Degradation of inositol 1,4,5-trisphosphate receptors during cell stimulation is a specific process mediated by cysteine protease activity.

Inositol 1,4,5-trisphosphate (InsP3) receptors are down-regulated in response to chronic activation of certain cell surface receptors because their degradation is accelerated. Studies on the nature of the down-regulatory process and the protease(s) responsible for receptor degradation are described here. InsP3 receptor down-regulation was not accompanied by parallel changes in the concentrations of several other relevant proteins (endoplasmic reticulum Ca2+-ATPase, 3-hydroxy-3-methylglutaryl-coenzyme A reductase, and protein kinases α and ε). Thus, the down-regulatory process selectively targets InsP3 receptors for degradation. Furthermore, down-regulation was unaffected by brefeldin A and NH4Cl, indicating that InsP3 receptor degradation occurs without removal of receptors from the endoplasmic reticulum and independently of functional lysosomes. Analysis of InsP3 receptor immunofluorescence confirmed that the receptors are not redistributed prior to or during down-regulation. Finally, of a range of protease inhibitors tested, only N-acetyl-Leu-Leu-norleucinal blocked down-regulation. Thus, cysteine protease activity accounts for InsP3 receptor degradation and analysis of proteolysis in permeabilized cells indicates that this activity is calpain. Thus, InsP3 receptor down-regulation appears to result from the highly selective calpain-mediated degradation of InsP3 receptors. Calpain activity may be stimulated by the high concentrations of Ca2+ that are thought to be found in the vicinity of activated InsP3 receptors.

Inositol 1,4,5-trisphosphate (InsP 3 ) receptors are down-regulated in response to chronic activation of certain cell surface receptors because their degradation is accelerated. Studies on the nature of the down-regulatory process and the protease(s) responsible for receptor degradation are described here. InsP 3 receptor down-regulation was not accompanied by parallel changes in the concentrations of several other relevant proteins (endoplasmic reticulum Ca 2؉ -ATPase, 3-hydroxy-3-methylglutaryl-coenzyme A reductase, and protein kinases ␣ and ⑀). Thus, the down-regulatory process selectively targets InsP 3 receptors for degradation. Furthermore, down-regulation was unaffected by brefeldin A and NH 4 Cl, indicating that InsP 3 receptor degradation occurs without removal of receptors from the endoplasmic reticulum and independently of functional lysosomes. Analysis of InsP 3 receptor immunofluorescence confirmed that the receptors are not redistributed prior to or during down-regulation. Finally, of a range of protease inhibitors tested, only N-acetyl-Leu-Leu-norleucinal blocked down-regulation. Thus, cysteine protease activity accounts for InsP 3 receptor degradation and analysis of proteolysis in permeabilized cells indicates that this activity is calpain. Thus, InsP 3 receptor downregulation appears to result from the highly selective calpain-mediated degradation of InsP 3 receptors. Calpain activity may be stimulated by the high concentrations of Ca 2؉ that are thought to be found in the vicinity of activated InsP 3 receptors.
Phosphoinositidase C (PC) 1 -mediated phosphatidylinositol 4,5-bisphosphate hydrolysis leads to the formation of two intracellular signaling molecules, inositol 1,4,5-trisphosphate (InsP 3 ) and 1,2-diacylglycerol (1,2). These molecules then bind to and activate, respectively, InsP 3 receptors and members of the protein kinase C (PKC) family (1,2). InsP 3 receptors play a crucial role in intracellular signaling as they form channels that conduct Ca 2ϩ in an InsP 3 -sensitive manner (1,(3)(4)(5). Of the InsP 3 receptors fully sequenced (types I, II, and III), the type I receptor is the most widely expressed and appears to be ubiquitous in animal tissues (4,5). The primary intracellular location of the type I receptor is the endoplasmic reticulum (ER) and its primary function appears to be to regulate Ca 2ϩ release from this organelle (3)(4)(5).
The importance of the type I InsP 3 receptor is reflected by the extent to which it is regulated; Ca 2ϩ , ATP, phosphorylation, and cytoskeletal elements can all regulate receptor function (4 -6). Furthermore, receptor concentration can be altered, since activation of certain PC-linked receptors reduces cellular type I InsP 3 receptor levels with half-maximal effects at ϳ1 h (7)(8)(9)(10). This down-regulation has been observed in a number of cell types stimulated with a variety of agonists; for example, in SH-SY5Y human neuroblastoma cells stimulated with carbachol, a muscarinic agonist (7,8), and in AR4 -2J rat pancreatoma cells stimulated with cholecystokinin (9). As InsP 3 receptor down-regulation suppresses Ca 2ϩ mobilization (7), the down-regulatory process may serve to limit Ca 2ϩ -mediated effects during chronic cell stimulation.
