Agonist-induced desensitization, internalization, and phosphorylation of the sst2A somatostatin receptor.

Cellular responsiveness to the inhibitory peptide somatostatin (SRIF) or its clinically used analogs can desensitize with agonist exposure. While desensitization of other seven-transmembrane domain receptors is mediated by receptor phosphorylation and/or internalization, the mechanisms mediating SRIF receptor (sst) desensitization are unknown. Therefore, we investigated the susceptibility of the sst2A receptor isotype to ligand-induced desensitization, internalization, and phosphorylation in GH-R2 cells, a clone of pituitary tumor cells overexpressing this receptor. A 30-min exposure of cells to either SRIF or the analog SMS 201-995 (SMS) reduced both the potency and efficacy of agonist inhibition of adenylyl cyclase. Internalization of receptor-bound ligand was rapid (t1/2 = 4 min) and temperature-dependent. SRIF and SMS increased the phosphorylation of the 71-kDa sst2A protein 25-fold within 15 min. Receptor phosphorylation was dependent on both the concentration and time of agonist exposure and was not affected by pertussis toxin pretreatment, indicating that receptor occupancy rather than second messenger formation was required. Receptor phosphorylation was also stimulated by phorbol 12-myristate 13-acetate activation of protein kinase C. Both ligand-stimulated and phorbol 12-myristate 13-acetate-stimulated receptor phosphorylation occurred primarily on serine. These studies are the first demonstration of agonist-dependent desensitization, internalization, and phosphorylation of the sst2A receptor and suggest that phosphorylation may mediate the homologous and heterologous regulation of this receptor.

cluding the brain, pituitary, pancreas, spleen, small intestine, and stomach (1,2), and the receptor protein has recently been shown to be widely distributed in the mammalian brain (3). Thus, this receptor isotype mediates many of the central and peripheral actions of SRIF.
Early studies on the signal transduction mechanisms activated by SRIF showed that sst receptors elicited their actions predominantly via pertussis toxin-sensitive G proteins (1,2,4). Thus, SRIF inhibition of adenylyl cyclase and Ca 2ϩ channels, as well as SRIF stimulation of K ϩ channels, phospholipase C, serine/threonine and tyrosine phosphatases, arachidonic acid release, and mitogen-activated protein kinases are inhibited by pertussis toxin treatment (5)(6)(7)(8)(9)(10)(11)(12). However, some actions of SRIF, such as stimulation of other tyrosine phosphatases as well as inhibition of Na/H exchange, are pertussis toxin-insensitive (13,14). The network of signaling pathways activated by individual sst receptor isotypes is largely unknown. Signaling mechanisms have been especially difficult to elucidate in the native environment of the receptors because most SRIF target cells express multiple sst receptor isotypes that cannot be individually activated with the analogs currently available.
For most G protein-coupled receptors, hormone treatment decreases receptor responsiveness (desensitization), receptor levels (down-regulation), or both. However, relatively little is known about sst receptor regulation, especially following acute hormonal challenge. Exposure to SRIF or to selective SRIF agonists such as SMS 201-995 (SMS) has been reported to lead to desensitization in pituitary cells over the course of hours, days, or weeks (15)(16)(17)(18). Desensitization occurring over both hours (19,20) and minutes (21) has been reported in the AtT20 corticotropic pituitary cell line, depending on the signaling pathway being examined. However, SRIF receptor desensitization was not detected in the mammotropic GH 4 C 1 pituitary cell line (22) and does not occur during long term treatment of many human pituitary tumors with SMS (23). SRIF receptors can be either down-regulated (24) or up-regulated (22) by hormone pretreatment depending on the cell type examined. The SRIF receptor isotypes involved in these varying effects are unknown although sst2 receptors are expressed, along with other sst receptor subtypes, in the normal pituitary, in pituitary tumors, and in the AtT20 and GH 4 C 1 cell lines (2,25). 2 Interestingly, for sst2A receptors exogenously expressed in Chinese hamster ovary cells, receptor binding has been reported to be either decreased (27,28) or increased (29) by SRIF pretreatment. However, no desensitization studies have been reported with this receptor isotype.
