Rapid down-regulation of the type I inositol 1,4,5-trisphosphate receptor and desensitization of gonadotropin-releasing hormone-mediated Ca2+ responses in alpha T3-1 gonadotropes.

Despite no evidence for desensitization of phospholipase C-coupled gonadotropin-releasing hormone (GnRH) receptors, we previously reported marked suppression of GnRH-mediated Ca(2+) responses in alphaT3-1 cells by pre-exposure to GnRH. This suppression could not be accounted for solely by reduced inositol 1,4,5-trisphosphate (Ins(1,4,5)P(3)) responses, thereby implicating uncoupling of Ins(1,4,5)P(3) production and Ca(2+) mobilization (McArdle, C. A., Willars, G. B., Fowkes, R. C., Nahorski, S. R., Davidson, J. S., and Forrest-Owen, W. (1996) J. Biol. Chem. 271, 23711-23717). In the current study we demonstrate that GnRH causes a homologous and heterologous desensitization of Ca(2+) signaling in alphaT3-1 cells that is coincident with a rapid (t((12)) < 20 min), marked, and functionally relevant loss of type I Ins(1,4,5)P(3) receptor immunoreactivity and binding. Furthermore, using an alphaT3-1 cell line expressing recombinant muscarinic M(3) receptors we show that the unique resistance of the GnRH receptor to rapid desensitization contributes to a fast, profound, and sustained loss of Ins(1,4,5)P(3) receptor immunoreactivity. These data highlight a potential role for rapid Ins(1,4,5)P(3) receptor down-regulation in homologous and heterologous desensitization and in particular suggest that this mechanism may contribute to the suppression of the reproductive system that is exploited in the major clinical applications of GnRH analogues.

The decapeptide gonadotropin-releasing hormone (GnRH) 1 is released from the hypothalamus of mammals in a pulsatile manner to regulate the exocytotic release of luteinizing hormone and follicle-stimulating hormone from pituitary gonadotropes. These hormones are central to the regulation of gonadal steroidogenesis and gamete maturation, and GnRH therefore plays a vital role in the control of vertebrate reproduction. GnRH acts on pituitary gonadotropes through a G-proteincoupled receptor that regulates phospholipase C (PLC) via Gproteins of the G␣ q/11 family (1). GnRH-mediated activation of PLC results in the hydrolysis of phosphatidylinositol 4,5-bisphosphate to generate both inositol 1,4,5-trisphosphate (Ins(1,4,5)P 3 ) and diacylglycerol. These second messengers are able to mobilize Ca 2ϩ from intracellular stores and activate protein kinase C, respectively (2), thereby propagating a signaling cascade that accounts for the biological effects of GnRH.
In contrast to nearly all other known PLC-coupled G-proteincoupled receptors (GPCRs), the GnRH receptor does not undergo rapid (seconds to minutes) desensitization following exposure to agonist (3)(4)(5)(6)(7)(8). Recent evidence has suggested that this lack of acute regulation is related to the lack of a C-terminal tail and the absence, therefore, of appropriate regulatory phospho-acceptor sites (8,9). From a functional perspective, the resistance of the GnRH receptor to rapid desensitization may serve to maintain cellular sensitivity and responsiveness during events such as the pre-ovulatory gonadotropin hormone surge and allow the frequency-encoded pattern of the hypothalamic pulsatile GnRH release (10,11) to be faithfully maintained at the level of the pituitary gonadotropes.
Despite the lack of acute regulation of the GnRH receptor, sustained exposure to GnRH is able to reduce GnRH-stimulated gonadotropin secretion, and this form of desensitization underlies the suppression of the reproductive system that is exploited in the major clinical applications of GnRH analogues (12). Given the importance of cytosolic Ca 2ϩ elevation in the mediation of GnRH-stimulated gonadotropin secretion (1,(13)(14)(15), we have previously explored the potential desensitization of this component of the GnRH receptor-mediated signaling pathway in an immortalized mouse pituitary cell line (␣T3-1). Despite no evidence for rapid desensitization of the GnRH receptor in these cells, pre-exposure to GnRH can cause a marked suppression of subsequent GnRH-mediated elevations of [Ca 2ϩ ] i . Both the spike phase of the response (which reflects Ins(1,4,5)P 3 -dependent mobilization of intracellular Ca 2ϩ ) and the sustained phase of the response (which is dependent upon Ca 2ϩ entry across the plasma membrane through voltage-operated Ca 2ϩ channels) were attenuated by GnRH pretreatment (4,5).
