Insulin Signal Transduction by a Mutant Human Insulin Receptor Lacking the NPEY Sequence

The cytoplasmic juxtamembrane domain of the human insulin receptor (hIR) contains a single copy of the tetrameric amino acid sequence Asn-Pro-Glu-Tyr (NPEY) (residues 969–972 in the exon 11-containing B-isoform), which exists in the context of NPXY. In this study, we examined the role of NPEY972 in mediating insulin signal transduction and cellular biological effects. Transfected Chinese hamster ovary cell lines expressing either the wild-type hIR-B isoform (hIR·WT) or a mutant receptor lacking the NPEY972 sequence (hIRΔNPEY) and control Chinese hamster ovary·Neo cells were used to comparatively analyze the following insulin effects: in vivo receptor tyrosine autophosphorylation and kinase activity, signal transduction to downstream signaling molecules, and stimulation of glycogen and DNA synthesis. The results showed that in comparison to hIR·WT, the hIRΔNPEY mutant demonstrated the following: (a) normal insulin-mediated receptor tyrosine phosphorylation, but ∼50% reduction in phosphorylation of p185-(insulin receptor substrate-1) and binding of the p85 subunit of phosphatidylinositol 3-kinase (PI 3-kinase), (b) an enhanced stimulation of PI 3-kinase enzymatic activity, (c) a complete inability to phosphorylate Shc, (d) minimal impairment of insulin sensitivity for glycogen synthesis, and (e) an augmented response to insulin-stimulated DNA synthesis via a high capacity, low sensitivity pathway. These results demonstrate the following: 1) the NPEY972 sequence is important but not absolutely essential for coupling of hIR kinase to insulin receptor substrate-1 and p85 or for mediating insulin’s metabolic and mitogenic effects, 2) the NPEY972 sequence is necessary for Shc phosphorylation, and 3) the absence of Shc phosphorylation releases the constraints on maximal insulin-stimulated mitogenic response, thus indicating that alternate signaling pathway(s) exist for this insulin action. This alternate pathway appears to be associated with enhanced activation of PI 3-kinase and is of high capacity and low sensitivity.

Although there has been considerable recent information on the structural basis of the interaction of IRS-1 and Shc with the NPEY motif of IR (5)(6)(7)(8)(9)(10), comparatively much less is known regarding the ultimate effects of such interactions on insulinstimulated biological responses such as the hormonal regulation of glucose metabolism and mitogenesis. Previously, it was thought that the NPEY motif served as the signal for IR endocytosis (14,15), in analogy with NPXY (where X represents any amino acid), the internalization signal of the low density lipoprotein receptor (16). However, we (17) and others (18) have subsequently shown that the intact JM domain rather than the NPEY motif per se is required for IR endocytosis. Mutation of Tyr 3 Phe in the NPEY of the A (exon 11Ϫ)-isoform of IR has been associated with impairment of IRS-1 phosphorylation and abrogation of downstream biological responses (18 -20). However, these effects were not reproduced in studies in which the entire JM domain together with the NPEY sequence was deleted (21,22). Thus, the effect on insulin biological response of specifically deleting the NPEY motif remains undetermined, especially for the B (exon 11ϩ)-isoform of the IR molecule.
Accordingly, in the present study, we have comparatively examined insulin signal transduction pathways and insulinstimulated metabolic and mitogenic bioresponses in transfected CHO cells stably expressing either the wild-type or mutant human IR (hIR) lacking the NPEY sequence. The results demonstrate that the mutant (hIR⌬NPEY) receptor undergoes normal insulin-stimulated autophosphorylation but has impaired ability to phosphorylate IRS-1 and a near-complete inability to phosphorylate Shc. However, the mutant receptor mediates enhanced insulin stimulation of PI 3-kinase activation. With respect to biological responses, the mutant receptor mediates a near-normal responsiveness of insulin stimulation of glycogen synthesis, whereas a maximal mitogenic response is paradoxically augmented, albeit with a lowered insulin sensitivity. Thus, these results indicate that removal of the NPEY sequence unmasks a high capacity, low sensitivity alternate insulin signaling pathway leading to mitogenesis and that this pathway is independent of Shc phosphorylation and occurs in a setting of increased PI 3-kinase activity. Products. Monoclonal antibody against phosphotyrosine (␣Y20) and polyclonal antibodies against Shc were obtained from Transduction Laboratories (Lexington, KY). Polyclonal antibody against IRS-1 was purchased from Upstate Biotechnology Inc. (Sarnac Lake, NY). Horseradish peroxidase-conjugated anti-mouse and anti-rabbit IgGs and the enhanced chemiluminescence kit were from Amersham Corp. Electrophoresis reagents were purchased from Bio-Rad. All other chemicals were reagent grade and purchased from Sigma.