Ligand-dependent and -independent desensitization of other G protein-coupled receptors is mediated by receptor phosphorylation (30). Examination of sst2A receptor phosphorylation requires a cell line expressing high levels of functional receptor protein. However, identification of appropriate cell lines for studies of exogenously expressed sst2A receptors has proven problematic. Although it is well established that SRIF inhibits adenylyl cyclase activity in native cells (1,2,4), the coupling of the sst2A receptor to adenylyl cyclase in heterologous cell lines varies with the cell model (1) suggesting that components required for faithful mimicry of the normal function of the sst2A receptor are not ubiquitously expressed. The GH 4 C 1 pituitary tumor cell line, which contains both the sst1 and sst2A receptors (25), 2 has been extensively used for studies of SRIF receptor signaling and regulation (4). Recently, a subclone of GH 4 C 1 cells was isolated following transfection with sst2A receptor cDNA (31). This transfected cell line (GH 4 -R2.20 cells) expresses approximately 100 times as many functional sst2A receptors as the parental GH 4 C 1 cells and has both the elevated sst2A receptor expression required for phosphorylation studies and an environment that allows normal receptor coupling. By taking advantage of the 10,000-fold greater affinity of SMS 201-995 for the sst2 over the sst1 receptor (1), we now demonstrate agonist-dependent desensitization, internalization, and phosphorylation of the sst2A receptor in this pituitary model cell line.

EXPERIMENTAL PROCEDURES
Hormones and Supplies-Cell culture medium and G418 were purchased from Life Technologies, Inc. The sst2A-receptor antiserum (R2-88) has been described. 2 Leupeptin, pepstatin A, phenylmethylsulfonyl fluoride, soybean trypsin inhibitor, bacitracin, cholesterol hemisuccinate, Nonidet P-40, and Protein A were obtained from Sigma. N-Dodecyl ␤-D-maltoside and pertussis toxin were purchased from Calbiochem and List Biological Laboratories, Inc. (Campbell, CA), respectively. CNBr-activated Sepharose 4B was from Pharmacia Biotech Inc. (Uppsala, Sweden). Bradford reagent and reagents for electrophoresis and Western blotting were obtained from Bio-Rad. Carrier-free Na 125 I was purchased from Amersham Corp. Phosphate-free Dulbecco's modified Eagle's medium and [ 32 P]orthophosphate were purchased from ICN Biomedicals (Irvine, CA). All other reagents were of the best grade available and purchased from common suppliers.
Cell Culture-The clonal GH 4 -R2.20 cell line (hereafter referred to as GH-R2 cells) was generated by transfecting GH 4 C 1 pituitary tumor cells with the rat sst2A receptor and was grown as described previously (31). Experimental cultures were plated in medium without G418 and used 2-7 days later with a medium change 18 -24 h prior to use. Experiments were carried out with cells plated in 100-mm dishes except whole cell binding experiments, which used cells in 35-mm wells.
Cell Membrane Preparation-GH-R2 cells were washed with and scraped into cold phosphate-buffered saline (PBS: 10 mM Na 2 HPO 4 , 150 mM NaCl, pH 7.4) containing protease and phosphatase inhibitors (1 mM phenylmethylsulfonyl fluoride, 10 mM sodium pyrophosphate, 10 mM sodium fluoride, 0.1 mM sodium orthovanadate, 100 nM okadaic acid). Following centrifugation, the cell pellet was resuspended in homogenization buffer (10 mM Tris-HCl, 5 mM EDTA, 3 mM EGTA, pH 7.6) containing protease and phosphatase inhibitors and membranes were prepared and stored as described previously (25). Membranes used for photoaffinity cross-linking were prepared in the presence of protease inhibitors alone.
Adenylyl Cyclase Measurements-GH-R2 cells were incubated in the absence or presence of 100 nM SMS (Sandoz Pharmaceuticals, Basel, Switzerland) or SRIF for 30 min in fresh growth medium. Membranes were then prepared as described above and assayed in triplicate (2-5 g of membrane protein/tube) for adenylyl cyclase activity as described previously (32). Cyclase activity was 30 -100 pmol/min/mg under basal conditions and 300 -800 pmol/min/mg in the presence of 100 nM VIP. Cyclase activity measured in the presence of both 100 nM VIP and varying concentrations of SMS was expressed as a percent of the VIPstimulated activity, and fitted values for maximal inhibition and EC 50 (the concentration required to produce half-maximal inhibition) were obtained by least squares nonlinear regression analysis of dose-response curves using the program D/R (Biomedical Computing, Inc. Houston, TX).
