Ubiquitination and proteasomal degradation of endogenous and exogenous inositol 1,4,5-trisphosphate receptors in alpha T3-1 anterior pituitary cells.

In alphaT3-1 mouse anterior pituitary gonadotropes, chronic activation of gonadotropin-releasing hormone (GnRH) receptors causes inositol 1,4,5-trisphosphate (InsP(3)) receptor down-regulation (Willars, G. B., Royall, J. E., Nahorski, S. R., El-Gehani, F., Everest, H. and McArdle, C. A. (2001) J. Biol. Chem. 276, 3123-3129). In the current study, we sought to define the mechanism behind this adaptive response. We show that GnRH induces a rapid and dramatic increase in InsP(3) receptor polyubiquitination and that proteasome inhibitors block InsP(3) receptor down-regulation and cause the accumulation of polyubiquitinated receptors. Thus, the ubiquitin/proteasome pathway is active in alphaT3-1 cells, and GnRH regulates the levels of InsP(3) receptors via this mechanism. Given these findings and further characterization of this system, we also examined the possibility that alphaT3-1 cells could be used to examine the ubiquitination of exogenous InsP(3) receptors introduced by cDNA transfection. This was found to be the case, since exogenous wild-type InsP(3) receptors, but not binding-defective mutant receptors, were polyubiquitinated in a GnRH-dependent manner, and agents that inhibited the polyubiquitination of endogenous receptors also inhibited the polyubiquitination of exogenous receptors. Further, we used this system to determine whether phosphorylation was involved in triggering InsP(3) receptor polyubiquitination. This was not the case, since mutation of serine residues 1588 and 1755 (the predominant phosphorylation sites in the type I receptor) did not inhibit polyubiquitination. In total, these data show that the ubiquitin/proteasome pathway is active in anterior pituitary cells, that this pathway targets both endogenous and exogenous InsP(3) receptors in GnRH-stimulated alphaT3-1 cells, and that, in contrast to the situation for many other substrates, phosphorylation does not trigger InsP(3) receptor polyubiquitination.

Hormone-induced secretion from anterior pituitary cells is modulated at many different levels, and among these is regulation of the activity and abundance of receptors involved in signal transduction (1,2). Indeed, recent studies on the ␣T3-1 mouse gonadotrope cell line have indicated that the suppression of secretion from gonadotropes in patients treated chronically with gonadotropin-releasing hormone (GnRH) 1 receptor agonists (2, 3) may result from a reduction in the expression of inositol 1,4,5-trisphosphate (InsP 3 ) receptors (2,4,5). InsP 3 receptors are a family of three proteins (termed type I, II, and III receptors) that form tetrameric ion channels in endoplasmic reticulum (ER) membranes, and upon binding of InsP 3 , the channels open, and Ca 2ϩ stored within the ER flows into the cytoplasm (6 -8). Thus, InsP 3 receptors play a pivotal role in linking G-protein-coupled receptor (GPCR)-mediated InsP 3 formation to increases in cytoplasmic free Ca 2ϩ concentration (9). A reduction in their expression (i.e. their down-regulation) would, therefore, be expected to suppress Ca 2ϩ mobilization (4,5,10) and secretion (2). InsP 3 receptor down-regulation in response to activation of certain GPCRs has also been seen in other cell types (11)(12)(13)(14)(15). This adaptive response is mediated by an increase in the rate of receptor degradation (11,13); is specific, since other ER and signaling proteins are not simultaneously affected (11,14); and appears to exist to allow chronically stimulated cells to reduce the sensitivity of their Ca 2ϩ stores to InsP 3 (4, 5, 10, 14 -16). The event that initiates receptor proteolysis appears to be InsP 3 binding, since only those GPCRs (e.g. GnRH, cholecystokinin, and muscarinic receptors) that persistently elevate InsP 3 levels cause InsP 3 receptor down-regulation (5,12,13), a binding-deficient mutant InsP 3 receptor is resistant to downregulation (17), and down-regulation in oocytes is elicited by microinjection of an InsP 3 analogue (18). Whether additional events (e.g. receptor phosphorylation) are required to trigger down-regulation remains to be resolved. Furthermore, it is not yet clear how InsP 3 receptors are degraded, and indeed, receptor proteolysis by calpain (12), caspase (19), and the ubiquitin/ proteasome pathway (14,15,20,21) have all been described.
