Human insulin-like growth factor I receptor 950tyrosine is required for somatotroph growth factor signal transduction.

Insulin-like growth factor I (IGF-I), a growth hormone (GH)-dependent growth factor exerts feedback regulation of GH by inhibiting GH gene expression. IGF-I inhibition of GH secretion is enhanced 3-5-fold in GC rat pituitary cells overexpressing the wild type 950Tyr human IGF-I receptor which autophosphorylates appropriately. To determine the critical amino acid sequence responsible for IGF-I signaling, insertion, deletion, and site-directed mutants were constructed to substitute for 950Tyr in exon 16 of the human IGF-I receptor beta-subunit transmembrane domain. All mutant transfectants bound IGF-I with a similar Kd to untransfected cells but had markedly increased (7-34-fold) IGF-I-binding sites. GH responsiveness to IGF-I was tested in mutant transfectants. Overexpressed site-directed and insertion mutant IGF-I receptors exhibited a modest suppressive effect on GH in response to the IGF-I ligand, similar to that observed in untransfected cells. Deletion mutant (IG-FIR delta 22) (amino acid 944-965) did not transduce the IGF-I signal to the GH gene. Site-directed and insertion mutants therefore did not enhance the IGF-I response of the endogenous rat receptor, unlike the 950Tyr wild type transfectants which enhanced the IGF-I signal. All mutant transfectants, except the deletion mutant, internalized radioactive ligand similarly to 950Tyr wild type transfectants. 950Tyr of the human IGF-I receptor is therefore required for IGF-I signal transduction in the pituitary somatotroph, but not for IGF-I-mediated internalization.

Human Insulin-like Growth Factor I Receptor B 6 0 T y r~~i n e Is Required for Somatotroph Growth Factor Signal Transduction* (Received for publication, February 14, 1992) Hironori Yamasaki, Diane Prager, Saba Gebremedhin, and Shlomo MelmedS Insulin-like growth factor I (IGF-I), a growth hormone (GH)-dependent growth factor exerts feedback regulation of GH by inhibiting GH gene expression. IGF-I inhibition of GH secretion is enhanced 3-5-fold in GC rat pituitary cells overexpressing the wild type s60Tyr human IGF-I receptor which autophosphorylates appropriately. To determine the critical amino acid sequence responsible for IGF-I signaling, insertion, deletion, and site-directed mutants were constructed to substitute for e60Tyr in exon 16 of the human IGF-I receptor &subunit transmembrane domain. All mutant transfectants bound IGF-I with a similar K d to untransfected cells but had markedly increased (7-34-fold) IGF-I-binding sites. GH responsiveness to IGF-I was tested in mutant transfectants. Overexpressed site-directed and insertion mutant IGF-I receptors exhibited a modest suppressive effect on GH in response to the IGF-I ligand, similar to that observed in untransfected cells. Deletion mutant (IG-FIR A22) (amino acid 944-965) did not transduce the IGF-I signal to the GH gene. Site-directed and insertion mutants therefore did not enhance the IGF-I response of the endogenous rat receptor, unlike the eeoTyr wild type transfectants which enhanced the IGF-I signal. All mutant transfectants, except the deletion mutant, internalized radioactive ligand similarly to ssoTyr wild type transfectants. s60Tyr of the human IGF-I receptor is therefore required for IGF-I signal transduction in the pituitary somatotroph, but not for IGF-I-mediated internalization.
Insulin-like growth factor I (IGF-I)' is a major growth factor which is regulated by pituitary growth hormone (GH) (1). IGF-I inhibits GH secretion in uiuo (2)(3)(4). This IGF-I feedback control occurs at the level of the pituitary GH gene by directly suppressing GH transcription (5). IGF-I signal transduction is initiated by binding of the ligand to the IGF-I receptor.
Overexpression of the human IGF-I receptor in GC rat pituitary cells enhanced the responsiveness of the somatotroph to IGF-I, manifested by augmented suppression of GH secretion as compared to untransfected cells (6). The IGF-I receptor, * This work was supported by National Institutes of Health Grants DK 33802 (to S. M.) and DK 02023 (to D. P.). 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.
