Protein Kinase C-mediated Phosphorylation of the Calcium-sensing Receptor Is Stimulated by Receptor Activation and Attenuated by Calyculin-sensitive Phosphatase Activity*

The agonist sensitivity of the calcium-sensing receptor (CaR) can be altered by protein kinase C (PKC), with CaR residue Thr888 contributing significantly to this effect. To determine whether CaRT888 is a substrate for PKC and whether receptor activation modulates such phosphorylation, a phospho-specific antibody against this residue was raised (CaRpT888). In HEK-293 cells stably expressing CaR (CaR-HEK), but not in cells expressing the mutant receptor CaRT888A, phorbol ester (PMA) treatment increased CaRpT888 immunoreactivity as observed by immunoblotting and immunofluorescence. Raising extracellular Ca2+ concentration from 0.5 to 2.5 mm increased CaRT888 phosphorylation, an effect that was potentiated stereoselectively by the calcimimetic NPS R-467. These responses were mimicked by 5 mm extracellular Ca2+ and abolished by the calcilytic NPS-89636 and also by PKC inhibition or chronic PMA pretreatment. Whereas CaRT888A did exhibit increased apparent agonist sensitivity, by converting intracellular Ca2+ (Ca2+i) oscillations to sustained plateau responses in some cells, we still observed Ca2+i oscillations in a significant number of cells. This suggests that CaRT888 contributes significantly to CaR regulation but is not the exclusive determinant of CaR-induced Ca2+i oscillations. Finally, dephosphorylation of CaRT888 was blocked by the protein phosphatase 1/2A inhibitor calyculin, a treatment that also inhibited Ca2+i oscillations. In addition, calyculin/PMA cotreatment increased CaRT888 phosphorylation in bovine parathyroid cells. Therefore, CaRT888 is a substrate for receptor-induced, PKC-mediated feedback phosphorylation and can be dephosphorylated by a calyculin-sensitive phosphatase.

The extracellular calcium-sensing receptor (CaR) 2 is a type III G protein-coupled receptor, whose primary function is to regulate parathyroid hormone (PTH) secretion and thus whole body calcium homeostasis (1,2). The parathyroid CaR acts by responding to elevated extracellular Ca 2ϩ (Ca 2ϩ o ) concentration in the blood to mobilize Ca 2ϩ i release, as well as other intracellular signals to suppress PTH production and release. In dispersed bovine parathyroid cells, the increase in Ca 2ϩ i concentration and suppression of PTH secretion elicited by high Ca 2ϩ o concentration can be inhibited by phorbol ester treatment (3)(4)(5).
The human CaR contains five predicted PKC consensus sites, two within its intracellular loops (Thr 646 and Ser 794 ) and three in its intracellular domain (Thr 888 , Ser 895 , and Ser 915 ) (6). Mutation of the two intracellular loop residues had little effect on Ca 2ϩ o sensitivity, whereas mutants T888V, S895A, and S915A each exhibit increased CaR sensitivity for Ca 2ϩ o , the most dramatic effect being with the mutation at Thr 888 (7). Substitution of Thr 888 with negatively charged amino acids mimics the effect of phorbol ester treatment on wild-type CaR (8). There is also evidence that CaR T888A elicits sustained Ca 2ϩ i mobilization rather than oscillations (9).
Whereas this suggests that PKC-mediated CaR phosphorylation regulates CaR function, no reagents have existed previously for investigating CaR phosphorylation directly. Thus, we have developed a phospho-specific antibody that recognizes the phosphorylated form of CaR T888 , this residue being chosen because its mutation elicited the greatest effect on CaR function of the putative PKC sites.