Materials-Cell
Cells and Cell Culture-Details of plasmid construction, transfection, and clonal selection of CHO cell lines have been presented previously (17). All lines were maintained in culture in Ham's F-12 media, supplemented with 10% fetal calf serum (v/v), 2 mM glutamine, 50 mg/ml gentamycin, and 400 g/ml G418. Receptor characterization including photoaffinity labeling and 125 I-insulin binding were performed as described previously (17). Cells were subcultured at 5-day intervals, and all studies were performed on cells at passage 15 or less. For the study of in situ tyrosine phosphorylation, PI 3-kinase activity, glucose incorporation into glycogen, and thymidine incorporation into DNA, cells were subcultured in 35-mm 6-well multidishes.
Insulin-stimulated in Vivo Receptor Tyrosine Autophosphorylation/ Kinase Studies-Confluent monolayers of the transfected CHO cell lines were incubated for 1 min at 37°C in medium (Eagle's minimal essential medium, 10 mM HEPES, 10 mg/ml BSA) without or with increasing concentrations of insulin. The cells were then solubilized in 3% SDS containing 1 mM N-ethylmaleimide, 2 mM phenylmethylsulfonyl fluoride, 2 mM sodium vanadate, and 1.0 g/ml aprotinin, and the proteins were analyzed by SDS-polyacrylamide (5-15% linear gradient) gel electrophoresis under disulfine nonreducing conditions. Proteins were transferred to nitrocellulose membrane by electroblotting, and phosphotyrosine-containing proteins were visualized by probing the membrane with ␣PY20 antibody followed by 125 I-Protein A and autoradiography according to established procedures (23).
Insulin Stimulation of Shc Phosphorylation-CHO cell lines expressing hIR⅐WT or hIR⌬NPEY receptors and control CHO⅐Neo cells were incubated for various times at 37°C with or without 100 nM insulin. The cells were then lysed in 4°C solubilizing buffer containing 1% Triton X-100, 50 mM HEPES (pH 7.6), 150 mM NaCl, 1 mM EDTA, 1 mM NaF, 10 g/ml aprotinin, 1 mM NaVO 4 , 10% glycerol. The cell lysates were clarified by centrifugation, and the supernatants were removed and incubated with anti-Shc antibody (5 g per dish) for 2 h at 4°C, followed by addition of 30 l of protein A-agarose suspension, and further incubated for an additional 120 min at 4°C. The immunoabsorbed complexes were collected by centrifugation, and the proteins in the pellet were released by heating in SDS-polyacrylamide gel electrophoresis sample buffer and resolved in 5-15% acrylamide gradient gel. The proteins were transferred to nitrocellulose sheets and Western blotted with anti-phosphotyrosine antibody (␣PY20). Labeled bands were detected using anti-mouse IgG conjugated with horseradish peroxidase and the enhanced chemiluminescence kit according to manufacturer's (Amersham Corp.) instructions.