Radioligand were radioiodinated using chloramine T and subsequently purified by reverse-phase high performance liquid chromatography as described previously (33). GH-R2 cells were incubated at 4°C in 1 ml of binding buffer (F10 medium containing 20 mM HEPES and 5 mg/ml lactalbumin hydrolysate, pH 7.4) containing approximately 100,000 cpm of [ 125 I-Tyr 3 ]SMS with or without 100 nM unlabeled SRIF (34). After 60 min, the cells were rinsed to remove unbound trace and then incubated in fresh 37°C buffer to allow internalization of the receptor-bound ligand. Cells were subsequently incubated on ice for 5 min in acidic glycinebuffered saline (100 mM glycine, 50 mM NaCl, pH 3.0) to release surfacebound ligand. After collection of the acidic buffer, the cells were dissolved in 0.1 N NaOH. The radioactivity in both the glycine buffer (representing surface-bound ligand) and in the cell lysates (representing internalized ligand) was measured (34). Specific binding was calculated as the difference between the amount of radioligand bound in each fraction in the absence (total binding) and presence of 100 nM SRIF (nonspecific binding).
Resolved proteins were transferred to PVDF membrane as described previously. 2 The membrane was then blocked for 2 h with "Blotto" (10 mM NaH 2 PO 4 , 10% nonfat dry milk, 10% glycerol, 0.2% Tween 20) and incubated overnight at 4°C with 1:20,000 dilution of anti-sst2A antibody R2-88 in Blotto. Following repeated washing, the membrane was incubated with 1:5000 dilution of goat-anti-rabbit antibody conjugated with horseradish peroxidase at room temperature for 1 h. Immunoreactive proteins were detected with the ECL chemiluminescent antibody detection system (Amersham Corp.).
Detection of Immunoprecipitated sst2A Receptor by Immunoblotting-GH-R2 cells were incubated in growth medium in the presence or absence of protein kinase activators for 30 min. Cell membranes were prepared as described above and solubilized by agitation at 4°C for 60 min in HEPES-buffered saline (150 mM NaCl, 20 mM Hepes, pH 7.4, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml soybean trypsin inhibitor, 10 g/ml leupeptin, 50 g/ml bacitracin, 5 mM EDTA, 3 mM EGTA, 10 mM sodium pyrophosphate, 10 mM sodium fluoride, 0.1 mM sodium orthovanadate) containing 4 mg/ml dodecyl ␤-maltoside, 200 g/ml cholesterol hemisuccinate (lysis buffer). After centrifugation at 100,000 ϫ g for 30 min, the supernatant was incubated overnight at 4°C with anti-sst2A-receptor antiserum covalently coupled to Protein A-Sepharose (final dilution of 1:200). The immunoprecipitated proteins were dissolved in sample buffer without mercaptoethanol for 15 min at 60°C. After removal of the Sepharose beads and the addition of 10% mercaptoethanol (v/v), proteins were resolved on 7.5% sodium dodecyl sulfate-polyacrylamide gels, transferred to PVDF membrane, and immunoblotted with anti-sst2A antiserum as described above.
Purification of the Phosphorylated sst2A Receptor-Metabolic labeling of cells and subsequent immunoprecipitation of the sst2A receptor was carried out as described previously (36,37). Briefly, cells (1 dish/ treatment) were incubated for 3 h in 3.5 ml of phosphate-free Dulbecco's modified Eagle's medium containing 1 mCi of [ 32 P]orthophosphate and either 1% newborn calf serum or 5 mg/ml lactalbumin hydrolysate. Hormones and pharmacological agents were then added directly to the labeling medium, and the cells were further incubated at 37°C under 5% CO 2 for the indicated times. The cells were then scraped into cold Hepes-buffered saline, pelleted, and solubilized in lysis buffer containing phosphatase inhibitors for 60 min at 4°C. The detergent lysates were centrifuged at 100,000 ϫ g for 30 min, and the protein content of the supernatants was assessed by the method of Bradford (35).
