 interaction causes marked increases in the following: (i) sensitivity of B cells to BCR ligation; (ii) BCR-mediated Ca 2؉ release from intracellular stores; (iii) Ca 2؉ entry from the extracellular compartment; and (iv) nuclear translocation of the Ca 2؉regulated nuclear factor of activated T cells. Hence, Rac-mediated stimulation of PLC␥ 2 activity serves to amplify B cell receptor-induced Ca 2؉ signaling.
Inositol phospholipid-specific phospholipases C (PLC) 4 catalyze the formation of inositol 1,4,5-trisphosphate (InsP 3 ) and diacylglycerol (DAG) from plasma membrane lipid substrate phosphatidylinositol 4,5-bisphosphate (PtdInsP 2 ) (1). Both the rise of the former two and the decline of the latter may serve as intracellular signals to regulate a myriad of cell functions (2). In B lymphocytes, receptors for cell surface immunoglobulins such as the B cell receptors (BCR), cleavage fragments of the third complement component (CD19/CD21) (3), bacterial, viral, or autoimmunity host DNA (toll-like receptors) (4), and even certain G-protein-coupled chemokine receptors (5) mediate activation of PLC␥ 2 , one of the two human PLC␥ isoforms. The activity of PLC␥ 2 controls many B cell functions, such as protein kinase signaling, nucleocytoplasmic trafficking of transcription factors, proliferation, differentiation, cytoskeletal reorganization, cell adhesion and migration, immunological synapse formation, affinity maturation, autoimmunity, homing to and retention in tissue microenvironments, survival, and susceptibility to transformation (6,7).
Inactivation of the PLC␥ 2 gene in the mouse caused specific defects in most cell types of hematopoietic origin, except for T cells (8,9). Mice lacking PLC␥ 2 showed reduced numbers of mature conventional B cells, a block in pro-B cell differentiation, B1 B cell deficiency, absence of IgM receptor-mediated Ca 2ϩ responses, and B cell-mitogen-induced cell proliferation. PLC␥ 2 also plays important roles in pre-BCR-mediated early B cell development, in BAFF receptor-mediated survival, and in activation of light-chain loci for recombination as well as recep-tor editing of self-reactive B cells (10 -12). Mutationally activated forms of PLC␥ 2 have been identified in mice subjected to N-ethyl-N-nitrosourea mutagenesis (Ali5 and Ali14) and, more recently, in patients with inherited forms of autoinflammation and immunodeficiency (13)(14)(15)(16). These defects also lead to deregulation of B cell functions.
Several of the changes described for PLC␥ 2 Ϫ/Ϫ B cells, e.g. defective Ca 2ϩ signaling, failure to proliferate in response to immunoglobulin receptor stimulation, impediment of B cell development, and failure to mount humoral responses to TD and TI antigens, were also observed in mice carrying deletions in all three genes encoding Vav guanine nucleotide exchange factors of Rho GTPases, Vav1, -2, and -3 (17). These results were difficult to interpret mechanistically because Vav proteins elicit both RhoGEF-dependent and -independent effects (18). However, some of the B cell defects were also observed in mice lacking either Rac2 (19) or both Rac1 and Rac2 (20), including a reduced ability of BCR or CD19 (co)ligation to increase [Ca 2ϩ ] i , suggesting that at least some of the B cell defects commonly observed in PLC␥ 2 and Vav1/2/3-null mice were due to loss of Rac activation. At that time, the available evidence suggested that Rac GTPases might activate PLC␥ 2 indirectly by enhancing the activity of phosphatidylinositol 4-phosphate 5-kinase (21), thus increasing the level of PtdInsP 2 , the substrate of both PLC␥ 2 and phosphoinositide 3-kinase (20). Enhanced availability of substrate to the former and enhanced formation of PtdInsP 3 by the latter were expected to activate PLC␥ 2 (22).
We have previously shown that Rac GTPases, but not Cdc42 or RhoA, activate PLC␥ 2 , but not PLC␥ 1 , by direct proteinprotein interaction (23). Neither enhanced formation of PtdInsP 2 nor PtdInsP 3 nor protein tyrosine phosphorylation are involved in this effect. Unlike activation of PLC␤ 2 , which is mediated by Rac interacting with the N-terminal PH domain of this effector, activation of PLC␥ 2 involves binding of Rac to the bipartite, split PH domain (spPH) juxtaposed between the two halves, X and Y, of the PLC␥ 2 catalytic domain (24). The threedimensional structures of the heterodimeric complex between PLC␥ 2 spPH and GTP␥S-activated Rac2, monomeric spPH, and monomeric Rac2 liganded with either GTP␥S or GDP allowed us to elucidate the conformational changes that accompany the formation of the signaling active PLC␥ 2 -spPH/ Rac2 heterodimer (25). A residue unique for spPH of PLC␥ 2 , but not PLC␥ 1 , Phe-897, was found to be particularly important for the functional and structural PLC␥ 2 -spPH/Rac2 interaction. Replacement of Phe-897 to the corresponding glutamine residue of PLC␥ 1 , F897Q, did not affect the overall three-dimensional structure of the PLC␥ 2 spPH domain, but it specifically blocked the interaction of the mutant domain with activated Rac2 (24,25).
Rac GTPases are present at many intracellular crossroads of B cell signaling. They receive inputs from numerous cell surface receptors to regulate and integrate a host of intracellular signaling proteins, including PLC␥ 2 and many proteins involved in cytoskeletal organization (26). This has made it intrinsically difficult to judge the pertinence of the PLC␥ 2 -Rac interaction observed in cell-free experiments and overexpression studies to cell signaling in more physiologically relevant cellular contexts. In this work, we have used DT40 B cells genetically deficient in PLC␥ 2 for functional reconstitution with the Rac-resistant PLC␥ 2 mutant F897Q to study the effects of a specific loss of the PLC␥ 2 -Rac interaction on BCR-mediated cell signaling. The results reveal that loss of this functional interaction causes a marked decrease of BCR-mediated Ca 2ϩ release from intracellular stores, Ca 2ϩ entry from the extracellular compartment, and nuclear translocation of the Ca 2ϩ -regulated transcription factor NFAT. Some of these changes have previously been observed in Rac-deficient mice (19,20). Hence, the results may provide a mechanistic explanation for findings on the role of Rac GTPases in B cell signaling obtained in vivo that have as yet remained largely unexplained. In addition, these insights into BCR-mediated cell signaling may also apply to the mechanisms of action of other B cell receptors such as CD19/CD21, to other cells of hematopoietic origin, e.g. platelets, and to human diseases, such as certain immunodeficiencies. To our knowledge, this is the first time that a Rho-resistant but otherwise normal Rho effector was reintroduced into a genetically Rho effectordeficient background to determine the relevance of the functional Rho effector interaction in a biologically highly relevant context.
cDNA Cloning-Because the 5Ј end of the mRNA encoding chicken PLC␥ 2 was unknown at the time, 5Ј rapid amplification of cDNA ends (27) was used to gather this information and produce full-length PLC␥ 2 cDNAs from reverse-transcribed DT40 cell mRNA. Two presumably allelic variants were found, which are identical at the protein level to each other and to database entry XP_414166, except for a Gln to His divergence at position 865. Based on the higher frequency (7/10) of His-865 among PCR products of DT40 cell mRNA, this haplotype was used herein for further studies. A histidine is present at this position in PLC␥ 2 of numerous species ranging from fish, such as coelacanth, to mammals, such as cattle or sheep. The plasmid NFAT1c-td-RFP611 encodes amino acids 1-400 of mouse NFAT1c fused to a pseudo-monomeric tandem dimer red fluorescent protein, td-RFP611 (28).