Thus, we examined the mechanism of InsP 3 receptor downregulation in GnRH-stimulated ␣T3-1 cells and show that it occurs via the ubiquitin/proteasome pathway. In characterizing this adaptive response, we identified major differences in the properties of commonly used proteasome inhibitors and found that deubiquitination of InsP 3 receptors occurs rapidly and is * This work was supported by National Institutes of Health Grant 5RO1DK49194, American Heart Association Grant 0256225T, and the Pharmaceutical Research and Manufacturers of America Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

EXPERIMENTAL PROCEDURES
Materials-␣T3-1 cells were kindly provided by Dr. P. Mellon (University of California, San Diego, CA) and were cultured as monolayers in Falcon Integrid tissue culture dishes in Dulbecco's modified Eagle's medium supplemented with 100 units/ml penicillin, 100 g/ml streptomycin, and 10% fetal calf serum; cells were subcultured every 3-7 days using 0.25% trypsin, 1 mM EDTA. Rabbit polyclonal antisera CT1h and CT1w were raised against the C terminus of the rat type I receptor and were affinity-purified and shown to specifically recognize endogenous type I InsP 3 receptors (13). CT1h immunoprecipitated both endogenous receptors and exogenous epitope-tagged type I receptors and was used in all immunoprecipitations. Surprisingly, however, this antiserum did not recognize epitope-tagged receptors in immunoblots. Thus, CT1w was used to probe for type I InsP 3 receptor expression in transfected cells, since this antiserum recognized both endogenous receptors and exogenous epitope-tagged receptors. Mouse monoclonal anti-ubiquitin (FK2), which recognizes both mono-and polyubiquitinated proteins, was purchased from Affiniti Research Products Limited, anti-hemagglutinin (HA) epitope (HA11) was from Babco, and anti-c-Myc (9E10) was from Roche Molecular Biochemicals. Peroxidase-conjugated antibodies, molecular mass markers, SDS, Triton X-100, protease inhibitors, N,N,NЈ,NЈ-tetrakis(2-pyridylmethyl)ethylenediamine (TPEN), and receptor agonists were obtained from Sigma; Protein A-Sepharose CL-4B (Protein A) was from Amersham Biosciences; dithiothreitol was from Bio-Rad; lactacystin and MG-132 were from Biomol; N-acetyl-Leu-Leu-norleucinal (ALLN) and thapsigargin were from Alexis; and epoxomicin was a kind gift from Dr. C. Crews (Yale University, New Haven, CT).
Electrophoresis and Immunoblotting-Samples were subjected to 5% PAGE and were immunoblotted as described (20). Immunoreactivity was detected with chemiluminescence using reagents from Pierce and was digitally imaged and quantitated with a Genegnome (Syngene), working within the nonsaturating range.

Measurement of InsP 3 Receptor Ubiquitination by
Immunoblotting-Control or stimulated cells in suspension were collected by centrifugation (750 ϫ g for 3 min at 4°C) and were solubilized by incubation for 30 min at 4°C with 1 ml of lysis buffer. Lysates were then centrifuged (16,000 ϫ g for 10 min at 4°C), supernatants were collected, and InsP 3 receptors were immunoprecipitated by incubation at 4°C with CT1h for 1h and then for a further 12-24 h with Protein A. Immune complexes were then isolated by centrifugation (500 ϫ g for 2 min), were washed three times with ice-cold lysis buffer, and in most experiments were resuspended in 2ϫ gel loading buffer and then immunoblotted with either CT1h or FK2. In additional experiments aimed at further characterizing the ubiquitinated species, washed immune complexes were released from Protein A and denatured by incubation at 100°C for 5 min in 100 l of 50 mM Tris, 2% SDS, 2 mM dithiothreitol, pH 7.4, were centrifuged (16,000 ϫ g for 1 min at 25°C), were diluted to 4 ml with lysis buffer, were pretreated with Protein A to remove residual CT1h, and finally were reimmunoprecipitated with FK2 and Protein A for 12-24 h and resuspended in 2ϫ gel loading buffer.
Measurement of InsP 3 Receptor Ubiquitination by Radiolabeling-Cell monolayers in 75-cm 2 Falcon flasks were incubated for 48 h with 100 Ci of [ 35 S]cysteine (NEG022T; PerkinElmer Life Sciences) in ␣T3-1 cell culture medium supplemented with sufficient nonradioactive cysteine (200 M) to allow for normal cell growth. Cells were then preincubated for 1 h with ALLN (20 g/ml), stimulated for 1 h with GnRH (2 M), and harvested in lysis buffer, and type I InsP 3 receptors were immunoprecipitated with CT1h and electrophoresed as described. Gels were stained with Coomassie Blue, and regions corresponding to unmodified and polyubiquitinated InsP 3 receptors were excised, homogenized, and assessed for radioactivity in 4 ml of scintillation fluid. Because ubiquitin does not contain cysteine, it does not become radiolabeled; thus, the percentage of receptors polyubiquitinated can be calculated from the amount of radioactivity migrating in the region corresponding to polyubiquitinated receptors relative to total receptor radioactivity.