The abbreviations used are: IGF-I, insulin-like growth factor I; GH, growth hormone; NPXY, aspargine-proline-X-tyrosine; HEPES, 4-(2-hydroxyethyl)l-piperazineethanesulfonic acid; IGFIR, insulinlike growth factor-I receptor; WT, wild type; GXLY, g1ycine-Xleucine-tyrosine. therefore, plays an important role in mediating IGF-I signaling to the GH gene. The IGF-I receptor is a disulfide-linked tetrameric transmembrane protein consisting of two extracellular a-and two cytoplasmic @-subunits (a2 @2) (7). Mutation of 960Tyr and '%er in the insulin receptor &subunit reduced insulin-mediated 2-deoxyglucose uptake and thymidine incorporation into Chinese hamster ovary cells (8). In addition, phosphorylation of IRS-I, a putative cellular message for insulin action was abolished by these mutations (9,10). Because of the high (84%) sequence homology between the intracellular domains of the insulin and IGF-I receptor (7), we constructed insertion, deletion, and site-directed mutants to substitute for '"Tyr in exon 16 of the human IGF-I receptor @-subunit transmembrane domain. GC rat pituitary cells were stably transfected with these mutant IGF-I receptor cDNAs and the suppression of GH secretion in response to IGF-I was assessed. In addition, the ability of mutant IGF-I receptors to internalize the IGF-I ligand was also tested because 950Tyr is situated within an NPXY motif which has been shown to be important for low density lipoprotein receptor internalization (11). The present study demonstrates that 950Tyr is required for IGF-I signaling but not for IGF-I internalization in the somatotroph.

EXPERIMENTAL PROCEDURES
Materials-Recombinant human IGF-I (Met-59) was kindly provided by Fujisawa Pharmaceutical Co (Osaka, Japan). "'I-IGF-I was purchased from Amersham Corp. a-IR3 was purchased from Oncogene (Manhasset, NY).
Cell Culture-GC rat pituitary cells secreting rat GH were grown at 37 "C in a humidified atmosphere of 95% air, 5% Cop in Dulbecco's modified Eagle's medium supplemented with 10% (v/v) fetal calf serum.
Stable Transfection into GC Rat Pituitary Cells-Semiconfluent GC cells in 100-mm dishes were cotransfected by the CaP04 method (12) using a 1O:l ratio of linearized mutant IGF-I receptor plasmid to pSV2neo. 16 h after transfection, the cells were shocked with 15% glycerol for 2 min, washed, and incubated in serum-containing me-

GGGAATGGAGTGCTGTATCGGGAGAAGATCACCATGAGC-3'
20953 dium for 24 h. Cells were then split 1:6, and neomycin (G418) was added in fresh medium at a concentration of 400 pg/ml. Medium was replenished every 72 h. G418-resistant colonies were subcloned and further characterized.
Southern Blot-Southern hybridization was performed as previously described (6). Briefly, genomic DNA of the mutant transfectants was extracted and digested with EcoRI. The DNA fragments were separated by electrophoresis on a 0.8% agarose gel and transferred to a nylon membrane. Membrane was baked, prehybridized, and hybridized with a 32P-labeled IGF-I receptor cDNA (specific activity 4 X lo8 cpmlpg). The membrane was washed and autoradiographed for 24 h at -70 C, using Fuji film and intensifying screens.
IGF-I Binding-Binding of radiolabeled IGF-I was performed in suspension. Cells (lo6) were incubated with lZ5I-IGF-I (50,000 cpm, specific activity, 2,000 Ci/mmol) and increasing concentrations of unlabeled IGF-I or a-IR3 (anti-human IGF-I receptor antibody) in a final volume of 1 ml of binding buffer (50 mM HEPES buffer (pH 8.0), 1% bovine serum albumin, 150 mM NaC1, 1.2 mM MgS04) at 15 "C. Nonspecific binding was defined as the binding observed in the presence of excess (100 nM) unlabeled IGF-I. At the end of the 3h incubation period, cells were centrifuged, and cell-associated radioactivity was separated from free lZ51-IGF-I by adding 300 pl of icecold dibutylphthalate. Cell-associated radioactivity of the samples was then determined by gamma-counting. Calculation of total bound ligand included the free-labeled ligand.
