Transforming Growth Factor- (cid:98) 1 Inhibits Type I Inositol 1,4,5-Trisphosphate Receptor Expression and Enhances Its Phosphorylation in Mesangial Cells*

A potentially important cross-talk characteristic of transforming growth factor- (cid:98) (TGF- (cid:98) ) is to inhibit plate-let-derived growth factor-induced intracellular calcium rise (Baffy, G., Sharma, K., Shi, W., Ziyadeh, F. N., and Williamson, J. R. (1995) Biochem. Biophys. Res. Commun. 210, 378–383) in murine mesangial cells. The present study examined the possible basis for this effect by eval- uating the regulation of the type I inositol 1,4,5-trisphos-phate receptor (IP 3 R) by TGF- (cid:98) . TGF- (cid:98) 1 down-regulates IP 3 R protein expression by > 90% with maximal and half- maximal effects after 8 and 2 h, respectively. TGF- (cid:98) 1 also decreased IP 3 R mRNA expression by 59% after 1 h. Phosphorylation of the IP 3 R was also demonstrated as early as 15 min after TGF- (cid:98) 1 exposure. Back phosphorylation assays of IP 3 R from TGF- (cid:98) 1-treated mesangial cells with protein kinase A (PKA), indicated that TGF-(cid:98) 1-induced phosphorylation of the IP 3 R occurs at simi- lar sites as for PKA. In vitro kinase assays using the known IP 3 R peptide substrates for PKA, RPS- GRRESLTSFGNP and ARRDSVLAAS, demonstrated that TGF- 3 PKA-induced phosphorylation of the purified nonneuronal short form of the type I IP 3 R primarily occurs on serine 1589 and on serine 1755 Our studies by PCR determined that, similar to other peripheral tissues previously examined (13, 14), mesangial cells and kidney tissue only contain the SII (cid:50) form of the receptor. The functional result of nonneuronal PKA-induced IP 3 R phosphorylation remains unclear. Our in vitro kinase assays demonstrate that forskolin treatment of mesangial cells induces PKA-mediated phosphorylation of both IP 3 R peptides that contain serine and serine 1755. Thus, both sites are potential phosphorylation sites on the nonneuronal IP 3 R by PKA. TGF- (cid:98) 1 treatment also resulted in significantly increased phosphorylation of both peptides, although the degree of phosphorylation on the peptide containing serine 1755 was twice as much as the peptide containing serine 1589. It should be noted that prior studies eval-uating the site of IP 3 R phosphorylation employed purified kinases and purified IP 3 R. Our findings employed mesangial cell-derived kinase preparations that were activated by agonists in the intact cell. Our studies do not exclude the possibil-ity that other sites on the type I IP 3 R may also be phosphorylated by TGF- (cid:98) 1, which may affect its function.

Transforming growth factor (TGF) 1 -␤1 has been implicated in a variety of inflammatory and noninflammatory kidney diseases (1). Our prior studies have demonstrated that TGF-␤1 is up-regulated in animal models of diabetic kidney disease (2) and inhibition of TGF-␤ activity by neutralizing antibodies reduces diabetic renal hypertrophy and gene expression of type IV collagen and fibronectin (3). Apart from its well described effects to stimulate matrix production and regulate cell growth, TGF-␤ also has a characteristic effect to modulate the phenotypic actions of other factors. In particular, TGF-␤ can inhibit the proliferative ability of PDGF and other mitogens in human mesangial cells (4). This is of particular relevance to many glomerular diseases in that up-regulation of multiple growth factors is observed concomitantly (5,6).
Studies in a variety of cell types have demonstrated a modulatory capacity of TGF-␤ to affect the cellular response to exogenous factors. TGF-␤ inhibits PDGF-induced proliferation and inositol 1,4,5-trisphosphate (IP 3 ) production in human bone marrow fibroblasts (7). In studies with cardiac fibroblasts (8) and vascular smooth muscle cells (9), pretreatment with TGF-␤1 for at least 30 min inhibits [Ca 2ϩ ] i release in response to isoproterenol and angiotensin II, respectively. In our studies with transformed murine mesangial cells, we demonstrated that TGF-␤ markedly inhibits PDGF-BB-induced increase in [Ca 2ϩ ] i (10).