The sst2A receptors were subjected to a two-step purification consisting of lectin affinity chromatography followed by immunoprecipitation with receptor antibody (36). 2 Briefly, equal amounts of lysate protein (ϳ2 mg/ml) were incubated at 4°C for 90 min with 100 l (packed volume) of washed wheat germ agglutinin-agarose (Vector Laboratories, Inc., Burlingame, CA). Following centrifugation, the wheat germ agglutinin-agarose was washed vigorously with 30 volumes of lysis buffer, and adsorbed glycoproteins were eluted at 4°C for 90 min with 250 l of lysis buffer containing 3 mM N,NЉ,Nٞ-triacetyl-chitotriose (Sigma) and 0.5% SDS (v/v). Eluted proteins were diluted 5-fold and incubated with a 1:200 dilution of the anti-sst2A receptor antibody R2-88 at 4°C for 90 min. 2 The samples were then incubated at 4°C for 90 min with 25 l (packed volume) of protein A-Sepharose. Following centrifugation, the beads were washed as described previously (36), and the immunoprecipitated proteins were solubilized in sample buffer (60°C, 15 min) and resolved on 7.5% sodium dodecyl sulfate-polyacrylamide gels.
For phosphoamino acid analysis, SDS-PAGE-resolved proteins were transferred to PVDF membrane, and the piece of membrane containing the 32 P-labeled receptor (detected by autoradiography) was excised and incubated in 100 l of 5.7 N HCl (Pierce) at 110°C for 0.5 or 2 h (37). Phosphoamino acids were resolved by two-dimensional thin layer electrophoresis on cellulose plates (37). Receptor phosphorylation and phosphoamino acid analyses were quantitated using a PhosphorImager (37).
Other Methods-Protein A (Sigma) was covalently coupled to CNBractivated Sepharose B according to the manufacturer's instructions (Pharmacia). Antireceptor IgG was covalently coupled to protein A-Sepharose as described previously. 2

Desensitization and Internalization of sst2A
Receptor Expressed in GH-R2 Cells-Whereas GH 4 C 1 cells contain low levels (ϳ0.1 pmol/mg of cell protein) of a mixture of the somatostatin receptor subtypes sst1 and sst2 (25, 39) 3 the transfected GH-R2 cell line expresses approximately 10 pmol of the sst2A receptor/mg of protein (data not shown). Incubation of intact GH-R2 cells for 2 h at 4°C with the sst2 selective ligand [ 125 I-Tyr 3 ]SMS and increasing concentrations of unlabeled peptides showed dose-dependent inhibition of radioligand binding with both SRIF (EC 50 ϭ 2.24 Ϯ 0.31 nM) and SMS (EC 50 ϭ 38.9 Ϯ 7.5 nM). The relative affinities of the rat sst2A receptor for these two ligands thus agrees well with those of the human sst2A receptor as determined in membrane binding studies (40).
To investigate the susceptibility of the sst2A receptor to desensitization, GH-R2 cells were incubated in the absence or presence of 100 nM SMS for 30 min at 37°C. Membranes were then prepared, and the effect of pretreatment on hormonal regulation of adenylyl cyclase activity was determined (Fig. 1). SMS pretreatment did not affect either basal or VIP-stimulated adenylyl cyclase activity (data not shown). In membranes from untreated cells SMS inhibited VIP-stimulated adenylyl cyclase activity with an EC 50 of 1.2 Ϯ 0.1 nM. Maximal inhibition was 63.7 Ϯ 1.5%. Preincubation of cells with SMS attenuated both the potency (EC 50 ϭ 7.0 Ϯ 0.4 nM) and efficacy of SMS inhibition (maximum inhibition ϭ 36.2 Ϯ 2.3%). Treatment of cells with 100 nM SRIF for 30 min had the same effect as SMS; maximal inhibition was reduced from 60.1 Ϯ 2.0 to 41.2 Ϯ 2.6% while the EC 50 for SMS was increased from 1.0 Ϯ 0.2 to 6.9 Ϯ 2.5 nM. Therefore, exposure to agonist results in homologous desensitization of the sst2A receptor.
To ascertain if peptide binding induced rapid internalization of the ligand-receptor complex, cells were incubated for 1 h at 4°C with [ 125 I-Tyr 3 ]SMS to occupy cell surface receptors, washed to remove unbound peptide, and then warmed to 37°C for different periods of time (Fig. 2). A rapid, time-dependent internalization of the receptor-bound ligand occurred at 37°C reaching a steady state by 90 min (Fig. 2, upper panel). The rate of internalization was fit to the sum of two first order reactions, giving a value of 4.0 Ϯ 0.7 min for the half-time of internalization of the receptor-ligand complex (Fig. 2, lower   panel). These results show that following binding to cell surface sst2A receptors, the bound ligand is rapidly internalized in a temperature-dependent manner.
