A synthetic peptide derived from a COOH-terminal domain of the insulin receptor specifically enhances insulin receptor signaling.

The role of the insulin receptor COOH-terminal domain in the regulation of insulin signal transduction was explored with a variety of synthetic peptides. One of the peptides, termed peptide HC, whose structure corresponds to residues 1293-1307 of the insulin proreceptor sequence, enhanced insulin-stimulated autophosphorylation of the insulin receptor in cell-free systems and in semipermeabilized Chinese hamster ovary (CHO) cells that had been transfected with an expression plasmid encoding the human insulin receptor (CHO/HIRc) at concentrations where there was no detectable effect on basal autophosphorylation levels or on receptor dephosphorylation. A lipophilic analogue of peptide HC, stearyl peptide HC, added to intact CHO/HIRc cells enhanced significantly insulin-stimulated insulin receptor autophosphorylation while having no effect on ligand-stimulated receptor phosphorylation in CHO cells overexpressing either the IGF-1 receptor or epidermal growth factor receptor. Addition of stearyl peptide HC to CHO/HIRc cells resulted in a 2.4 ± 0.3-fold increase in the amount of insulin-stimulated phosphatidylinositol 3-kinase detected in anti-IRS-1 immunoprecipitates and a 2.1 ± 0.6-fold increase in the levels of tyrosine phosphorylation of mitogen-activated protein kinase in response to insulin. Finally, a derivative of peptide HC coupled to a biotin moiety was prepared and showed to bind with the β-subunit of the wild-type insulin receptor and a truncated receptor that lacks 43 amino acids from its carboxyl terminus. However, there was little binding, if any, of the peptide with the IGF-1 receptors or the epidermal growth factor receptors. Taken together, our data demonstrate that a pentadecapeptide related to the carboxyl terminus of the insulin receptor binds to the insulin receptor β-subunit and that this interaction may contribute to the increased receptor's intrinsic activity and signal transduction.

The insulin receptor is structurally related to the insulin-like growth factor-1 (IGF-1) 1 receptor (1-3), both having tyrosine kinase function with very similar properties. Highly overlapping yet different responses are elicited upon activation of these receptors with their specific ligand (4). Both insulin and IGF-1 can bind to the receptor for the other with a much lower affinity. Although controversial (5), it is thought that the differences in metabolic activities and mitogenic effects of insulin and of IGF-1 are correlated with differences in the carboxylterminal domains of insulin and IGF-1 receptor ␤-subunits, whereby their association with distinct signaling proteins may contribute to specific cellular activities. Indeed, in cells expressing a chimeric IGF-1 receptor in which its carboxyl-terminal domain is replaced by that of the insulin receptor, the stimulation of glycogen synthesis and mitogen-activated protein (MAP) kinase pathway are correlated with the activities of the insulin receptor (6). Likewise, substitution of the insulin receptor carboxyl terminus with that of the IGF-1 receptor severely affects insulin-stimulated mitogenic responses (7,8).
Thus, it appears that kinase regulation by sequences within the COOH terminus may, in part, be involved in defining the specificity of downstream signaling events. Of interest is the fact that antipeptide antibodies directed against epitopes in the region 1294 -1317 2 of the insulin receptor inhibit in vitro insulin receptor kinase activity toward an exogenous substrate (9). This region in particular is poorly conserved between the two receptors, and the insulin receptor sequence contains a serine residue at position 1315; phosphorylation of that residue after exposure to phorbol ester (10,11) provides the potential for modulation of insulin receptor signaling through its carboxylterminal tail. These studies thus suggest that the differences in this region of the carboxyl-terminal domains of the insulin receptor and IGF-1 receptor may be functionally significant.
As an initial step in delineating those features of the insulin receptor that are responsible for the specific nature of insulin signaling, we have synthesized a series of peptides whose sequences correspond to COOH-terminal regions that are found in human and rodent insulin receptors but not in the IGF-1 receptors (12,13) and showed that peptide HC (amino acid residues 1293-1307) was able to bind to the insulin receptor, thereby potentiating the ability of insulin to activate specifically receptor tyrosine kinase and diverse signaling pathways.

MATERIALS AND METHODS
Transfection and Cell Culture-CHO cell lines used in this study have been previously described (14 -17). These include the CHO cell lines stably co-transfected with a plasmid containing a neomycin resistance gene driven by a SV40 promoter and a plasmid encoding the normal human insulin receptor without exon 11 (CHO/HIRc), the trun-* 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.