A common pathway of raising [Ca 2ϩ ] i with PDGF, isoproterenol, and angiotensin II is the activation of phospholipase C␥ or phospholipase C␤ and the generation of IP 3 from phosphatidylinositol 4,5-bisphosphate (11). The raised IP 3 levels bind to IP 3 receptors (IP 3 Rs) in the endoplasmic reticulum to release stored calcium into the cytoplasmic space (11). This is thought to be a crucial step in allowing the cell to respond to these agonists. There are at least three isoforms of the IP 3 receptor, derived from three distinct genes (reviewed in Ref. 12). The type I isoform is abundant in cerebellum but is also present in many peripheral tissues. Alternative splicing of the type I IP 3 R results in deletion of the SII segment in nonneural tissues (13,14). The SII segment is present between two serines (serine 1589 and serine 1755) (13,14), which are phosphorylated by protein kinase A (PKA) (14,15). Phosphorylation of the cerebellar type I IP 3 R by PKA impairs [Ca 2ϩ ] i mobilization by IP 3 (16,17). In addition, muscarinic receptor activation impairs IP 3 -induced [Ca 2ϩ ] i mobilization by enhancing the degradation of the type I isoform in neuroblastoma cells (18,19). The type III isoform of the IP 3 R has been recently found to be upregulated during apoptosis of T-lymphocytes (20). The type I isoform of the IP 3 R is present in the glomerulus of the kidney (21), primarily in mesangial cells (22), as well as in the renal vascular system (23). Other isoforms of the IP 3 R have not been identified in the glomerulus (21). Since TGF-␤ appears to modulate IP 3 -induced [Ca 2ϩ ] i mobilization in a variety of cell types, we postulate that TGF-␤ may mediate some of its effects via regulation of the IP 3 R.
Our present study demonstrates that TGF-␤ inhibits the protein expression of the type I isoform of the IP 3 R in mesangial cells and that this may be partly due to a decrease of the steady-state mRNA level. In addition, TGF-␤ rapidly phosphorylates the IP 3 R. The likely consequence of these effects is to modulate IP 3 R function and thus affect the cellular responsiveness to agents that act via activation of the IP 3 R. Cell Culture-An SV40-transformed murine glomerular mesangial cell line (MMC) that has been previously described (24) was primarily used in these studies. These cells retain many of the differentiated characteristics of mesangial cells in primary culture (24). To assure that our findings were not influenced by transformation, we performed a limited series of experiments in rat glomerular mesangial cells that were conducted between passages 4 and 10. Isolation of rat mesangial cells was performed as detailed previously (25).
Time Course of TGF-␤1 Regulation of IP 3 R Protein-MMC grown in Dulbecco's modified Eagle's medium (DMEM) (Life Technologies, Inc.) with 10% fetal calf serum were harvested and plated onto 100-mm dishes with growth media. After reaching 80% confluence, cells were incubated in serum-free DMEM for 24 h. During the subsequent final 8 h of incubation, cells were treated with TGF-␤1 (10 ng/ml) for the last 1, 2, 4, and 8 h, washed with PBS three times, and harvested in lysis buffer that contained 50 mM Tris-HCl (pH 7.2), 150 mM NaCl, 1% (w/v) Triton X-100, 1 mM EDTA, 1 mM PMSF, and 5 g/ml each of aprotinin and leupeptin. All samples, including the control samples, were harvested after the same overall duration of incubation. Protein concentrations of samples were quantitated, and equal amounts of protein were run on a 7% SDS-PAGE gel, transferred to nitrocellulose, and immunoblotted with an antibody raised to the C terminus of the type I IP 3 R from brain (26). The primary antibody was then removed, and the membrane was incubated with horseradish peroxidase-conjugated secondary antibody. Immunoreactive bands were detected using enhanced chemiluminescence (Amersham). Densitometric analysis of scanned images was performed on a Macintosh 7600/132 computer using the public domain NIH Image program. Measurements in control samples were assigned a relative value of 100%.