Cell Culture and Analysis of Inositol Phosphate Formation-Chicken WT (RCB1464) and PLC␥ 2 -deficient (PLC␥ 2 Ϫ/Ϫ ; RCB1469) DT40 B cells (all from Cell Bank, RIKEN BioResource Center, Japan) were cultured at 37°C in a humidified atmosphere of 90% air and 10% CO 2 in RPMI 1640 medium containing supplements. DT40 B cells stably reconstituted with either WT or mutant PLC␥ 2 were cultured in RPMI 1640 medium containing supplements and 0.5 g/ml puromycin. COS-7 cells were maintained at 37°C in a humidified atmosphere of 90% air and 10% CO 2 in Dulbecco's modified Eagle's medium (Invitrogen, catalogue no. 41965-039) supplemented with 10% (v/v) FCS (Invitrogen, catalogue no. 10270-106), 2 mM glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin (all from PAA Laboratories). Transfection of COS-7 cells and analysis of inositol phosphate formation to determine PLC activity were carried out as described previously (24).
Preparation of DT40 Cell Lysates-DT40 cells (15 ϫ 10 6 ) were lysed in 240 l of ice-cold buffer A containing 20 mM Tris/HCl, pH 7.5, 2 mM EDTA, 2 g/ml soybean trypsin inhibitor, 3 mM benzamidine, 0.1 mM PMSF, 1 M pepstatin, 1 M leupeptin, and 1 g/ml aprotinin by forcing the suspension eight times through a 0.45 ϫ 25-mm needle attached to a disposable syringe. Nuclei and unbroken cells were removed from the cell lysate by centrifugation at 300 ϫ g for 10 min at 4°C. Fifty g of cell lysate protein were subjected to SDS-PAGE, and immunoblotting was performed using antibodies reactive against the c-Myc epitope and ␤-actin.
Expression of Proteins in Sf9 Cells and Measurement of PLC Activity in Vitro-The production of crude preparations of isoprenylated, recombinant human Rac2 and chicken PLC␥ 2 in Sf9 insect cells and the determination of PLC activity in vitro were carried out as described previously (24).
Determination of the Intracellular Calcium Concentration-To monitor the changes in [Ca 2ϩ ] i in individual DT40 B cells in real time, 3 ϫ 10 5 DT40 B cells per channel were seeded into six channel microscopy slides (uncoated ibidi -slide VI 0.4 , ibidi, catalogue no. 80601), which had been coated with poly-L-lysine (0.01% (w/v)), and incubated for 45 min at 37°C and 10% CO 2 to allow for adhesion to the substrate. Next, unbound cells were flushed out with the culture medium, and the adherent cells were loaded with 2 M fluo-4 AM, 0.02% (v/v) Pluronic F-127, by incubation at room temperature for 30 min in RPMI 1640 medium containing supplements. Subsequently, cells were washed twice with buffer B (20 mM HEPES/NaOH, pH 7.4, 143 mM NaCl, 6 mM KCl, 1 mM MgSO 4 , 5.6 mM glucose). To discriminate between mobilization of Ca 2ϩ from intra-and extracellular sources, BCR stimulation was either performed in buffer B or in buffer B supplemented with 1 mM CaCl 2 . The experiments were performed at room temperature.
Spinning Disc Confocal Fluorescence Microscopy-Fluorescence images were taken with the acquisition software Andor iQ 1.6 using a spinning disk confocal microscope assembled in our laboratory from individual components as described earlier (30). Briefly, its main components are a CSU10 scan head (Yokogawa, Tokyo, Japan), an inverted microscope (Axio Observer, Zeiss) with oil immersion objective (UPlanSApo 60ϫ/1.35 NA, Olympus), environmental control system (PeCon), an image splitter unit (OptoSplit II, Cairn Research), and an EMCCD camera (DV-887, Andor). Diode-pumped solid-state lasers were employed for fluorescence excitation at 473 nm (HB-Laser LSR473-100) and 532 nm (HB-Laser LSR532-250). Timelapse confocal fluorescence images consisting of 512 ϫ 512 pixels were acquired in a field of 1024 ϫ 1024 m 2 with a depth resolution of ϳ2 m, time resolution of 1.5 frames/s, and exposure time for each frame of 400 ms. The 473 nm excitation laser power measured at the optical fiber output to the spinning disc module was set to ϳ4 milliwatts throughout the study.
Quantitative Analysis of the Ca 2ϩ Responses-The fluorescence intensity fluctuations of the Ca 2ϩ -sensitive dye fluo-4 were baseline-corrected and normalized according to the maximal intensity in the presence of ionomycin to account for the differences in fluorescent dye loading between individual cells ( Fig. 4A and supplemental Video S1). Then, baseline-corrected and normalized Ca 2ϩ response traces were subjected to further analysis using algorithms developed in-house and implemented in MATLAB (MathWorks). A Ca 2ϩ response peak was defined by a transient positive divergence from the baseline with a minimum of 10% in intensity of the maximum observed in the presence of ionomycin. The number of responding cells was calculated as the number of traces with at least one Ca 2ϩ response peak and expressed as the percentage of the total number of traces in each experiment. Latency was defined as a mean time lapse between the time of ligand addition and the time corresponding to the maximum of the first Ca 2ϩ response peak. Cells that did not show any detectable Ca 2ϩ spike during the initial observation time for the release of Ca 2ϩ from internal stores, 365.1 s, were considered to display maximal latency equal to the time of observation. The integrated intensity was calculated as the mean area under the curve of all individual traces corresponding to single cells. The peak frequency was calculated for cells showing discrete Ca 2ϩ spikes with baseline resolution, i.e. after addition of anti-IgM at concentrations of 4, 40, and 400 ng/ml, and expressed in millihertz (10 Ϫ3 Hz). The same cells were used to calculate the peak amplitude as the average intensity of the Ca 2ϩ peaks in percentage of the ionomycin maximum. The integrated intensities, peak frequencies, and peak amplitudes of nonresponding cells were set to zero.