Interaction of the p85 Subunit of PI 3-Kinase with Insulin-activated hIR⅐IRS-1 Complex-The insulin-induced association of the regulatory (p85) subunit of PI 3-kinase with hIR⅐IRS-1 complexes was assessed in the different cell lines by measuring the ability of a Sepharose-coupled p85-glutathione S-transferase (GST) fusion protein to precipitate the complexes from cell lysates. The Sepharose-coupled p85-GST (24) and Sepharose-GST reagents were kindly provided by Dr. Alan Saltiel (Parke-Davis). The different CHO cell lines grown in 100-mm dishes were serum-starved overnight and then incubated with or without 100 nM insulin for 1 min at 37°C in medium (pH 7.5) consisting of minimal essential medium, 10 mM HEPES, and 10 mg/ml bovine serum albumin. The cells were then lysed in 4°C solubilizing buffer (described above under Shc phosphorylation studies). The lysates were clarified by centrifugation, and equal aliquots of the supernatants were incubated for 120 min at 4°C with 50 l of p85-GST-Sepharose or GST-Sepharose beads. The beads were then pelleted and washed, and the associated proteins were released in Laemmli's sample buffer and analyzed by SDS-polyacrylamide gel electrophoresis under reducing conditions. The proteins were then transferred to nitrocellulose sheets and Western blotted with anti-phosphotyrosine (␣PY20) antibody. The immunoblotted protein bands were visualized after probing with 125 I-labeled Protein A followed by autoradiography.
Phosphatidylinositol 3-Kinase Activity-PI 3-kinase activity was measured on the IRS-1 associated fraction of the enzyme. 100 g of cell extract protein was incubated with 2 g of polyclonal IRS-1 antibody for 16 h at 4°C. Immunocomplexes were collected with Protein A-agarose. The procedure for washing of immunoprecipitates and enzyme assay was that of Kelley and Ruderman (25). The reaction was for 20 min and the final [ATP] was 50 M (20 Ci of [␥-32 P]ATP). Reactions were halted by addition of acidified CHCl 3 :CH 3 OH (1:1) and chromatographed on TLC plates in CHCl 3 :CH 3 OH:H 2 O:NH 4 OH (60:47:11.3:2). Autoradiography was performed on the dried plates, and bands were quantitated by densitometry.
Glucose Incorporation into Glycogen-Cells were serum-starved for 24 h and then washed 2 ϫ with Eagle's minimal essential medium, 0.1% BSA (pH 7.4). The reaction was started by addition of fresh media, varying concentrations of insulin, and D-[U-14 C]glucose (1 Ci/well). The assay proceeded for 2 h in the CO 2 incubator. The reaction was terminated by washing cells rapidly 5 ϫ with 4°C phosphate-buffered saline and solubilizing in 1 N NaOH and precipitating glycogen as described previously (26). Results are presented as nanomoles of glucose incorporated into glycogen normalized to cell number or protein.
Thymidine Incorporation into DNA-A modification of the method described in McClain et al. (27) was used. Confluent cells were refed with serum-free medium for 24 h. Cells were then treated with varying insulin concentrations for 16 h. The media were replaced with Eagle's minimal essential medium, 0.1% BSA (pH 7.4), together with any treatments, and [ 3 H]thymidine (0.5 mCi) was added to each well. The cells were incubated for 1 h at 37°C. Cells were then rinsed twice with 5 ml of chilled phosphate-buffered saline, once with 1 ml of methanol, twice with 5 ml of chilled 5% trichloroacetic acid (w/v), followed by 5 ml of ethanol at 4°C. The cells were then dissolved in 0.5 ml of 1 N NaOH and neutralized with an equal volume of 1 N HCl. Aliquots were removed for liquid scintillation counting and protein determination by the Bradford method (28), using BSA as the standard. Results are presented as % of the total thymidine added incorporated into DNA, normalized to cell number or protein.

RESULTS
Cell Lines Expressing Wild-type and Mutant hIRs-In this study, we utilized transfected CHO cell lines that stably express either the wild-type hIR B-isoform or a deletion mutant of this isoform lacking the NPEY sequence (17). The characteristics of the WT and mutant receptor are summarized in Fig. 1. The hIR⌬NPEY mutant receptor, in which the single copy of the NPEY sequence that exists in the submembranous domain of the ␤-subunit has been deleted, is overexpressed on the surface of the transfected CHO cells as 440-kDa ␣ 2 ␤ 2 heterotetramer and binds insulin with similar high affinity (EC 50 ϭ 4 nM) as the wild-type hIR (Fig. 1). The transfected CHO cells express 9.1 ϫ 10 5 and 9.0 ϫ 10 5 receptors per cell of hIR⅐WT and hIR⌬NPEY types, respectively. This is as compared with the control cells transfected with neo alone which express only 2000 endogenous rodent IR/cell (17). Thus, our experimental cells provide a valid system for comparative analysis of the role of the NPEY sequence in insulin signal transduction. To control for the possible influence of clonal variation on the results, two different cell lines each expressing similar numbers of either wild-type or hIR⌬NPEY receptors were tested for biological responses. Since the different clones for each receptor showed similar behavior with regard to insulin sensitivity and responsiveness (not shown), a single clone of each receptor was selected for further study.