Desensitization of voltage-operated Ca 2ϩ channels may account for the desensitization of the plateau phase of the GnRHmediated response in ␣T3-1 cells (4), but the mechanism underlying attenuated mobilization of Ca 2ϩ from intracellular stores is unclear. Although pre-exposure to GnRH reduces both the number of plasma membrane GnRH receptors and the ability of GnRH to generate Ins(1,4,5)P 3 , these effects are insufficient to account for the reduced release of intracellular Ca 2ϩ (5). Indeed, this desensitization is heterologous and therefore most probably reflects post-receptor modification(s). Because desensitization of GnRH-stimulated Ca 2ϩ mobiliza-tion from intracellular stores cannot be attributed to attenuation of Ins(1,4,5)P 3 generation or depletion of hormone-mobilizable intracellular Ca 2ϩ pools, it appears to reflect a reduction in the efficiency with which Ins(1,4,5)P 3 mobilizes Ca 2ϩ from intracellular stores (5). There is now accumulating evidence that GPCR-mediated activation of PLC causes down-regulation of Ins(1,4,5)P 3 receptors (16 -23), most probably through proteolysis that is initiated as a consequence of activation by Ins(1,4,5)P 3 (20,22,23). However, this down-regulation often requires several hours of agonist stimulation (17,23), whereas GnRH desensitizes Ca 2ϩ responses in ␣T3-1 cells with prestimulation periods of 10 -30 min (5). Thus, if Ins(1,4,5)P 3 receptor down-regulation underlies desensitization of GnRHstimulated Ca 2ϩ mobilization, it would have to occur unusually rapidly in ␣T3-1 cells. The current study was therefore undertaken to establish whether GnRH is able to cause Ins(1,4,5)P 3 receptor down-regulation and whether any such effect underlies desensitization of Ca 2ϩ mobilization in these cells. ␣T3-1 cells stably transfected with recombinant human M 3 muscarinic receptors (7) were used in these experiments to enable comparison of responses to PLC-activating GPCRs that do (M 3 ) and do not (GnRH) show rapid homologous desensitization (7).

EXPERIMENTAL PROCEDURES
Materials and Cell Culture-Reagents of analytical grade were obtained from suppliers listed previously (5, 24 -26), unless stated, or alternatively from Sigma. Antibodies against PLC isoforms and G␣ q/11 were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA), with the exception of PLC␦1, which was from Upstate Biotechnology (Lake Placid, NY). The ␣T3-1 gonadotrope cell line was originally a gift from Dr. P. Mellon, University of California, San Diego, CA, and in the current study we used a cell line (␣T3-1/M 3 ) derived from this, which also expresses the recombinant human muscarinic M 3 receptor. Like the endogenously expressed GnRH receptor, this GPCR also couples to the activation of PLC in this cell line (7), and we have demonstrated that muscarinic M 3 receptors are subject to rapid but partial desensitization, whereas the endogenously expressed GnRH receptors show no evidence of such regulation (7). Cells were cultured in Dulbecco's modified Eagle's medium supplemented with 50 IU/ml penicillin, 50 g/ml streptomycin, 2 mM glutamine, and 10% (v/v) fetal calf serum. Cultures were maintained at 37°C in 5% CO 2 , humidified air and passaged weekly. For experiments, cells were harvested with 10 mM HEPES, 154 mM NaCl, 0.54 mM EDTA (pH 7.4) and re-seeded for use 1-2 days later. Cells were always maintained, and the experimental manipulations were always performed, at 37°C unless stated otherwise.
Dynamic Video Imaging of Cytosolic Ca 2ϩ -Video imaging of fura-2loaded cells was performed as described previously (5). Briefly, cells grown on glass coverslips were loaded with the acetoxymethyl ester of fura-2 (2 M) for 30 min at 37°C in 1 ml of buffer (pH 7.4, composition (mM): NaCl 127, CaCl 2 1.8, KCl 5, MgCl 2 2, NaH 2 PO 4 0.5, NaHCO 3 5, glucose 10, HEPES 10 with 0.1% bovine serum albumin). Cells were then washed several times and placed within a heated (37°C) perfusion stage of a Nikon Diaphot inverted microscope. Image capture was performed using MagiCal hardware following alternate excitation at 340 and 380 nm with emission recorded at 510 nm. Values were averaged from 16 or 32 video frames, and background fluorescence was subtracted prior to ratioing. The ratio of fluorescence at 340 and 380 nm was calculated on a pixel-by-pixel basis using maximum and minimum values defined by treatment with 5 M ionomycin in medium with either 10 mM CaCl 2 or 10 mM EGTA and assuming a dissociation constant of 225 nM for fura-2 and Ca 2ϩ at 37°C, as described (14).