¶ To whom correspondence should be addressed: Diabetes Section, Rm. 2B-01, Box 23, Gerontology Research Center, National Institute on Aging, 4940 Eastern Ave., Baltimore, MD 21224. Tel.: 410-558-8198; Fax: 410-558-8381; E-mail: Bernierm@vax.grc.nia.nih.gov. 1 The abbreviations used are: IGF-1, insulin-like growth factor-1; IRS-1, insulin receptor substrate-1; EGF, epidermal growth factor; PI 3-kinase, phosphatidylinositol 3-kinase; MAP kinase, mitogen-activated protein kinase; CHO, Chinese hamster ovary; BSO-COES, bis [ Derek LeRoith (National Institutes of Health, Bethesda, MD). CHO cells overexpressing both the insulin receptors and the EGF receptors (CHO/EI) were generated as follows. The expression plasmid containing the cDNA for the human insulin receptor, pCVSVHIRc, was kindly provided by Dr. Peter A. Wilden (University of Missouri, Columbia, MO). The plasmid containing hygromycin resistant gene pSVHPH (ATCC, Bethesda, MD) was linearized with BamHI. Five g of PvuI-(Life Technologies Inc.) linearized pCVSVHIRc and 0.5 g of linearized pSVHPH were cotransfected into CHO/EGFR cells (2.75 ϫ 10 6 cells/ml) by electroporation using a Gene Pulser (Bio-Rad) at 300 mV and 960 microfarads. Aliquots of cells were then diluted and plated in multiwell culture dishes. After 24 h, cells were exposed to 300 g/ml hygromycin, which has been tested to be 100% lethal after 5 days in nontransfected controls. Colonies resistant to hygromycin were identified after 5 days and allowed to grow for another 10 days. Cells were then passaged and tested for the expression of insulin receptor. Four independent CHO/EI clones were obtained that expressed large numbers of both the EGF and insulin receptors. All cell lines were grown on tissue culture plates in F-12 medium containing 10% fetal bovine serum and maintained in a humidified incubator with 5% CO 2 at 37°C. Cell culture reagents were purchased from Life Technologies, Inc. The approximate number of receptors for insulin, EGF, or IGF-1 in their respective cell lines was determined by Scatchard analyses of competitive binding studies with labeled ligand and the homologous unlabeled peptide.
Peptide Synthesis-All peptides used in this study (see Fig. 1A for amino acid sequences) were prepared by solid phase synthesis employing Fmoc chemistry on an Applied Biosystems 430A automatic peptide synthesizer as described (18). Peptide HC was modified to increase its lipophilicity by incorporation of the C-18 aliphatic stearic acid (Sigma) at the amino terminus using the following conditions for automation on the peptide synthesizer: a portion of the protected peptide resin was treated with 20% piperidine to remove the NH 2 -terminal Fmoc protection group, washed extensively with N-methyl-pyrrolidone, and dried. Stearic acid, which had been converted into the symmetrical anhydride, reacted with the peptide-bound resin, as described (18). The release from the solid phase support matrix and deprotection of amino acid side chains were accomplished by treating the fatty acylated peptide on the resin with 83.5% trifluoroacetic acid containing 4.5% phenol, 4% thioanisole, and 2% ethanethiol. Diethyl ether was used at 0°C to remove the unreacted reagents. Stearyl peptide HC was loaded onto a C18 column equilibrated in 20% solvent B for reverse phase high pressure liquid chromatography and purified as follows. Solvent A contained 10 mM trifluoroacetic acid, and solvent B contained 10 mM trifluoroacetic acid, 80% acetonitrile. A linear gradient was run over 60 min of 20 -80% solvent B. The flow rate was 1 ml/min. The eluant was monitored at 210 nm. The composition of stearyl peptide HC was confirmed by mass spectral analysis. Biotinylated HC peptide (SSHCQREEAGGRDGG-6K(biotin)-amide; where 6 represents amino caproic acid) was synthesized by Research Genetics (Huntsville, AL), and its composition was confirmed by mass spectral analysis. Finally, 3S-peptide I (TRDIY(S)ETDY(S)Y(S)RK-amide) is a tris-tyrosine-O-sulfated (Y(S)) peptide that was chemically synthesized in our laboratory, as described previously (18). It involved the solid phase tyrosine sulfation of resinbound peptides where all three tyrosine residues were incorporated as Fmoc-tyrosine-OH.
In Vitro Phosphorylation of the Insulin Receptor-The preparation of membranes from CHO/HIRc cells and the partial purification of the insulin receptors by chromatography on WGA-bound agarose were essentially as described (19). WGA-purified insulin receptors were preincubated with 100 nM insulin and a range of concentrations of peptides in a buffer containing 100 mM Hepes, pH 7.4, 5 mM MnCl 2 , and 1 mM dithiothreitol for 10 min at room temperature. The phosphorylation reaction was initiated by the addition of 40 M [␥-32 P]ATP (5 cpm/fmol) and continued for 5 min. The reaction was stopped by the addition of an equal volume of 2 ϫ Laemmli sample buffer (20). The proteins were resolved by electrophoresis on precast 4 -12% gradient polyacrylamide gel (Novex; San Diego, CA) under reducing conditions followed by autoradiography of the dried gels at Ϫ70°C using enhancing screens and Hyperfilm-MP films (Amersham Corp.). The levels of 32 P associated with the receptor ␤ subunit were quantitated using a Betascope 603 blot analyzer (Betagen; Waltham, MA).