In Vivo Tyrosine Autophosphorylation and Kinase Activities-An early step in transmembrane insulin signaling involves rapid autophosphorylation of the receptor ␤-subunit on specific tyrosine residues, a process that activates the IR kinase and initiates a further phosphorylation cascade involving downstream signal transduction molecules, the predominant one being insulin receptor substrate 1, IRS-1 (1)(2)(3). To comparatively analyze this signaling function, cells expressing hIR⅐WT or hIR⌬NPEY receptors were incubated for 1 min at 37°C without or with increasing insulin concentrations. Cell lysates were then subjected to Western blot analysis with ␣PY monoclonal antibody to assess the in vivo tyrosine-phosphorylated cellular proteins. The resulting autoradiogram is shown in Fig. 2 and demonstrates a similar dose-dependent insulin stimulation in tyrosine phosphorylation of the hIR⅐WT and hIR⌬NPEY receptors. Similarly, insulin stimulation of phosphorylation of the endogenous substrate pp185 (IRS-1) is also observed although the magnitude of the response is lower in the hIR⌬NPEY-expressing cells at each insulin concentration examined. In contrast, the control CHO⅐Neo cells exhibit no insulin stimulation in tyrosine phosphorylation of cellular proteins. Fig. 3 shows quantitative analysis of the tyrosine autophosphorylation/kinase data and demonstrates that despite the similarity in receptor tyrosine phosphorylation, the magnitude of IRS-1 phosphorylation is reduced by ϳ50% in cells expressing hIR⌬NPEY at all insulin concentrations tested.
Insulin Stimulation of Shc Phosphorylation-In addition to phosphorylation of IRS-1, insulin activation of the IR kinase also leads to phosphorylation of Shc, an Src homology 2 (SH2)domain containing protein, and this process has been implicated in mitogenic signaling by insulin (1-3, 12, 13). The phosphorylated Shc binds to the JM domain of the IR ␤-subunit that contains Tyr 972 (5)(6)(7)9). Accordingly, we examined the effect of the deletion of the NPEY 972 motif on insulin stimulation of Shc phosphorylation. Fig. 4 shows that in hIR⅐WT-expressing cells insulin rapidly stimulates phosphorylation of primarily the 52-kDa Shc isoform and to a much lesser extent the 46-and 66-kDa Shc species. Insulin stimulation of the 52-kDa Shc phosphorylation in hIR⅐WT-expressing cells is very rapid, occurring within 5 s of insulin addition and reaching maximal levels by 1-5 min. In contrast, the cells expressing the hIR⌬NPEY 972 receptor exhibit essentially no increase in Shc phosphorylation. Similarly, the control CHO⅐Neo cells exhibit no insulin stimulation of Shc phosphorylation.
Insulin-induced Association of the hIR⅐IRS-1 Complex with Phosphatidylinositol 3-Kinase and Stimulation of Its Enzymatic Activity-A major signaling event downstream of the activated IR kinase is stimulation of the lipid kinase activity of phosphatidylinositol 3-kinase (1-3, 11, 25). This heterodimeric enzyme associates with phosphorylated IRS-1 via the SH2 domain of its regulatory (p85) subunit, and this is followed by stimulation of the kinase activity of the catalytic (p110) subunit of the enzyme (1-3, 11, 25). Since the hIR⌬NPEY receptor exhibits a reduction in insulin stimulation of IRS-1 phosphorylation (Fig. 3), we next assessed the functional consequences of this defect by comparatively analyzing the abilities of the different receptors to couple with and activate PI 3-kinase. coupling to p85 that can occur via insulin activation of the endogenous rodent IR kinase. Quantification of the radioactivities in the p85-GST precipitated IRS-1 and hIR ␤-subunit phosphoprotein bands revealed that in comparison to hIR⅐WTexpressing cells, the hIR⌬NPEY-expressing cells exhibit a 60% reduction in the amount of IRS-1 and a 44% reduction in the amount of ␤-subunit precipitation (data not shown).