Immunodetection of sst2A Receptor in GH-R2 Cells-To determine whether the sst2A receptor was phosphorylated, we next developed and validated methods for the detection and purification of the sst2A receptor protein. The immunoblot in Fig. 3 (left panel) shows that a receptor antibody that specifically recognizes the sst2A receptor isotype reacted with a broad 71-kDa protein band in GH-R2 cell extracts. 2 This band was not detected in immunoblots incubated with preimmune sera or with immune sera in the presence of 1 M antigen peptide.
To determine whether this 71-kDa protein was the sst2A receptor, GH-R2 membranes were photoaffinity-labeled with [Leu 8 ,D-Trp 22 , 125 I-Tyr 25 ]SRIF-28 and N-5Ј-azido-2Ј-nitrobenzoyl-N-oxysuccinimide in the presence or absence of 100 nM SRIF (38). Membranes were then either directly solubilized with sample buffer (Fig. 3, middle panel) or solubilized with a non-denaturing detergent and then immunoprecipitated with preimmune or immune sst2A receptor antisera (Fig. 3, right  panel). Subsequent analysis by SDS-PAGE and autoradiography showed that a 71-kDa protein was photoaffinity-labeled in membranes and that radiolabeling was effectively competed with SRIF, as expected for a high affinity sst receptor (middle panel). The photoaffinity-labeled protein was immunoprecipitated by receptor antiserum but not by preimmune antiserum nor by immune serum in the presence of 1 M antigen peptide (right panel). Therefore, the receptor antibody effectively precipitated the 71-kDa sst2A receptor protein expressed in GH-R2 cells.
Agonist 100 nM SRIF for 15 min. Following detergent solubilization and partial purification by lectin affinity chromatography, the sst2A receptor was immunoprecipitated with receptor antibody and analyzed by SDS-PAGE and autoradiography. A low level of basal receptor phosphorylation was detectable with long film exposure (Figs. 6 and 8) or by analysis with a PhosphorImager, although it is difficult to discern in Fig. 4. Treatment of cells with SRIF increased the phosphorylation of the 71-kDa sst2A receptor protein 22 Ϯ 6-fold over basal (n ϭ 5). The 71-kDa phosphoprotein was not immunoprecipitated with either preimmune serum or with immune serum in the presence of antigen peptide (data not shown). Interestingly, it was necessary to solubilize the immunoprecipitated receptor by heating in sample buffer with 6 M urea to dissociate receptor aggregates. When either phosphorylated (Fig. 4, lane 3) or photoaffinitylabeled (data not shown) receptors were immunoprecipitated and then solubilized in SDS sample buffer under reducing conditions but without urea and heating, a higher molecular weight band was observed in addition to the 71-kDa receptor. This receptor aggregate appeared to be generated during the immunoprecipitation procedure since it was not present in GH-R2 cell extracts as analyzed by immunoblotting (Fig. 3, left panel) or in solubilized photoaffinity-labeled membranes (data not shown).
If sst2A receptor phosphorylation plays a role in either agonist-stimulated sst2A receptor desensitization or internalization, phosphorylation should be increased within the time frame of these regulatory events. As shown in Fig. 5, stimulation of sst2A receptor phosphorylation was half-maximal after a 2-min incubation with 100 nM SRIF, was maximal by 5 min, and was then maintained for at least 30 min. SMS also increased receptor phosphorylation by 2 min (data not shown) and was as effective as SRIF (Fig. 5).
Agonist stimulation of sst2A receptor phosphorylation was also concentration-dependent (Fig. 6). Phosphorylation of the sst2A receptor was significanctly elevated by 3 nM SRIF and was further increased with higher doses of SRIF up to a maximal effect with 100 nM peptide. Incubation of 32 P-labeled GH-R2 cells with concentrations of SRIF greater than 100 nM did not produce any additional stimulation (data not shown). Thus, phosphorylation of the sst2A receptor increases in response to ligand stimulation in a time-and concentration-dependent manner.