To date, investigation of the mechanism of InsP 3 receptor down-regulation has shown that it results from accelerated receptor degradation (10). This acceleration is not mediated by PKC and appears to require persistent receptor activation by InsP 3 (7,8). It also requires InsP 3 receptor-mediated Ca 2ϩ -flux, since thapsigargin, which discharges ER Ca 2ϩ stores independently of InsP 3 receptors (1, 3), blocks carbachol-induced InsP 3 receptor down-regulation in SH-SY5Y cells (10). An additional link with Ca 2ϩ comes from a study showing that a Ca 2ϩ -dependent cytosolic cysteine protease activity, perhaps calpain (11), can cleave purified type I InsP 3 receptor (12). As yet, however, it is not known whether the InsP 3 receptor degradation seen in intact cells (7-10) reflects a general increase in proteolysis or is specific, whether InsP 3 receptors are relocated prior to their degradation, and which activity is responsible for InsP 3 receptor degradation. Here we report on the specificity and site of InsP 3 receptor degradation in intact cells and provide evidence that a cysteine protease activity, most likely calpain, is responsible for InsP 3 receptor down-regulation.

EXPERIMENTAL PROCEDURES
Cell Culture and Pretreatment-SH-SY5Y cells, generously provided by either Dr. J. L. Beidler (Sloan-Kettering Cancer Center, New York) or Dr. S. K. Fisher (University of Michigan), and AR4 -2J cells (obtained from ATCC) were cultured routinely as described (9). When appropriate, stimuli or inhibitors were added directly to culture medium and cells were maintained at 37°C for the time of pretreatment required.
Analysis of InsP 3 Receptor Down-regulation-Control or pretreated cell monolayers were harvested in 155 mM NaCl, 10 mM Hepes, 1 mM EDTA, pH 7.4, and were collected by centrifugation at 400 ϫ g for 2 min. Cell pellets were then resuspended in ice-cold homogenization buffer (10 mM Tris, 1 mM EGTA, 0.2 mM phenylmethylsulfonyl fluoride, 1 mM dithiothreitol, 10 M leupeptin, 10 M pepstatin, 0.2 M soybean trypsin inhibitor, pH 7.4) and were disrupted with a Polytron homogenizer (1 ϫ 10 s). Homogenates were then centrifuged (38,000 ϫ g for 10 min at 4°C) and microsomal preparations obtained were resuspended in homogenization buffer and frozen until required. For analysis of type I InsP 3 receptor, Ca 2ϩ -ATPase, and PKC isozymes, microsomal preparations were denatured at 100°C for 5 min in SDS gel loading buffer (14); for HMG-CoA reductase analysis, samples were incubated at 20°C for 45 min in SDS gel loading buffer (14) plus 6 M urea. Samples (10 g of protein/lane) and prestained molecular mass markers were then subjected to electrophoresis in polyacrylamide gels, were transferred to nitrocellulose, incubated with antisera, peroxidase-conjugated secondary antibodies, ECL detection reagents (Amersham), and x-ray film (14).
Analysis of InsP 3 Receptor Degradation in Permeabilized Cells-For permeabilization, pellets of harvested SH-SY5Y cells were resuspended in 120 mM KCl, 1 mM EDTA, 20 mM Tris, pH 7.4 (Buffer A). The suspension was then treated with digitonin (75 g/ml) for 10 min at 20°C and was centrifuged (1000 ϫ g for 2 min). The pellet of permeabilized cells was then washed and finally resuspended in ice-cold Buffer A. To prepare cytosol, pellets of harvested SH-SY5Y cells were resuspended in ice-cold 1 mM EGTA, 1 mM dithiothreitol, 20 mM Tris, pH 7.4, and were disrupted with 25 strokes of a Dounce tissue grinder. The homogenate was then centrifuged at 40,000 ϫ g for 10 min at 4°C and the supernatant was used as a cytosol preparation.