Insulin stimulation of the IRS-1-associated PI 3-kinase enzymatic activities were also compared in the different cell lines, and the data are presented in Fig. 6. The insulin doses used for each cell line were selected to correspond to those that subsequently gave maximal biological responses (see below). The autoradiograms in Fig. 6A demonstrate that the insulin-stimulated PI 3-kinase activities (as reflected by formation of phosphatidylinositol phosphate) are augmented in hIR⅐WT and hIR⌬NPEY-expressing cells as compared with the CHO⅐Neo cells. The autoradiographic bands were also quantified by densitometric scanning and the PI 3-kinase activities presented relative to the basal activity for each cell line in each experiment (Fig. 6B). In CHO⅐Neo cells, insulin increased PI 3-kinase activity in a dose-dependent manner, but maximal stimulation (746% basal) did not occur until 83.3 nM (500 ng/ml) insulin. By comparison, the cells expressing hIR⅐WT exhibited markedly enhanced insulin sensitivity of PI 3-kinase stimulation, with maximal effect (1185% basal) occurring at only 0.83 nM (5 ng/ml) insulin. Stimulation was somewhat reduced at 83.3 nM insulin. The cells expressing hIR⌬NPEY exhibited the highest responsiveness of PI 3-kinase stimulation. Maximal stimulation that occurred at 1.67 nM (10 ng/ml) insulin (3057% basal) was also maintained at 83.3 nM insulin and represented ϳ2.5 ϫ the maximal stimulation attained by hIR⅐WT-expressing cells.
Insulin Stimulation of Glycogen Synthesis and Thymidine Incorporation into DNA-The results described thus far have shown that the hIR⌬NPEY receptor manifests at least three The CHO cell lines were incubated for the indicated times in the absence (Ϫ) or with (ϩ) 100 nM insulin. The cells were then solubilized, and the lysates were subjected to immunoprecipitation with anti-Shc antibody followed by electrophoresis and Western blot analysis using anti-phosphotyrosine antibody. alterations in insulin signaling: decreased ability to phosphorylate IRS-1, inability to phosphorylate Shc, and augmented stimulation of PI 3-kinase activity. To determine the ultimate biological consequences of these alterations in signaling events, we measured both metabolic and mitogenic insulin responses in the various cell lines. The metabolic response selected was glucose incorporation into glycogen, which in fibroblasts is reflective of insulin effects on both glucose uptake and glycogen synthase (26,29). Thymidine uptake into DNA was taken as a measure of mitogenesis, and preliminary studies revealed that 24 h of serum starvation of confluent cells were sufficient to express maximal insulin responses (not shown). Fig. 7A shows that basal glycogen synthesis was similarly increased in hIR⅐WT and hIR⌬NPEY-expressing cells as compared with CHO⅐Neo cells but that insulin responsiveness was generally similar in the three cell lines. However, there were significant differences in insulin sensitivity among the different cell lines (Fig. 7B). Overexpression of wild-type hIR greatly increased sensitivity (EC 50 ϭ 36.7 Ϯ 10.0 pM) compared with CHO⅐Neo cells (EC 50 ϭ 1.67 Ϯ 0.60 nM). Cells expressing hIR⌬NPEY receptors displayed insulin sensitivity (EC 50 ϭ 106 Ϯ 17 pM) far greater than CHO⅐Neo cells but less than hIR⅐WT-expressing cells. Thus the complement of endogenous rodent IR in CHO⅐Neo cells is sufficient to maximally stimulate glycogen synthesis. Overexpression of hIR does not change responsiveness but serves to increase sensitivity, with the hIR⌬NPEY receptor signaling with nearly the same efficiency as the wild-type hIR.