Western Blotting-Cells were grown to confluence in 6-well multiwell dishes. Medium was removed, and the cells were washed (2 ϫ 1 ml) with medium containing 0.1% bovine serum albumin and incubated in a further 1 ml. GnRH was then added at the appropriate concentration, and the cells were incubated at 37°C in 5% CO 2 , humidified air. For immuno-detection of Ins(1,4,5)P 3 receptors, medium was aspirated after the required time, the cell monolayer was washed once with ice-cold Krebs/HEPES (7), and 1 ml of ice-cold TE buffer (10 mM Tris, 10 mM EDTA, pH 7.4) was added. Cells were left for 5 min on ice and then scraped from the surface of the plate. Following trituration through a fine gauge needle the resulting suspension was centrifuged (13,000 ϫ g, 4°C, 15 min). The supernatant was then aspirated, and the pellet was resuspended in 50 l of TE buffer. An equal volume of sample buffer (100 mM Tris-HCl, 2% SDS, 10% glycerol, 0.1% bromphenol blue, and 200 mM dithiothreitol) was then added, and the samples were boiled for 3 min. Proteins were resolved by 5% SDS-polyacrylamide gel electrophoresis, transferred to nitrocellulose, and probed using isotype-specific Ins(1,4,5)P 3 receptor antibodies. Immunoreactive bands were detected with ECL reagents and exposure to Hyperfilm-ECL (Amersham Pharmacia Biotech). Where required, densitometric analysis of the resulting bands was performed using a Bio-Rad GS-670 imaging densitometer with Molecular Analyst version 1.2 software. For immuno-detection of other proteins (G␣ q/11 and PLC isoforms), Western blotting was carried out as above with the exception that the cell monolayers were solubilized in 200 l of solubilization buffer (10 mM Tris, 10 mM EDTA, 500 mM NaCl, 1% Nonidet P-40, 0.1% SDS, 0.5% deoxycholate, 1 mM phenylmethylsulfonyl fluoride, 100 g/ml iodoacetamide, 100 g/ml benzamidine) for 30 min on ice before being processed as described above. Details and use of the polyclonal antibody against the type I Ins(1,4,5)P 3 receptor and of the monoclonal antibodies against the type II and type III Ins(1,4,5)P 3 receptors have been described previously (26). Other antibodies were at dilutions according to the instructions of the suppliers. (1,4,5)P 3 -Ins(1,4,5)P 3 receptor binding was determined in membranes of ␣T3-1 cells as described (16). Incubations were for 45 min at 4°C in 200 l of incubation buffer with 2-10 g of membrane protein, 10 nCi of [ 3 H]Ins(1,4,5)P 3 , and 0 or 10 Ϫ9 -10 Ϫ5 M unlabeled Ins(1,4,5)P 3 . The incubations were terminated by centrifugation and removal of supernatants by aspiration. Pellets were then solubilized in NaOH and transferred to scintillant for ␤-counting.