Transfection of ␣T3-1 Cells-Cells were harvested using 0.25% trypsin/1 mM EDTA, were seeded into six-well Falcon plates at a density of 2 ϫ 10 6 /well, and were transfected 24 h later by adding 1 ml of fresh culture medium containing a complex of cDNA and 9 l of Superfect (Qiagen), prepared according to the manufacturer's instructions. The cDNAs used were as follows: pCW7, which encodes His 6 , c-Myc epitopetagged yeast ubiquitin (Myc-ubiquitin) and was a kind gift from Dr. R. R. Kopito (Stanford University) (22); pcDNA3 (empty vector); pc-WIHA (17), which encodes wild-type mouse type I InsP 3 receptor tagged at the C terminus with an HA epitope (InsP 3 RHA wt ); pcWIHA⌬ (17), which encodes a binding-defective, HA-tagged mutant mouse type I receptor (InsP 3 RHA⌬) that lacks residues 316 -352; and pcWIHA A/A , which encodes an HA-tagged receptor (InsP 3 RHA A/A ) with serine 3 alanine mutations at positions 1588 and 1755. This mutant was created using the QuikChange TM kit (Stratagene). In brief, pcWIHA was first mutated to introduce alanine at position 1755 using primer pair 5Ј-G-GAAGAAGAGAGGCGCTTACCAGCTTTGG-3Ј and 5Ј-CCAAAGCTGG-TAAGCGCCTCTCTTCTTCC-3Ј. This mutant was then further mutated to introduce alanine at position 1588 using primer pair 5Ј-CGCA-GAGACGCTGTACTGGCAGCTAGCAGAGACTAC-3Ј and 5Ј-GTAGTC-TCTGCTAGCTGCCAGTACAGCGTCTCTGCG-3Ј. The first and second primer pairs also introduced HaeII and NheI sites, respectively, to facilitate screening. The correct introduction of the desired mutations into the polymerase-generated region was confirmed by sequencing. 48 h after transfection, the cells were exposed to stimuli or inhibitors and were harvested and solubilized by incubation for 30 min at 4°C in 1 ml of lysis buffer. After centrifugation (16,000 ϫ g for 10 min at 4°C), type I InsP 3 receptors were immunoprecipitated with CT1h (to purify endogenous and exogenous receptors) or HA11 (to purify exogenous HA-tagged receptors only), and immunoprecipitates were immunoblotted with HA11, 9E10, CT1h, or CT1w.
Measurement of InsP 3 Concentration-InsP 3 concentration in suspensions of ␣T3-1 cells incubated at 37°C was measured with a radioreceptor assay exactly as described (20).
Miscellaneous-Data shown are representative of at least two independent experiments. Combined data are mean Ϯ S.E. (n Ն 3) or mean range (n ϭ 2).

InsP 3 Receptor Down-regulation in ␣T3-1 Cells-Initial measurements of type I InsP 3 receptor levels in lysates from
GnRH-stimulated cells showed that receptor down-regulation in response to GnRH (0.1 M) was half-maximal at 15 Ϯ 2 min (Fig. 1A). This is considerably more rapid than that seen in other cell types (11)(12)(13)(14)(15), most likely because GnRH receptors are refractory to desensitization and thus elevate InsP 3 concentration profoundly and persistently (2,4).
To confirm that persistent GnRH receptor activation and InsP 3 formation were needed for type I InsP 3 receptor downregulation, we utilized the GnRH receptor antagonist antide, which blocks GnRH-induced InsP 3 formation when added simultaneously with or after GnRH (Fig. 3B). As expected, antide blocked down-regulation when added simultaneously with GnRH (Fig. 1B, lane 3). However, we also observed that the down-regulation seen after a 60-min exposure to GnRH (lane 2) was not mimicked by exposure to GnRH alone for 5 min, followed by a further 55-min incubation in antide-supplemented medium (lane 5). This shows that acute GnRH receptor activation is not sufficient to program the cells to subsequently downregulate InsP 3 receptors and is consistent with the view (13,20) that persistent elevation of InsP 3 concentration is a prerequisite for down-regulation.
Proteasome Inhibitors Block Down-regulation and Cause the Accumulation of Ubiquitinated Receptors-In order to determine whether or not GnRH-induced InsP 3 receptor down-regulation is via the ubiquitin/proteasome pathway, we exposed ␣T3-1 cells to GnRH in the absence or presence of a range of proteasome inhibitors and monitored type I InsP 3 receptor levels and associated ubiquitin immunoreactivity. Three of these inhibitors, ALLN, MG-132, and lactacystin, are widely employed, the first two being peptides that are reversibly acting transition state analogues and the latter being a structurally different pseudosubstrate that covalently modifies the active site (23). The remaining inhibitor used, epoxomicin, is a novel, highly potent, irreversible inhibitor (24). In the absence of inhibitor, incubation with GnRH for 1 h caused InsP 3 receptor down-regulation ( Fig. 2A, lane 2, lower panel) but did not cause the accumulation of ubiquitinated species (Fig. 2A, lane 2, upper panel). In contrast, when the cells were preincubated with inhibitors for 2 h, GnRH-induced InsP 3 receptor downregulation was blocked, and a parallel increase in the level of ubiquitin immunoreactivity associated with InsP 3 receptors was observed ( Fig. 2A, lanes 3-10). Whereas they exhibited different potencies (see legend to Fig. 2A), the four inhibitors were equally efficacious in causing the accumulation of ubiquitinated species and completely blocked down-regulation at maximal concentration (Fig. 2, A and E, lanes 2, 4, 6, and 8).