To assess the importance of sst2A receptor-G i/o coupling for SRIF-induced receptor phosphorylation, GH-R2 cells were pretreated with 100 ng/ml pertussis toxin for 24 h. This treatment abolished SRIF inhibition of VIP-stimulated cAMP accumulation (data not shown). However, SRIF-induced phosphorylation of the sst2A receptor was unaffected by pertussis toxin pretreatment (Fig. 7). Therefore, functional interaction of the sst2A receptor with pertussis toxin-sensitive G proteins is not necessary for agonist-induced sst2A receptor phosphorylation. GH 4 C 1 rat pituitary cells express m2 muscarinic and A 1adenosine receptors that couple to the same pertussis toxinsensitive effector pathways as SRIF (41). However, incubation of 32 P-labeled GH-R2 cells with either the muscarinic agonist carbachol or with the adenosine receptor agonist PIA did not cause any increase in sst2A receptor phosphorylation (Fig. 8,  top panel). Thus heterologous activation of SRIF-stimulated second messenger cascades is not sufficient to stimulate sst2A receptor phosphorylation.
The Effect of Protein Kinase Activators on sst2A Receptor Phosphorylation-To assess the effect of second messengerregulated protein kinases on sst2A receptor phosphorylation, 32 P-labeled GH-R2 cells were incubated for 15-min with either no additions (Control), 100 nM SRIF, 200 nM phorbol 12-myristate 13-acetate (PMA), or 10 M forskolin prior to sst2A receptor purification. As shown in Fig. 8 (bottom panel), a 15 min incubation with the protein kinase C activator PMA stimulated a 32-fold increase in sst2A receptor phosphorylation whereas stimulation of cAMP synthesis with forskolin had no effect. Incubation with the Ca 2ϩ ionophore ionomycin (10 nM for 15 min) also did not alter sst2A receptor phosphorylation (data not shown). Therefore, heterologous activation of protein kinase C increased sst2A receptor phosphorylation whereas stimulation of either protein kinase A or Ca 2ϩ -dependent protein kinases apparently did not.
The signal intensity observed following immunoprecipitation of the 32 PO 4 -labeled receptor is determined by 1) the stoichiometry of receptor phosphorylation and 2) the receptor concentration. The latter depends, in turn, on receptor immunoprecipitation efficiency. The sst2A antibody used in this study recognizes a region in the sst2A receptor cytoplasmic tail containing potential phosphorylation sites. Hence phosphorylation in this region could decrease the efficiency with which the sst2A receptor is immunoprecipitated and thereby prevent the detection of 32 PO 4 labeling as was shown to occur with the bombesin receptor (37). To test whether receptor phosphorylation affected the ability of the antibody to recognize the receptor protein, GH-R2 cells were incubated for 15 min with no additions or with either 100 nM SRIF, 200 nM PMA, or 10 M forskolin. Membranes were then prepared, and the ability of the antibody to recognize the different phosphorylated forms of the receptor was assessed both by immunoblotting (Fig. 9, upper panel) and by immunoprecipitation (Fig. 9, lower panel). Incubation of GH-R2 cells with SRIF, PMA, or forskolin for 15 min did not affect cellular sst2A receptor levels and did not alter the immunoprecipitation efficiency of sst2A receptor (Fig.  9). Therefore, the 32 PO 4 incorporation into the purified receptor (Figs. 3-8) accurately reflects its phosphorylation state.
Phosphoamino Acid Analysis of Phosphorylated sst2A-To identify the phosphorylated residues in the sst2A receptor, phosphoamino acid analysis was carried out with receptor from control cells and from cells incubated with either SRIF or PMA. Following acid hydrolysis of the receptor for 30 min, phosphoserine and phosphotyrosine residues were detected (Fig. 10, top  panel). Hydrolysis for 2 h facilitated detection of phosphoserine and phosphothreonine residues (Fig. 10, bottom panel). Under all three treatment conditions, the most heavily labeled residue was phosphoserine. However, all three phosphoamino acid species were detectable following analysis of sst2A receptor from unstimulated cells, and phosphorylation of all three residues was increased in a concentration-dependent manner in response to agonist stimulation (data not shown). These studies show that basal and SRIF-and PMA-stimulated sst2A receptor phosphorylation occur primarily on serine residues, although an increase in the phosphorylation of threonine and tyrosine also occurs.

DISCUSSION
The studies presented here demonstrate for the first time that hormone binding leads to rapid desensitization, internalization, and phosphorylation of the sst2A receptor.