Determination of Ins(1,4,5)P 3 Receptor Density by Binding of [ 3 H]Ins
Ins(1,4,5)P 3 -mediated Release of 45 Ca 2ϩ from Intracellular Stores of Permeabilized Cells-45 Ca 2ϩ release assays were performed in cytosol-like buffer (composition (mM): KCl 120, KH 2 PO 4 2, (CH 2 COONa) 2 5, MgCl 2 2.4, HEPES 20, ATP 2, pH 7.2) using a previously described method (27). The [Ca 2ϩ ] of the cytosol-like buffer was determined using fura-2 (28) and buffered to 120 -190 nM with EGTA. Cells (2 ml containing 4 -6 mg of protein) were permeabilized by the addition of ␤-escin (25 g/ml). The suspension was then centrifuged (500 ϫ g, 2 min) and resuspended in 6 ml of cytosol-like buffer, and 45 Ca 2ϩ was added (equivalent to 0.4 l/ml 45 Ca 2ϩ at 1.98 mCi/ml). After gentle vortexing, the cells were left for 15-20 min at room temperature. To initiate release of loaded 45 Ca 2ϩ , 50 l of cells were added to 50 l of Ins(1,4,5)P 3 . After 60 s, 500 l of silicon oil was added, and the cells were centrifuged at 16,000 ϫ g for 2 min. The aqueous phase and most of the silicon oil phase were aspirated, the tubes were inverted, and the remaining oil was allowed to drain. The pellets were solubilized in scintillant, and the unreleased 45 Ca 2ϩ was determined. Release was calculated as a percentage of the total 45 Ca 2ϩ loaded. The size of the rapidly releasable pool was also determined using 10 M ionomycin to indicate the amount of 45 Ca 2ϩ loaded into the intracellular stores. This was ϳ80% of the total 45 Ca 2ϩ .

RESULTS
Pretreatment of ␣T3-1/M 3 cells for 1 h with maximal concentrations (7) of either GnRH (1 M) or methacholine (1 mM) resulted in both homologous and heterologous desensitization of agonist-mediated Ca 2ϩ signaling (Fig. 1, a and b). The homologous and heterologous loss of Ca 2ϩ signaling as a result of GnRH or methacholine pretreatment were maximal following FIG. 2. a-c, agonist-mediated down-regulation of type I Ins(1,4,5)P 3 receptor immunoreactivity in ␣T3-1/M 3 cells. Cells were challenged with 1 M GnRH, 100 nM GnRH, or 1 mM methacholine for the times indicated, and Western blots for the type I Ins(1,4,5)P 3 receptor were performed. Western blots, representative of three experiments, show the type I Ins(1,4,5)P 3 receptor in the absence of agonist treatment or following treatment with either 1 M GnRH (a) or 1 mM methacholine (b) for the indicated times. The density of the band representing the type I Ins(1,4,5)P 3 receptor was quantified and expressed as a percentage of that in the absence of agonist (c). Data are mean Ϯ S.E.; n ϭ 3. q, 1 M GnRH; E, 100 nM GnRH; f, 1 mM methacholine. d and e, time course of the recovery of type I Ins(1,4,5)P 3 receptor immunoreactivity following down-regulation with GnRH. Following treatment with GnRH (1 M) for 1 h, cells were washed, and the incubation was continued in the presence of an antagonist of the GnRH receptor (antide, 1 M). Cells were then either solubilized immediately or allowed the indicated recovery time before solubilization. Western blotting for the type I Ins(1,4,5)P 3 receptor was then performed. A representative Western blot is shown (d); below the blot are the mean densitometric data (e). The density of the band representing the type I Ins(1,4,5)P 3 receptor was quantified and expressed as a percentage of that under basal (no agonist treatment) conditions. The data are the mean Ϯ S.E.; n ϭ 3. 30 -60 min of pretreatment and were sustained at this level for at least 24 h of pretreatment (Fig. 1c).
Expression of type I, II, and III Ins(1,4,5)P 3 receptors was examined by Western blotting of solubilized ␣T3-1/M 3 cells using isoform-specific antibodies (26). Expression of all three isoforms was detected, although bands representing type II and III Ins(1,4,5)P 3 receptors were faint and then only apparent after exposure of the Western blot to film for longer periods of time (data not shown). Although we are unable to quantitate precisely the relative levels of the different types of Ins(1,4,5)P 3 receptors, the predominant expression of type I is consistent with that in the pituitary, which expresses type I, II, and III Ins(1,4,5)P 3 receptors at 73, 24, and 3% of the total receptor population, respectively (19).