Autophosphorylation and Dephosphorylation of Insulin Receptors in
Permeabilized Cells-The phosphorylation of the insulin receptors in semipermeabilized CHO/HIRc cells was carried out, as recently described (21). In brief, confluent monolayers of CHO/HIRc cells were serum-starved for 4 h, semipermeabilized with 35 g/ml digitonin for 20 min at room temperature, and transferred to 6°C for 5 min. Cells were treated with insulin (0 or 100 nM) for 15 min at 6°C, and the phosphorylation reaction was initiated by the addition of 100 M [␥-32 P]ATP and 4 mM MnCl 2 . Three min later, 50 M peptide HC or stearyl peptide HC was added for 5 min. Cell extracts were prepared and clarified; cell lysates were immunoprecipitated with a monoclonal anti-insulin receptor antibody (␣IR; clone 29B4, Oncogene Science, Uniondale, NY) adsorbed on protein A/protein G-agarose (Oncogene Science). Proteins were analyzed by SDS-polyacrylamide gel electrophoresis under reducing conditions followed by autoradiography. The effect of peptide HC or 3S-peptide I on receptor dephosphorylation was studied by preincubating semipermeabilized CHO/HIRc cells with 100 nM insulin and [␥-32 P]ATP/Mn 2ϩ for 5 min at 6°C before the peptide addition. Five min later, the dephosphorylation reaction was then initiated by adding a chase solution that contained at final concentrations 4 mM unlabeled ATP and 20 mM EDTA. The reaction was stopped by the rapid removal of incubation medium and immersion of the culture dishes in liquid nitrogen. The cells were lysed, and the insulin receptors were recovered by chromatography on WGA-bound agarose. The eluted proteins were analyzed, and the levels of 32 P associated with the receptor ␤-subunit were quantitated as described above.
Receptor Autophosphorylation in Intact Cells-Confluent cells grown in 35-mm culture dishes were incubated in F-12 medium devoid of serum for 16 h at 37°C, after which the medium was replaced with 0.5 ml of F-12 medium containing a range of concentrations of stearyl peptide HC or 1 mM vanadate. Cell monolayers were incubated for 1 h at 37°C before the addition of insulin (100 nM), IGF-1 (100 nM), or EGF (1 nM) to CHO/HIRc cells, CHO/IGF-1R cells, or CHO/EGFR cells, respectively. CHO/EI cells were treated with both 100 nM insulin and 1 nM EGF. One minute later, the incubation was terminated by removing the fluid and immersing the cell dishes in liquid nitrogen. Cells were then scraped into radioimmune precipitation buffer (20 mM Tris (pH 7.5) containing 137 mM NaCl, 1 mM Na 3 VO 4 , 100 mM NaF, 0.1% SDS, 0.5% sodium deoxycholate, 1% Triton X-100, 0.02% NaN 3 , 0.2 mM phenylmethylsulfonyl fluoride, 1 mM benzamidine, 8 g/ml aprotinin, 2 g/ml leupeptin), and the cell lysates were incubated for 1 h on ice. After centrifugation (17,000 ϫ g, 20 min) to remove cell debris, the clarified supernatant from CHO/HIRc cell lysates was incubated with ␣IR antibody, while monoclonal anti-phosphotyrosine (␣PY; clone 4G10, Upstate Biotechnology Inc. (UBI), Lake Placid, NY) antibody was used with clarified CHO/IGF-1R, CHO/EGFR, and CHO/EI cell lysates. After an overnight incubation at 4°C, the immune complexes were precipitated with protein A/protein G-agarose beads; the beads were washed, and the antibody receptor complexes were eluted with Laemmli sample buffer containing 5% 2-mercaptoethanol. After SDSpolyacrylamide gel electrophoresis, proteins were electrotransferred to polyvinylidene difluoride (PVDF) membranes (Novex) and probed with polyclonal ␣PY antibody (UBI). The immunoprecipitates from CHO/EI cells were immunoblotted with a polyclonal antibody raised against a peptide corresponding to residues 1293-1306 of the insulin receptor (␣CT-IR; Ref. 21) or ␣EGFR antibody (PC19 -2; Oncogene Science). The immunoblots were visualized by using anti-rabbit immunoglobulin G coupled to horseradish peroxidase and the enhanced chemiluminescence (Amersham) detection kit. Quantification was performed using ImageQuant TM software (version 3.3) on a Molecular Dynamics Densitometer (Sunnyvale, CA).
Phosphatidylinositol 3Ј-Kinase Assay-In vitro phosphorylation of phosphatidylinositol was carried out in immune complexes as described previously (22). Confluent cultures of CHO/HIRc cells were serumstarved for 16 h, incubated for 1 h with 50 M stearyl peptide HC, and then treated without or with 3 nM insulin for 1 min. The cells were lysed in 20 mM Tris-Cl, pH 8.0, containing 137 mM NaCl, 1 mM MgCl 2 , 1 mM CaCl 2 , 10% glycerol, 1% Nonidet P-40, 150 M orthovanadate, 1 mM phenylmethylsulfonyl fluoride, and 2 mM dithiothreitol. The cell lysates were clarified by centrifugation and incubated with polyclonal IRS-1 antibody (UBI) overnight at 4°C. Following precipitation and washes, the immune complexes were incubated with 20 g of phosphatidylinositol and [␥-32 P]ATP (40 M, 10 cpm/fmol) in a buffer containing 20 mM Tris-Cl, pH 7.5, 150 mM NaCl, 0.4 mM EGTA, and 10 mM MgCl 2 for 20 min at room temperature. Reactions were stopped with 20 l of 6 N HCl and 160 l of chloroform/methanol (1:1) and centrifuged. Aliquots of the lower organic phase were applied to a silica gel TLC plate (Merck) that had been treated with 1% potassium oxalate in 40% methanol. TLC plates were developed in chloroform/methanol/water/ammonia (60:47: 11.3:2 (v/v/v/v)) and dried, and the lipids were detected with iodine vapor. 32 P-Labeled spots were visualized by autoradiography and quantitated on a Packard InstantImager.