Polymerase Chain Reaction-based Analysis of Type I IP 3 R in MMC and Regulation of Type I IP 3 mRNA by TGF-␤1-First strand cDNA was prepared from total RNA isolated from mouse cerebellum, kidney, and MMC using Moloney murine leukemia virus reverse transcriptase and oligo(dT) primers. Polymerase chain reaction (PCR) was performed using specific primers for the type I IP 3 R isoform (5Ј-CGT GGA TGT TCT ACA CAG ACC AG-3Ј) and (5Ј-TTG GAA CTT CTT GAA GAG ACT A-3Ј) (13). These primers are on either side of the SIIϪ splice domain of the type I IP 3 R (13). Each reaction mixture contained 50 mM KCl, 10 mM Tris-HCl (pH 8.3), 3 mM MgCl 2 , oligonucleotide primers at 0.5 M, all four dNTPs (each at 0.2 mM), 2.5 units of Taq DNA polymerase (Perkin-Elmer), and 15% of the first strand cDNA products in a final volume of 50 l. After an initial cycle of 4 min at 94°C, the reaction was cycled 40 times for 60 s at 55°C, 3 min at 72°C, and 60 s at 94°C. The reaction was completed with one cycle at 72°C for 5 min. For initial analysis, the PCR product was run on 2% agarose gel stained with ethidium bromide. The PCR product was cloned into the pCRII TA cloning system (Invitrogen, La Jolla, CA) and sequenced to confirm its identity.
To assess whether IP 3 R mRNA was affected by TGF-␤1 treatment, MMC incubated in serum-free media was treated with TGF-␤1 (10 ng/ml) for 1, 2, and 4 h, washed with ice-cold PBS and total RNA isolated using acid guanidinium thiocyanate-phenol-chloroform (27). Poly(A) mRNA was isolated from total RNA by an oligo(dT) affinity column (Promega, Madison, WI). 3 g of poly(A) mRNA were loaded onto a 1.2% agarose gel containing 2.2 M formaldehyde, electrophoresed, and transferred onto nylon membrane. Cerebellum RNA was run as a positive control. The probe for the type I IP 3 R was obtained by running the PCR product (see above) from MMC on a low melt agarose gel and cutting out the band corresponding to the type I IP 3 R. The probe was purified and labeled via the random prime method (Boehringer Mannheim). Hybridization and washing conditions were performed as described previously (2). To standardize for loading, membranes were stripped and reprobed with a ␤-actin cDNA probe (kindly provided by Dr. P. Norton). Densitometry was performed as described above, and mRNA levels were calculated relative to those of ␤-actin.
Phosphorylation of Type I IP 3 R-MMC incubated in serum-free DMEM for 24 h were washed with phosphate-free DMEM and incubated in the same buffer for 10 min. [ 32 P]Inorganic phosphate was added to give a final concentration of 0.3 mCi/ml, and the incubation was continued for a further 90 min. The cells were exposed to various agonists for the desired times, and the incubations were then quenched by the addition of 4 ml of ice-cold PBS containing 200 M sodium orthovanadate. Cells were washed twice with the PBS/vanadate solution and scraped off the dish. Following centrifugation at 150 ϫ g for 15 s, the cells were washed twice in PBS/vanadate and finally resuspended in lysis buffer that contained 50 mM Tris-HCl (pH 7.2), 150 mM NaCl, 1% (w/v) Triton X-100, 1 mM EDTA, 1 mM sodium orthovanadate, 50 mM tetrasodium pyrophosphate, 100 mM NaF, 5 nM okadaic acid, 1 mM PMSF, and 5 g/ml each of aprotinin and leupeptin. The cells were solubilized on ice for 30 min. Insoluble material was removed by centrifugation for 10 min at 25,000 ϫ g. All of the extracts were precleared with 25 l of a 50% (v/v) slurry of Staphylococcus aureus cell wall (Pansorbin, Calbiochem). IP 3 R was immunoprecipitated from the labeled extracts by overnight incubation with 100 l of 20% protein A-Sepharose beads and 100 g of the antibody to the type I IP 3 R. The immunoprecipitates were washed three times in lysis buffer, and the phosphorylated proteins were analyzed by SDS-PAGE. The polypeptides in the gel were transferred to nitrocellulose and then autoradiographed. Membranes were subsequently immunoblotted with antibody to IP 3 R to localize the phosphorylated IP 3 R.