Challenge of ␣T3-1/M 3 cells with 1 M GnRH resulted in a rapid and marked loss of type 1 Ins(1,4,5)P 3 receptor immunoreactivity (Fig. 2, a and c), which had a half-time of ϳ20 min, was maximal by 60 min (Ͻ20% of immunoreactivity remaining), and sustained for at least 24 h in the continued presence of agonist (Fig. 2, a and c). Comparable data were obtained using 100 nM GnRH (Fig. 2c). Challenge of ␣T3-1/M 3 cells with 1 mM methacholine also resulted in a marked loss of type I Ins(1,4,5)P 3 receptor immunoreactivity, albeit with a rate and magnitude that was less than that observed with GnRH (Fig. 2,  b and c). In contrast to treatment with 1 M GnRH, there was some recovery of type I Ins(1,4,5)P 3 receptor immunoreactivity during 24-h treatment with methacholine (Fig. 2, b and c). The GnRH-mediated loss of type I Ins(1,4,5)P 3 receptor immunoreactivity was concentration-dependent, with an EC 50 of Ϫ9.91 Ϯ 0.61 (log 10 , M; n ϭ 4; 0.12 nM) (all data with errors are mean Ϯ S.E.) at 60 min of treatment (data not shown).
In experiments designed to examine the rate of recovery of type I Ins(1,4,5)P 3 receptor immunoreactivity, cells were first treated with 1 M GnRH for 1 h to induce down-regulation. Agonist activation of GnRH receptors was then stopped by washing the cell monolayer and continuing the incubation in the presence of the GnRH receptor antagonist, antide (1 M).
Type I Ins(1,4,5)P 3 receptor immunoreactivity was then examined over the subsequent 24 h. The 1-h pretreatment with GnRH resulted in a marked loss of type I Ins(1,4,5)P 3 receptor immunoreactivity that was further reduced after 1 h of "recovery" time (Fig. 2, d and e). Levels of receptor immunoreactivity then increased back to basal levels by 24 h (Fig. 2, d and e). Similar results were obtained when cells were treated with 1 M GnRH and washed, but antagonist was not added (data not shown).
Thimerosal has been reported to increase Ins(1,4,5)P 3 receptor sensitivity (29). In naive ␣T3-1/M 3 cells, 100 M thimerosal increased the spike [Ca 2ϩ ] i response to a sub-maximal (5 nM) (Fig. 5a) but not maximal (1 M) (Fig. 5c) concentration of GnRH when cells were challenged in the absence of extracellular Ca 2ϩ (to assess the effects on Ca 2ϩ release only). This suggests that the Ins(1,4,5)P 3 receptor does not limit the magnitude of the Ins(1,4,5)P 3 -mediated Ca 2ϩ release in naive cells stimulated with a maximal concentration of GnRH but can limit the response to submaximal agonist concentrations. When cells were pretreated for 1 h with 100 nM GnRH (in the presence of extracellular Ca 2ϩ ), thimerosal had little effect on the subsequent response (again in the absence of extracellular Ca 2ϩ ) to a submaximal concentration of GnRH but markedly potentiated the response to a maximal concentration (Fig. 5, b  and d), suggesting that, in GnRH-pretreated cells, Ins(1,4,5)P 3 receptor activation is rate-limiting for GnRH-stimulated Ca 2ϩ mobilization.
Although challenge of cells with 1 M GnRH resulted in a dramatic reduction in type I Ins(1,4,5)P 3 receptor immunoreactivity (see above), exclusion of Ca 2ϩ from the extracellular buffer prevented the GnRH-mediated loss of Ins(1,4,5)P 3 receptors but had no significant effect on the basal (nonagoniststimulated) levels over a 1-h period (Fig. 6a). Furthermore, the absence of extracellular Ca 2ϩ partially prevented the homologous desensitization of GnRH-mediated spike and plateau Ca 2ϩ signaling (Fig. 6, b and c). Incubation of cells for 4 h with the cysteine protease (and proteasome) inhibitor N-acetyl-Leu-Leu-norleucinal (ALLN, 100 g/ml) (20,22,23) prior to and during a 1-h incubation with 1 M GnRH also markedly attenuated the agonist-induced loss of type I Ins(1,4,5)P 3 receptor immunoreactivity (Fig. 7a). The effects of ALLN on the desensitization of the agonist-mediated Ca 2ϩ response were, however, difficult to interpret. Thus, even in the absence of GnRH pretreatment, ALLN markedly inhibited the spike and plateau [Ca 2ϩ ] i responses to GnRH (data not shown), and we therefore used the more specific proteasome inhibitor lactacystin (10 M, 4 h) (22). Lactacystin also markedly protected type I Ins(1,4,5)P 3 receptor immunoreactivity against GnRH-medi- After control pretreatments, thimerosal increased the response to 5 nM GnRH but not that to 1 M GnRH, whereas in GnRHpretreated (desensitized) cells, thimerosal increased the response to 1 M GnRH but not that to 5 nM GnRH. ϩ lactacystin, 102.5 Ϯ 34.1%). Furthermore, lactacystin attenuated, but did not completely prevent, GnRH-mediated desensitization of [Ca 2ϩ ] i mobilization (Fig. 7, b and c). There was also some inhibition of GnRH-mediated spike [Ca 2ϩ ] i signaling in the presence of lactacystin (Fig. 7b), although the plateau was unaffected (data not shown).