Tyrosine Phosphorylation of MAP Kinases-Cells were serumstarved for 4 h and subsequently incubated in the absence or presence of 50 M stearyl peptide HC for 1 h prior to the addition of 100 nM insulin for 1 min. After lysis of the cells in radioimmune precipitation buffer, clarified cell extracts were subjected to immunoprecipitation with monoclonal ␣PY antibody. Immunoprecipitates were then electrophoresed on SDS-polyacrylamide gels under reducing conditions, and Western immunoblotting was performed on a PVDF membrane using 5% (w/v) fat skimmed milk in 20 mM Tris, pH 7.4, containing 150 mM NaCl and 0.1% Tween 20 to block nonspecific sites. The membrane was incubated with a polyclonal MAP kinase antibody that reacts with the 44-kDa Erk-1 and, to a lesser extent, the 42-kDa Erk-2 protein (sc-94, Santa Cruz Biotechnology, Santa Cruz, CA) followed by detection with the ECL chemiluminescence system.
Chemical Cross-linking Protocol-Semipermeabilized cell monolayers were incubated with insulin (100 nM), IGF-1 (100 nM) or both insulin (100 nM) and EGF (5 nM), as indicated, in the presence of 50 M biotinylated peptide HC with or without a 10-fold excess of unmodified peptide HC for 15 min at 6°C. The phosphorylation reaction was initiated by the addition of 100 M ATP and 4 mM MnCl 2 . Fifteen min later, 0.2 mM of the homobifunctional cross-linking reagent bis[2-(succinimidooxycarbonyloxy)ethyl]sulfone (BSO-COES; Pierce) was added for 15 min. Cell extracts were prepared, and the clarified cell lysates were immunoprecipitated with ␣IR, monoclonal anti-IGF-1 receptor (␣IGF-1R, clone 3B7, Santa Cruz), or polyclonal anti-EGF receptor (␣EGFR; UBI) antibodies adsorbed on protein A/protein G-agarose. The immunoprecipitated proteins were electrophoresed on SDS-polyacrylamide gels under reducing conditions; the proteins were electrotransferred to PVDF membrane and probed with enzyme-linked streptavidin (Vector Laboratories Inc., Burlingame, CA), or subjected to immunoblotting with polyclonal ␣PY, ␣CT-IR, ␣EGFR, or a polyclonal antibody raised against a peptide corresponding to the major autophosphorylation domain of both the insulin receptor and IGF-1 receptor (␣-G86.2, kindly provided by C. Ramachandran, Merck Frosst Canada). The blots were visualized using the ECL chemiluminescence detection system.
Statistical Analysis-Significant differences were determined by an analysis of variance coupled to Fisher's PLSD test for multiple mean comparison using StatView 4.01 (Abascus Concepts, Inc.).

RESULTS
Peptide HC Enhances Insulin-stimulated Autophosphorylation of the Insulin Receptor-In order to examine the ability of the COOH-terminal domain of the insulin receptor to modulate insulin receptor autophosphorylation in vitro, the peptides HC, HC scramble, HC-N5, and CT-24 were prepared, based on the corresponding sequences that are present in the insulin receptor but not in the receptor for IGF-1 (Fig. 1A). A 98-amino acid COOH-terminal fragment of the insulin receptor (residues 1245-1343) has previously been shown to stimulate the autophosphorylation activity of the insulin receptor nearly 3-fold (23). Using partially purified insulin receptors from CHO/HIRc cells, peptide HC (based on the corresponding amino acid residues 1293-1307) increased the insulin-stimulated receptor autophosphorylation by nearly 2-fold (Fig. 1A), while having no effect on basal receptor autophosphorylation activity (data not shown). As shown in Fig. 1B, preincubation with peptide HC increased insulin-stimulated receptor autophosphorylation in a concentration-dependent manner, with a maximum ranging between 50 and 100 M. Two structural variants of peptide HC (HC scramble and HC-N5) were used and have been found to be inactive. In contrast to the enhancing effect of peptide HC on autophosphorylation, CT-24 peptide exerted an inhibitory activity.
Next, peptide HC was tested for its ability to enhance insulin-stimulated insulin receptor autophosphorylation in semipermeabilized cells. This treatment gives small molecules free access to the intracellular milieu and yet retains much of the membrane architecture of the whole cells (17). In control cells, insulin (100 nM) alone caused a 2.5 Ϯ 0.2-fold increase in receptor phosphorylation (Fig. 2). Similar to the effect in cell-

FIG. 1. Effect of synthetic peptides on insulin-stimulated autophosphorylation of partially purified insulin receptors.
