Role of Phospholipase C (cid:98) 3 Phosphorylation in the Desensitization of Cellular Responses to Platelet-activating Factor*

Platelet-activating factor (PAF) stimulates a diverse array of cellular responses through receptors coupled to G proteins that activate phospholipase C (PLC). Trun-cation of the cytoplasmic tail of the receptor to remove phosphorylation sites (mutant PAF receptor, mPAFR) results in enhancement of PAF-stimulated responses. Here we demonstrate that PAF or phorbol 12-myristate 13-acetate (PMA) pretreatment inhibited wild type PAFR-induced PLC-mediated responses by (cid:59) 90%, whereas these responses to the phosphorylation-defi-cient mPAFR were inhibited by (cid:59) 50%, despite normal G protein coupling, suggesting a distal inhibitory locus. PAF and PMA, as well as a membrane permeable cyclic AMP analog, stimulated phosphorylation of PLC (cid:98) 3. A protein kinase C (PKC) inhibitor blocked phosphorylation of PLC (cid:98) 3 stimulated by PAF and PMA but not by cAMP. Activation of protein kinase A (PKA) by cAMP did not result in inhibition of Ca 2 (cid:49) mobilization stimulated by PAF. In contrast, cAMP did inhibit the response to formylpeptide chemoattractant receptor. These data suggest that homologous desensitization of PAF-medi-ated responses is regulated via phosphorylation at two levels in the signaling pathway, one at the receptor and the other at PLC (cid:98) 3 mediated by PKC but not by PKA. Phosphorylation Western Blotting and Immunoprecipitation of PLC (cid:98) 3— For Western blotting, lysate (150 (cid:109) g protein) from RBL-2H3 cells, and for compari-son, lysates from bovine brain, COS cells, and HL-60 cells were resolved on a 6% SDS-PAGE gel. The proteins were transferred to nitrocellulose membrane, blocked with 3% milk, and incubated with 1.5 (cid:109) g/ml of different anti-PLC (cid:98) antibodies. The immunoreactive proteins were visualized using ECL Western blotting detection system. Phosphoryla- tion of PLC (cid:98) 3 was performed essentially as described for the epitope-tagged chemoattractant receptors (19). Briefly, RBL-2H3 cells (3 (cid:51) 10 6 ) were subcultured overnight in 60-mm tissue culture dishes. The following day, cells were washed twice with 5 ml of phosphate-free Dulbecco’s modified Eagle’s medium and incubated in the same medium plus [ 32 P]orthophosphate (150 (cid:109) Ci/dish) for 90 min. The labeled cells were stimulated, and PLC (cid:98) 3 was immunoprecipitated from cell lysate with anti-PLC (cid:98) 3 antibody. The proteins were resolved on a 6% SDS-PAGE gel and visualized by autoradiography.


plays important roles in inflammation
and has physiological effects on cardiovascular, reproductive, and central nervous systems (1). PAF mediates its effects via the activation of a seven transmembrane domain G proteincoupled receptor (2,3). One of the consequences of PAF receptor activation is the phospholipase C (PLC)-mediated hydrolysis of phosphatidylinositol 4,5-bisphosphate leading to the activation of protein kinase C (PKC) and mobilization of intracellular Ca 2ϩ (4). Molecular cloning has revealed three classes of PLC: PLC␤, PLC␥, and PLC␦, and each of these occur in several isoforms (5,6). All four PLC ␤ isoforms are activated to varying extent by ␣ subunit of a class of G proteins known as G q (7)(8)(9)(10). PLC␤2 and PLC␤3 are also activated by the ␤␥ subunit of the G i family of G proteins (11)(12)(13). Tyrosine phosphorylation of PLC␥ is required for its activation, but the mechanism by which PLC␦ is activated is unknown (5). PAFR couples to both pertussis toxin (Ptx)-sensitive and Ptx-insensitive G proteins in a variety of cells (4,14,15). In B cells and platelets, PAF activates PLC␥ (16,17), but the identity of PLCs activated by PAF in phagocytic leukocytes is unknown.
In phagocytes, PAF-stimulated G protein activation and Ca 2ϩ mobilization are desensitized by prior treatment of cells with PAF and other chemoattractants such as the formylated peptides, C5a and interleukin-8 (18). This laboratory has developed stably transfected RBL-2H3 cells to study molecular mechanisms of leukocyte chemoattractant receptor regulation (15, 19 -21). Using wild type and phosphorylation-deficient PAFR expressed in this cell line, we now describe the novel finding that homologous desensitization of PAF-mediated phosphoinositide hydrolysis and Ca 2ϩ mobilization is mediated via two processes; one at the level of receptor phosphorylation and the other at the level of PLC activation apparently by PKC-dependent phosphorylation of PLC␤3.