Basal and maximally insulin-stimulated thymidine uptake were similar in CHO⅐Neo and hIR⅐WT-expressing cells (Fig. 8). Basal uptake was also similar in hIR⌬NPEY-expressing cells. However, maximal insulin stimulation in these cells was augmented, representing ϳ2.5 ϫ that of the Neo and hIR⅐WT cells whether the uptake data were expressed on the basis of cell number (Fig. 8A) or cellular protein level (Fig. 8B). The augmented responsiveness for thymidine uptake appeared to be specific for insulin as the maximal stimulation in response to IGF-1 was similar in hIRWT (552% basal) and hIR⌬NPEY cells (502% basal). Serum stimulation of thymidine uptake was also comparable in hIRWT (2490%) and hIR⌬NPEY cells (2436%) and slightly lower in Neo (1447%) cells.
The insulin dose-response data for mitogenesis showed that the augmented maximal responsiveness in hIR⌬NPEY cells (1527% basal uptake) occurred at 1.67 nM insulin (Fig. 9A). The corresponding maximal responsiveness values in Neo cells (424% basal) and hIR⅐WT cells (489% basal) were attained at 4.2 and 0.33 nM insulin, respectively. The insulin sensitivities for mitogenesis are better represented in Fig. 9B and show that the hIR⅐WT-expressing cells had a markedly increased sensitivity (EC 50 ϭ 60 Ϯ 13 pM) as compared with CHO⅐Neo cells (EC 50 ϭ 1.15 Ϯ 0.28 nM), whereas the hIR⌬NPEY cells had an intermediate sensitivity (EC 50 ϭ 401 Ϯ 67 pM). DISCUSSION The juxtamembrane domain of the insulin receptor contains the tetrameric sequence, NPEY 972 , which exists as NPXY motif that was shown to be required for internalization of certain receptors such as that for low density lipoprotein (16). However, analysis of cells transfected with deletion mutants of IR lacking this sequence has revealed that insulin binding and internalization, as well as receptor processing, are normal (17). The NPEY sequence has also been shown to play a role in linking the insulin-activated IR kinase to signal transduction molecules. The results of the current study show that deletion of the single copy of the NPEY sequence that exists in the hIR ␤-subunit results in major alterations in insulin signaling and in mediation of biological responses.
The insulin-induced tyrosine phosphorylation of the IR ␤-subunit leads to conformational changes in the receptor (1,30) that facilitate interaction of the activated IR kinase with downstream signal transduction molecules, of which the interactions of IRS-1 and Shc have been extensively studied (1-3, 5-9). Both of these molecules interact with the NPEY region of the IR ␤-subunit (5-9), and the impaired phosphorylation of IRS-1 and Shc by the hIR⌬NPEY receptor is consistent with this fact. However, it is of interest to note that while IRS-1 FIG. 7. Insulin stimulation of glycogen synthesis. Insulin treatment of the CHO cells and measurement of glycogen synthesis were performed as described under "Experimental Procedures." A, absolute rates of glucose incorporation into glycogen in the basal and maximally insulin-stimulated states. B, insulin sensitivity of glycogen synthesis. Results are presented as % of the difference between basal and maximally insulin-stimulated activity for each cell line in each experiment. The results shown are the mean Ϯ S.E. from three to five separate experiments, each performed in triplicate. phosphorylation is reduced only partially (ϳ50%), Shc phosphorylation is almost totally inhibited. This suggests that IRS-1 and Shc interact with the IR by similar but not identical mechanisms. It has previously been proposed that IRS-1 and Shc bind to phosphorylated NPXY motifs via their respective non-SH2 phosphotyrosine binding (PTB) domains, regions that have also been termed SAIN (Shc and IRS-1 NPXY-binding) domains (6). In addition to its SAIN or PTB domain, IRS-1 also utilizes a second domain, the pleckstrin homology domain, for its interaction with IR (10). In fact, there are many receptors that mediate IRS-1 phosphorylation yet do not contain NPEY sequences (31). Therefore, it is possible that the higher level of IRS-1 phosphorylation (in comparison to Shc phosphorylation) mediated by the hIR⌬NPEY receptor may be attributable to such differences in the binding properties of IRS-1 and Shc. The stoichiometry of IRS-1 phosphorylation was not determined in the present studies so it is unknown if lower IRS-1 phosphorylation represents fewer molecules phosphorylated or less phosphorylation per IRS-1 molecule. The fact that there is still significant IRS-1 phosphorylation suggests that Tyr 972 is not absolutely essential for this event.