GH Secretion-5 X lo5 cells were plated on 9-cm2 multiwells and grown for 24 h in the growth medium. Medium was then aspirated and replenished with 1.5 ml of serum-free defined medium (13) with or without 6.5 nM IGF-I. Aliquots of medium were removed at the indicated time points and assayed for rat GH using radioimmunoassay reagents supplied by the National Hormone and Pituitary Program, National Institute of Diabetes and Digestive and Kidney Diseases (Bethesda, MD).
ZGF-I Internalization-Cells (lo6) were incubated with Iz5I-IGF-I (40,000 cpm) in a final volume of 0.5 ml of Dulbecco's modified Eagle's medium (pH 7.8) with 1% bovine serum albumin and 10 mM HEPES. At the end of the 30-min incubation period at 37 "C, the suspension was acidified by the addition of 1 N HCl(30 pl) (14). After 6 min of incubation at 4 "C, intracellular (acid-resistant) or cellassociated '2SI-IGF-I was determined as described in IGF-I binding.
Degraded Iz5I-IGF-I was assessed by trichloroacetic acid precipitation (10%) of 300 p1 of incubation buffer. After 4 min at 4 "C, the degraded IZ5I-IGF-I in the supernatant was determined by gamma-counting (15).
Tyrosine Phosphorylation-Cells labeled for 16 h with [35S]methionine were washed with phosphate-buffered saline and incubated with or without IGF-I (6.5 nM) in serum-free defined medium for 1 min at 37 "C. Cells were then lysed with lysis buffer (1% Triton X-100,30 mM sodium pyrophosphate, 10 mM Tris (pH 7.6), 5 mM EDTA (pH 8), 50 mM NaCl, 0.1% bovine serum albumin, 2 mM sodium orthovanadate, 200 mM phenylmethylsulfonyl fluoride) and centrifuged for 15 min at 3000 X g. Immunoprecipitation was performed using a monoclonal anti-phosphotyrosine antibody (Ab-2, Oncogene Science, Manhasset NY) at 4 "C for 3 h. Protein A-Sepharose (Pharmacia LKB Biotechnology Inc.) and anti-mouse IgG (Sigma) were incubated with the monoclonal anti-phosphotyrosine antibody cell lysate complex for 2 h at room temperature. Immunoprecipitates were washed six times in lysis buffer, resuspended in sample buffer, boiled for 5 min, and electrophoresed on 7.5% sodium dodecyl sulfatepolyacrylamide gels.

RESULTS AND DISCUSSION
Mutant human IGF-I receptors are depicted in Fig. 1. For these experiments, 950Tyr was replaced with either Cys, Ser, Ala, Leu, or Thr (site-directed mutations). IGFIR-CRH consisted of three additional amino acids (Cys-Arg-His) inserted at position 950.22 amino acids (from 944 to 965) were deleted in IGFIR A22. GC rat pituitary cells were stably transfected with these mutant IGF-I receptor cDNAs (6). After 1 month of neomycin selection, binding of lZ51-IGF-I was assessed in  untransfected cells and in all mutant stable transfectants. Only clones exhibiting at least "-fold increase in maximal specific binding were selected for further study. Fig. 2. 4 shows that increasing amounts of unlabeled IGF-I displaced cellassociated lZ5I-IGF-I binding to all cells, with 50% displacement of maximum binding achieved by 0.6-2 nM IGF-I. When these binding displacement data were subjected to Scatchard analysis, a linear plot was obtained for all mutant transfectants, indicating the presence of a single class of high affinity receptors in these pituitary cells (Fig. 2B). The derived association constant ( K d ) for lZ5I-IGF-I binding was similar in all transfectants, ranging from 0.25 to 0.66 nM. This similarity of ligand affinity for the different mutant receptors is not surprising as the configuration of the extracellular domain (a-subunit) is expected to remain intact when the P-subunit was mutated. The derived number of mutant IGF-I receptors present on each transfectant was increased from 7-to 34-fold compared to untransfected cells, as shown in Fig. 2C.