Desensitization to native SRIF peptides as well as to the clinically used analog SMS or octreotide has been observed in some tissues and cancers but not in others, suggesting that susceptibility to desensitization depends on the sst receptor subtypes present. For example, in GH 4 C 1 cells, which express both the sst1 and sst2 receptors (4,25), prolonged exposure to SRIF does not attenuate subsequent SRIF inhibition of adenylyl cyclase (22). Since SRIF binds with high affinity to all sst receptor isotypes, the lack of effect of SRIF pretreatment on GH 4 C 1 cell responsiveness represents the sum of the functional effects on all of the SRIF receptors present. Thus, the behavior of the individual receptor isotypes cannot be deduced from these experiments. However, the analog SMS binds with high affinity only to the sst2 and sst5 receptor subtypes (1). As reverse transcriptase-polymerase chain reaction analysis showed that sst5 receptor mRNA is not expressed in GH-R2 cells 4 the diminished potency and efficacy of SMS to inhibit adenylyl cyclase activity following pretreatment of cells with this analog represents homologous desensitization of sst2A receptor function. Taken together, our present and previous (22) data indicate that in GH 4 C 1 cells sst1 and sst2 receptors act redundantly to mediate SRIF inhibition of adenylyl cyclase and that the sst1-receptor isotype is resistant to desensitization. The potential for sst2A receptor desensitization must thus be recognized not only for the physiological actions of SRIF but also when analogs such as SMS/octreotide are used clinically for the treatment of acromegaly and other sst2 receptor-posi-4 Y. Wang and A. Schonbrunn, unpublished observations. tive tumors (42).
As in the case of desensitization, most studies of SRIF receptor internalization have utilized cell lines expressing multiple sst receptors whose composite behavior was monitored using SRIF analogs that bound to several sst receptor isotypes. Hormone binding did not lead to receptor-mediated internalization in either GH 4 C 1 pituitary cells or in RINm5F insulinoma cells (34,43). In contrast, variable amounts of receptor-bound peptide were internalized in AtT-20 pituitary cells, human pituitary tumor cells, islet cells, and pancreatic acinar cells (44 -47). In these early studies the internalization of the hormone-receptor complex could not be attributed to a specific sst receptor isotype. We show here that agonist binding to the sst2A receptor triggers rapid receptor-mediated internalization. Very recently the sst2 receptor was also reported to mediate internalization of bound [Leu 8 ,D-Trp 22 , 125 I-Tyr 25 ]SRIF-28 when expressed in Chinese hamster ovary cells (29). However, in Chinese hamster ovary cells only about 20% of the receptorbound ligand was internalized even after 60 min. The explanation for the differences in the extent of receptor-mediated internalization in the two studies is not clear; internalization of the receptor-ligand complex could be influenced by both the cellular environment and the nature of the bound ligand.
The mechanisms mediating sst2A receptor desensitization and internalization are not known but, as has been postulated for other G-protein coupled receptors, may involve receptor phosphorylation (30,48,49). Indeed, this hypothesis is consistent with our findings that 1) agonist stimulation of sst2A receptor phosphorylation (t1 ⁄2 Ϸ 2 min, Fig. 5) occurs concurrently with receptor internalization in GH-R2 cells (t1 ⁄2 Ϸ 4 min, Fig. 2) and that 2) maximal sst2A receptor phosphorylation is evident under conditions used to elicit sst2A receptor desensitization.
Phosphorylation of other G protein-coupled receptors is catalyzed by two types of kinases: second messenger-activated kinases and G protein-coupled receptor kinases (GRKs). Our results suggest that agonist-stimulated phosphorylation of the sst2A receptor preferentially involves GRKs. SRIF inhibition of adenylyl cyclase, as well as regulation of other signaling pathways, is blocked by pertussis toxin, which prevents coupling of sst receptors to G i/o (5)(6)(7)(8)(9)(10)(11)(12). However, pertussis toxin pretreatment did not affect SRIF stimulation of sst2A receptor phosphorylation. These results with the sst2A receptor agree with observations with m2 muscarinic receptor, another inhibitory G protein-coupled receptor known to be phosphorylated by GRKs (50). Hormone stimulation of sst2A receptor phosphorylation by a second messenger-independent mechanism thus suggests the involvement of G protein-coupled receptor kinases, athough our experiments do not rule out the possibility that second messenger cascades activated via pertussis toxin-insensitive G proteins mediate SRIF stimulation of sst2A receptor phosphorylation. However, two other observations argue against this possibility. First, even though PIA, carbachol, and SRIF produce the same inhibitory effect on hormone secretion, adenylyl cyclase activity, and intracellular calcium in GH cells (41,51) neither PIA nor carbachol increased sst2A receptor phosphorylation. Thus activation of adenosine or muscarinic receptors does not produce the same effect as does occupancy of the sst2A receptor with agonist. Second, stimulation of sst2A receptor phosphorylation occurred at relatively high concentrations of hormone. By analogy to the ␤-adrenergic receptor, where much lower agonist concentrations are required to stimulate receptor phosphorylation by cAMP-dependent protein kinase than by GRKs (30), the necessity for high SRIF concentrations for sst2A receptor phosphorylation suggests that the agonist-occupied receptor is being preferentially phosphoryl-ated by GRKs. Indeed, Mayor et al. (52) reported that homologous desensitization of S49 mouse lymphoma cells with SRIF occurred concurrently with the translocation of a G proteincoupled receptor kinase from the cytoplasm to the plasma membrane. While these results are highly suggestive, additional experiments will be neccessary to directly demonstrate a role for G protein-coupled receptor kinases in sst2A receptor phosphorylation.