The augmented insulin stimulation of PI 3-kinase activity in cells expressing hIR⌬NPEY receptors was unexpected in view of the impaired IRS-1 phosphorylation and decreased association of p85 with the activated IR⅐IRS-1 complex that was observed. As IRS-1 contains 9 YXXM sequences, which can bind PI 3-kinase after phosphorylation (1-4), selective phosphorylation of these sites could result in full PI 3-kinase association and activation even as net IRS-1 phosphorylation is reduced. There may be a threshold of IRS-1 phosphorylation beyond which there is no further increase in PI 3-kinase activity. Indeed, Wilden and Broadway (32) have shown that 4-and 10-fold increases over control in IRS-1 phosphorylation in CHO cells gave equivalent increases in PI 3-kinase activity. Similarly, Yamaguchi and Pessin (33) have shown that the expression of signaling molecules can have biphasic effects on downstream responses. The same thing could be happening in hIR⅐WT compared with hIR⌬NPEY cells; the extent and nature of IRS-1 phosphorylation in hIR⌬NPEY cells could be sufficient for full PI 3-kinase activation, whereas the additional phosphorylation in hIR⅐WT cells, either by steric hindrance or lower affinity competition by other phosphorylated sequences for p85, could be interfering with signaling. Another explanation for the discordance between IRS-1 phosphorylation and PI 3-kinase activation could be that the mutant receptors may be mediating insulin stimulation of an especially sensitive pool of PI 3-kinase. In this regard, it is known that insulin activation of PI 3-kinase occurs in a subcellular membrane compartment (25), and it is possible that there may be pools of PI 3-kinase with different insulin sensitivities. The occurrence of different isoforms of both p85 and p110 (the catalytic subunit of PI 3-kinase) (11) is also consistent with the possibility that the different isoforms may participate in signaling in a pathwayand/or cell-specific manner. In our experiments it is not possible to distinguish between different isoforms or pools of PI 3-kinase. Regardless of the specific mechanism(s) by which hIR⌬NPEY receptors mediate insulin stimulation of PI 3-kinase, it is highly significant that augmented stimulation of the enzyme occurred in the presence of reduced IRS-1 phosphorylation and p85 association. This novel observation suggests that the linking of insulin stimulation of IR to PI 3-kinase may utilize alternate pathways with differential coupling efficiencies depending on the complement of IRS-1 and p85 that associate with the activated receptor.
The biological significance of the altered insulin signaling exhibited by the hIR⌬NPEY receptor was assessed by measuring two major bioeffects, stimulation of glucose incorporation into glycogen and thymidine incorporation into DNA. The increased rates of basal glycogen synthesis exhibited by the hIR⅐WT and hIR⌬NPEY-expressing cells may, at least in part, be reflective of the increased basal glucose uptake that is usually observed in CHO cells overexpressing hIRs (15, 18 -20, 34). However, the maximal insulin effect is generally similar regardless of the type of receptor expressed, suggesting that the native complement of receptors in Neo cells is sufficient to manifest the full response. What overexpression of receptors does is increase insulin sensitivity. Yet the dose-response curve is biphasic and at higher insulin levels, where IRS-1 would be most highly phosphorylated, there may be interference with signaling. The only noticeable effect of the NPEY deletion is to reduce insulin sensitivity compared with WT cells. Chen et al. (35) have suggested that the NPXY domain contributes to insulin sensitivity but is not essential for signaling at high levels of effector molecules. The normal insulin response of glycogen synthesis in hIR⌬NPEY cells, in the absence of Shc phosphorylation, also shows that Shc is not necessary for this metabolic response.