Permeabilized cells and cytosol were then co-incubated without or with Ca 2ϩ and protease inhibitors in 45 mM KCl, 0.375 mM EDTA, 0.375 mM EGTA, 0.375 mM dithiothreitol, 15 mM Tris, pH 7.4 (final volume, 40 l). After 25 min at 37°C, reactions were quenched with SDS gel loading buffer (14) and samples were probed for type I receptor immunoreactivity.
Immunofluorescence Microscopy-SH-SY5Y cells, grown on coverslips, were rinsed with Dulbecco's phosphate-buffered saline supplemented with Ca 2ϩ /Mg 2ϩ and then fixed in the same buffer plus 3.6% paraformaldehyde (Acros Organics). All subsequent steps were performed in Ca 2ϩ /Mg 2ϩ -free phosphate-buffered saline; cells were washed once for 2 min, permeabilized with 0.2% Triton X-100 for 10 min, washed 3 times (10 min/wash), incubated with type I InsP 3 receptorspecific antibody plus 10% fetal calf serum for 1 h, washed 3 times, incubated with rhodamine-labeled donkey anti-rabbit IgG (Chemicon) plus 10% fetal calf serum for 1 h and finally washed 3 times. To define nonspecific fluorescence, cells were incubated with antibody preadsorbed with peptide antigen (10 g/ml). Coverslips were then rinsed with water, mounted in 90% glycerol, 0.1% p-phenylenediamine and immobilized with nail polish. Photomicrographs were generated from 10 s exposures using a Nikon Microphot-FXA microscope equipped for analysis of rhodamine fluorescence.

Specificity of InsP 3 Receptor Degradation-To gain insight
into the down-regulatory process, other relevant proteins were quantified. Initially examined were ER Ca 2ϩ -ATPase, which like the type I InsP 3 receptor is an ER membrane protein involved in Ca 2ϩ homeostasis (3,13), and PKC isozymes ␣ and ⑀, which like InsP 3 receptors can be down-regulated during cell stimulation because of accelerated proteolysis (16 -19). Fig. 1a shows that while type I InsP 3 receptor immunoreactivity in SH-SY5Y cells was dramatically reduced by exposure to carbachol (half-maximal effect at ϳ1 h), Ca 2ϩ -ATPase and PKC␣ and ⑀ immunoreactivity remained unchanged for up to 4 h. Thus, InsP 3 receptor down-regulation does not reflect a general in-crease in ER membrane protein degradation and is discrete from the process that leads to PKC degradation.
Next, we examined whether HMG-CoA reductase depletion accompanied InsP 3 receptor down-regulation, since this ER membrane enzyme can also down-regulate (in response to sterols) because of accelerated degradation (20 -23). These studies were conducted in AR4 -2J cells, in which HMG-CoA reductase was up-regulated 2-3-fold by incubation of cells in serum-free medium for 18 h. 2 Fig. 1b shows that while cholecystokinin was very effective at down-regulating the type I InsP 3 receptor it had no effect on HMG-CoA reductase levels (lanes 3 and 4). Conversely, sterols down-regulated HMG-CoA reductase, but did not alter InsP 3 receptor levels ( lanes 5 and 6). Thus, the proteolytic mechanisms that control the levels of InsP 3 receptors and other ER membrane proteins are highly selective. Fig. 1 rule out wholesale destruction of the ER as the basis of InsP 3 receptor down-regulation, the possibility remains that downregulation results from the selective destruction of a small portion of ER that is enriched in InsP 3 receptors. This is plausible because InsP 3 receptors may reside in specialized regions of the ER (24,25) and Ca 2ϩ mobilizing agents can cause ER fragmentation and can regulate lysosome-mediated ER autophagy (26 -28). Autophagic consumption of an ER subfraction would be expected to require both vesicularization of that subfraction and lysosomal activity. However, neither brefeldin A, which blocks vesicle and protein transfer from the ER (29), nor NH 4 Cl, which inhibits lysosomal proteases (30), inhibited the effects of carbachol in SH-SY5Y cells (Fig. 2).