Back Phosphorylation of the Type I IP 3 R by PKA-Back phosphorylation of IP 3 R immunoprecipitates was performed as detailed previously (26). MMC were rested in serum-free DMEM for 24 h and then incubated with various agonists for 15 min. The cells were centrifuged and solubilized in lysis buffer, and the extracts were obtained were immunoprecipitated with IP 3 R antibody as described above. The immunoprecipitates were washed three times in a phosphorylation buffer that contained 120 mM KCl, 50 mM Tris-HCl (pH 7.2), 0.1% Triton X-100 (w/v), 0.3 mM MgCl 2 , 0.5 mM PMSF, and 10 g/ml each aprotinin and leupeptin. Aliquots of protein A-Sepharose beads were incubated in 50 l of the phosphorylation buffer containing 100 units/ml of the catalytic subunit of PKA and 1 Ci of [␥-32 -P]ATP (3000 Ci/mmol). The immunoprecipitates were phosphorylated for 15 min at 30°C, and the reaction was terminated by washing the protein A-Sepharose beads three times with phosphorylation buffer containing 1 mM unlabeled MgATP. The phosphorylated proteins in the immunoprecipitates were separated on 5% SDS-PAGE gels and transferred to nitrocellulose, which was then autoradiographed. The membrane was later immunoblotted to locate the receptor. In all experiments the IP 3 R was the only phosphorylated band above the prestained myosin molecular mass marker (200 kDa).
In Vitro Kinase Assay with IP 3 R Peptides-The peptides RPS-GRRESLTSFGNP and ARRDSVLAAS, which include the major phosphorylation site for PKA (serine 1755 and serine 1589) (14,15) were selected for in vitro kinase assays. MMC incubated in serum-free DMEM were treated with agonists for various periods of time, washed with PBS, harvested with cold extraction buffer (25 mM Tris-HCl, pH 7.4, 0.5 mM EDTA, 0.5 mM EGTA, 10 mM ␤-mercaptoethanol, 1 g/ml leupeptin, 1 g/ml aprotinin, 1 mM PMSF), and homogenized with a Dounce homogenizer. The lysate was centrifuged for 5 min at 4°C at 14,000 ϫ g, and the supernatant was saved. The protein concentrations of the supernatants were quantitated, and equal amounts of protein were added to a reaction mixture containing 40 mM Tris-HCl, pH 7.4, 20 mM MgCl 2 , 0.1 mg/ml bovine serum albumin, 200 M IP 3 R peptide substrate (RPSGRRESLTSFGNP or ARRDSVLAAS), and 3000 Ci/ mmol [␥-32 P]ATP, and 0.5 mM ATP per reaction. Experiments were performed in parallel with the addition of a PKA inhibitor peptide, TTYADFIASGRTGRRNAIHD (1 M) (Promega). The reaction was allowed to proceed for 5 min at 30°C and then terminated with the addition of 2.5 M guanidine hydrochloride. 10 l of sample was spotted onto phosphocellulose filter paper (1 ϫ 1 cm) and washed repeatedly with 1 M NaCl and subsequently with 1 M NaCl in 1% H 3 PO 4 . The papers were then dried in an oven and placed in scintillation vials for radioactive counting. The stoichiometry of phosphorylation was assessed from the specific activity of [ 32 P]ATP and from the amount of peptide used. The concentration of the peptide substrate and the duration of the in vitro kinase reaction were varied to define the concentration and time dependence of peptide phosphorylation. 