Control experiments (e.g. Fig. 2D, lane 3) showed that the inhibitors alone did not increase basal InsP 3 receptor levels or cause the accumulation of ubiquitinated species. Additional controls showed that the ubiquitinated species were indeed modified type I receptors, since when purified, they were clearly immunoreactive with type I receptor antiserum (Fig.  2B, lower panel, lane 2). The inability of the same antiserum to detect ubiquitinated receptors in crude type I receptor immunoprecipitates ( Fig. 2A, lower panel) is most likely explained by the low abundance of ubiquitinated receptors relative to unmodified receptors. To address this issue, the proportion of type I InsP 3 receptors ubiquitinated was defined in experiments in which cells were radiolabeled with [ 35 S]cysteine, and in maxi-mally stimulated cells it was found to be 9 Ϯ 1% of total (n ϭ 3). The ubiquitinated receptors migrated as a "smear" (ϳ275-380 kDa) ( Fig. 2A, upper panel, and Fig. 2B) slightly less rapidly than unmodified type I receptor (ϳ260 kDa) ( Fig. 2A, lower panel), indicative of the formation of a spectrum of polyubiquitinated species and typical of the migration of other polyubiquitinated proteins (25,26). In total, the finding that the four structurally and mechanistically different proteasome inhibitors all have the same effect shows that the ubiquitin/proteasome pathway mediates InsP 3 receptor down-regulation in ␣T3-1 cells. This conclusion is supported by findings that specific inhibitors of other candidate proteolytic pathways (20 M benzyloxycarbonyl-Asp-Glu-Val-Aspfluoromethyl ketone, a caspase inhibitor, and 20 M PD150606, a calpain inhibitor) did not block GnRH-induced InsP 3 receptor down-regulation. 2 Consistent with this conclusion and the rapidity of downregulation (Fig. 1A), analysis of the time dependence of polyubiquitination (Fig. 2C) revealed that in the absence of ALLN, polyubiquitinated receptors accumulated very rapidly (peaking at 5 min) and were detectable only transiently, presumably because they are degraded rapidly by the proteasome; this also explains why polyubiquitinated receptors were not detected after incubation with GnRH alone for 1 h (Fig. 2A, lane 2). In contrast, when ALLN was present, polyubiquitinated receptor accumulation peaked at ϳ20 min and thereafter did not decline (Fig. 2C). Surprisingly, ALLN also suppressed the initial rate of receptor polyubiquitination (Fig. 2C). This was not due to a reduction in the potency of GnRH, which was half-maximally effective at ϳ5 nM in the absence or presence of ALLN, 2 and indicates that as well as inhibiting the degradation of polyubiquitinated species, proteasome inhibitors may also reduce the rate of polyubiquitination.
Given the mechanistic differences between the proteasome inhibitors, we also analyzed their kinetics. Fig. 2D shows that the effects of ALLN are very rapid in onset; ALLN was maximally effective with a preincubation time of 1 h or more (lanes [5][6][7][8] and was close to being maximally effective when added simultaneously with GnRH (lane 4). Fig. 2E shows that when used at maximally effective concentrations (5-10 times higher than half-maximal values defined in the legend to Fig. 2A), MG-132 (lanes 4 and 5), like ALLN (lanes 2 and 3), acted rapidly, being similarly effective with 0-or 2-h preincubation. In contrast, epoxomicin (lanes 8 and 9) and particularly lactacystin (lanes 6 and 7) were slower acting, being much less effective when added simultaneously with GnRH as compared with 2-h preincubation.
InsP 3 Receptor Deubiquitination-Since proteasome inhibitors completely block GnRH-induced InsP 3 receptor downregulation, it would be expected that a large proportion of cellular InsP 3 receptors would accumulate as polyubiquitinated species when the proteasome was inhibited. However, this was not the case, since only 9 Ϯ 1% of receptors were polyubiquitinated in the presence of ALLN plus GnRH, and maximal accumulation of polyubiquitinated receptors in the presence of ALLN was only approximately twice that seen in its absence (Fig. 2C). Thus, we examined whether deubiquitination might be counteracting the accumulation of polyubiquitinated receptors. Antide was used for these studies, since it blocks InsP 3 formation and InsP 3 receptor polyubiquitination when added simultaneously with GnRH (Fig. 3B, left  panel, and Fig. 3A, lane 6) and rapidly (within 10 min) returns InsP 3 concentration to basal levels when added to GnRH-stimulated cells (Fig. 3B, right). shows that the addition of antide to ALLN-preincubated, GnRH-stimulated cells results in a rapid decline in the level of polyubiquitinated InsP 3 receptors, indicating that they are being deubiquitinated. Thus, deubiquitinating enzymes (27,28) are active in ␣T3-1 cells and appear to participate in suppressing the build-up of polyubiquitinated InsP 3 receptors.