A, WGApurified insulin receptors were preincubated with insulin (100 nM) and 50 M of the indicated peptides for 10 min at room temperature. The phosphorylation reaction was then initiated by adding [␥-32 P]ATP for 5 min as described under "Materials and Methods." Proteins were analyzed by SDS-polyacrylamide gel electrophoresis under reducing conditions, and the levels of 32 P associated with the receptor ␤-subunit were determined by Betagen counting of the dried gels. Each value is expressed relative to that determined in the absence of peptide. Numbers in parentheses refer to the number of independent observations. *, p Ͻ 0.01 versus no peptide. B, a dose-response curve for the stimulation of insulin receptor autophosphorylation by peptide HC was generated by the experimental protocol described above. Bars represent the mean Ϯ range of a representative experiment performed in duplicate. The autoradiogram of the 32 P-labeled receptor ␤-subunit is shown in the inset. Comparable results were obtained in a separate experiment. free systems, 50 M peptide HC doubled the insulin-stimulated insulin receptor autophosphorylation when compared with semipermeabilized cells incubated with insulin alone. This potentiating effect was maintained after incorporation of a stearyl moiety at the NH 2 terminus of peptide HC. Neither peptide had any effect on the insulin receptor autophosphorylation in unstimulated cells (data not shown).
To explore the possibility that peptide HC enhances the insulin receptor autophosphorylation by causing a decrease in the dephosphorylation of the receptor ␤-subunit, we have employed a pulse-chase technique using semipermeabilized CHO/ HIRc cells. In this cell model, it has been previously demonstrated that the active, phosphotyrosine-containing insulin receptor is dephosphorylated rapidly by protein-tyrosine phosphatases (17). As shown in Fig. 3, less than 45% of the total phosphotyrosyl insulin receptors present remained after a 3-min dephosphorylation reaction. Using 25 M peptide HC, it was found that peptide HC failed to prevent the loss of 32 P from the phosphorylated receptors (Fig. 3). This lack of effect was observed in the presence of peptide HC up to 100 M (data not shown). These studies were extended with the use of 3S-peptide I, a tris-sulfotyrosyl peptide, known to inhibit proteintyrosine phosphatases (18). In contrast to our findings with peptide HC, 3S-peptide I caused a 85% decrease in the dephosphorylation of the insulin receptor (Fig. 3). This suggests that the increase in insulin receptor autophosphorylation by peptide HC cannot be accounted for by an altered receptor dephosphorylation.
Effect of Stearyl Peptide HC on Receptor Tyrosine Phosphorylation in Intact Cells-The incorporation of a fatty acid moiety at the NH 2 terminus of peptide HC dramatically increased its lipophilicity (data not shown). Indeed, this approach has been successfully used in our laboratory, where we showed that stearyl 3S-peptide I was a potent inhibitor of insulin receptor dephosphorylation in intact cells (18,24).
To investigate the effect of peptide HC in intact cells, stearyl peptide HC was prepared and added to CHO/HIRc cells. The extent of ligand-stimulated tyrosine phosphorylation of the insulin receptors was then compared with that of vanadate, a known inhibitor of protein-tyrosine phosphatases. CHO/HIRc cells were exposed to a range of concentrations of stearyl peptide HC or vanadate (1 mM) for 1 h, followed by stimulation with insulin. Exposure of cells to stearyl peptide HC caused a concentration-dependent increase in insulin-stimulated insulin receptor autophosphorylation, with a maximal effect at 25 M (Fig. 4). Consistent with previously published data, vanadate enhanced greatly the steady-state levels of phosphorylation of Three minutes later, 50 M peptide HC or stearyl peptide HC was added for 5 min. Cell extracts were prepared, and clarified cell lysates were immunoprecipitated with ␣IR and protein A/protein G-agarose. Proteins were analyzed by SDS-polyacrylamide gel electrophoresis, visualized, and quantitated with a Betagen counter. The data are presented as Betagen counts and represent the average Ϯ S.E. from 4 -6 independent observations. *, p Ͻ 0.01 versus insulin alone.

FIG. 3. Lack of inhibition of insulin receptor dephosphorylation by peptide HC.
Semipermeabilized CHO/HIRc cells were treated with 100 nM insulin for 15 min at 6°C, and the autophosphorylation with [␥-32 P]ATP was carried out for 5 min. Thereafter, cells were incubated without or with peptide HC (25 M) or 3S-peptide I (100 M) for an additional 5 min. The dephosphorylation reaction was initiated by the addition of a chase solution (unlabeled ATP and EDTA) for 3 min at 6°C. The cells were lysed, and the insulin receptors were recovered by WGA chromatography and separated by SDS-polyacrylamide gel electrophoresis. The levels of 32 P associated with the receptor ␤-subunit were determined by Betagen counting of the dried gels. The data plotted are expressed as the percentage of phosphorylated receptors remaining after dephosphorylation in the presence or absence of peptide and represent the average Ϯ range of a representative experiment where each point was determined using two dishes.

FIG. 4. Insulin-stimulated autophosphorylation of insulin receptor in cells treated with stearyl peptide HC.
Confluent monolayers of CHO/HIRc cells were serum-starved for 16 h, treated with a range of concentrations of stearyl peptide HC or 1 mM vanadate for 1 h, and stimulated with 100 nM insulin for 1 min. The phosphorylation reaction was terminated with liquid nitrogen. The cells were lysed, and the insulin receptors were immunoprecipitated with ␣IR antibody. The immunoprecipitates were resolved by SDS-polyacrylamide gel electrophoresis, transferred to PVDF membrane, and immunoblotted using a polyclonal ␣PY antibody. The data are presented as -fold stimulation above control (treated with insulin alone) and represent the average Ϯ S.E. from two to three separate experiments where each point was determined using two dishes. *, p Ͻ 0.01 versus insulin alone. the insulin receptor by insulin. Neither the basal nor insulinstimulated receptor autophosphorylation was affected by stearic acid (data not shown).