Using commercially available antibodies against the PLC isoforms ␤ 1-4 , ␥ 1-2 , and ␦ 1-2 , the expression of PLC␤ 1 , ␤ 3 , ␥ 1 , and ␥ 2 was demonstrated in ␣T3-1/M 3 cells (Fig. 8). The immunoreactivity of those antibodies that did not detect proteins in ␣T3-1/M 3 cells was confirmed using extracts from either SH-SY5Y neuroblastoma cells or rat brain (data not shown). Given that GnRH receptor-mediated responses are via G␣ q/11 and most likely, therefore, via PLC␤ isoforms, we examined the influence of GnRH or methacholine treatment on the expression of G␣ q/11 , PLC␤ 1 , and ␤ 3 . Exposure of ␣T3-1/M 3 cells for up to 1 h with maximal concentrations of either GnRH (1 M) or methacholine (1 mM) had no consistent effects on the levels of G␣ q/11 or the PLC isoforms ␤ 1 and ␤ 3 (Fig. 8). DISCUSSION It has been known for over two decades that sustained stimulation of gonadotropes with GnRH causes desensitization of GnRH-stimulated gonadotropin secretion (30), an effect that can be either exploited or avoided in clinical applications of GnRH analogues (12). The recent discovery that mammalian GnRH receptors do not show rapid homologous desensitization (3)(4)(5)(6)(7)(8) reveals that desensitization of GnRH-stimulated gonadotropin secretion must reflect changes distal to the receptor, as implied by earlier work showing that GnRH receptor regulation does not explain desensitization of gonadotropin secretion (5). Our work has revealed that pretreatment of ␣T3-1 cells with GnRH causes a pronounced desensitization of GnRHstimulated mobilization of Ca 2ϩ from intracellular stores (4, 5), which is of particular interest in the light of the established importance of Ca 2ϩ mobilization in mediation of GnRH-stimu-lated gonadotropin secretion. Because this desensitization is heterologous (cross-desensitization is seen with responses to other PLC-activating stimuli), it most likely reflects changes in the amount or activity of effector proteins distal to the GnRH receptor.
Activation-dependent down-regulation of effector proteins is emerging as an important mechanism for post-receptor adaptive responses, and such responses have already been observed in response to GnRH. In ␣T3-1 cells, GnRH causes a loss of G␣ q (31) and of regulatory and catalytic subunits of protein kinase A (32). It also causes the apparent proteolysis of protein kinase C ␦ and ⑀ (33). Here we have focused on the possible effects of GnRH on effector proteins involved in Ca 2ϩ metabolism and found that a 60-min pretreatment with GnRH causes pronounced desensitization of GnRH-stimulated Ca 2ϩ mobilization without measurably altering cellular levels of PLC␤1, PLC␤3, or G␣ q . These data are consistent with earlier studies demonstrating that GnRH reduces G␣ q levels extremely slowly (half-time Ͼ 6 h) (31) and that reduced Ins(1,4,5)P 3 generation alone does not account for desensitization of Ca 2ϩ mobilization in these cells (5). In contrast, we show that GnRH pretreatment causes a pronounced down-regulation of Ins(1,4,5)P 3 receptors (as demonstrated by radioligand binding and immunological quantification of type I Ins(1,4,5)P 3 receptors), that this effect is functionally significant (as demonstrated by a reduction in Ins(1,4,5)P 3 -stimulated 45 Ca 2ϩ mobilization in permeabilized cells), and that the onset of this effect, and recovery from the effect, have similar kinetics to the onset of, and recovery from, desensitization of Ca 2ϩ mobilization (4,5).