Analysis of Nuclear NFAT1c Translocation-To study the BCR-mediated, Ca 2ϩ -dependent translocation of the transcription factor NFAT into the nucleus, 30 g of cDNA encoding NFAT1c-td-RFP611 was transferred into 10 7 PLC␥ 2 Ϫ/Ϫ DT40 B cells per cuvette that had or had not been reconstituted with either WT or F897Q mutant PLC␥ 2 by nucleofection (Amaxa Nucleofector Technology, Cell Line Nucleofector kit T, program B-023). Forty eight h after nucleofection, 6 ϫ 10 5 DT40 B cells per channel were allowed to adhere to poly-Llysine-coated ibidi 6-channel microscopy slides. After two washing steps with buffer B, the cells were treated for 60 min with either 40 g/ml of anti-IgM or 2 M ionomycin in buffer B containing 1 mM Ca 2ϩ . The translocation of NFAT1c-td-RFP611 was analyzed at room temperature on a single cell level by spinning disc confocal microscopy over a time period of 60 min using a 532-nm excitation laser for fluorescence imaging.
Indirect Immunofluorescence Staining-For immunofluorescence staining, 3 ϫ 10 5 DT40 cells per channel were allowed to adhere to tissue culture-treated 6-channel microscopy slides (ibiTreat, ibidi -slide VI 0.4 , ibidi, catalogue no. 80606,) as described above. The cells were then fixed by incubation for 15 min in 4% (w/v) paraformaldehyde in phosphate-buffered saline (PBS), pH 7.4. After washing the cells twice with PBS, the nonreacted aldehyde groups were quenched by treatment for 10 min with 50 mM NH 4 Cl in PBS. The cells were then washed four times with PBS, and nonspecific protein-binding sites in the channels were blocked by incubation for 1 h with buffer C (PBS, pH 7.4, 5% (v/v) FCS, 0.02% (w/v) sodium azide, 0.01% (w/v) saponin, 0.1% (v/v) Triton X-100). The cells were then incubated for 45 min with buffer C containing the primary antibody reactive against c-Myc epitope (1:2000). After four washing steps with buffer D (PBS, pH 7.4, 0.01% (w/v) saponin, 0.1% (v/v) Triton X-100), the cells were incubated for 45 min with buffer C containing the secondary goat anti-mouse IgG antibody conjugated with Alexa Fluor 488 (1:1000). After four further washing steps with buffer D, the expression of PLC␥ 2 protein was analyzed on a single cell level by spinning disk confocal microscopy employing a 473-nm excitation laser for fluorescence and a halogen lamp for transmission light imaging. All steps were done at room temperature.
FRAP Experiments-FRAP studies (32, 33) were conducted as described earlier (34). The experiments were performed 24 -26 h post-transfection on COS-7 cells transfected with PLC␥ 2 -GFP derivatives. All experiments were conducted at 22°C, in Hanks' balanced salt solution supplemented with 20 mM HEPES, pH 7.2. The monitoring argon ion laser beam (488 nm, 1.2 microwatts; Innova 70C, Coherent) was focused through the microscope (AxioImager.D1, Carl Zeiss MicroImaging) to a Gaussian spot with a radius ϭ 0.77 Ϯ 0.03 m (ϫ63/1.4 NA oil-immersion objective) or 1.17 Ϯ 0.05 m (ϫ40/1.2 NA water immersion objective). Experiments were conducted with each beam size (beam size analysis is described in Refs. 35,36). The ratio between the illuminated areas ( 2 (ϫ40)/ 2 (ϫ63)) was 2.28 Ϯ 0.17 (n ϭ 59). After a brief measurement at the monitoring intensity, a 5-milliwatt pulse (4 -6 ms or 10 -20 ms for the ϫ63 and ϫ40 objectives, respectively) bleached 50 -70% of the fluorescence in the spot. Fluorescence recovery was followed by the monitoring beam. The apparent characteristic fluorescence recovery time () and the mobile fraction (R f ) were derived from the FRAP curves by nonlinear regression analysis, fitting to a lateral diffusion process with a single value (37). The significance of differences between values measured with the same beam size was evaluated by Student's t test. To compare ratio measurements ((ϫ40)/ (ϫ63) and 2 (ϫ40)/ 2 (ϫ63)), we employed bootstrap analysis, which is preferable for comparison between ratios (38), as described by us earlier (36), using 1000 bootstrap samples.
Miscellaneous-SDS-PAGE and immunoblotting were performed according to standard protocols using antibodies reactive against the c-Myc epitope for wild-type and mutant PLC␥ 2 . Immunoreactive proteins were visualized using the ECL Western blotting detection system (GE Healthcare). 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 S.E. of triplicate determinations, if not stated otherwise in the figure legends. Unless stated otherwise, the significance of differences was assessed by using either the unpaired t test with two-tailed p values or repeated measures analysis of variance with Tukey's post test (Fig. 7B), both contained in GraphPad InStat, version 3.10. Statistically significant effects are denoted by ***, p Ͻ 0.001; **, 0.001 Ͻ p Ͻ 0.01; and *, 0.01 Ͻ p Ͻ 0.05. Nonsignificant (ns) changes are denoted by ns, 0.05 Ͻ p. In Figs. 1, D and E, and 6, A-C, the data were fitted by nonlinear least squares curve fitting to three or four parameter dose-response equations using GraphPad Prism, version 5.04. In certain cases, the global curve fitting procedure contained in Prism was used to determine whether the best fit values of selected parameters differed between data sets. The simpler model was selected unless the extra sum of squares F-test had a p value of less than 0.05. In Fig.  6A, right panel, the data were fitted to the equation of a bellshaped dose response curve provided by Prism, with manual adjustments.

Results
Point Mutation F897Q Renders the DT40 Cell PLC␥ 2 Orthologue Specifically Resistant to Stimulation by Activated Rac2-Mutational and structural analyses have shown that the stimulation of human PLC␥ 2 by activated Rac GTPases is due to an interaction of the Rac switch I and II regions with specific residues in the C-terminal half of the PLC␥ 2 spPH domain (24,25). Fig. 1A illustrates the importance of the PLC␥ 2 residue Phe-897, contained within the C-terminal ␣ helix of spPH, in this respect. Phe-897 forms the core of a hydrophobic pocket on PLC␥ 2 spPH, which interacts with several hydrophobic residues of the Rac2 switch I and II regions, such as Val-36 (I), Phe-37 (I), Trp-56 (II), Leu-67 (II), and Leu-70 (II). Rac2-PLC␥ 2 complex formation is thought to have two consequences as follows: (i) conversion of GTP-bound Rac2 from conformational state 1 to state 2. These states were initially described for H-Ras as having low versus high affinity, respectively, toward Ras effectors (39). (ii) Stabilization of the Phe-897 side chain was in one of the two conformational states observed in the crystal structures of free spPH (25). Replacement of Phe-897 in human PLC␥ 2 spPH by the corresponding human PLC␥ 1 residue glutamine is expected to reduce the hydrophobic momentum of the Rac2 binding pocket on PLC␥ 2 spPH and, possibly, interfere with the reorientation of the Leu-67 side chain upon complex formation (Fig. 1A). Consistent with this view, the F897Q substitution blocked activation of human PLC␥ 2 by constitutively active Rac2 and abolished binding of GTP␥S-activated Rac2 to PLC␥ 2 spPH, while leaving the overall fold of PLC␥ 2 spPH unaffected (24,25).