To determine the specificity of stearyl peptide HC, CHO cells expressing large numbers of receptors for IGF-1 (CHO/IGF-1R cells) or EGF (CHO/EGFR cells) were incubated with 50 M stearyl peptide HC for 1 h, followed by the addition of their cognate ligand. In contrast to the insulin receptor, there was no enhancement in the ligand-stimulated autophosphorylation of the IGF-1 receptor (Fig. 5A) or the EGF receptor (Fig. 5B). Inhibition of cellular protein-tyrosine phosphatases by vanadate led to an increase in ligand-stimulated receptor autophosphorylation in both cell lines.
To gain further insight into the selective activation of insulin receptor functions by peptide HC, CHO cells expressing large numbers of both insulin receptors and EGF receptors were generated (CHO/EI cells) and treated with 50 M stearyl peptide HC; then receptor phosphorylation was assessed following the addition of insulin and EGF. The results shown in Fig. 6 indicate that stearyl peptide HC caused a 1.92 Ϯ 0.07-fold increase in insulin receptor autophosphorylation in insulinstimulated CHO/EI cells but was without effect on ligandstimulated EGF receptor phosphorylation.
Activation of Phosphatidylinositol 3-Kinase Activity-An immediate effect of insulin receptor activation is the phosphorylation of IRS-1, one of the major substrates of insulin receptor kinase. Tyrosine-phosphorylated IRS-1 can associate with and stimulate PI 3-kinase activity (25,26). An analysis of PI 3-kinase activity in anti-IRS-1 immunoprecipitates revealed that insulin increased by 5.7 Ϯ 0.4 times the incorporation of 32 P into phosphatidylinositol when compared with unstimulated CHO/HIRc cells (Fig. 7). In agreement with our findings with the insulin receptor autophosphorylation, stearyl peptide HC caused a further 2.4 Ϯ 0.3-fold increase in PI 3-kinase activity in insulin-stimulated cells (Fig. 7) but was without effect on basal PI 3-kinase activity (data not shown). Neither the basal nor insulin-stimulated PI 3-kinase activity was affected by 50 M stearic acid (data not shown).
Insulin-stimulated Tyrosine Phosphorylation of MAP Kinase-MAP kinase activity is rapidly stimulated in response to insulin and other growth factors via a mechanism that involves both tyrosine and serine/threonine phosphorylation of the enzyme itself (27,28). In order to determine whether stearyl peptide HC affects tyrosine phosphorylation of MAP kinase, ␣PY immunoprecipitates from CHO/HIRc cells incubated under various conditions were analyzed by Western blot with an anti-MAP kinase antibody. As shown in Fig. 8, insulin alone led to a 2.3 Ϯ 0.3-fold increase in phosphorylation of MAP kinase (and its activity thereof), when compared with control unstimulated cells, while the addition of 50 M stearyl peptide HC caused a further 2.1 Ϯ 0.6-fold increase in MAP kinase phosphorylation in cells stimulated with insulin.
Binding of Biotinylated Peptide HC to the Insulin Receptors-To determine whether the effect of peptide HC on insulin stimulation of receptor autophosphorylation and signaling were due to a direct association between peptide HC and a domain of the receptor, semipermeabilized CHO/EI cells were incubated in the presence of 50 M biotinylated peptide HC and then exposed to 0.2 mM BSO-COES, a bifunctional cross-linking reagent. Extracts were prepared from insulin-and EGFstimulated CHO/EI cells and subjected to immunoprecipitation with either ␣IR or ␣EGFR antibodies. Immunoprecipitates were analyzed for biotinylation by blotting with enzyme-conjugated streptavidin. A clear signal could be detected in ␣IR immunoprecipitates (Fig. 9A). We observed a ϳ95-kDa biotinylated protein, suggesting binding of biotinylated peptide HC to the ␤-subunit of the insulin receptor with no modification of the receptor ␣-subunit (Fig. 9A, lane 1). Cross-linking of the bioti-

FIG. 5. Ligand-stimulated tyrosine phosphorylation of growth factor receptors in cells treated with stearyl peptide HC. Confluent monolayers of CHO/IGF-1R cells (A) and CHO/EGFR cells (B)
were serum-starved for 16 h, treated with (ϩ) or without (Ϫ) 50 M stearyl peptide HC or 1 mM vanadate for 1 h, and stimulated with 100 nM IGF-1 or 1 nM EGF for 1 min. Cell extracts were prepared and subjected to immunoprecipitation with a monoclonal ␣PY antibody. The immunoprecipitated proteins were separated by SDS-polyacrylamide gel electrophoresis, transferred to PVDF membrane, and immunoblotted with a polyclonal ␣PY antibody. The data are presented as -fold stimulation above control (treated with ligand alone) and represent the average of two independent experiments.