EXPERIMENTAL PROCEDURES
Materials-Bisindolylmaleimide or GF 109203X (BIM), 8-(4-chlorophenylthio)-adenosine 3Ј-cyclic monophosphate (Cpt-cAMP) and the tyrosine kinase inhibitor PP1 were purchased from Calbiochem. Affinity purified polyclonal antibodies against PLC␤1, PLC␤2, and PLC␤3 were obtained from Santa Cruz Biotechnology. ECL Western blot analysis kit was purchased from Amersham Corp. Mouse monoclonal IgE and antigen (dinitrophenylated BSA) were generously provided by Drs. Juan Rivera and Henry Metzger (National Institutes of Health). All other materials were obtained from sources previously described (15).
Cell Culture, Phosphoinositide Hydrolysis, and Ca 2ϩ Mobilization-RBL-2H3 cells stably expressing ϳ3 ϫ 10 4 receptors for PAFR and mPAFR were used (21). The cells were maintained in Dulbecco's modified Eagle's medium supplemented with 15% fetal bovine serum, glutamine (2 mM), penicillin (100 units/ml), and streptomycin (100 mg/ml) (19). For phosphoinositide hydrolysis, cells (5 ϫ 10 4 cells/well) were subcultured overnight in 96-well tissue culture plate with 1 Ci/ml of [ 3 H]inositol in an inositol-free medium supplemented with 10% dialyzed fetal bovine serum. Cells were washed with HEPES-buffered saline containing 10 mM LiCl and 0.1% BSA and stimulated, and the generation of total [ 3 H]inositol phosphates was determined. For Ca 2ϩ mobilization, cells (3 ϫ 10 6 ) were loaded with 1 M indo-1/AM in the presence of 1 M pluoronic acid for 30 min at room temperature. Cells were washed and resuspended in 1.5 ml HEPES-buffered saline, and intracellular Ca 2ϩ mobilization was determined as described previously (19). Western Blotting and Immunoprecipitation of PLC␤3-For Western blotting, lysate (150 g protein) from RBL-2H3 cells, and for comparison, lysates from bovine brain, COS cells, and HL-60 cells were resolved on a 6% SDS-PAGE gel. The proteins were transferred to nitrocellulose membrane, blocked with 3% milk, and incubated with 1.5 g/ml of different anti-PLC␤ antibodies. The immunoreactive proteins were visualized using ECL Western blotting detection system. Phosphorylation of PLC␤3 was performed essentially as described for the epitopetagged chemoattractant receptors (19). Briefly, RBL-2H3 cells (3 ϫ 10 6 ) were subcultured overnight in 60-mm tissue culture dishes. The following day, cells were washed twice with 5 ml of phosphate-free Dulbecco's modified Eagle's medium and incubated in the same medium plus [ 32 P]orthophosphate (150 Ci/dish) for 90 min. The labeled cells were stimulated, and PLC␤3 was immunoprecipitated from cell lysate with anti-PLC␤3 antibody. The proteins were resolved on a 6% SDS-PAGE gel and visualized by autoradiography.

RESULTS AND DISCUSSION
The mechanism of PAF receptor desensitization was studied using the rat basophilic leukemia (RBL-2H3) cells expressing wild type PAFR and a mPAFR in which the cytoplasmic tail was truncated to delete potential phosphorylation sites (21). In contrast to PAFR, mPAFR was resistant to both ligand and PKCinduced receptor phosphorylation and desensitization of G protein coupling, as measured by [ 35 S]GTP␥S binding to membranes (21,22). However, when indo-1-loaded cells expressing mPAFR were stimulated with PAF (10 nM), washed, and restimulated with PAF (10 nM), the peak Ca 2ϩ response was still homologously desensitized by ϳ50% (Table I). PMA also inhibited mPAFRmediated Ca 2ϩ mobilization by ϳ50%, suggesting a locus of inhibition distal to G protein activation. In cells expressing wild type PAFR, both PAF and PMA inhibited PAF-induced Ca 2ϩ mobilization by Ͼ90%. PMA also caused substantial inhibition of mPAF-induced generation of inositol phosphates and completely blocked the response to PAF (Fig. 1). To further test the role of PKC on PAFR desensitization, cells expressing mPAFR and PAFR were treated with PMA (100 nM) overnight to deplete PKC or incubated with the PKC-inhibitor bisindolylmaleimide for 10 min to block the activity of the enzyme. PAF-stimulated generation of inositol phosphates was then determined. Under both of these conditions, PAF-induced responses were enhanced 2-3-fold in cells expressing mPAFR (Fig. 1). Overnight PMA treatment did not enhance the response to PAF in cells expressing PAFR, but bisindolylmaleimide did potentiate the response to PAF. Because PMA-induced PKC depletion involves activation of the enzyme, it was possible that the differential effect of PMA on mPAFR and PAFR cells resulted from PKC-induced phosphorylation and down-regulation of PAFR but not mPAFR. Indeed, treatment of cells with PMA (100 nM, overnight) resulted in ϳ50% decrease of the wild type PAFR expression but had no effect on mPAFR as measured by the PAFR antagonist [ 3 H]WEB 2086 binding to intact cells (data not shown). These data suggest that homologous desensitization of PAF-mediated cellular responses is mediated by two processes. One involves receptor phosphorylation, presumably by PKC and a G protein-coupled receptor kinase, which uncouple the receptor from G proteins (21). In addition, a new mechanism was identified as a consequence of PKC-induced modification of a component distal to G protein activation but proximal to the generation of inositol phosphates.