To minimize the differences between the cellular system to be functionally reconstituted and analyzed here and unmodified DT40 B cells, a cDNA encoding chicken PLC␥ 2 was produced using mRNA from DT40 cells as a template. The encoded protein shares 83% identical residues with human PLC␥ 2 . Within the C-terminal helices of the two spPH domains, 14 of 17 residues, including Phe-897, are identical, and three are conserved. The experiments shown in Fig. 1, B and C, examined the consequences of the F897Q mutation on the ability of the DT40 PLC␥ 2 orthologue (henceforth referred to as PLC␥ 2 ) to interact with activated Rac2. Fig. 1B shows that constitutively active Rac2 G12V , but not WT Rac2, caused a marked (ϳ19-fold) stimulation of PLC␥ 2 activity in COS-7 cells, but it was ineffective in control cells lacking PLC␥ 2 and in cells expressing the F897Q mutant. These observations were confirmed in a cell-free system (Fig. 1C). The F897Q mutation did not affect activation of PLC␥ 2 by calcium ions or by loss of autoinhibition (Fig. 1, D and E) and had no influence on the extent of PLC␥ 2 protein tyrosine phosphorylation in response to H 2 O 2 (Fig. 1F).
To measure the effect of the F897Q mutation on the interactions of PLC␥ 2 with the plasma membrane in intact cells, FRAP experiments were performed on GFP-tagged WT and the F897Q mutant PLC␥ 2 . PLC␥ 2 -GFP fluorescence recovery occurred with a fluorescence recovery time () of 0.079 s ( Fig.  2A, left panel). Coexpression of constitutively active Rac2 G12V enhanced the interaction of the enzyme with the plasma membrane, resulting in an ϳ1.5-fold increase in of PLC␥ 2 -GFP to 0.115 s ( Fig. 2A, right panel). Interestingly, FRAP beam size analysis experiments showed that the WT enzyme and the variant F897Q did not differ in their modes of membrane interaction, as the fluorescence recovery of both PLC␥ 2 -GFP and PLC␥ 2 F897Q -GFP occurred by a mixture of binding to and dissociation from the membrane, referred to as exchange, and of stable association with the membrane, resulting in lateral diffusion. In the presence of Rac2 G12V , the fluorescence recovery of WT PLC␥ 2 -GFP was shifted toward lateral diffusion, as the ratio between the fluorescence recovery times with the two beam sizes (ϫ40 and ϫ63 objectives) was 2.3 ( Fig. 2, B and C). This ratio is indistinguishable from that expected for recovery by pure lateral diffusion (2.28 Ϯ 0.17, the ratio between the areas illuminated by the two beam sizes employed). Consequently, in the presence of Rac2 G12V , the interaction of PLC␥ 2 -GFP with the plasma membrane is dominated by lateral diffusion, suggesting that the exchange rate becomes much slower than the lateral diffusion rate (i.e. a shift to stable membrane interactions). In marked contrast, Rac2 G12V had no effect at all on the mode of membrane association of the mutant PLC␥ 2 F897Q -GFP, which correlates well with our observations that Rac2 G12V cannot interact with and fails to stimulate PLC␥ 2 F897Q . We conclude that PLC␥ 2 is not compromised in its overall activity and in its mode of membrane interaction by the F897Q mutation, in contrast to its specific loss of regulation by activated Rac.
Functional Reconstitution of Wild-type and F897Q Mutant PLC␥ 2 into PLC␥ 2 Ϫ/Ϫ DT40 B Cells-Next, genetically PLC␥ 2deficient DT40 cells were stably reconstituted with either isogenic WT PLC␥ 2 or the PLC␥ 2 F897Q mutant such that the resultant cell clones were indistinguishable in terms of enzyme expression and subcellular distribution (Fig. 3). Spinning disc confocal fluorescence microscopy was then used to determine   Ϫ/Ϫ cells, before and after stable reconstitution of the latter with either WT or F897Q mutant PLC␥ 2 , for their ability to respond to BCR ligation with increases in [Ca 2ϩ ] i . In accordance with earlier studies (40), cells were first treated with 40 ng/ml anti-IgM in the absence of extracellular Ca 2ϩ to measure the BCR-mediated Ca 2ϩ release from intracellular stores. After some time, the effect of the same concen- tration of anti-IgM was determined in the presence of 1 mM extracellular Ca 2ϩ to also allow for the entry of extracellular Ca 2ϩ into the cells. Finally, the Ca 2ϩ ionophore ionomycin was added to normalize the fluorescence intensities obtained for the individual cells (Fig. 4A). The latter values were subsequently used to normalize the individual intensity time traces. Sixty cells from each cell type were analyzed individually in each experiment (Fig. 4B). Each of the curves shown represents the fluorescence intensity time trace of a single cell in the observation area. Fig. 4B, upper left panel, shows that addition of anti-IgM to WT DT40 caused oscillations of [Ca 2ϩ ] i in most but not all cells. In responding cells, the oscillations commenced after a lag time of Նϳ2 min. Although some cells displayed a less oscillatory and more monophasic increase in [Ca 2ϩ ] i upon addition of extracellular Ca 2ϩ accompanied by a loss of spiking, most responding cells showed a relatively homogeneous oscillatory behavior in the absence and presence of extracellular Ca 2ϩ . In PLC␥ 2 Ϫ/Ϫ cells, by contrast, none of the cells responded to addition of anti-IgM (400 ng/ml), regardless of whether Ca 2ϩ was absent from or present in the incubation medium (Fig. 4B,  upper right panel). Addition of 4 ng/ml trypsin to these cells in the presence of extracellular Ca 2ϩ caused the appearance of a single [Ca 2ϩ ] i peak, presumably by activation of G-proteincoupled PAR2 receptors endogenously present in DT40 cells, followed by activation of endogenous PLC␤ (41). In this case, almost all cells responded, and the lag time was much shorter (ϳ10 s). Fig. 4B, lower panels, shows that although both PLC␥ 2 and PLC␥ 2 F897Q were able to reconstitute oscillatory Ca 2ϩ responses to PLC␥ 2 Ϫ/Ϫ DT40 cells following BCR ligation, substantial quantitative differences were apparent already at the level of visual assessment of the two sets of time traces. Specifically, the proportion of cells responding to anti-IgM as well as both the peak amplitudes and the frequencies of the spikes were considerably lower, and the mean latencies of the overall responses were distinctly longer for cells expressing PLC␥ 2 F897Q rather than WT PLC␥ 2 . A more detailed, quantitative characterization of this difference will be presented below in Figs. 5 and 6. A quantitative comparison of the [Ca 2ϩ ] i transients observed in PLC␥ 2 ϩ/ϩ DT40 cells versus PLC␥ 2 Ϫ/Ϫ cells reconstituted with WT PLC␥ 2 in response to 400 ng/ml anti-IgM is shown in Fig. 4C (upper panels). There were no significant differences between the two cell types in the proportions of responding cells, integrated peak intensities, peak frequencies, and amplitudes (Fig. 4C, lower panels). Thus, the PLC␥ 2 -deficient cells reconstituted with WT PLC␥ 2 bear a close resemblance to their native counterparts. Fig. 4D shows that PLC␥ 2 Ϫ/Ϫ cells reconstituted with WT versus F897Q mutant PLC␥ 2 do not appreciably differ in [Ca 2ϩ ] i responses triggered by addition of thapsigargin, an inhibitor of the sarco-/endoplasmic reticulum Ca 2ϩ -ATPase (40,41). Thus, the F897Q mutation and, by extension, interaction of PLC␥ 2 with activated Rac have no effect on BCR-independent [Ca 2ϩ ] i mobilization in DT40 B cells.