3 R Protein by TGF-␤1-An antibody raised to the C terminus of the rat brain type I IP 3 R recognized a 240-kDa polypeptide from the SV40-transformed MMC (Fig. 1A). The same sized band was noted from protein derived from mouse and rat cerebellum (data not shown). The addition of TGF-␤1 (10 ng/ml) down-regulated the expression of the IP 3 R if administered for 2-8 h (Fig. 1A). Fig. 1B shows the cumulative data assembled from densitometric scans of several experiments. TGF-␤1 reduced IP 3 R protein expression to 42% by 2 h and 25% at 4 h, with respect to control values. The maximal effect was seen at 8 h, at which point IP 3 R protein expression was reduced to 3% of control values. A similar pattern was noted with early passage nontransformed rat glomerular mesangial cells. Type I IP 3 R immunoreactivity was reduced to 57% of control at 2 h of TGF-␤1 treatment and to 49% of control at 4 h of TGF-␤1 treatment (Fig. 2, A and B).

Down-regulation of Type I IP
TGF-␤1 Inhibits the Expression of the IP 3 R mRNA--To determine if the effects of chronic exposure to TGF-␤1 on inhibition of the IP 3 R protein expression may be due to an effect on synthesis of new protein, we evaluated possible regulation by TGF-␤1 at the messenger RNA level of type I IP 3 R. Northern analysis of poly(A) mRNA from MMC revealed a single Ͼ10kilobase mRNA band (Fig. 3A). A band of identical size was noted with mouse and rat cerebellum RNA (data not shown). MMC were treated with TGF-␤1 for 1 and 4 h, and the results from a representative Northern analysis are shown in Fig. 3A. Quantitative analysis demonstrates a decrease to 41% of control type 1 IP 3 R mRNA expression as early as 1 h of exposure to TGF-␤1, and it remains suppressed at 4 h of exposure to TGF-␤1 (31%) (Fig. 3B). Thus, the reduction in mRNA for the type I IP 3 R precedes the reduction in protein.
Mesangial Cells and Mouse Kidney Only Express the SIIϪ Form of the IP 3 R-It has been previously demonstrated that the cerebellum contains the long form (insertion of the SII segment) of the type I IP 3 R, whereas peripheral tissues may contain exclusively the alternatively spliced short form (SIIϪ) or both the SIIϪ and SIIϩ of the type I IP 3 R (13,14). To evaluate the length of the IP 3 R in mesangial cells and mouse kidney, we chose primers that surround the SII domain of the IP 3 R cDNA (13). Fig. 4 shows that mesangial cells and kidneys from mice only express a 580-bp DNA segment, whereas cerebellum cDNA expresses a 700-bp segment. As described previously (13), the cerebellum fragment corresponds to the SIIϩ of the IP 3 R, and the 580-bp fragment corresponds to the SIIϪ of the IP 3 R that has the SII segment deleted. This was confirmed by sequencing of the 580-bp DNA segment.
Phosphorylation of the IP 3 R by TGF-␤1-It has been previously demonstrated that phosphorylation of the IP 3 R may affect its function; therefore, we evaluated whether TGF-␤1 could affect the phosphorylation status of the type I IP 3 R. MMC labeled with [ 32 P]orthophosphate were treated with TGF-␤1 for variable times, and the IP 3 R was immunoprecipitated and resolved by SDS-PAGE. A time course demonstrating the peak effect of TGF-␤1 on phosphorylation of the IP 3 R is shown in Fig. 5A. Quantitative analysis demonstrates that phosphorylation was increased by 2-fold at 15 min, 5-fold at 30 min, and 4-fold at 60 min of TGF-␤1 treatment (Fig. 5B). As shown in Fig. 6, TGF-␤1 and forskolin treatment for 15 min enhanced phosphorylation of the IP 3 R to a similar degree.