Although most G protein-coupled receptors activate PLC␤, angiotensin receptors in smooth muscle cells and PAF receptors in B cells and platelets cause tyrosine phosphorylation and activation of PLC␥ (16,17,23). The molecular mechanism by which G protein-coupled receptors activate PLC␥ has yet to be determined. PMA, which causes serine phosphorylation of PLC␥ (24), leads to reduction of both tyrosine phosphorylation of PLC␥ and generation of inositol phosphates stimulated by cross-linking of cell surface IgE receptors with antigen (25,26). This suggests that if PAF activates PLC␥, a similar modification of PLC␥ by PKC could be involved in its homologous desensitization. However, the tyrosine kinase inhibitor PP1 (27), which blocked antigen (IgE)-stimulated Ca 2ϩ mobilization  6 ) expressing mPAFR or PAFR were loaded with indo-1 and stimulated with PAF (10 nM), and peak Ca 2ϩ mobilization was determined as described (19). For homologous desensitization, PAF-stimulated cells were washed, and 5 min later they were restimulated with PAF (10 nM) and peak Ca 2ϩ mobilization was determined. Cells were also preincubated with PMA (100 nM), and 5 min later they were stimulated with PAF (10 nM) and peak Ca 2ϩ mobilization was determined. The basal unstimulated intracellular Ca 2ϩ concentration was 150 Ϯ 10 and was subtracted from the values shown, which are the means Ϯ S.E. of three different experiments.

Phosphorylation of PLC␤3
11707 by Ͼ70%, presumably by inhibiting tyrosine phosphorylation of PLC␥, had no effect on PAF-mediated responses (Table II). Elevation of intracellular cAMP has also been shown to cause serine phosphorylation of PLC␥ and inhibition of antigen-stimulated responses (24). As shown in Table II, the membrane permeable cAMP analog 8-(4-chlorophenylthio)-adenosine 3Јcylic monophosphate (cpt-cAMP) substantially inhibited IgEmediated Ca 2ϩ mobilization but had no effect on the PAFmediated response. These data suggest that unlike B cells and platelets (16,17) PAF does not activate PLC␥ in RBL-2H3 cells and that PKA does not down-regulate PAF-mediated responses.
We therefore sought to determine whether a PLC␤ isoform is activated by PAF in RBL-2H3 cells. To identify the PLC␤ isoforms expressed in RBL-2H3 cells, their lysates and lysates from bovine brain, COS and HL-60 cells were separated by SDS-PAGE, and the presence of PLC␤1, PLC␤2, and PLC␤3 was determined by Western blotting using PLC␤ isoform-specific antibodies. All cells tested, including RBL-2H3 cells, expressed PLC␤3 in different amounts (Fig. 2). Although PLC␤1 and PLC␤2 were expressed at high levels in brain and HL-60 cells, respectively, neither could be detected in RBL-2H3 cells (Fig. 2).