FIG. 6. Selective effect of stearyl peptide HC on ligand-stimulated receptor tyrosine phosphorylation in CHO/EI cells. Confluent monolayers of CHO/EI cells were serum-starved for 16 h, treated with (ϩ) or without (Ϫ) 50 M stearyl peptide HC for 1 h, and stimulated with 100 nM insulin and 1 nM EGF for 1 min. Cell extracts were prepared and subjected to immunoprecipitation with a monoclonal ␣PY antibody. The immunoprecipitated proteins were separated by SDSpolyacrylamide gel electrophoresis, transferred to PVDF membrane, and immunoblotted with ␣CT-IR or ␣EGFR antibodies. The data are presented as -fold stimulation above control (treated with both insulin and EGF) and represent the average Ϯ S.E. from three separate experiments. *, p Ͻ 0.01 versus insulin alone. nylated peptide HC to the ␤-subunit was specific because it can be blocked by excess of unmodified peptide HC (Fig. 9A, lane 2). Further evidence for the specific nature of this association was provided by the absence of complex between biotinylated peptide HC and the EGF receptor (Fig. 9A, lane 3). To determine whether both the insulin receptor and EGF receptor were activated upon ligand stimulation, immunoblotting analysis with ␣PY antibodies or with receptor-specific antibodies was performed on the same immunoprecipitates. Stimulation of semipermeabilized CHO/EI cells with insulin and EGF induced a marked increase in the tyrosine phosphorylation of ϳ95and 170-kDa proteins, corresponding to the apparent size of the insulin and EGF receptors (Fig. 9B). Immunoblotting of the same samples with specific antireceptor antibodies confirmed their identities (Fig. 9, C and D). The binding of peptide HC to the insulin receptors that was observed in semipermeabilized cells can be recreated in vitro using a semipurified preparation of insulin receptors (data not shown).
When semipermeabilized CHO cells expressing a mutant insulin receptor lacking 43 amino acids from the carboxyl terminus of the ␤-subunit (⌬43) were incubated with biotinylated peptide HC, binding of the peptide to the truncated receptors was observed at a level that was higher than in CHO/EI cells (Fig. 10, lane 1 and 2). In contrast, there were little binding, if any, of the peptide to the ␤-subunit of the IGF-1 receptor in CHO/IGF-1R cells (Fig. 10, lane 3). Immunoblotting studies, with an antibody directed against the major autophosphorylation domain of both the insulin and IGF-1 receptors, showed that the three cell lines expressed similar numbers of receptors (data not shown). Taken together, these data demonstrate that peptide HC binds specifically to the insulin receptor ␤-subunit in a region other than the carboxyl-terminal 43 amino acids. DISCUSSION Earlier studies point to the fact that the specificity of insulin and IGF-1 signaling is generated at the receptor level, where defined receptor subdomains interact with specific sets of regulatory molecules, some of which are common to both receptors (29 -36) whereas others are possibly unique. There are regions of the ␤-subunit carboxyl-terminal domain of the insulin receptor that are strikingly different from the corresponding sequences of the IGF-1 receptor (3) and, therefore, indicate the potential for the generation of specific cellular responses. Indeed, recent studies showed that a peptide equivalent to the carboxyl-terminal 98 amino acids of the insulin receptor (sequence 1245-1343 of the proreceptor) stimulates the kinase FIG. 7. Effect of stearyl peptide HC on activation of PI 3-kinase by insulin. Serum-starved monolayers of CHO/HIRc cells, pretreated with (ϩ) or without (Ϫ) 50 M stearyl peptide HC for 1 h, were stimulated or not stimulated with 3 nM insulin for 1 min. Cell lysates were immunoprecipitated with an anti-IRS-1 antibody, the immune pellets were washed, and the associated PI 3-kinase activity was assayed in vitro by 32 P incorporation into phosphatidylinositol. The resulting phosphatidylinositol 3-phosphate was resolved by thin layer chromatography (inset). Data from the inset are presented as -fold stimulation above control (treated with vehicle alone) and represent the average Ϯ range from a typical experiment where each point was determined using two dishes. Similar results were obtained in a separate experiment.
FIG. 8. Effect of stearyl peptide HC on insulin-stimulated tyrosine phosphorylation of MAP kinase. Serum-starved monolayers of CHO/HIRc cells, pretreated with (ϩ) or without (Ϫ) 50 M stearyl peptide HC for 1 h, were stimulated or not stimulated with 100 nM insulin for 1 min and then lysed as described under "Materials and Methods." The cell lysates were immunoprecipitated with ␣PY antibodies. The immunocomplexes were separated by SDS-polyacrylamide gel electrophoresis followed by Western immunoblotting using a MAP kinase-specific antibody. The data are presented as -fold stimulation above control (treated with vehicle alone) and represent the average Ϯ range from two independent observations. activity of purified insulin receptors but has no effect on the IGF-1 receptor activity (23). The role of the carboxyl-terminal domain of the insulin receptor in transducing insulin signals has been restricted, in most part, to the use of deletion mutant receptors. Several studies have reported that a region within the carboxyl terminus downstream of residue 1274 of the insulin proreceptor sequence has defined effects on kinase function by promoting both catalytic efficiency and stability of the receptor tyrosine kinase activity (37)(38)(39)(40)(41)(42)(43).