Concentration Dependence on Anti-IgM of [Ca 2ϩ ] i Response Impairment in DT40 B Cells Expressing Wild-type Versus Racinsensitive F897Q Mutant PLC␥ 2 -To determine the influence of the Rac-PLC␥ 2 interaction on [Ca 2ϩ ] i changes in DT40 B cells in response to increasing extents of BCR ligation, PLC␥ 2
Ϫ/Ϫ DT40 cells stably reconstituted with either WT or F897Q mutant PLC␥ 2 were treated with increasing concentrations of anti-IgM in the absence of extracellular Ca 2ϩ , followed by the same concentration of anti-IgM in the presence of 1 mM Ϫ/Ϫ ), and from three independent clones PLC␥ 2 Ϫ/Ϫ DT40 cells stably expressing similar quantities of either wild-type (A-C) or F897Q mutant PLC␥ 2 (a-c) were subjected to SDS-PAGE and immunoblotting (upper panel). The same membrane was subsequently probed with an anti-␤-actin antibody to control for equal loading of samples (lower panels). All six clones were used for experimentation in this study, with no differences detected between clones A-C and a-c, respectively. B, cells from the clones A and c were analyzed by indirect fluorescence staining (left panels). Right panels, corresponding phase contrast images. C, mean fluorescence intensities of the three images each, as shown in B, were corrected for background staining. Similar results were obtained for other pairs of clones. extracellular Ca 2ϩ . Cells were treated with 4 M ionomycin at the end of each experiment for baseline correction and normalization of the [Ca 2ϩ ] i responses in single cells. Fig. 5, left panels, shows that oscillatory [Ca 2ϩ ] i responses developed in cells reconstituted with WT PLC␥ 2 at anti-IgM concentrations between 4 and 400 ng/ml. At 40 -400 ng/ml anti-IgM, the responses were similar in both their frequencies and intensities, in the absence or presence of extracellular Ca 2ϩ . At 4 g/ml anti-IgM (Fig. 5, next-to-lowest left panel), individual Ca 2ϩ oscillations appeared to coalesce into single major asymmetric peaks in many traces. These peaks reached their maximum intensities within seconds after anti-IgM addition in the absence of extracellular Ca 2ϩ to gradually decline to base levels over the next 6 min. Addition of extracellular Ca 2ϩ to anti-IgM at this time point led to a second, more symmetric [Ca 2ϩ ] i wave, reaching a maximum after about 1.5 min and declining with similar kinetics thereafter. At still higher concentrations of anti-IgM, 40 g/ml, the pattern was qualitatively similar but somewhat reduced in quantitative terms, both in the absence and presence of extracellular Ca 2ϩ . In cells expressing the Racinsensitive mutant F897Q of PLC␥ 2 (Fig. 5, right panels), there was a striking loss of the [Ca 2ϩ ] i responses, in particular at low concentrations of anti-IgM, e.g. 40 ng/ml. In each case, this effect was clearly evident both with and without extracellular Ca 2ϩ . At high (4 and 40 g/ml) concentrations of anti-IgM, there also was a loss of coalescence of individual oscillations in responding cells and a decrease in the overall duration of the [Ca 2ϩ ] i response to BCR ligation, in particular in the absence of extracellular Ca 2ϩ .
The results of further quantitative analyses of the differences between the [Ca 2ϩ ] i responses of cells expressing WT versus F897Q PLC␥ 2 are shown in Fig. 6. Thus, in the absence of extra- Ϫ/Ϫ DT40 cells restores the quantitative features of B cell-receptor-mediated increases in cytosolic Ca 2ϩ to the phenotype observed in wild-type DT40 cells. Top panels, unmodified DT40 cells (PLC␥ 2 ϩ/ϩ ) or PLC␥ 2 Ϫ/Ϫ DT40 cells stably expressing wild-type PLC␥ 2 (PLC␥ 2 Ϫ/Ϫ ϩ WT) were sequentially treated as in Fig. 4B, panel a, except that anti-IgM was used at 400 ng/ml. Bottom panels, the intensity traces obtained for both groups of cells were quantitatively analyzed as indicated for the percentage of responding cells, integrated peak intensity, peak frequency, and peak amplitude. A total number of 213 and 196 cells, respectively, was analyzed in three independent experiments (n ϭ 3). D, PLC␥ 2 Ϫ/Ϫ DT40 cells stably expressing wild-type or F897Q mutant PLC␥ 2 show similar Ca 2ϩ responses to thapsigargin. PLC␥ 2 Ϫ/Ϫ cells stably expressing wild-type (left panel) or F897Q mutant PLC␥ 2 (right panel) (78 cells each) were treated at the times indicated by the arrowheads in the following sequence: 100 nM thapsigargin; no Ca 2ϩ 3 100 nM thapsigargin; 1 mM Ca 2ϩ 3 4 M ionomycin; 1 mM Ca 2ϩ . cellular Ca 2ϩ , there was a marked reduction in the anti-IgM sensitivity of the latter cells by an order of magnitude, but no change in the maximal extent, when the proportion of responding cells was used as a response parameter (Fig. 6A, left panel).
In the presence of extracellular Ca 2ϩ , the major difference was a distinct loss of the proportion of responding cells, which was similar (ϳ55%) at low and intermediate anti-IgM concentrations, 4 -400 ng/ml, and even more striking (ϳ80%) at high anti-IgM concentrations (Fig. 6A, right panel). There was no apparent change in the sensitivity of the cells to anti-IgM. Integration of the fluorescence intensities in single cells over time and evaluation of their concentration dependence on anti-IgM at the level of the means showed a reduction of the maximal mean intensity in cells expressing the F897Q mutant by 64% in the absence of extracellular Ca 2ϩ (Fig. 6B). In its presence, this loss amounted to 82% (Fig. 6C). Fig. 6D shows that resistance of PLC␥ 2 to regulation by Rac also resulted in a longer latency of the [Ca 2ϩ ] i response to addition of anti-IgM. This became evident at low anti-IgM concen-trations (40 ng/ml) and more prominent at all higher concentrations. [Ca 2ϩ ] i peak frequencies and amplitudes were only analyzed for the three lowest ligand concentrations, 4 -400 ng/ml, and in the absence of extracellular Ca 2ϩ , mostly because of peak coalescence at higher anti-IgM concentrations. By assigning a peak frequency of zero to nonresponding cells, we determined a reduction of this parameter by ϳ86 and ϳ55% in cells expressing the F897Q mutant rather than WT PLC␥ 2 (Fig.  6E, left panel). Likewise, decreases of peak amplitudes amounting to ϳ84 and ϳ58% were observed for the cells harboring the mutant at 40 and 400 ng/ml anti-IgM, respectively (Fig. 6E,  right panel). Collectively, these results indicate that the regulation of PLC␥ 2 by Rac is not an absolute requirement for BCRmediated Ca 2ϩ release at most BCR ligand concentrations tested herein, both in the absence and presence of extracellular Ca 2ϩ . However, although the qualitative patterns of [Ca 2ϩ ] i responses to BCR ligation are similar in the absence and presence of PLC␥ 2 regulation by Rac, there are striking quantitative differences. These are readily evident at the level of the proportion of responding cells and their sensitivity to anti-IgM, the latency of the response after addition of anti-IgM, as well as both the peak frequency and amplitude of the oscillatory [Ca 2ϩ ] i responses.