Studies with purified G proteins and PLC revealed that PLC␤1 and PLC␤3 are equally responsive to activation by G␣ q (28). It was therefore possible that in RBL-2H3 cells, which do not express PLC␤1, PAF activates PLC␤3 and that PKC-induced modification of this enzyme accounts for the distal component of PAFR homologous desensitization. To test this hypothesis, cells were labeled with [ 32 P]orthophosphate and stimulated with PAF, and cell lysate was immunoprecipitated with PLC␤3-specific antibody. As shown in Fig. 3 (A and B), PAF caused phosphorylation of PLC␤3 in a dose-and time-dependent manner. PLC␤3 phosphorylation was detectable at physiologic PAF concentration of 0.3 nM and was maximum by 10 nM. PAF-induced PLC␤3 phosphorylation was rapid, reached a maximum by 15 s, and remained elevated for 10 min. To determine whether there was a correlation between PKCmediated PLC␤3 phosphorylation and inhibition of PAF-stimulated responses, cells were treated with or without the PKC inhibitor bisindolylmaleimide, stimulated with PAF (10 nM) and PMA (100 nM), and PLC␤3 phosphorylation was determined. As shown in Fig. 3C, PAF-and PMA-induced PLC␤3 phosphorylation was almost completely inhibited by bisindolylmaleimide. Cpt-cAMP also phosphorylated PLC␤3 presumably through activation of PKA, and this phosphorylation was not inhibited by bisindolylmaleimide. PLC␤3 phosphorylation was also tested in two other cell lines that endogenously express PAF receptors. These were the human monocyte-like U937 cells and the murine macrophage-like J774.1 cells. In both cases, PAF, PMA, and cpt-cAMP stimulated phosphorylation of PLC␤3 (Fig. 3D).
The effect of cAMP elevation on PLC activation was determined using the RBL-2H3 cells coexpressing PAF and formylpeptide receptors. In such cells, cpt-cAMP did not inhibit Ca 2ϩ mobilization to PAF, although it did inhibit the response of the formylpeptide receptor (Table II). Furthermore, in cells expressing mPAFR, cpt-cAMP (1 mM, 5 min) had no effect on PAF-stimulated generation of [ 3 H]inositol phosphates, whereas PMA (100 nM, 5 min) blocked this response by ϳ50%.   (21) were incubated with IgE (0.2 g/ml) overnight to saturate cell surface IgE receptors. The following day cells were loaded with indo-1 for 30 min in the absence (control) and the presence of the tyrosine kinase inhibitor, PP1 (10 M), and Ca 2ϩ mobilization in response to PAF (0.2 nM), fMLP (1.0 nM), and antigen, dinitrophenylated (DNP) BSA (1 ng/ml) was determined. Indo-1-loaded cells were also preincubated with the membrane permeable cAMP analog, cpt-cAMP (1 mM, 5 min), and its effect on peak Ca 2ϩ mobilization was tested. Values are the means Ϯ S.E. of four experiments. The numbers in parenthesis represent the percentage of inhibition of response as compared with control. Phosphorylation of PLC␤3 11708 Preincubation of cells with cpt-cAMP had no effect on PMAmediated inhibition of PAF-stimulated response (data not shown). These data support the hypothesis that PLC␤3 is likely activated by PAFR and that PLC␤3 phosphorylation by PKC, but not by PKA, is responsible for one component of PAFR homologous desensitization. PLC␤1 is phosphorylated by PKC (29), and PLC␤2 is phosphorylated by PKA (30). The present study revealed that both PKC and PKA phosphorylate of PLC␤3 in vivo (Fig. 3C). More importantly, phosphorylation of PLC␤3 by the different protein kinases appears to be associated with distinct functional consequences. Although phosphorylation of PLC␤3 by PKC resulted in the inhibition of responses stimulated by PAF, its phosphorylation by PKA was not associated with such inhibition. On the other hand, PKA activation by cpt-cAMP inhibited formylpeptide receptor-mediated responses. In RBL-2H3 cells PAF receptors activate PLC predominantly via a Ptx-insensitive G protein, whereas formylpeptide-mediated response is completely blocked by Ptx (15,19). Therefore, differential inhibition of responses to PAF and formylpeptide by PKC and PKA is likely due to the different G protein subunits these receptors utilize to activate PLC. It was demonstrated in vitro that G␣q and G␤␥ released from Ptx-sensitive G protein interact with distinct regions of PLC␤3 to activate the enzyme (28). This led Smrcka and Sternweis to propose that different receptors that utilize G q or G i could activate PLC␤3 by producing G␣q or G␤␥ (28). The data presented herein support this hypothesis and provide a molecular basis for the differential regulation of G␣qand G␤␥-mediated activation of PLC␤3. It is likely that PKC and PKA phosphorylate PLC␤3 on distinct sites that block its activation by G␣q and G␤␥, respectively. Similar to RBL-2H3 cells, many human cell lines and tissues express PLC␤3 but not PLC␤1 or PLC␤2 (9,11,28,31). Given the duality of PLC␤3 activation by both G␣ q and G␤␥ and its ubiquitous tissue distribution, phosphorylation of PLC␤3 by PKC and PKA may be general mechanisms by which functions of different G protein/PLC-coupled receptors are regulated.