Interestingly, pharmacological activation of several second messenger-regulated kinases showed that phosphorylation of the sst2A receptor was specifically stimulated following a 15min incubation with the protein kinase C activator, phorbol 12-myristate 13-acetate (Fig. 8). Although heterologous activation of protein kinase C can regulate sst2A receptor function, it is unlikely that this kinase catalyzed the SRIF-stimulated phosphorylation because incubation with SRIF did not increase phospholipase C activity in GH-R2 cells as assayed by inositol trisphosphate accumulation. 4 Moreover, in preliminary studies we found that the amount of 32 PO 4 incorporated into the sst2A receptor following incubation with maximal concentrations of both SRIF (100 nM) and PMA (200 nM) was close to the sum of the incorporation achieved with either agent alone. This observation further argues against the involvement of protein kinase C in ligand-stimulated receptor phosphorylation. Although it is quite likely that the sst2A receptor is a substrate for protein kinase C because the receptor protein contains consensus sequences for protein kinase C phosphorylation (53), it is also possible that protein kinase C indirectly influences the phosphorylation state of the receptor by activating another kinase and/or by inactivating a protein phosphatase.
The functional consequences of protein kinase C-stimulated receptor phosphorylation are unknown. Previous studies showed that a 2-4-h exposure of pancreatic acinar cells (54) or GH 4 C 1 cells (55) to phorbol ester decreased SRIF binding, with no change in binding affinity observed in the acinar cells (54). Incubation with phorbol ester also attenuated SRIF inhibition of adenylyl cyclase activity in GH 4 C 1 cells (26). However, from these early studies one cannot assess the direct impact, if any, of protein kinase C phosphorylation on the sst2A receptor. Our observation that protein kinase C activation leads to a dramatic increase in sst2A receptor phosphorylation paves the way for the critical analysis of the mechanism by which this phosphorylation occurs as well as its biological consequences.
Although the sst2A receptor also contains putative consensus sequences for protein kinase A phosphorylation (53), stimulation of cAMP synthesis with forskolin did not increase 32 PO 4 incorporation into the sst2A receptor (Fig. 8). The ineffectiveness of forskolin in stimulating sst2A receptor phosphorylation raised the possibility that phosphorylation was not detected because of a potential experimental problem. As was the case for the bombesin receptor (37), phosphorylation of residues within the region of the sst2A receptor recognized by our antibody could inhibit receptor immunoprecipitation. Therefore, it was prudent to assess the ability of the antibody to recognize the receptor both before and after kinase activation. As shown in Fig. 9, neither SRIF, PMA, nor forskolin altered sst2A receptor immunoprecipitation efficiency. Thus, the signal intensity of phosphorylated sst2A receptor in these studies accurately reflects the incorporation of 32 PO 4 into the purified receptor.
Agonist-and PMA-stimulated sst2A receptor phosphorylations occur at multiple sites as 32 PO 4 labeling of phosphoserine, phosphothreonine, and phosphotyrosine residues was increased (Fig. 10). However, phosphoserine was the predominantly labeled residue under all hydrolysis conditions tested, suggesting that the sst2A receptor is phosphorylated primarily on serine(s). Identification of these phosphorylation sites within the intracellular regions of the receptor will allow a critical examination of the causal relationship between sst2A receptor phosphorylation, desensitization, and internalization.