Dependence of BCR-mediated Nuclear Translocation of the Transcription Factor NFAT1c on Functional Rac-PLC␥ 2
Interaction-Within minutes after BCR-mediated B cell activation, several transcription factors, including the Ca 2ϩ -dependent transcription factor NFAT1c, are translocated into the nucleus to induce transcription of regulatory genes involved in B cell-fate decisions (6). To determine the role of a functional Rac-PLC␥ 2 interaction in this response, a C-terminally truncated NFAT1c construct consisting of the N-terminal transactivation and the calcineurin-binding domain (amino acids 1-400) fused to the red fluorescent protein td-RFP611 was introduced into PLC␥ 2 Ϫ/Ϫ DT40 B cells and PLC␥ 2 Ϫ/Ϫ DT40 B cells stably expressing either WT or F897Q mutant PLC␥ 2 , and its nuclear translocation following BCR ligation was analyzed by spinning disk fluorescence microscopy (Fig. 7). Treatment of cells with ionomycin, resulting in PLC␥ 2 -independent nuclear translocation of NFAT1c-td-RFP611 in almost all transfected cells, was used as a positive control (Fig. 7, A, lower panels, and  B, right panel). In PLC␥ 2 -deficient cells, there was little, if any, nuclear translocation of the fluorescent reporter protein in response to anti-IgM (Fig. 7A, upper panels), consistent with the concept that this event is absolutely dependent on PLC␥ 2induced Ca 2ϩ mobilization. In cells expressing WT PLC␥ 2 , NFAT1c-td-RFP611 translocated into the nuclei in about 80% of the cells (Fig. 7B, left panel). Importantly, only 16% of the cells expressing the Rac-insensitive PLC␥ 2 mutant F897Q displayed nuclear NFAT1c-td-RFP611, such that uncoupling of PLC␥ 2 from Rac caused an 80% decrease of a cellular response directly related to altered gene transcription and, by extension, cell fate decisions in B cells.

Discussion
This study shows that specific interference with a functional Rac-PLC␥ 2 interaction in intact DT40 B cells precipitates several clear alterations in BCR-mediated Ca 2ϩ signaling. First, the

DT40 cells. PLC␥ 2
Ϫ/Ϫ cells stably expressing wild-type (left column) or F897Q mutant PLC␥ 2 protein (right column) were treated with increasing concentrations of anti-IgM. The treatment was performed as indicated by the arrowheads in the following sequence: anti-IgM; no Ca 2ϩ 3 anti-IgM; 1 mM Ca 2ϩ 3 4 M ionomycin; 1 mM Ca 2ϩ . Forty nine to 60 cells were analyzed in each single experiment.
sensitivity of Ca 2ϩ release from internal stores to the BCR ligand anti-IgM is reduced by an order of magnitude. Second, the extent of the Ca 2ϩ signal in the presence of extracellular Ca 2ϩ , also involving entry of extracellular Ca 2ϩ into the cells, is greatly diminished. Third, Ca 2ϩ -regulated, NFAT-mediated transcriptional regulation is largely decreased.
Several lines of evidence have already suggested that PLC␥ 2 and Rac are both activated by BCR ligation, although a direct interaction of the two signaling molecules has received relatively little attention (6,26,42,43). Thus, several elements of the canonical B cell receptor signaling cascade, e.g. Syk (44), Btk (45), and BLNK (46,47), are known to physically interact with and activate the Rac activator Vav. BCR cross-linking caused activation of both Rac1 and Rac2 within minutes (48). Rac cooperated with PLC␥ 2 in BCR-mediated activation of c-Jun N-terminal kinase, p38 mitogen-activated protein kinase, SRF, and NFAT (49,50). Dominant negative Rac1 suppressed these cooperative relationships. These results were interpreted to suggest that the signaling pathways controlling the activities of Rac and PLC␥ 2 are activated by BCR ligation in parallel to converge at points distal to enhanced formation of InsP 3 and DAG. Btk, BLNK, Vav, and PLC␥ 2 form highly coordinated microsignalosomes in a process dependent on Syk and Lyn (51). Formation of these complexes is important for amplification of signaling and, hence, for appropriate B cell activation, in particular at low antigen concentrations. The results presented here strongly suggest that signal amplification is caused within or in the vicinity of these sites by convergence of signals emanating from activated BCR through a direct interaction of activated Rac with PLC␥ 2 .
Calcium oscillations arise from a complex interplay between PLC-mediated generation of InsP 3 and DAG, InsP 3 receptormediated release of intracellular Ca 2ϩ , and entry of extracellular Ca 2ϩ mediated by channels that are either store-operated, activated by DAG, or subject to other regulatory controls (52). These incremental changes may be rapidly overturned by InsP 3 metabolism and Ca 2ϩ efflux from the cytosol (6, 53). All three InsP 3 receptors are present in DT40 B cells and display differential sensitivities to InsP 3 in these cells (54). Expression of only one of the three in InsP 3 receptor-deficient cells resulted in very distinct oscillatory responses to similar degrees of BCR activation. This finding suggests that InsP 3 receptors are a major regulatory site of Ca 2ϩ oscillations in these cells. In addition, feedback mechanisms may exist in B cells giving rise to spatiotemporal dynamics of the levels of InsP 3 (55) and, by extension, of DAG. For example, the transient receptor potential cation channel type 3-mediated increases in [Ca 2ϩ ] i were shown in B cells to promote translocation of PLC␥ 2 to the plasma membrane and further activation of the enzyme, by a process involving physical interaction with TRPC3 cation channels (56). Interestingly, some aspects of the latter interaction appeared to be independent of the phospholipase C activity of PLC␥ 2 (40,56). Selective and direct inhibition of TRPC3 channels by pyrazole-3 eliminated the Ca 2ϩ influx-dependent PLC␥ 2 plasma membrane translocation and the late oscillatory phase of the BCR-induced Ca 2ϩ response (57). The same TRPC3 inhibitor also suppressed Rac1 activation without affecting total Rac1 protein abundance in cardiomyocytes (58), whereas activated Rac1 enhanced the rapid vesicular translocation and membrane insertion of another transient receptor potential cation channel, TRPC5 (59). Thus, the interaction of PLC␥ 2 with Rac may influence BCR-mediated Ca 2ϩ oscillations in several ways.