The situation is entirely different with respect to insulin effects on mitogenesis in that basal activity was unchanged but maximal responsiveness was markedly enhanced in hIR⌬NPEY-expressing cells. That this effect roughly parallels the increase in insulin stimulation of PI 3-kinase activity in these cells is consistent with the proposed role of this enzyme in insulin regulation of mitogenesis and cell growth (1). Additionally, the augmented insulin stimulation of mitogenesis in the absence of Shc phosphorylation in the hIR⌬NPEY-expressing cells further suggests that the IR can utilize an alternate pathway that bypasses this step. This pathway that is unmasked by the ⌬NPEY mutation is of high capacity but low insulin sensitivity. Normal responsiveness to IGF-1 and serum suggests that this alternative pathway may be specific for insulin, albeit with a reduced sensitivity.
Certain differences and similarities are apparent between the current results and those reported by other investigators regarding the influence of various mutations of the NPEY region on insulin signaling and responses. For example, the impaired ability of the hIR⌬NPEY receptor to mediate phosphorylation of IRS-1 is qualitatively similar to previous reports that various point mutations of the tyrosine residue of the NPEY sequence (18 -20) or deletion of a 12-amino acid segment containing this sequence (20) impaired insulin stimulation of IRS-1 phosphorylation. On the other hand, Thies et al. (21) reported no impairment of pp185 (IRS-1) phosphorylation upon deletion of the entire exon 16-encoded JM region containing the NPEY sequence. With respect to mediation of glycogen synthesis by the hIR⌬NPEY receptor, the modest decrease in insulin sensitivity without appreciable alteration in maximal response is similar to findings on the ⌬exon 16 mutant of hIR reported by McClain (22). However, our results do differ from those of Backer et al. (20) and Kaburagi et al. (18) who showed that point mutations of Tyr 960 (corresponding to the tyrosine in NPEY) impaired both insulin sensitivity and responsiveness of glycogen synthesis. Finally, the decreased insulin sensitivity for mitogenesis exhibited by the hIR⌬NPEY-expressing cells is similar to findings in many of the earlier reports involving different JM domain mutations (18,20,22). Novel findings in the current study include augmented maximal insulin responsiveness of mitogenesis in hIR⌬NPEY cells in association with enhanced PI 3-kinase stimulation and absent Shc phosphorylation. There are some plausible explanations for the differences between our results and those of the earlier reports. First, although the NPXY motif is implicated in coupling activated IR kinase with signal transduction molecules, most of the earlier studies evaluated either only point mutation of the tyrosine residue (18 -20) or large deletions of the JM domain containing this motif (20 -22). Our study evaluated the effect of deleting just the NPXY motif as a unit. As this region of the receptor contains a tyrosine/␤-turn (15) which, when phosphorylated, binds to an L-shaped cleft in the PTB domain of IRS-1 (8 -10), the three-dimensional structure of a deletion of the NPXY sequence may differ considerably from a tyrosine substitution. The NPXY deletion may have induced a favorable conformational change of the activated hIR kinase that facilitated its interaction with an alternate signaling pathway. Second, all of the earlier studies that examined the role of the JM domain in signaling utilized the A-isoform of the hIR (18 -22), whereas the B-isoform was employed in the current study. In this regard, the A-and B-isoforms of hIR are known to differ in their insulin binding, internalization, and signaling properties (36,37), and it is possible that such factors may have also contributed to the observed differences.
In summary, the results of this study have demonstrated that the ⌬NPEY 972 mutation of the hIR B-isoform leads to modulation of insulin signaling consisting of the following: 1) decreased ability to phosphorylate IRS-1; 2) an inability to phosphorylate Shc; 3) enhanced insulin stimulation of PI 3-kinase activity; 4) a minimally altered stimulation of glycogen synthesis; but 5) augmented maximal responsiveness of mito-genesis with reduced insulin sensitivity. These findings lead us to conclude that the absence of the NPEY 972 sequence facilitates coupling of the activated hIR kinase to a high capacity, low sensitivity alternate signaling pathway for mitogenesis that is associated with enhanced activation of PI 3-kinase but has minimal influence on glycogen synthesis pathway(s). Thus, the insulin receptor contains the information necessary to engage multiple signaling pathways and maintains a redundancy for signal transduction that can be differentially activated.