To confirm integration of exogenous human IGF-I receptor into rat GC cell genomic DNA, Southern blot analysis of transfected cell DNA was performed. The results of DNA hybridizations obtained from the mutant-transfected and untransfected cells are shown in Fig. 3. The human receptor probe hybridized appropriately to all transfectant DNA samples but only hybridized minimally to DNA extracted from untransfected GC cells, confirming that the exogenous human IGF-I receptor cDNAs were integrated into the genomic DNA of these cells. The major 5-kilobase DNA fragment which was expected to be released from pRSV-IGFIR by digestion with EcoRI was visualized in each transfectant, in addition to randomly sized DNA bands.
Synthesis of the mature a2 P2 tetrameric IGF-I receptor is initiated by dimerization of the a0 proreceptor precursors (7, 17). Overexpressing the human IGF-I receptor in rat cells, therefore, may facilitate the formation of hybrids between the endogenous rat aP receptor and the exogenous human a@ half-receptor (18,19). In an attempt to quantitate endogenous rat IGF-I receptors present in the transfectants, we performed IZ5I-IGF-I binding in the presence of increasing amounts of the monoclonal antibody for the human IGF-I receptor, a- IR3. a-IR3 has been shown to specifically inhibit '2sI-IGF-I binding to human IGF-I receptors, but not rat IGF-I receptors (20). Using excess concentrations of a-IR3 (>1 p~) , a residual 7% nondisplaceable I2'I-IGF-I binding was still present on transfected cells (Fig. 4). This suggests that despite overexpression of the transfected human IGF-I receptor, a small population of rat/human hybrid receptors and/or endogenous rat/rat holotetrameric receptors are also present. Although a-IR3 failed to compete for labeled IGF-I binding to endogenous receptors in untransfected heterologous Chinese hamster ovary cells (21), we cannot definitively conclude that residual nondisplaceable binding in 950Tyr wild type (WT) GC transfectants represents human/rat hybrid receptors. We have recently demonstrated that most of the endogenous rat IGF- I receptors present on transfected GC cells formed hybrids with the overexpressed human IGF-I receptors.' Therefore, the residual nondisplaceable binding may represent a predominant population of human/rat hybrid receptors. Although the a-IR3 displacement experiment was performed using 95nTyr (WT) transfectants, similar hybrid formation presumably occurred in the other mutant receptor transfectants. 2 To analyze the capability of the mutant IGF-I receptors to transduce the IGF-I signal to the GH gene, cells were treated with IGF-I (6.5 nM) and GH secretion was measured. Previous dose-response experiments had demonstrated that 6.5 nM IGF-I maximally attenuates somatotroph GH secretion with no further GH suppression observed by increasing concentrations of IGF-I (6). Furthermore binding studies showed that 6.5 nM IGF-I occupies more than 90% of the available binding sites on these mutant cells. Incubation of the cells overexpressing '"Tyr (WT) receptor with IGF-I (6.5 nM) resulted in a 61% suppression of GH observed after 24 h of IGF-I treatment compared to a modest suppression of GH observed in untransfected cells ( p < 0.001) (Fig. 5). Mutant transfectants responded to IGF-I similarly to untransfected cells, with GH secretion suppressed only by 20%. A22 transfectants, however, exhibited no suppression of GH, despite the presence of the endogenous rat or hybrid IGF-I receptors. Therefore although overexpression of intact WT IGF-I receptor in GC cells enhanced IGF-I signaling, all mutant receptor transfectants failed to augment the IGF-I signal. 950Tyr, therefore, is required for IGF-I signal transduction to the GH gene.
Most endogenous rat IGF-I receptors in these cells may have formed hybrids with the overexpressed transfected human IGF-I receptors.2 The deletion mutant (IGFIR A22) failed to inhibit GH secretion in response to IGF-I despite the presumed presence of intact rat/A22 human hybrids. Interestingly, truncated platelet-derived growth factor receptors have been demonstrated to inhibit receptor autophosphorylation when forming a dimer with intact platelet-derived growth factor receptor (22). Conclusions regarding hybrid receptor function from the data shown should be interpreted with caution, as the IGFIR A22 construct results in a rela- tively gross mutation, compared with the site-specific receptor mutations.