Immunoblotting and Immunocytochemistry-Immunoblotting was performed as described previously (10,11). Anti-CaR mouse monoclonal antibody, raised to amino acids 214 -235 (ADD) of the extracellular domain of the human parathyroid CaR was from Affinity Bioreagents (Golden, CO) and anti-protein phosphatase 2A (catalytic subunit) monoclonal antibody was from Upstate (Millipore, Chandlers Ford, Hampshire, UK). Phosphorylation of CaR residue Thr 888 was studied using a custom-generated (Sigma Genosys) polyclonal antibody raised to the phosphorylated form of Thr 888 contained within amino acids 882-896 of the human CaR sequence (KVAARA(pT) LRRSNVSR). The antibody was affinity purified using the nonphosphorylated peptide to remove non-phosphospecific antibody, followed by collection of the phosphospecific antibody (1.1 mg/ml) on a column prepared using the phosphorylated peptide. For immunoprecipitation experiments, cell lysates prepared using SDS-free RIPA buffer were precleared with Protein A-Sepharose, then mixed with anti-CaR pT888 antibody (1:100) followed by Protein A-Sepharose overnight (4°C with constant rotation). The immunoprecipitates were collected by centrifugation and washed 3 times in SDS-free RIPA buffer prior to immunoblotting as before.
For immunofluorescence, cells were grown on coverslips to 50 -80% confluence and fixed with 10% (w/v) paraformaldehyde solution at room temperature for 30 min and permeabilized with 0.075% (w/v) saponin in phosphate-buffered saline for 10 min. Indirect immunofluorescence was performed using the anti-CaR pT888 polyclonal antibody (1:200 dilution) and an Alexa 488conjugated goat anti-rabbit secondary antibody (1:1000; Molecular Probes). Alternatively, anti-CaR monoclonal (ADD; 1:200) was used with a donkey anti-mouse secondary antibody conjugated with Alexa 594 (1:1000; Molecular Probes). In some experiments the primary antibody was preincubated overnight at 4°C with an excess of antigenic peptide (either phosphorylated or non-phosphorylated) prior to incubation with the cells. Immunofluorescence was examined using a Zeiss Axioplan 2 fluorescence microscope with images acquired using a Hamamatsu digital camera, with each sample imaged under identical exposure conditions. Images were processed using the software package KS300 version 3.0 (Carl Zeiss Ltd., Hertfordshire, UK).
Intracellular Ca 2ϩ Assay-CaR-HEK cells were cultured on glass coverslips and loaded with Fura-2/AM (1 M for 1 h) at room temperature in the dark in Ca 2ϩ assay buffer (20 mM HEPES, pH 7.4, 125 mM NaCl, 4 mM KCl, 1.2 mM CaCl 2 , 0.5 mM MgCl 2 , 5.5 mM glucose) supplemented with 0.1% bovine serum albumin. Non-absorbed Fura-2/AM was removed by washing and the cells were equilibrated for 10 min in Experimental Buffer containing the baseline [Ca 2ϩ ] o appropriate for the ensuing experiment. The cells were mounted in a perfusion chamber (Warner Instruments, Hamden, CT) and observed through a ϫ40 oil-immersion objective. Dual-excitation wavelength microfluorometry was then performed using a Nikon Diaphot inverted microscope (Cairn Research Ltd., Kent, UK). Experiments were performed at room temperature in Ca 2ϩ assay buffer containing various concentrations of CaCl 2 (0.5 mM unless otherwise stated).
Statistical Analysis-Data are presented as mean Ϯ S.E. and statistical significance was determined by one-way analysis of variance (Tukey post hoc test) or by unpaired t test as appropriate. , lysed, and solubilized in Laemmli buffer and then processed for immunoblotting using either ADD monoclonal antibody (total CaR) or anti-CaR pT888 antibody as stated. In A, panel ii, solubilization occurred either under non-reducing (NR) or reducing (␤-mercaptoethanol, ␤Me) conditions. In A, panel iii, the CaR pT888 antibody was used to immunoprecipitate the fraction of CaR proteins phosphorylated on residue Thr 888 with the resulting immunoprecipitate pellets and supernatants immunoblotted using the total CaR antibody. Where indicated, some cells were cotreated with PMA (1 M) and calyculin (100 nM) prior to lysis. B, cells, preincubated for 10 min in the absence or presence of PMA, were processed for immunoblotting (i) or immunofluorescence (ii) using anti-CaR pT888 antibody preincubated in the absence or presence of phosphorylated (ϩphos) or nonphosphorylated (ϩnonphos) immunizing peptide. All data are representative of a minimum of three independent experiments.