Back Phosphorylation of the Type I IP 3 R by Protein Kinase A-Given that the IP 3 R has been previously demonstrated to be phosphorylated by PKA (15,17), and based on our prior data that cyclic nucleotide dependent kinases may mediate TGF-␤1 inhibition of [Ca 2ϩ ] i mobilization (10), we asked if PKA might be playing a role in TGF-␤1 phosphorylation of the IP 3 R. Triton X-100 extracts of unlabeled control or TGF-␤1-treated mesan-  Fig. 1  gial cells were immunoprecipitated, and the immunoprecipitates were phosphorylated in vitro after incubation with [ 32 P]ATP and the catalytic subunit of PKA (Fig. 7A, upper  panel). The major polypeptide phosphorylated in the immunoprecipitates could be shown to be IP 3 R by immunoblotting (Fig.  7A, lower panel). The enhanced phosphorylation of the protein in TGF-␤1-treated mesangial cells in vivo markedly lowered the incorporation of 32 P in the in vitro phosphorylation assay.

FIG. 2. Time course of TGF-␤1 inhibition of type I IP 3 R immunoreactivity in rat mesangial cells. Shown is a study similar to that in
Quantitative analysis demonstrates that the degree of in vitro phosphorylation of the IP 3 R by PKA was decreased to 34% of control with 15 min of TGF-␤1 treatment (Fig. 7B). This experiment suggests that TGF-␤1 stimulates phosphorylation of the IP 3 R at sites that are phosphorylated by protein kinase A.
TGF-␤1 Phosphorylates the Type I IP 3 R Peptides RPS-GRRESLTSFGNP and ARRDSVLAAS via PKA-The sites on the type I IP 3 R from cerebellum that have been demonstrated to be phosphorylated by PKA are serine 1755 and serine 1589 (14,15,17). Synthetic peptides that contain these two serine sites, RPSGRRESLTSFGNP and ARRDSVLAAS, respectively, were used as substrate in an in vitro kinase assay with crude cell lysates prepared from control and TGF-␤1-treated MMC.  3-8). After washing, the cells were solubilized with Triton X-100, and the IP 3 R was immunoprecipitated as described under "Experimental Procedures." The immunoprecipitates were analyzed on 5% SDS-PAGE, and the polypeptides were transferred to nitrocellulose. Panel A shows the autoradiogram of a representative experiment. Panel B shows the histogram of mean Ϯ S.E. of data derived from three separate experiments. *, p Ͻ 0.01 versus control.
FIG. 6. Phosphorylation of the type I IP 3 R by TGF-␤1 and forskolin. MMC labeled with [ 32 P]orthophosphate were incubated with TGF-␤1 (10 ng/ml, 15 min) or forskolin (10 M, 15 min). Immunoprecipitated IP 3 R was analyzed on 5% SDS-PAGE, transferred to nitrocellulose, and exposed to autoradiography. phosphorylated peptide (Fig. 8B). Prolonged incubation (Ͼ15 min) decreased kinase activity, possibly due to endogenous phophatase activity present in the lysate. TGF-␤1 treatment for 15 min demonstrated enhanced kinase activity at all concentrations of the peptide and at all durations of the in vitro kinase reaction. Varying the concentration and time of the kinase reaction of the peptide ARRDSVLAAS gave similar relationships (data not shown) as noted for the peptide RPS-GRRESLTSFGNP. For the subsequent experiments, the concentration of the peptide was 200 M, and the duration of the kinase reaction was 5 min. TGF-␤1 treatment of MMC for 5 and 15 min stimulated phosphorylation of the peptide RPS-GRRESLTSFGNP by 150 and 158% of control values, respectively (Table I and Fig. 9). Forskolin treatment resulted in slightly greater phosphorylation (201%) at 5 min but phosphorylation similar to that of TGF-␤1 at 15 min (157%). The addition of the peptide inhibitor of PKA (PKI) completely prevented the enhanced phosphorylation of this peptide by both forskolin and TGF-␤1 (Fig. 9). Based on the specific activity of [ 32 P]ATP and the amount of IP 3 R peptide added to the reaction mixture, under control conditions 0.038 Ϯ 0.003 mol of phos-phate was incorporated per mol of the IP 3 R peptide RPS-GRRESLTSFGNP. TGF-␤1 treatment (10 ng/ml for 15 min) increased the phosphorylation to 0.059 Ϯ 0.001 mol of phosphate/mol of the IP 3 R peptide RPSGRRESLTSFGNP. Although dilute, crude cell lysates do not appear to contain enough kinase activity to demonstrate stoichiometric phosphorylation on the IP 3 R peptides, this method does allow for comparison of relative kinase activities from treated and untreated cells.