In several systems, activation of PLC-coupled GPCRs has been shown to down-regulate Ins(1,4,5)P 3 receptors, an effect attributed to Ca 2ϩ -dependent proteolysis of active Ins(1,4,5)P 3 receptors (17, 20, 22, 23). Activation of these receptors has been shown to cause their ubiquitination while still in the endoplasmic reticulum membrane (22). This is thought to target Ins(1,4,5)P 3 receptors for proteasomal degradation, as demonstrated by the fact that proteasome inhibitors can prevent GPCR-mediated down-regulation (20,22,23). Our data are in accord with this model, because we have found that GnRHmediated down-regulation of type I Ins(1,4,5)P 3 receptor immunoreactivity is prevented in Ca 2ϩ -free medium and by the two protease inhibitors ALLN and lactacystin. Interestingly, we have found that the down-regulation of type I Ins(1,4,5)P 3 receptors caused by GnRH is more rapid, more pronounced, and more slowly reversed than that caused by methacholine (muscarinic M 3 receptor activation). This is despite the fact that both stimuli cause comparable increases in [Ca 2ϩ ] i in these cells and occur even when the concentrations of GnRH and methacholine are matched to give comparable maximal increases in Ins(1,4,5)P 3 levels in these cells (100 nM GnRH, 1 mM methacholine) (7). However, the muscarinic M 3 receptor undergoes a partial rapid homologous desensitization and therefore causes a transient increase in Ins(1,4,5)P 3 mass, reducing to a sustained plateau after a peak at 10 s, whereas the GnRH receptor does not rapidly desensitize and therefore causes a sustained increase in Ins(1,4,5)P 3 mass, which reaches maximal levels within 20 -30 s (7). This clearly implies that the Ins(1,4,5)P 3 receptor down-regulation is sensitive not just to the magnitude of the Ins(1,4,5)P 3 response but also to its duration, precisely as expected if it is the Ins(1,4,5)P 3 occupied (active) receptor conformation that is sensitive to proteolysis (22). Thus, the lack of GnRH receptor desensitization may contribute to the unusual rapidity of Ins(1,4,5)P 3 receptor down-regulation in these cells. Typically, Ins(1,4,5)P 3 receptor down-regulation occurs with a half-time of 4 -24 h (17, 23), as compared with Ͻ20 min in GnRH-stimulated ␣T3-1/M 3 cells (Fig. 2). Presumably with other PLC-activating GPCRs, receptor desensitization attenuates Ins(1,4,5)P 3 responses and thereby reduces the rapidity and/or magnitude of Ins(1,4,5)P 3 receptor down-regulation. It should be noted, however, that bombesin and cholecystokinin reduce Ins(1,4,5)P 3 receptor levels with a half-time of Ͻ30 min in AR4 -2J cells (19) and that methacholine caused Ins(1,4,5)P 3 receptor down-regulation with a half-time of Ͻ60 min in ␣T3-1/M 3 cells, demonstrating that relatively rapid down-regulation can occur, even with receptors that do desensitize.
The major question raised by our data is whether downregulation of Ins(1,4,5)P 3 receptors contributes to or underlies desensitization of Ca 2ϩ mobilization. Our investigations of response kinetics are entirely compatible with this possibility because we have found a) that the time-course of Ins(1,4,5)P 3 receptor down-regulation in response to GnRH is comparable with that for the onset of desensitization (Figs. 2 and 1c, respectively) and b) that both effects are maintained as GnRH pretreatment is extended to 24 h. Further support for the possible causal relationship is provided by the demonstrations a) that the GnRH-mediated Ins(1,4,5)P 3 receptor loss and desensitization of Ca 2ϩ signaling are associated with reduced Ins(1,4,5)P 3 -stimulated mobilization of 45 Ca 2ϩ from permeabilized cells (directly establishing the functional significance of Ins(1,4,5)P 3 receptor regulation in this system), b) that Ca 2ϩfree medium prevents and attenuates GnRH-mediated Ins(1,4,5)P 3 receptor down-regulation and desensitization of Ca 2ϩ mobilization, respectively, c) that lactacystin prevents and attenuates GnRH-mediated Ins(1,4,5)P 3 receptor downregulation and desensitization of Ca 2ϩ mobilization, respectively, and d) that thimerosal partially reverses desensitization of Ca 2ϩ mobilization. Because thimerosal increases the affinity of Ins(1,4,5)P 3 receptors for Ins(1,4,5)P 3 (29), it would only be expected to influence responses to Ins(1,4,5)P 3 under conditions where Ins(1,4,5)P 3 receptor activation is limiting. Thus, the ability of thimerosal to amplify Ca 2ϩ mobilization by 5 nM GnRH, but not by 1 M GnRH, implies that Ins(1,4,5)P 3 receptor activation is limiting for the response to the low concentration of GnRH but not to the high concentration. In desensitized cells, however, thimerosal increased the response to the high concentration of GnRH, demonstrating that in these cells Ins(1,4,5)P 3 receptor activation has become limiting. This is precisely what would be expected if Ins(1,4,5)P 3 receptor loss leaves the desensitized cells with insufficient Ins(1,4,5)P 3 receptors for efficient mobilization of Ca 2ϩ even in the face of sufficient GnRH-stimulated Ins(1,4,5)P 3 levels.