Thapsigargin, TPEN, and Glycerol Inhibit Polyubiquitination and Down-regulation-The effects of potential inhibitors that might provide insight into the mechanism of polyubiquitination and down-regulation were also tested (Fig. 4). Thapsigargin inhibits Ca 2ϩ -ATPases that pump Ca 2ϩ into the ER, reduces intraluminal Ca 2ϩ concentration, and disrupts ER function (29). Fig. 4 (A, lane 4, and B, lane 2) shows that thapsigargin inhibits GnRH-induced InsP 3 receptor down-regulation and polyubiquitination without affecting InsP 3 formation (Fig. 4C), suggesting that Ca 2ϩ binding to intraluminal regions of the type I InsP 3 receptor (6 -8) or to other ER proteins that interact with the type I InsP 3 receptor (29 -31) is required for this process. TPEN chelates Zn 2ϩ with high affinity and has been shown to inhibit the activity of purified RING domain-containing E3 ubiquitin-protein ligases, presumably by removing the Zn 2ϩ that is normally complexed with the RING domain (28,(32)(33)(34). Fig. 4 (A, lane 6, and B, lane 3) shows that TPEN inhibits InsP 3 receptor down-regulation and polyubiquitination, and Fig. 4C shows that this is not due to inhibition of InsP 3 formation. 3 These data suggest that a RING

FIG. 2. Proteasome inhibitors block InsP 3 receptor down-regulation and cause the accumulation of polyubiquitinated receptors.
␣T3-1 cells in suspension were incubated with or without 0.1 M GnRH and proteasome inhibitors and were harvested, lysates were prepared, InsP 3 receptors were immunoprecipitated with CT1h, and ubiquitin and type I receptor immunoreactivity was assessed in immunoblots with FK2 (upper panels) and CT1h (lower panels), respectively. The 220 -420-kDa regions of gels are shown, and the arrows and brackets mark the respective positions of unmodified type I receptor (ϳ260 kDa) and polyubiquitinated type I receptor (ϳ275-380 kDa). A, dose dependence of proteasome inhibitor effects. Cells were preincubated for 2 h with ALLN (lanes 3-6) or MG-132 (lanes 7-10) at the concentrations indicated and were then exposed to GnRH for 1 h (lanes 2-10). Halfmaximal accumulation of polyubiquitinated species and inhibition of down-regulation occurred at 4 g/ml ALLN, 0.3 g/ml MG-132, 0.5 M lactacystin, and 0.04 M epoxomicin (mean, n Ն 2 independent experiments). B, immunoreactivity of purified polyubiquitinated species. Extracts of CT1h-derived immunoprecipitates from control cells (lane 1) or 20 g/ml ALLN-pretreated, GnRH-stimulated cells (lane 2) were reimmunoprecipitated with FK2 to purify ubiquitinated species and were then probed with FK2 or CT1h. C, time course of GnRH-induced type I InsP 3 receptor polyubiquitination. Cells were preincubated without or with 20 g/ml ALLN for 2 h and were then incubated with GnRH for 0 -60 min. The ubiquitin immunoreactivity of immunoprecipitated receptors was then assessed and quantitated (mean Ϯ S.E., n Ն 3). D, time dependence of ALLN effects. Cells were preincubated for 0 -4 h with 20 g/ml ALLN (lanes 3-8) and were then exposed to GnRH for 1 h (lanes 2 and 4 -8). E, time dependence of proteasome inhibitor effects on ubiquitination. Cells were preincubated for 2 or 0 h with 20 g/ml ALLN (lanes 2 and 3), 2 g/ml MG132 (lanes 4 and 5), 3 M lactacystin (lanes 6 and 7), or 0.3 M epoxomicin (lanes 8 and 9) and were then exposed to GnRH for 1 h.  1-5). Ubiquitin immunoreactivity associated with immunoprecipitated type I InsP 3 receptors was then assessed as in Fig. 2 domain-containing E3 mediates InsP 3 receptor ubiquitination. Finally, glycerol has been proposed to act as a "chemical chaperone," acting to enhance the proper folding and suppress the degradation of either misfolded ER-associated proteins or proteins destined for ER-associated degradation (35,36). Glycerol did inhibit down-regulation and polyubiquitination of InsP 3 receptors (Fig. 4, A, lane 8, and B, lane 4). However, it also completely inhibited InsP 3 formation (Fig. 4C), making it impossible to draw conclusions related to its action as a chaperone. (Figs. 1-4), we next examined whether exogenous receptors, introduced by transient transfection, could be polyubiquitinated in a GnRH-dependent manner, since this would provide a system for the analysis of mutant receptors. Pilot studies utilizing cDNA encoding green fluorescent protein and a variety of transfection reagents revealed that 5-10% of ␣T3-1 cells could be transfected and that expression of exogenous wild-type type I receptor was insufficient to reproducibly increase the InsP 3 receptor or polyubiquitin content of CT1h-derived immunoprecipitates probed as in Figs. 2-4. 2 Thus, we sought to selectively measure exogenous InsP 3 receptor ubiquitination in the limited population of transfectable cells, by coexpressing Myc-ubiquitin and HA-tagged InsP 3 receptors. Fig. 5A illustrates the feasibility of this approach, since when Myc-ubiquitin was expressed alone, GnRH-induced Mycpolyubiquitination of endogenous receptors (upper panel, lanes 1-3) paralleled the profile of endogenous ubiquitin incorporation into type I receptors (Fig. 2C), indicating that Myc-ubiquitin is incorporated into polymeric chains capable of mediating proteasomal degradation. Further, co-expression of InsP 3 RHA wt (lanes 4 -6) at levels insufficient to increase the total InsP 3 receptor content of CT1h immunopre- cipitates (lower panel, compare lanes 1-3 with lanes 4 -6) considerably enhanced Myc-polyubiquitination (upper panel,  lanes 4 -6), indicating that exogenous receptors are effi-ciently polyubiquitinated. It is noteworthy that whereas endogenous receptors were down-regulated by GnRH (lower panel), the exogenous HA-tagged receptors were not (middle panel, lanes 4 -6). This is consistent with other studies showing that the HA tag inhibits InsP 3 receptor down-regulation (17).