Previous studies have shown that the interaction of PLC␥ 2 with Rac2 G12V markedly enhances its enzymatic activity in vitro and in intact cells and coincides with a translocation of PLC␥ 2 from the soluble to the particulate fraction of intact cells (23). Here, we demonstrate that binding of activated Rac to WT PLC␥ 2 also changes the mode of plasma membrane association of the latter from a mixture of exchange and lateral diffusion to almost pure lateral diffusion (Fig. 2). This suggests that Rac-PLC␥ 2 interaction reduces the exchange rate between membrane-associated and cytoplasmic WT PLC␥ 2 such that it becomes negligible relative to its lateral diffusion rate, demonstrating a stronger membrane association. Hence, in intact cells, Rac-activated WT PLC␥ 2 mainly travels along the plasma membrane, allowing for a spatio-temporal pattern of encounter with both its lipid substrate and its protein interaction partners, e.g. TRPC3, that is different from that of its unliganded or Racresistant counterparts.
The dependence of the Ca 2ϩ oscillations in cells expressing WT PLC␥ 2 on low to intermediate concentrations of anti-IgM (Յ4 g/ml) suggests that signal transduction occurs mostly through frequency modulation, although there is also a clear increase in the amplitude with increasing BCR ligation. It is also evident that, in this concentration range of anti-IgM, the spiking behavior of the cells is largely independent of the presence of extracellular Ca 2ϩ . The relative independence of oscillation frequency on external Ca 2ϩ suggests that DT40 B cells are highly efficient at recycling their internal Ca 2ϩ (53). In aggregate, these findings indicate that a minimal model based on reversible desensitization of InsP 3 receptors may suffice to explain the oscillatory pattern observed at low to intermediate anti-IgM concentrations (53,60). According to the minimal model, [Ca 2ϩ ] i oscillations are driven by a coupled process of Ca 2ϩ -induced activation and obligatory intrinsic inactivation of the Ca 2ϩ -sensitized state of InsP 3 receptors. Ca 2ϩ spikes are initiated by the Ca 2ϩ -mediated conversion of low affinity InsP 3 receptors from their low to the high affinity type (60). In this paradigm, the decrease in latency and the increase in oscillation frequency observed in Fig. 6, D and E, would be due to the decrease in the time required to generate a sufficient Ca 2ϩ trigger signal to initiate the first or next Ca 2ϩ release spike, respectively. Enhanced and accelerated translocation of PLC␥ 2 to the plasma membrane containing its phospholipid substrate by activated Rac would allow for a decrease in the time necessary for reaching this threshold level, such that considerably lower levels of activated BCR are required to elicit a given [Ca 2ϩ ] i response. Of note, enhanced association of PLC␥ 2 with the plasma membrane following interaction with activated Rac is indeed observed in the FRAP beam size analysis experiments (Fig. 2), and the absence of this effect on the PLC␥ 2 mutant defective in Rac interactions (F897Q) is in accord with the much lower quantitative effect of anti-IgM in cells expressing this mutant on the Ca 2ϩ oscillations (Fig. 6E). In addition to temporal changes of PLC␥ 2 activation by Rac, spatial changes may come into play. Given that the Ca 2ϩ conductances are similar for the three InsP 3 -R subtypes (61), the increase in amplitude observed with increasing anti-IgM concentrations may be due to expression of the receptors at different cellular levels in DT40 B cells, with higher abundance of subtypes with lower InsP 3 sensitivity (InsP 3 -R1 (4.7 M); InsP 3 -R2 (0.35 M); InsP 3 -R3 (18.6 M); EC 50 values for InsP 3 -mediated Ca 2ϩ release in permeabilized cells expressing a single InsP 3 -R subtype in parentheses (54)).
When extracellular Ca 2ϩ is provided to cells incubated without Ca 2ϩ for several minutes at high concentrations of anti-IgM, the main effect of a loss of Rac regulation by PLC␥ 2 is a marked reduction in the number of responding cells, which cannot be recuperated by increasing the concentration of anti-IgM. Several scenarios may explain this observation: (i) the interaction of PLC␥ 2 with activated Rac could strengthen the productive interaction of PLC␥ 2 with plasma membrane channels such as TRPC3 allowing entry of extracellular Ca 2ϩ , either by altered temporal or spatial interaction of PLC␥ 2 with the membrane (cf. Fig. 2). If TRPC3 channels were involved in Rac activation in B cells, as they appear to be in cardiomyocytes (see above), this would further reinforce this process. (ii) Cells could be desensitized to BCR-mediated Ca 2ϩ signaling with continued anti-IgM exposure. They can become less sensitive to this course of action if the Rac-PLC␥ 2 interaction is intact. Both homologous and heterologous desensitization of B cell membrane-immunoglobulin-mediated Ca 2ϩ mobilization have long been known to occur within minutes of anti-immunoglobulin exposure (62). Interestingly, anti-IgM-treated B cells are hyperresponsive to AlF 4 Ϫ and mastoparan (63). Although the two reagents are known as activators of heterotrimeric G proteins, they also appear to activate Rho GTPases, including Rac (64,65). If this was to occur in B cells, the results presented here and by Cambier et al. (63) may indicate that anti-IgM-mediated B cell activation affects a function(s) of the macromolecular complex involving PLC␥ 2 , without impinging upon the activation of the enzyme by Rac, and that the complex is protected, at least in part, when the Rac-PLC␥ 2 interaction is intact. The markedly diminished effect of high concentrations of anti-IgM on [Ca 2ϩ ] i observed in cells lacking the functional Rac-PLC␥ 2 interaction does not appear to require exposure of B cells to anti-IgM in the absence of extracellular Ca 2ϩ , because the inhibitory effect was also evident at the level of nuclear NFAT1c-td-RFP611 translocation, which was assayed in the continued presence of extracellular Ca 2ϩ (Fig. 7).