Site of InsP 3 Receptor Degradation-While the data in
Whether or not InsP 3 receptors are redistributed prior to their degradation was investigated more directly by analyzing receptor immunofluorescence. In unstimulated SH-SY5Y cells (Fig. 3a), a complex pattern of expression was seen; relatively low InsP 3 receptor levels distributed evenly throughout the cytoplasm and higher concentrations in discrete areas at the cell margins. However, both the cytoplasmic and punctate staining fell at similar rates during stimulation with carbachol ( Fig. 3, b-d). Together with the data in  4). Cells were then harvested and probed with antisera against type I InsP 3 receptor, Ca 2ϩ -ATPase, PKC␣, or PKC⑀. Bands corresponding to these proteins were detected at ϳmolecular mass 260, 110, 85, and 92 kDa, respectively. b, AR4 -2J cells were maintained in serum-free culture medium for 18 h and were then incubated for 6 h without stimulus (lanes 1 and 2), with 0.5 M cholecystokinin (lanes 3 and 4), or with 15 g/ml cholesterol plus 1.5 g/ml 25-hydroxycholesterol (lanes 5 and 6). Cells were then harvested and probed with antisera against type I InsP 3 receptor or HMG-CoA reductase. Bands corresponding to these proteins were detected at ϳ260 and 97 kDa, respectively. Data from duplicate culture dishes are shown as the reduction in HMG-CoA reductase immunoreactivity was relatively minor. cells were preincubated with a range of protease inhibitors and were then stimulated with carbachol for 2 h (Fig. 4). The agents tested were the cysteine protease inhibitor N-acetyl-Leu-Leunorleucinal (ALLN, also known as calpain inhibitor I), the metalloprotease inhibitors phosphoramidon and N-carbobenzoxy-Gly-Phe, the aspartic protease inhibitor pepstatin, the serine protease inhibitors 4-amidinophenylmethanesulfonyl fluoride and 3,4-dichloroisocoumarin, and the serine/cysteine protease inhibitor leupeptin. While these agents do not readily cross the plasma membrane, previous studies have shown that ALLN (31), phosphoramidon (32), pepstatin (33), 3,4-dichloroisocoumarin (34), and leupeptin (35) can to an extent enter cells and inhibit proteolysis. Fig. 4 shows that the downregulatory effect of carbachol (compare lanes 1 and 2) was inhibited by ALLN (lanes 3-5), but not by the other agents (lanes 6 -12), and that ALLN alone had no effect on InsP 3 receptor immunoreactivity (lane 13). Thus, cysteine protease activity appears to degrade the type I InsP 3 receptor. To exclude the possibility that the inhibitory effect of ALLN resulted from nonspecific cytotoxicity, Ca 2ϩ mobilization was measured in fura-2 loaded control and ALLN-treated SH-SY5Y cells as described previously (36). Pretreatment with ALLN (100 g/ ml) for 4 h did not significantly alter carbachol-induced in-creases in cytosolic Ca 2ϩ concentration, 2 indicating that ALLN does not interfere with InsP 3 generation or its Ca 2ϩ -mobilizing potential. Thus, inhibition of down-regulation by ALLN is indeed a consequence of cysteine protease inhibition.
To define the protease activity we examined, in permeabilized cells, the effects of calpain inhibitor peptide, a 27-residue derivative of calpastatin (37), the endogenous inhibitor of calpains (11). Calpain inhibitor peptide specifically inhibits the activity of purified -calpain with half-maximal effect at ϳ0.1 M (37). Fig. 5 shows that Ca 2ϩ was effective in stimulating InsP 3 receptor degradation in digitonin-permeabilized cells only when cytosol was present (compare lanes 3 and 6) and that calpain inhibitor peptide inhibited this proteolysis with halfmaximal effect at 0.1-1 M (lanes 11-13). Thus, calpain activity is responsible for InsP 3 receptor degradation under these conditions. As ALLN also inhibited InsP 3 receptor proteolysis (lanes 7-9) with similar potency to that seen in intact cells (Fig.  4), it is likely that the inhibitory effect of ALLN in intact cells is also due to inhibition of calpain activity. It remains a possibility, however, that a cysteine protease activity other than calpain degrades the InsP 3 receptor in intact cells, since ALLN is also known to inhibit other cysteine proteases (38). DISCUSSION The finding that InsP 3 receptors are degraded specifically in stimulated intact cells (Fig. 1) immediately places constraints on possible mechanisms that might account for InsP 3 receptor down-regulation, since activation of a protease with freedom to degrade a broad range of substrates is clearly untenable. Moreover, as neither brefeldin A nor NH 4 Cl inhibited down-regulation and the type I InsP 3 receptor was not redistributed during cell stimulation (Figs. 2 and 3), it appears that InsP 3 receptors are degraded while occupying their normal intracellular sites (i.e. the membranes of intracellular Ca 2ϩ stores). These findings, together with evidence that InsP 3 binding initiates receptor degradation (7,8), restricts the plausible mechanisms that could account for down-regulation to two possibilities; either (i) a protease is stimulated specifically in the vicinity of activated InsP 3 receptors, or (ii) a conformational change in InsP 3 receptors during InsP 3 binding makes them susceptible to proteolysis. Support for the first possibility comes from the findings that Ca 2ϩ mobilization via InsP 3 receptors is required for accelera- (lanes 1 and 2), 50 mM NH 4 Cl (lanes 3 and 4) or 10 g/ml brefeldin A (lanes 5 and 6) and were then incubated for a further 20 h without stimulus (lanes 1, 3, and 5) or with 1 mM carbachol (lanes 2, 4,  and 6). Cells were then harvested and probed with antiserum against type I InsP 3 receptor.  lanes 11 and 12). Cells were then further incubated for 2 h without carbachol (ϪC, lanes 1 and 13) or with 1 mM carbachol (ϩC, lanes 2-12) and were then harvested and probed with antiserum against the type I InsP 3 receptor. tion of InsP 3 receptor degradation (10) and that certain cytosolic proteases, such as the calpains, are activated by Ca 2ϩ (11) and appear to mediate down-regulation (Figs. 4 and 5). Thus, the Ca 2ϩ mobilized during InsP 3 receptor activation may provide the signal that stimulates proteolysis.