GnRH-induced Ubiquitination of Exogenous InsP 3 Receptors in Transfected Cells-Having characterized the polyubiquitination of endogenous receptors
In order to assess the ubiquitination of just the exogenous receptors, we immunoprecipitated with HA11 (Fig. 5B), which purifies only HA-tagged receptors (see Fig. 5C). Initial analysis of InsP 3 RHA wt revealed, somewhat surprisingly, the existence of a Myc-immunoreactive band very similar in size to unmodified InsP 3 RHA wt (Fig. 5B, upper panel, lane 4), even in the absence of GnRH. This is likely to be InsP 3 RHA wt modified by one or a very small number of Myc-ubiquitin residues, which hereafter is referred to as "monoubiquitinated" receptor. Importantly, however, exposure of these cells to GnRH led to a large increase in Myc-polyubiquitination (upper panel, lanes 5 and 6), confirming that exogenous InsP 3 RHA wt is polyubiquitinated. In contrast, in cells expressing InsP 3 RHA ⌬ , which does not bind InsP 3 (17), the level of Myc-ubiquitination was unaffected by GnRH (upper panel, lanes 7-9), indicating that InsP 3 RHA ⌬ is not polyubiquitinated. Parallel analysis of HAtagged receptor content (lower panel) showed that both InsP 3 RHA wt and InsP 3 RHA ⌬ were expressed and that InsP 3 RHA ⌬ (ϳ265 kDa) migrated slightly more rapidly than InsP 3 RHA wt (ϳ270 kDa). To demonstrate that only the HAtagged receptors were purified and that endogenous receptors did not co-immunoprecipitate, we probed the HA11-derived immunoprecipitates with CT1w, which recognizes both HAtagged and endogenous receptors in immunoblots, and with CT1h, which recognizes only endogenous receptors in immunoblots. Fig. 5C (upper panel, lane 2) shows that InsP 3 RHA wt migrates more slowly than endogenous receptor (lane 5), as previously described (17), and that no endogenous receptor was co-immunoprecipitated. This was confirmed by the observation that CT1h (lower panel) did not immunoreact with lane 2. Likewise, CT1h did not immunoreact with immunoprecipitated  Fig. 1. B, inhibition of polyubiquitination. Cells preincubated with 20 g/ml ALLN for 2 h were incubated for 20 min, and ubiquitin immunoreactivity was assessed as in Fig  2. C, effects on InsP 3 formation. Cells were incubated for 20 min, and InsP 3 concentration was assessed as in Fig. 3. Data shown are mean Ϯ S.E. of triplicate samples (*, p Ͻ 0.02).
do not co-immunoprecipitate with the exogenous receptors. 4 Importantly, these data rule out the possibility that endogenous receptors might contribute to the ubiquitination seen in Fig. 5B. Finally, Myc-ubiquitin immunoreactivity in lysates, which was predominantly in high molecular mass conjugates, was the same in cells expressing InsP 3 RHA wt and InsP 3 RHA ⌬ , 2 indicating that the lack of Myc-polyubiquitination in the latter was indeed due to the deletion in the InsP 3 receptor. Thus, polyubiquitination of transiently expressed exogenous receptors is mediated by InsP 3 binding, indicating that they interact appropriately with InsP 3 and are subject to the same regulatory processes as stably expressed receptors (37). Furthermore, their processing paralleled that of endogenous receptors, since GnRH-induced Myc-polyubiquitination of InsP 3 RHA wt was inhibited by thapsigargin and antide (Fig. 5D).