Under these conditions, the insensitivity of PLC␥ 2 to activated Rac led to a marked reduction in BCR-mediated nuclear translocation of NFAT1c-td-RFP611 (Fig. 7), strongly suggesting that the changes observed at the level of [Ca 2ϩ ] i are effectively transduced to the transcriptional level. In B lymphocytes, the activity of the Ca 2ϩ -regulated transcriptional regulators c-Jun N-terminal kinase (JNK), NFB, and NFAT are differentially regulated by the amplitude and duration of oscillatory Ca 2ϩ signals (66). NF-B and JNK are selectively activated by a large transient [Ca 2ϩ ] i rise, whereas NFAT is activated by a low, sustained Ca 2ϩ plateau. These results and findings obtained in T lymphocytes (67) and basophilic leukemia cells (68) are consistent with the notion that nuclear translocation of NFAT functions as a working memory of Ca 2ϩ signals by decoding Ca 2ϩ oscillations (69). In general, this process appears to be more cost-effective than translation of continuous Ca 2ϩ signals. In this study, nuclear translocation and, hence, activation of NFAT1c-td-RFP611 was studied at 40 g/ml anti-IgM, allowing for similar numbers of responders (ϳ92%) when cells expressing either WT or F897Q mutant PLC␥ 2 were analyzed for BCR-mediated Ca 2ϩ release from internal stores (Fig. 6A), with an overall reduction of the integrated fluorescence intensity in the latter cells by 64% (Fig. 6B). In the presence of extracellular Ca 2ϩ , both the proportion of responding cells and the mean integrated single cell fluorescence intensity were reduced by more than 80% (Fig. 6, A and C). Because NFAT activation is highly dependent on both InsP 3 -induced intracellular Ca 2ϩ release and store-operated Ca 2ϩ entry mediated by CRAC channels (50,70), it is likely that reductions of both responses participate in reducing nuclear NFAT translocation. Furthermore, analysis of the Ca 2ϩ oscillatory behavior in the absence of extracellular Ca 2ϩ at 40 g/ml anti-IgM revealed that the majority of responses (Ͼ75%) consisted of more than one Ca 2ϩ spike in cells expressing WT PLC␥ 2 , whereas this parameter was reduced to 35% in cells expressing the F897Q mutant, showing only a single Ca 2ϩ spike in most cases (cf. Fig. 5, lower  panels). Thus, reduced [Ca 2ϩ ] i spiking from intracellular sources may also contribute to the substantial decrease in nuclear NFAT translocation caused by resistance of PLC␥ 2 to activation by Rac.
Because Rac GTPases are activated in B cells by quite a number of cell surface receptors in addition to the BCR, the signaling pathway convergence shown in this work is very well suited to detect coincident extracellular signals and to ensure signaling reliability by correctly interpreting the emergence of signals according to the particular cellular context and the specific (patho)physiological circumstances (71). Thus, agonist activation of receptors for integrins, chemokines, and pathogen-derived ligands has also been shown to activate Rac, which provides the GTPase with the potential to act as a central B cell signaling hub. For integrins, it is interesting to note that coliga-tion of the ␤ 2 ␣ L integrin LFA-1 and BCR has been shown to cause an ϳ10-fold reduction in the amount of BCR ligand required for B cells to make a tight contact with target membranes containing the receptor ligands (72). Using a similar experimental design, Henderson et al. (73) showed that Rac2 plays an important role in outside-in signaling from LFA-1 that leads to firm adhesion of murine B cells to ICAM-1. Interestingly, the Rac2 deficiency could be bypassed by treatment of B cells with phorbol ester and ionomycin, thus mimicking enhanced activation of PLC␥ 2 . Although the molecular mechanisms of Rac activation by LFA-1 in B cells are unknown, they may be similar to those encountered in T cells (74 -76), although a non-Vav RacGEF(s) may be involved in this case (73). Hence, simultaneous activation of both BCR and LFA-1, e.g. at the immunological synapse, may give rise to locally enhanced Rac activation, allowing for an increased potency of BCR ligand to mediate integrin activation, at least in part by enhanced activity of PLC␥ 2 (73,77).
A positive interaction between Rac and PLC␥ 2 may also be (patho)physiologically relevant in platelets, where PLC␥ 2 is activated downstream of other immunoreceptor tyrosinebased activation motif-coupled receptors, such as the major platelet collagen receptor glycoprotein VI or CLEC-2 (78), and by integrin ␣ 2 ␤ 1 (79). Inactivation of the Rac1 gene in the mouse caused defective thrombus formation on collagen under flow conditions (78), and pharmacological inhibition of Rac in human platelets led to reduced PLC␥ 2 activity and impaired ␣ IIb ␤ 3 fibrinogen receptor stimulation (79). Intriguingly, the functional consequences of Rac1 deficiency were particularly striking at low and intermediate concentrations of GPVI or CLEC-2 receptor agonists (78), suggesting that Rac may enhance the sensitivity of PLC␥ 2 stimulation to extracellular ligands in platelets as well. Hence, pharmacological targeting of Rac1 could be an interesting approach in the development of future antiplatelet drugs.
The regulation of B cells by BCR ligation is intertwined with their regulation by certain chemokines, such as the CXCR4 and CXCR5 agonists CXCL12 and CXCL13, respectively (5). Activated CXCR4 and CXCR5 appear to be coupled to both Btk and PLC␥ 2 and, through a yet unidentified RacGEF (possibly DOCK2), to activation of Rac. Recently, Rac2 has been shown to be critical for LFA-1-mediated adhesion of mouse B cells in response to CXCL12 or CXCL13 (73). It remained unclear whether Rac2 causes enhanced adhesion directly by augmenting inside-out activation of integrins or indirectly by fostering receptor-mediated activation of PLC␥ 2 . Recently, pathogenderived signals have been suggested to activate Rac in murine B cells via Toll-like receptor TLR4 and MyD88 (80). Activated TLR4 activation also enhanced CXCR5-mediated Rac activation. Of note, TLR4 was found in other leukocytes to be coupled to activation of PLC␥ 2 (81,82), raising the possibility that Rac is of regulatory importance in that respect as well.
The enhanced BCR ligand sensitivity observed for cells expressing WT PLC␥ 2 versus its F897Q mutant is reminiscent of the reduced threshold for antigen receptor stimulation observed in human peripheral blood B lymphocytes upon coligation of CD19 with BCR (3). In mouse B cells, the costimulatory effect of CD19 on BCR-mediated increases in [Ca 2ϩ ] i was clearly reduced in B cells from rac2 Ϫ/Ϫ mice (19). Previously, these effects were mainly ascribed to the known interactions of CD19 with the RhoGEF Vav and to the positive interaction of Vav, presumably via activated Rac, with phosphatidylinositol 4-phosphate 5-kinase and/or PI3K, followed by indirect activation of PLC␥ 2 (6,42). Our current results strongly suggest that Rac enhances BCR-mediated Ca 2ϩ signaling in B cells by direct interaction with and activation of PLC␥ 2 .
Several families have been described with members affected by homozygous mutations in the CD19 gene causing undetectable or substantially decreased levels of CD19 in B cells. In these patients, suffering from an antibody-deficiency syndrome, there were marked alterations in Ca 2ϩ mobilization in B cells following anti-IgM treatment (83). Similar defects of Ca 2ϩ mobilization were observed in patients with defective CD81 and CD21, which function in a complex with CD225 and CD19 and cooperate with BCR to mediate antigen recognition (84,85). Thus, the stimulatory interaction of PLC␥ 2 with Rac may not only play a pivotal role in determining the sensitivity of the BCR to stimulation by antigen, but may also contribute to BCR coreceptor signaling and to functional alterations of the latter in human disease such as certain forms of monogenetic common variable immunodeficiency (86). The very recent observation of a RAC2 loss-of-function mutation in two siblings with characteristics of a common variable immunodeficiency is consistent with this view (87).