RESULTS
CaR-HEK cells were treated with the phorbol ester, phorbol 12-myristate 13-acetate (PMA) to stimulate PKC activity and the resulting change in CaR T888 phosphorylation was investigated by immunoblotting and immunofluorescence. The affinity purified, phospho-specific anti-CaR pT888 antibody detected a PMA-induced increase in immunoreactivity in two protein bands (Fig. 1A, panel i) corresponding to the high-mannose "core"-glycosylated CaR (140 kDa) and the mature glycosylated CaR (160 kDa) (12). To confirm the specificity of the signal, the following experiments were performed. First, to demonstrate that the protein bound to the anti-CaR pT888 antibody possesses the same biochemical properties as CaR, the PMA-pretreated CaR-HEK cell lysates were solubilized in Laemmli buffer in the presence or absence of reducing agent. Because native CaR is a disulfide-linked dimer (13,14), it exhibits dimer-like electrophoretic mobility in the absence of reducing agent but is solubilized to monomers in its presence. Here, both anti-CaR pT888 polyclonal and anti-CaR monoclonal (ADD, i.e. total CaR) antibodies detected CaR-like monomers where the lysates were solubilized with ␤-mercaptoethanol and dimeric immunoreactivity under non-reducing conditions ( Fig. 1 A,  panel ii). Second, the anti-CaR monoclonal ADD antibody immunodetected CaR-sized proteins in immunoprecipitated pellets generated using the anti-CaR pT888 antibody (Fig. 1A, panel iii). CaR-reactive signal was observed in the samples pretreated with PMA in the absence or presence of calyculin A (PP1/PP2A protein phosphatase inhibitor), but was absent in unstimulated cells. Next, to confirm that the CaR pT888 antibody was indeed phospho-specific, the PMAinduced immunoreactivity detected in CaR-HEK cells was ablated by pretreating the anti-CaR pT888 antibody with phosphorylated immunizing peptide but not when using non-phosphorylated peptide at an identical concentration (Fig. 1B, panel i). Similar results were obtained when cells grown on glass coverslips where incubated in the presence or absence of PMA and then fixed and stained with anti-CaR pT888 antibody. PMA treatment increased the resulting immunofluorescence but this staining was ablated by the phosphorylated peptide ( Fig. 1B, panel ii). At very high peptide concentrations (Ն1 M), both phosphorylated and non-phosphorylated peptides ablated antibody binding, whereas at very low peptide concentrations (Ͻ1 nM) neither ablated the signal (data not shown). Finally, CaR-like, phorbol-sensitive immunoreactivity was observed in the lysates of HEK cells stably transfected with wild-type CaR but not with a mutant receptor in which Thr 888 was replaced with alanine (CaR T888A ; Fig. 2A). Similar observations were made by immunofluorescence in which PMA elicited anti-CaR pT888 immunoreactivity in CaR-HEK cells but not in CaR T888A -HEK cells despite both cell types exhibiting equivalent immunoreactivity to the ADD monoclonal antibody (Fig. 2B).
To examine whether CaR T888 phosphorylation is altered by receptor stimulation, CaR-HEK cells were incubated for 10 min in buffer containing increasing Ca 2ϩ o concentrations in the presence or absence of the calcimimetic NPS R-467 (CaR positive allosteric modulator). Increasing Ca 2ϩ o concentration from 0.5 to 2.5 mM induced a small rise in CaR T888 phosphorylation (Fig. 3A), whereas addition of NPS 467-R (1 M) increased it further (Fig. 3, A and B; ϩ42 Ϯ 10%, p Ͻ 0.05). Cotreatment with the PKC inhibitor GF109203X inhibited the signal substantially (Ϫ60 Ϯ 3%, p Ͻ 0.001). These responses were qualitatively similar to the immunofluorescence observed in situ in CaR-HEK cells grown on cover-  slips and stained as before. Under control conditions (0.5 mM Ca 2ϩ o -containing buffer), NPS R-467 and GF109203X were without effect (data not shown). None of the above responses were observed in HEK cells stably transfected with empty vector (data not shown).