In the present study, we have observed that a specific pentadecapeptide whose amino acid sequence is derived from the carboxyl-terminal region of the insulin receptor (residues 1293-1307) enhances insulin-stimulated receptor autophosphorylation under a number of experimental conditions. This peptide, termed peptide HC, has been used along with other peptides related to the COOH terminus of the receptor to characterize its "specific" effect on kinase function. A truncated peptide (HC-N5) and an analog of peptide HC with a scrambled sequence lack the ability to enhance autophosphorylation, and on this basis we conclude that primary sequence requirements do participate in peptide HC response. Consistent with our results, Kaliman et al. (9) have recently shown that binding of antipeptide antibodies to the COOH-terminal domain of the insulin receptor modulated receptor-mediated substrate phosphorylation. They reported that antipeptide antibodies against the sequence 1294 -1317 (encompassing peptide HC sequence) inhibit receptor kinase activity, possibly because binding of antipeptide antibodies causes perturbation of the COOH-terminal domain and affects exogenous substrate phosphorylation. Thus, this region of the receptor appears to be involved in regulation of insulin receptor kinase activity.
CT-24 peptide (residues 1300 -1323) exerts an inhibitory effect when assayed in vitro. It is interesting to note that the latter peptide contains two tyrosines, corresponding to residues 1316 and 1322 of the second major autophosphorylation domain. It is known that the rate of autophosphorylation with respect to activation of the intrinsic protein kinase can be inhibited by exposure of the insulin receptor to substrate before autophosphorylation is initiated (44,45). Therefore, one of the contributing factors in the inhibition of insulin receptor autophosphorylation by CT-24 peptide may result from its addition to unphosphorylated receptors and occupation of the substrate binding site. Interestingly, cells expressing a mutated insulin receptor, where tyrosines 1316 and 1322 are replaced by Phe (YF2 receptor), exhibit enhanced substrate phosphorylation when compared with wild-type receptor (46,47). Additional evidence for the modulatory role of these two COOH-terminal tyrosines is provided by Kaliman et al. (9), who show that YF2 receptors have a 2-fold higher kinase activity than wild-type receptors and possess an insulin-insensitive conformation that corresponds to an active insulin receptor form. A plausible explanation for these observations is that in the basal state tyrosines 1316 and 1322 help maintain cis-inhibition of the receptor kinase by occupying the substrate binding site. Upon insulin addition, these two tyrosines are disengaged (by conformational change), and trans-autophosphorylation of the autocatalytic domain occurs (48). Hence, insulin-induced conformational change of the COOH terminus has been recently correlated with the ligand's ability to stimulate receptor autophosphorylation (49).
Here we have introduced stearyl peptide HC in intact CHO/ HIRc cells and evaluated its influence on insulin receptor functions. Our results show that the addition of stearyl peptide HC in intact cells has an initial effect of enhancing insulin signaling at the level of the insulin receptor itself as well as at several postreceptor targets. This includes the association and activation of PI 3-kinase to phosphorylated IRS-1 and increased levels of tyrosine phosphorylation of MAP kinase in response to insulin. Based on the PI 3-kinase data, one can assume that exposure of cells with stearyl peptide HC results in an increase in the insulin-stimulated tyrosine phosphorylation of IRS-1. Although the effects of peptide HC on kinase function are quantitatively modest (ϳ2-fold), the significance of changes of this magnitude is relevant, since small changes in insulin receptor activity would probably have major consequences on integrated glucose homeostasis and cellular regulation. Because of its lack of effect in the absence of insulin, peptide HC does not appear to mimic the stimulation by insulin but rather affects a sequence of events that is involved in modulation of the in vitro and in vivo insulin receptor activity and functions. It is noteworthy that stearyl peptide HC has no effect on the autophosphorylation of two closely related growth factor receptors. Peptide HC corresponds to a noncatalytic segment of the insulin receptor that contains serine residues 1293 and 1294, thought to be targets for interaction with serine kinase tightly associated to the insulin receptor kinase in an insulin-dependent manner (50) and to be substrates for phorbol ester-stimulated protein kinase C (51). The phosphorylation of this domain is closely linked to inactivation of the insulin receptor kinase (52). These data and ours indicate that this defined region within the carboxyl terminus plays an important role in the molecular regulation of insulin receptor activity and functions.
In this report, we have also shown specific binding of peptide HC with the insulin receptor but not with the IGF-1 or EGF receptors, suggesting that an association between peptide HC and the insulin receptor ␤-subunit may be required for the increase in receptor kinase activity. One possible consequence of this association is the inhibition of dephosphorylation of the activated insulin receptor, whereby peptide HC occupies a subdomain critical for the interaction of the ␤-subunit with protein-tyrosine phosphatase(s) involved in receptor dephosphorylation. This appears unlikely because of the fact that peptide HC increases insulin-stimulated autophosphorylation of the insulin receptor in cell-free systems and in semipermeabilized cells without an apparent change in receptor dephosphorylation. These results suggest that residues 1293-1307 of the insulin receptor interact with another noncatalytic region of the receptor ␤-subunit to regulate the phosphotransferase activity. Our findings with the ⌬43 mutant receptors support the notion that the important structural determinants that are required for direct binding of peptide HC to the insulin receptor still remain present. Additional studies with a number of cell lines expressing truncated receptors lacking larger deletions will be needed in order to determine where on the receptor peptide HC binds. Finally, experiments are in progress to identify which amino acid residues in the peptide HC are required for the binding to the receptor ␤-subunit.