TGF-␤1 treatment of MMC for 5 and 15 min also stimulated phosphorylation of the peptide ARRDSVLAAS by 121 and 123% of control values (Table I and Fig. 10). Forskolin treat- ment of MMC resulted in a greater degree of phosphorylation of this peptide (153 and 133% at 5 and 15 min, respectively). The enhanced phosphorylation was reversed if PKI was added to the reaction mix (Fig. 10). These results suggest that TGF-␤1stimulated phosphorylation of the IP 3 R peptides is mediated via PKA activation in murine mesangial cells. DISCUSSION The main conclusion based on the results of our study is that the type I IP 3 R is a target of regulation by TGF-␤1. TGF-␤1 has a potent ability to inhibit the protein expression of the type I IP 3 R after 2-4 h of TGF-␤1 exposure in glomerular mesangial cells. The decreased protein expression of the type I IP 3 R may be partly due to diminished synthesis of the IP 3 R as we found a decrease in mRNA levels prior to observing a decrease in protein levels. Our finding that TGF-␤1 decreases steady-state IP 3 R mRNA levels can be explained either by an effect on transcription of the type I IP 3 R gene or by enhancement of the degradation rate of the mRNA and will require further studies.
Our results differ somewhat from those of Wojcikiewicz et al. (19) in that carbachol decreased IP 3 R protein expression due mainly to enhanced degradation of the IP 3 R in human neuroblastoma cells. However, this group also reported a decrease in mRNA levels after 3 h of exposure to carbachol, which may have contributed to the decreased protein levels. The presence of several ATTTA sites in the 3Ј-untranslated region of the type I IP 3 R cDNA in both the mouse and rat (28,29) suggests that mRNA stability and/or translation may be subject to regulation (30,31). Apart from regulation by TGF-␤1 of the mRNA for the type I IP 3 R, it is likely that degradation of the protein is also enhanced by TGF-␤1, since we have previously found the halflife of type I IP 3 R protein to be 8 -11 h (32).
Our observation that TGF-␤1 inhibits type I IP 3 R expression also demonstrates that factors that do not directly induce a [Ca 2ϩ ] i flux may affect IP 3 R expression. Carbachol-induced IP 3 R degradation has been linked to regulation of functional [Ca 2ϩ ] i stores (19), although the mechanism underlying this linkage has not been demonstrated. TGF-␤1 has not been found to directly affect [Ca 2ϩ ] i in many cell types evaluated (33,34), including mesangial cells (10). Therefore, it is likely that TGF-␤ regulation of the IP 3 R differs from the down-regulation observed with [Ca 2ϩ ] i -mobilizing agonists.