The effect of NPS R-467 on CaR T888 phosphorylation was stereoselective as 1 M NPS S-467 was without effect (Fig. 4). Replacement of NPS R-467 with an additional 2.5 mM CaCl 2 (i.e. 5 mM final Ca 2ϩ o concentration) also elicited additional CaR T888 phosphorylation and this effect was abolished by cotreatment with the calcilytic NPS-89636 (1 M; CaR negative allosteric modulator). High Ca 2ϩ o -induced CaR T888 phosphorylation was also observed in cells treated at room temperature (data not shown).
Previous studies have shown that the mutant receptor CaR T888A exhibits greater sensitivity to Ca 2ϩ o than the wildtype receptor (8) and that stimulation of CaR T888A with 3 mM Ca 2ϩ o elicits non-oscillatory, sustained Ca 2ϩ i mobilization compared with the oscillatory behavior observed with the wildtype receptor (9). To examine this further, we compared the responses of wild-type CaR and CaR T888A to various concentrations of Ca 2ϩ o using single-cell epifluoresence Ca 2ϩ i imaging (Fig. 5). In wild-type CaR-HEK cells, increasing Ca 2ϩ o concentration from 0.5 to 2 mM induced a transient response in some cells (16%), whereas most of the others were non-responsive ( Fig. 5   via chronic phorbol ester pretreatment has not been reported in this context. Here we found that in wild-type CaR-HEK cells, chronic PMA pretreatment appears to increase the sensitivity of the receptor and elicits non-oscillatory, sustained responses in most cells (Fig. 6). In CaR T888A cells, chronic PMA pretreatment had little overall effect on Ca 2ϩ o sensitivity but resulted in responses that were mostly sustained.
Next we examined the effect on CaR-induced Ca 2ϩ i oscillations of a non-selective PKC inhibitor (GF109203X) compared with the effects of two conventional PKC (i.e. Ca 2ϩ -dependent)-selective inhibitors namely Gö6976 and an anilinomonoindolylmaleimide ␤-selective PKC inhibitor (15). Addition of GF109203X to cells exhibiting Ca 2ϩ i oscillations produces sustained responses in most (Ն80%) of the cells (Fig.  7). Where cells exhibiting oscillations were first treated with either 300 nM Gö6976 or 500 nM PKC␤ inhibitor, conversion from oscillatory to sustained Ca 2ϩ i mobilization was observed in 40.1 and 27.5%, respectively, of the cells sensitive to subsequent GF109203X treatment. This concentration of PKC␤ inhibitor should also inhibit PKC␣ and PKC␥, whereas no change in the Ca 2ϩ i oscillations were seen in response to 50 nM PKC␤ inhibitor that should be selective for that isozyme (15).
Next, to identify a candidate phosphatase for the dephosphorylation of CaR T888 , CaR-HEK cells were pre-treated acutely with PMA (10 min) and then incubated for a further 5 min in the presence or absence of various phosphatase inhibitors. In each case, GF109203X (250 nM) was included during the 5-min incubation to prevent any further phosphorylation. In the absence of any phosphatase inhibitors, substantial dephospho-rylation of CaR T888 was observed within 5 min, however, in the presence of the PP1/PP2A inhibitor calyculin A (100 nM), dephosphorylation was prevented (Fig. 8A). In contrast, tautomycin (PP1 inhibitor) and FK506 (PP2B inhibitor) had little effect on CaR T888 dephosphorylation (not shown). To examine this further, the colocalization of PP2A and CaR pT888 was investigated by dual-fluorophore immunofluorescence. CaR-HEK cells were incubated acutely in PMA and then fixed and stained with the phospho-specific anti-CaR pT888 antibody and a monoclonal anti-PP2A catalytic subunit antibody, followed by goat anti-rabbit (Alexa 488) and donkey antimouse (Alexa 594) antibodies, respectively. Merging of the resulting green (CaR pT888 ) and red (PP2A) channels reveals partial colocalization (yellow) of the 2 antigens (Fig. 8B). Finally, to determine whether phosphatase inhibition affects CaR function, Fura-2-loaded CaR-HEK cells were exposed to 2.5 mM Ca 2ϩ o to induce Ca 2ϩ i oscillations and then switched to an identical buffer supplemented with 100 nM calyculin, 500 M endothall thioanhydride, or 5 nM tautomycin. Calyculin and endothall both inhibited the Ca 2ϩ i oscillations reversibly, whereas tautomycin was without effect.