Exon 16 of the insulin receptor interacts with IRS-1, a putative cellular messenger for transduction of the insulin signal (10). Furthermore, phosphorylation of IRS-I was also reduced by mutations of 960Tyr and '"Ser in the insulin receptor (8-10). Interestingly, this mutant insulin receptor underwent @-subunit autophosphorylation in response to insulin binding, indicating that receptor autophosphorylation is not in and of itself sufficient for phosphorylation of cytoplasmic protein (8, 9). Therefore, the transmembrane domain around 960Tyr in the insulin receptor appears to be important for interaction with the IRS-I molecule to transduce the insulin signal. IGF-I has also been shown to induce phosphorylation of IRS-I in intact cells (23)(24)(25), and exon 16 of the IGF-I receptor may therefore interact with IRS-I to facilitate IGF-I signaling.
To evaluate whether or not the transfected receptor nevertheless retained its ability to autophosphorylate as well as to phosphorylate endogenous pp183 (lo), cells were treated with IGF-I for 1 min after metabolic labeling. Fig. 6 shows that '"Tyr does in fact contain phosphorylated tyrosine residues (97 kDa) which are immunoprecipitated by monoclonal antiphosphotyrosine antibody, as well as a larger phosphorylated protein corresponding to the predicted size of pp183. As expected, the 135-kDa a-subunit is also immunoprecipitated, indicating the integrity of the covalent bonding of the receptor subunits prior to resolution. The site-directed mutant, ' "Ala, also demonstrated a similar phosphorylation pattern, albeit of lesser intensity. 950Tyr of the IGF-I receptor @-subunit is situated within an asparagine-proline-X-tyrosine (NPXY) motif. This sequence has been determined to be required for low density lipoprotein receptor internalization (11). We therefore tested the ability Cells expressing either WT or ' "Ala IGF-I receptor were metabolically labeled with [35S]methionine and then exposed to IGF-I (6.5 nM) for 1 min at 37 "C. Cells were lysed and immunoprecipitated with monoclonal anti-phosphotyrosine antibody as described and analyzed on 7.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis.
Internalized '"I-IGF-I was determined by stripping ligand bound on the cell membrane by acidification of incubation buffer. Degraded 12sI-IGF-I was determined by trichloroacetic acid precipitation. After substracting the values derived from untransfected cells, % internalization was calculated as described under "Experimental Procedures." Each value represents mean & S.E. of at least three separate experiments, each performed in duplicate.
of IGF-I receptor mutants to internalize '2sI-IGF-I. Sitedirected and insertion mutants as well as the gsoTyr (WT) receptors internalized 46-65% of exogenous lZ5I-IGF-I, while IGFIR A22 failed to internalize labeled IGF-I ligand (Fig. 7). This suggests that gs"Tyr is not required for IGF-I receptor internalization. Recently, the NPXY motif present in the insulin receptor has been shown not to contribute to ligandinduced receptor internalization (26), while the GXLY motif present upstream from the NPXY motif is critical for insulin receptor internalization (27). Although the GXLY motif is intact in the IGFIR A22, the deleted region (944-965) in IGFIR A22 may contain additional critical amino acid sequences for IGF-I internalization other than this motif.
In this study, we demonstrate that 950Tyr of the human IGF-I receptor @-subunit transmembrane domain is required for IGF-I signaling to the GH gene. The inhibitory effects of IGF-I on GH gene expression require a relatively long lagperiod, unlike the early metabolic effects of IGF-I, such as thymidine incorporation, glycogen synthesis, and glucose uptake which manifest within minutes (1). Although no pituitary cells lacking IGF-I receptors have been described, the cells employed for these studies provide a unique model to study IGF-I ligand-mediated signaling to the nucleus, independent of the metabolic actions of IGF-I. This model of polypeptide hormone secretion allows dynamic testing of the IGF-I signal in a physiologically relevant cell type.