FIG. 2. Effects of NH 4 Cl and brefeldin A on type I InsP 3 receptor down-regulation. SH-SY5Y cells were pretreated for 2 h with either vehicle
But, would Ca 2ϩ concentration be raised sufficiently to stimulate calpain activity and would such activation lead to selective degradation of InsP 3 receptors? The two known forms of calpain, -calpain and m-calpain, are activated in vitro by 1-100 M and 0.1-1 mM Ca 2ϩ , respectively, but in vivo may be an order of magnitude more sensitive to Ca 2ϩ due to the presence of co-factors, such as phosphatidylinositol 4,5-bisphosphate (11,39). Furthermore, as total and free Ca 2ϩ concentrations in the ER lumen are, respectively, ϳ2 mM and ϳ10 M or above (40,41), and InsP 3 receptors form high-conductance channels (40), concentrations of Ca 2ϩ greater that micromolar should be generated in the cytosol in the vicinity of channel pores. Indeed, experimental estimates of local Ca 2ϩ concentrations near sites of store mobilization are 5-6 M (42) and may be considerably higher in the immediate vicinity of channels. Such concentrations would be sufficient to elevate calpain activity (11). These values contrast with "global" increases in cytosolic Ca 2ϩ concentration which would be unlikely to activate calpains. For example, global cytosolic Ca 2ϩ concentration during the sustained phase of Ca 2ϩ mobilization in carbacholstimulated SH-SY5Y cells is only ϳ0.2 M (36).
With regard to the specificity of InsP 3 receptor degradation, Ca 2ϩ concentration declines rapidly with increasing distance from channel pores, because of the high buffering capacity of cytosol (40). Thus, Ca 2ϩ -dependent protease activation should be confined to the immediate vicinity of InsP 3 receptors. Additionally, if InsP 3 -induced conformational changes in the type I InsP 3 receptor (43) make the receptor more susceptible to proteolysis, such structural changes could contribute to the selectivity of the down-regulatory process.
In summary, the data presented show that persistent activation of PC-linked receptors leads to the selective proteolysis of InsP 3 receptors without changes in ER integrity or prior InsP 3 receptor redistribution, and that a cysteine protease activity, probablyand/or m-calpain, mediates this adaptive response. We suggest that Ca 2ϩ released via InsP 3 receptors generates local concentrations of Ca 2ϩ sufficient to stimulate calpain(s). Such a role for calpain(s) concurs with the known calpain-mediated degradation of other targets in response to activation of PC-linked receptors, namely PKC⑀ in pituitary cells in response to thyrotropin-releasing hormone (19), and PC-␤3 in platelets in response to thrombin and collagen (44).
As yet, however, the factors which govern that these proteins should be targets for calpain(s) are unknown. Perhaps the factor that unifies InsP 3 receptors, PKC⑀ and PC-␤3 is their localization to membranes; the latter two proteins become membrane-bound upon cell stimulation (19,44). Membranes may provide either a lipid co-factor that facilitates calpain activation (39) or, because of the presence of Ca 2ϩ channels, may be the only sites at which Ca 2ϩ concentration is sufficiently high to raise calpain activity.