To examine whether the effects of PMA and calyculin on CaR function are specific to Thr 888 , CaR T888A -HEK cells were stimulated with 2 mM Ca 2ϩ o to elicit either oscillatory or sustained Ca 2ϩ i mobilization and then cotreated with either drug. Despite the absence of Thr 888 , PMA still suppressed receptor-mediated Ca 2ϩ i mobilization (Fig. 9A). Increasing the Ca 2ϩ o concentration to 3 and then to 5 mM overcame the inhibitory effect of PMA. Interestingly, the PMA tended to ablate the Ca 2ϩ i oscillations while having little effect on the rate of decline of the "sustained" responses. Similarly, cotreatment with calyculin ablated CaR T888A -induced Ca 2ϩ i oscillations but had little effect on cells exhibiting a sustained response (Fig. 9B). Following the removal of calyculin, 5 mM Ca 2ϩ o elicited a potent response in all cells demonstrating their continued viability.
Finally, bovine parathyroid cells were incubated in the absence or presence of PMA/calyculin, fixed, and then examined for their CaR pT888 content as before. The effect of PMA/ calyculin cotreatment was to increase endogenous CaR T888 phosphorylation in the parathyroid cells (Fig. 10A). In addition, in Fura-2-loaded bovine parathyroid cells, the rise in [Ca 2ϩ ] i elicited by increasing [Ca 2ϩ ] o from 0.5 to 2 mM, was suppressed by cotreatment with PMA (Fig. 10B).

DISCUSSION
Given the key role of the CaR in controlling extracellular calcium homeostasis, the regulation of its function has important consequences for whole animal physiology (1,2). A role for PKC-mediated phosphorylation in the regulation of CaR activity has been proposed before, based on the use of pharmacologic modulators of PKC activity and the expression of mutant receptors lacking certain PKC consensus sites (3)(4)(5)(7)(8)(9). However, to prove conclusively that CaR T888 can be phosphorylated in vivo and then to determine the agonist sensitivity of such phosphorylation, as well as to identify a candidate CaR phosphatase, we raised a phospho-specific polyclonal antibody to residue Thr 888 of the human CaR (6).
The CaR pT888 -reactive protein bands visible by immunoblotting correspond in size to the high mannose and fully mature forms of CaR (12) and behave as disulfide-linked dimers (13,14) in the absence of reducing agent. In addition, the monoclonal antibody (ADD) recognizing total CaR detected the same bands as above in immunoprecipitates pulled-down using the phosphospecific antibody, whereas no anti-CaR pT888 immunoreactivity was observed in empty vector-transfected HEK cells. Together, these data demonstrate that the anti-CaR pT888 antibody detects CaR as opposed to an endogenous HEK cell protein.
Next, ablation of the signal with the phosphorylated immunizing peptide but not the non-phosphorylated peptide demonstrates the phosphospecificity of the CaR pT888 antibody. This is supported by the observation that cells stably expressing CaR T888A exhibited no immunoreactivity with anti-CaR pT888 but normal immunoreactivity with the ADD monoclonal antibody.