The implication of decreased IP 3 R protein expression is that [Ca 2ϩ ] i mobilization would be impaired, as demonstrated previously (18) in neuroblastoma cells. Theoretically, the TGF-␤1treated cell should be less sensitive to mobilize [Ca 2ϩ ] i when exposed to any agonist that stimulates IP 3 -mediated [Ca 2ϩ ] i mobilization via the IP 3 R. Thus, our findings may explain the previous observations that long term exposure (Ͼ1 h) to TGF-␤1 would inhibit [Ca 2ϩ ] i mobilization by isoproterenol (8) and PDGF. 2 Short term (15-30 min) treatment of TGF-␤1 in mesangial and smooth muscle cells are also sufficient to block agonistinduced [Ca 2ϩ ] i mobilization (9,10). This effect would not be explained by decreased expression of type I IP 3 R, since it requires at least 1 h of exposure to TGF-␤. The short term effect may be mediated via TGF-␤1-induced phosphorylation of the IP 3 R. It has been demonstrated that phosphorylation of the IP 3 R may affect its ability to release [Ca 2ϩ ] i upon exposure to IP 3 (17). This issue is complex, since there are different lengths of the type I IP 3 R in neuronal and nonneuronal tissues that may affect the site and consequence of phosphorylation (14). Cerebellar type I IP 3 R is noted to be in the long form with insertion of an SII domain, which is situated between the two major serines (serines 1589 and 1755) that are phosphorylated 2 K. Sharma, unpublished observations.  9. In vitro kinase assays with IP 3 R peptide substrate RPS-GRRESLTSFGNP. MMC were treated with TGF-␤1 (10 ng/ml) for 5 and 15 min or forskolin (10 M) for 5 and 15 min. In vitro kinase assays were performed as described under "Experimental Procedures" and as noted in Table I by PKA. Using cerebellar derived IP 3 R in liposomes, Cameron et al. (16) recently demonstrated that PKA-induced phosphorylation impairs IP 3 -mediated [Ca 2ϩ ] i , whereas protein kinase C-induced phosphorylation of the cerebellar IP 3 R enhances IP 3 -mediated [Ca 2ϩ ] i release. PKA-induced phosphorylation of the purified nonneuronal short form of the type I IP 3 R primarily occurs on serine 1589 and on serine 1755 (14). Our studies by PCR determined that, similar to other peripheral tissues previously examined (13,14), mesangial cells and kidney tissue only contain the SIIϪ form of the receptor. The functional result of nonneuronal PKA-induced IP 3 R phosphorylation remains unclear. Our in vitro kinase assays demonstrate that forskolin treatment of mesangial cells induces PKA-mediated phosphorylation of both IP 3 R peptides that contain serine 1589 and serine 1755. Thus, both sites are potential phosphorylation sites on the nonneuronal IP 3 R by PKA. TGF-␤1 treatment also resulted in significantly increased phosphorylation of both peptides, although the degree of phosphorylation on the peptide containing serine 1755 was twice as much as the peptide containing serine 1589. It should be noted that prior studies evaluating the site of IP 3 R phosphorylation employed purified kinases and purified IP 3 R. Our findings employed mesangial cell-derived kinase preparations that were activated by agonists in the intact cell. Our studies do not exclude the possibility that other sites on the type I IP 3 R may also be phosphorylated by TGF-␤1, which may affect its function.
The interaction of TGF-␤1 with the type I IP 3 R probably leads to important modulatory influences on glomerular mesangial cells in vivo. Chronic glomerular overexpression of TGF-␤ has been demonstrated in experimental diabetic kidney disease, experimental glomerulonephritis, and puromycin-induced nephrosis (1). In experimental diabetes, other factors including PDGF-BB, fibroblast growth factor, endothelin, and the renin-angiotensin II system are also found to be up-regulated (6,35). The diabetic kidney demonstrates chronic vasodilation of the afferent arteriole, mesangial cell stretching, and glomerular hypertrophy without a great degree of mesangial cell proliferation. Theoretically, TGF-␤-induced IP 3 R downregulation may impair angiotensin-and endothelin-induced smooth muscle cell and mesangial cell contraction and impair PDGF-and fibroblast growth factor-induced mesangial cell proliferation. Whether this property of TGF-␤ would be beneficial or deleterious in contributing to chronic renal disease progression remains to be investigated.