Although our data are largely consistent with the possibility that GnRH-mediated Ins(1,4,5)P 3 receptor down-regulation underlies desensitization of Ca 2ϩ mobilization, several lines of evidence might argue against this interpretation. Thus, stimulation of muscarinic receptors results in a time course of heterologous desensitization of GnRH-mediated [Ca 2ϩ ] i elevation similar to the homologous desensitization caused by GnRH pretreatment. This is despite the finding that GnRH causes a more rapid and greater loss of Ins(1,4,5)P 3 receptor immunoreactivity than muscarinic receptor stimulation. Furthermore, the retention of ϳ50% of Ins(1,4,5)P 3 receptors (Fig. 3) and the fact that maximal Ins(1,4,5)P 3 -stimulated 45 Ca 2ϩ mobilization is only reduced by ϳ42% (Fig. 4) stand in contrast to the almost complete loss of the spike phase [Ca 2ϩ ] i response to GnRH in desensitized cells (Fig. 1). Similarly, complete inhibition of type I Ins(1,4,5)P 3 receptor down-regulation by pretreatment in Ca 2ϩ -free medium, or in the presence of ALLN or lactacystin, contrasts to only partial inhibition, or no measurable inhibition, of desensitization. These apparent inconsistencies could reflect contributions from other as yet unidentified mechanisms of desensitization or may reflect differences in the relative contributions of Ins(1,4,5)P 3 receptor subtypes, or of receptors in different cellular locations, to the end points quantified. Thus, type II Ins(1,4,5)P 3 receptors, which are relatively resistant to down-regulation in other systems (19), may contribute disproportionally to the 45 Ca 2ϩ mobilization response, and local down-regulation of Ins(1,4,5)P 3 receptors in the immediate vicinity of GnRH receptors may be more extreme than that revealed by global measurements of all Ins(1,4,5)P 3 receptors. Alternatively, it is possible that the 50% loss of Ins(1,4,5)P 3 receptors and the consequent increase in mean distance between functional Ins(1,4,5)P 3 receptors are sufficient to prevent propagation of Ca 2ϩ mobilization by calcium-induced calcium release (34) and therefore have a disproportionately large effect on Ca 2ϩ responses in intact cells (as compared with permeabilized cells or membrane preparations). It is equally possible, however, that other modifications of Ins(1,4,5)P 3 receptors (e.g. phosphorylation, ATP binding, ubiquitination) inhibit Ins(1,4,5)P 3 receptor signaling in the desensitized cells without altering immunoreactivity or radioligand binding in membrane preparations. Some of these modifications, particularly ubiquitination, appear to be involved in the targeting of Ins(1,4,5)P 3 receptors for degradation (22). That such targeting occurs is demonstrated by our finding that Ins(1,4,5)P 3 receptor immunoreactivity continues to decline following removal of GPCR activation (Fig. 2).
Whereas a number of studies have demonstrated the principle of agonist-induced Ins(1,4,5)P 3 receptor down-regulation (16 -23), the current study provides evidence of a setting in which such regulation may be functionally relevant. Thus, loss of Ins(1,4,5)P 3 receptors following either pre-ovulatory surges in GnRH or, in particular, the clinical use of GnRH agonists may play a part in the suppression of gonadotrope function. It should be noted that such a mechanism would also result in a compromised function of other Ins(1,4,5)P 3 -dependent, Ca 2ϩmobilizing receptors expressed on pituitary cells (e.g. pituitary adenylyl cyclase-activating polypeptide receptors). Such heterologous loss of function by this mechanism may be less apparent in other systems in which GPCR desensitization may serve to limit the down-regulation of signaling components shared with other receptors.