Having validated the antibody and demonstrated phorbolsensitive CaR T888 phosphorylation, it was then possible to demonstrate that acute calcimimetic treatment elicited CaR T888 phosphorylation in a stereoselective manner. This response was mimicked by an additional increase in Ca 2ϩ o concentration, an effect that was inhibited by cotreatment with the calcilytic NPS-89636 (11,16,17). Therefore, CaR activation elicits feedback phosphorylation of its own intracellular domain residue Thr 888 . As CaR T888 phosphorylation could be induced by   4). B, PMA-stimulated CaR-HEK cells were fixed and costained with the polyclonal anti-CaR pT888 antibody and a monoclonal anti-PP2A catalytic subunit antibody, followed by goat anti-rabbit (Alexa 488) and donkey anti-mouse (Alexa 594) antibodies. Merging of the resulting green (CaR pT888 ) and red (PP2A) signal reveals partial colocalization (yellow) of the 2 antigens. C, Fura-2-loaded CaR-HEK cells were exposed to 2.5 mM Ca 2ϩ o to induce Ca 2ϩ i oscillations and then switched to an identical buffer supplemented with 100 nM calyculin A, 500 M to 1 mM endothall, or 5 nM tautomycin. Cells were then washed in Experimental Buffer and stimulated with 3 mM Ca 2ϩ o -containing buffer to demonstrate continued viability of the responses. Data representative of at least 90 cells from a minimum of three independent experiments are shown. MAY 18, 2007 • VOLUME 282 • NUMBER 20 acute PMA treatment and inhibited by GF109203X it is highly likely that the phosphorylation is PKC-dependent.

CaR T888 Phosphorylation
CaR pT888 immunoreactivity was observed in a 160-kDa band, most likely the fully mature receptor, and a 140-kDa band, presumed to be the high mannose, core-glycosylated protein.
Whereas feedback phosphorylation of the mature, membranelocalized receptor could regulate receptor function or Ca 2ϩ i oscillation generation/maintenance, it is less clear what purpose phosphorylating the immature receptor would serve. For example, it could regulate the maturation or forward trafficking of the CaR to the membrane. It will be interesting to determine whether under certain conditions, phosphorylation of the two bands can occur differentially, for example, by different PKC isozymes or with non-identical time courses. Thus, the current study establishes that PKC-mediated phosphorylation of the mature CaR can occur, although possibly in addition to phosphorylation of intracellular or immature receptors.
In contrast to a previous report (9), we did not find that CaR T888A produced exclusively sustained Ca 2ϩ i responses. Indeed, in 2 mM Ca 2ϩ o , more cells exhibited oscillations than sustained responses, although in 3 mM Ca 2ϩ o most cells did exhibit sustained responses similar to the data reported previously (9). Nevertheless, even in 3 mM Ca 2ϩ o , some CaR T888A -HEK cells still exhibited oscillations and this should not be possible if dynamic changes in CaR T888 phosphorylation account exclusively for Ca 2ϩ i oscillations. It should be noted, however, that the oscillations induced by CaR T888A here are much slower than those elicited by the wild-type CaR. The reason for the apparent discrepancy between the two studies is unclear. In the previous study (9), transient transfection was employed, whereas here we studied HEK cells stably expressing CaR T888A . Given the heightened sensitivity of CaR T888A it is possible that in generating the stable clone, the most responsive cells died and thus slightly less responsive cells were selected. In support of this, we noted during the selection process that the culture of CaR T888A -HEK cells in Dulbecco's modified Eagle's medium (containing 1.8 mM CaCl 2 ) resulted in substantial cell death necessitating the use of RPMI media instead (data not shown).