However, it must be noted that the processing of exogenous and endogenous receptors was not identical, since exogenous receptors were "monoubiquitinated." This was not dependent on GnRH stimulation or InsP 3 binding, since both InsP 3 RHA wt and InsP 3 RHA ⌬ were modified (Fig. 5B), was not blocked by thapsigargin or antide, and, significantly, became more prominent at InsP 3 receptor cDNA levels Ͼ0.1 g, 2 indicating that it may result from the overexpression of exogenous protein. Nevertheless, by expressing relatively low amounts of exogenous InsP 3 receptor, the contribution of monoubiquitination to the overall ubiquitination signal could be minimized, and it was possible to use transient receptor expression to probe the events that trigger polyubiquitination.
The Role of Phosphorylation in Ubiquitination-Phosphorylation has been shown to trigger the polyubiquitination of many proteins (27,28) and could contribute to triggering InsP 3 receptor polyubiquitination, since the type I InsP 3 receptor is phosphorylated by protein kinase A (PKA) (6 -8) under conditions that lead to down-regulation (38). PKA-mediated phosphorylation of the mouse type I receptor occurs at serine residues 1588 and 1755 (39); thus, we created and analyzed a phosphorylation-resistant mutant receptor (InsP 3 RHA A/A ) in which both sites are converted to alanine. Fig. 5B (lanes 10 -12) shows that exogenous InsP 3 RHA A/A was Myc-polyubiquitinated in an identical manner to InsP 3 RHA wt (lanes 4 -6), indicating that PKA-dependent phosphorylation does not contribute to the triggering of InsP 3 receptor polyubiquitination.

DISCUSSION
In summary, the data presented show that GnRH-induced InsP 3 receptor down-regulation in ␣T3-1 cells is mediated by the ubiquitin/proteasome pathway, that transiently expressed exogenous InsP 3 receptors are polyubiquitinated similarly to endogenous receptors, and that InsP 3 binding, but not PKAmediated InsP 3 receptor phosphorylation, is a key event in the process that leads to polyubiquitination. In addition, we demonstrate that deubiquitination limits the accumulation of polyubiquitinated InsP 3 receptors and that both thapsigargin and TPEN inhibit polyubiquitination.
Importantly, to the best of our knowledge, this study represents the first analysis of the ubiquitin/proteasome pathway in anterior pituitary cells and the first demonstration that a hypothalamic releasing factor, such as GnRH, can utilize this pathway to regulate protein levels. Indeed, ubiquitin/proteasome pathway-mediated InsP 3 receptor down-regulation is likely to contribute to the mechanism by which long term administration of GnRH and its analogues to patients suppresses luteinizing hormone/follicle-stimulating hormone secretion and produces a hypogonadal state (2,3). Further, these data raise the possibility that other proteins, perhaps those involved signal transduction (26,40,41), might also be targeted by the ubiquitin/proteasome pathway in anterior pituitary cells upon GPCR activation.
These studies also show that it is the ubiquitin/proteasome pathway, and not other candidate proteolytic systems (12,19), that accounts for GnRH-mediated InsP 3 receptor down-regulation. Thus, in response to GnRH receptor activation, InsP 3 receptors are targeted by members of the E2/E3 enzyme family FIG. 5. Ubiquitination of exogenous InsP 3 receptors. ␣T3-1 cell monolayers were transfected with 0.5 g of pCW7 (encoding Mycubiquitin) and 0.05 g of either pcDNA3 (empty vector), pcWIHA (encoding InsP 3 RHA wt ), pcWIHA ⌬ (encoding InsP 3 RHA ⌬ ), or pcWIHA A/A (encoding InsP 3 RHA A/A ) and were incubated with 0.1 M GnRH under the conditions indicated. Cells were then harvested, lysates were prepared, InsP 3 receptors were immunoprecipitated (IP) with CT1h or HA11, and immunoblots (IB) were probed for Myc-ubiquitin immunoreactivity with 9E10 or for InsP 3 receptor immunoreactivity with HA11, CT1w, or CT1h, as indicated. The 220 -420-kDa regions of gels are shown, and the migration positions of endogenous receptor (ϳ260 kDa; arrows), unmodified or monoubiquitinated HA-tagged exogenous receptors (ϳ265 or ϳ270 kDa; arrowheads), and Myc-polyubiquitinated receptor (ϳ275-380 kDa; brackets) are indicated. A and B, Myc-ubiquitin and InsP 3 receptor immunoreactivity of immunoprecipitates from GnRH-stimulated transfected cells. C, InsP 3 receptor immunoreactivity of immunoprecipitates from unstimulated transfected cells (lanes 1-4) and of a sample of endogenous ␣T3-1 cell type I receptor (lane 5). D, inhibitory effects of 1 M thapsigargin and 3 M antide on Myc-polyubiquitination in cells co-transfected with pCW7 and pcWIHA and stimulated with GnRH for 60 min. that conjugate ubiquitin to proteins (27,28). In analyzing this response in ␣T3-1 cells, we have extended our understanding of InsP 3 receptor polyubiquitination and the ubiquitin/proteasome pathway in general in several ways. First, our data show that the accumulation of polyubiquitinated InsP 3 receptors in the presence of proteasome inhibitors is countered by deubiquitination. Currently, virtually all work on deubiquitination has been done on purified proteins or disrupted cells (27,42), and very little is known about the role of this activity in intact cells, apart from a recent study showing that a novel enzyme specifically deubiquitinates and stabilizes p53 (43). Whereas the nature of the activity that deubiquitinates InsP 3 receptors in intact cells was not examined, its existence explains why the accumulation of polyubiquitinated species in proteasome inhibitor-treated cells is relatively minor, amounting to only 9 Ϯ 1% of the total receptor complement. It also suggests that in cells not exposed to proteasome inhibitors, polyubiquitinated InsP 3 receptors will be subject to the competing effects of deubiquitinating enzymes (causing stabilization) and the proteasome (causing degradation). These and other findings (43) raise the possibility that this situation is the norm for all polyubiquitinated proteins. Second, with regard to the mechanism of ubiquitination, the ability of TPEN to inhibit InsP 3 receptor polyubiquitination implicates a RING domain-containing E3 in this process, perhaps one of the recently identified E3s involved in the degradation of ER proteins (36,44). Mechanistic insight was also obtained using thapsigargin, which depletes ER Ca 2ϩ (29,30) and completely inhibited InsP 3 receptor polyubiquitination. This indicates that intraluminal Ca 2ϩ plays a role in InsP 3 receptor polyubiquitination. Intriguingly, ubiquitin/proteasome pathway-mediated processing of other ER proteins is also inhibited by depletion of ER Ca 2ϩ (45,46). Thus, InsP 3 receptor polyubiquitination appears to be via a pathway common to all ER proteins targeted via the ubiquitin/proteasome pathway that is dependent on the normal storage of Ca 2ϩ in the ER and perhaps on Ca 2ϩ binding to one or more of the many Ca 2ϩ -binding ER proteins (29 -31). Finally, comparison of the effects of different proteasome inhibitors showed that they varied considerably in their rate of action, with ALLN and MG-132 acting much more rapidly than lactacystin and epoxomicin. This kinetic variation most likely reflects the mechanistic differences between the inhibitors (23,24) and clearly should be taken into account when these inhibitors are used.
We have also been able to show that exogenous transiently expressed receptors are polyubiquitinated in response to GPCR activation. This is significant, because in order to study the triggering and site specificity of polyubiquitination, it will be necessary to assess a large number of mutant receptors, which can be most easily accomplished by expressing them transiently. In summary, we found that exogenous InsP 3 RHA wt was processed in response to GnRH receptor activation similarly to endogenous InsP 3 receptor but with the exception that it was also constitutively "monoubiquitinated" (modified with one or a very small number of Myc-ubiquitin residues). Significantly, whereas InsP 3 RHA wt was both poly-and monoubiquitinated, InsP 3 RHA ⌬ , which does not bind InsP 3 (17), was only monoubiquitinated, indicating that only polyubiquitination occurs in response to InsP 3 binding. Furthermore, InsP 3 RHA wt polyubiquitination, but not monoubiquitination, was blocked by thapsigargin and antide. Thus, monoubiquitination appears to be a process mechanistically discrete from that which mediates polyubiquitination and may be a response to the overexpression of exogenous receptors, a view supported by the observation that the prominence of monoubiquitination increased as exogenous receptor expression increased. Thus, it is possible that the capacity of ␣T3-1 cells to accommodate and correctly process transiently expressed exogenous InsP 3 receptors is relatively limited, and above that capacity, the receptors are monoubiquitinated. The fact that exogenous receptors did not associate to a detectable extent with endogenous receptors lends credence to this view. An alternative explanation is that InsP 3 receptor monoubiquitination is a normal cellular event that has been revealed because of the high sensitivity of the Myc epitope antibody. In this regard, it has recently been shown that other receptors and their associated proteins can be monoubiquitinated (40,48) as a prelude to their trafficking to lysosomes. It is an intriguing possibility that InsP 3 receptors could be similarly processed.
With regard to events that trigger type I InsP 3 receptor polyubiquitination, the analysis of InsP 3 RHA ⌬ shows that InsP 3 binding, and presumably the conformational changes that result from this binding (49), cause the receptor to become polyubiquitinated. The possibility that phosphorylation might be involved in triggering polyubiquitination was also examined, since the type I InsP 3 receptor is phosphorylated stoichiometrically in response to PKA activation in intact cells (6 -8) and, indeed, in response to activation of the G q -linked GPCRs that lead to InsP 3 receptor down-regulation (38). However, PKA-dependent phosphorylation of the receptor was clearly not required for polyubiquitination, since InsP 3 RHA A/A was polyubiquitinated equivalently to InsP 3 RHA wt .
In conclusion, our studies show that the ubiquitin/proteasome pathway is active in ␣T3-1 anterior pituitary cells and mediates InsP 3 receptor degradation in response to activation of GnRH receptors. Since transiently expressed exogenous receptors are polyubiquitinated similarly to endogenous receptors, use of this cell type will allow for the analysis of a range of mutant receptors and the dissection of the molecular events that lead to InsP 3 receptor polyubiquitination.