In CaR T888A -HEK cells stimulated with 2 mM Ca 2ϩ o , the addition of either PMA or calyculin had little effect on the sustained responses, but instead inhibited Ca 2ϩ i oscillations. Increasing the Ca 2ϩ o concentration, in the continued presence of PMA, overcame the inhibitory effect of the phorbol ester. Because PMA and calyculin act to increase or sustain Ser/Thr phosphorylation, respectively, these data could be explained by a PKCmediated decrease in CaR sensitivity that occurs despite the absence of Thr 888 . If true, these data could suggest that there is a further signaling determinant within CaR or its associated signaling machinery for the establishment and maintenance of Ca 2ϩ i oscillations. In support of this is the fact that chronic PMA pretreatment virtually abolished Ca 2ϩ i oscillations in  CaR T888A -HEK cells (Ͼ98%) and did completely abolish oscillations in wild-type CaR-HEK cells. In the current study, chronic PMA pretreatment elicited heightened Ca 2ϩ o sensitivity in wild-type CaR-HEK cells in a manner similar to the effects of PKC inhibitors published previously (7,9). Chronic PMA pretreatment most likely down-regulates conventional and novel PKC isoforms (18) and PKC␣ and -⑀ have been reported to be activated upon CaR stimulation in parathyroid and CaR-HEK cells (19). These data are consistent with the previous observation (20) that CaR-induced inositol 1,4,5-trisphosphate formation was heightened in bovine parathyroid cells pre-exposed overnight to PMA (1 M). However, in that study the Ca 2ϩ o -induced Ca 2ϩ i mobilization dose-response curve was not leftward shifted and neither was the suppressive effect of 2 mM Ca 2ϩ o on PTH secretion significantly enhanced, although its inhibitory effect may have already been close to maximal under those conditions (20). PKC inhibitors selective for the conventional PKC isozymes partially mimicked the effect of the non-selective PKC inhibitor GF109203X in converting oscillatory Ca 2ϩ i responses into sustained responses, albeit in no more than 40% of the cells. Thus, together with the data from the chronic PMA pretreatment experiments, these data could suggest that a combination of conventional and novel PKC isozymes contribute to CaR regulation.
Spontaneous dephosphorylation of CaR T888 occurred in the 5 min following a 10-min PMA pretreatment indicating the presence of an active phosphatase at the CaR intracellular domain. As this effect was largely unaltered by selective inhibitors of the phosphatases PP1 and PP2B, but ablated by the PP1/PP2A inhibitor calyculin (21) and the PP2A-selective inhibitor endothall, this suggests that PP2A, or at least a calyculin-sensitive phosphatase, is responsible for this dephosphorylation. Indeed, calyculin decreased CaR-induced oscillation frequency and in some cells suppressed them entirely, consistent with the idea that CaR T888 phosphorylation is inhibitory to Ca 2ϩ i mobilization. Larger concentrations of calyculin gave much greater inhibitory effects (not shown). However, these cells also tended to detach rapidly from the coverslip as described for other cell types (22,23). Thus, only experiments in which normal oscillations were recoverable following the removal of calyculin are presented. The inhibitory effect of calyculin was mimicked by 500 M endothall thioanhydride, a protein phosphatase inhibitor (24) that exhibits greater selectivity for PP2A over PP1 (IC 50 , PP2A, 90 nM; PP1, 5 M). In contrast, the PP1-selective inhibitor tautomycin (5 nM) was without effect (IC 50 , PP1, 1 nM; PP2A 10 nM). In further support of the involvement of PP2A in CaR T888 dephosphorylation, the catalytic domain of PP2A colocalized, partially, with CaR pT888 . Together, these data implicate a calyculin-sensitive phosphatase, most likely PP2A, as the phosphatase responsible for CaR T888 dephosphorylation. Thus it will be interesting to investigate CaR signaling in cells underexpressing or overexpressing the various protein phosphatases.
Finally, anti-CaR pT888 immunoreactivity was also observed in bovine parathyroid cells treated with PMA in the presence of calyculin indicating the potential utility of this antibody for studying CaR phosphorylation in endogenous expression systems. For example, the antibody could be used to examine whether the loss of parathyroid Ca 2ϩ o sensitivity seen in conditions such as renal disease or aging involves CaR phosphorylation as opposed to down-regulation. Also, it could be used to test whether circadian changes in PTH secretion might be explained by changes in CaR T888 phosphorylation and thus Ca 2ϩ o sensitivity, over the course of the day. Together these data confirm residue Thr 888 plays a predominant role in the regulation of Ca 2ϩ i oscillations but that a further PKC-dependent mechanism could be involved, perhaps even a second phosphorylation site. In addition, we demonstrate for the first time that CaR activation induces feedback phosphorylation of the intracellular domain residue Thr 888 and that this residue is a substrate for a calyculin-sensitive protein phosphatase.