Tyr624 and Tyr628 in Insulin Receptor Substrate-2 Mediate Its Association with the Insulin Receptor*

In addition to the pleckstrin homology domain and the phosphotyrosine binding domain in insulin receptor substrate (IRS)-1 and IRS-2, a region between amino acids 591 and 786 in IRS-2 (IRS-2-(591–786)) binds to the insulin receptor. Based on peptide competition studies, this region interacts with the phosphorylated regulatory loop of the insulin receptor; we designate this region the kinase regulatory loop binding (KRLB) domain. Two tyrosine residues in the KRLB domain at positions 624 and 628 are crucial for this interaction. Phosphorylation of tyrosine residues in the KRLB domain by the insulin receptor inhibits the binding to the receptor. These results reveal a novel mechanism regulating the interaction of the insulin receptor and IRS-2 that may distinguish the signal of IRS-2 from IRS-1.

The insulin receptor (IR) 1 mediates tyrosine phosphorylation of several cellular substrates, including IRS-1, IRS-2, and Gab-1 (1)(2)(3). These IRS (insulin receptor substrate) proteins provide an interface between the activated insulin receptor and various signaling proteins. IRS proteins are composed of a COOH terminus containing multiple tyrosine phosphorylation sites in various amino acid sequence motifs that bind to the Src homology-2 domain in certain enzymes and adapter molecules (4 -7). In addition to the phosphorylation sites, IRS proteins contain other domains to engage activated membrane receptors. At the extreme NH 2 terminus, the IRS proteins contain a pleckstrin homology (PH) domain (IH1 PH ). The IH1 PH is essential in IRS-1 for the interaction with a physiological level of insulin receptor (8); this domain plays a similarly important role in IRS-2 and Gab-1. In addition to the PH domain, IRS-1 and IRS-2 contain a phosphotyrosine binding (PTB) domain (IH2 PTB ), which binds to the phosphorylated NPEY motif in the cytoplasmic region of the receptors of insulin, insulin-like growth factor-I, and interleukin-4 (9 -15).
A third region between residues 591 and 786 in IRS-2 engages the activated insulin receptor (13,16). Using a yeast two-hybrid analysis, Tyr 624 and Tyr 628 in IRS-2 were found to contribute significantly to this interaction. The amino acid sequence in the region does not reveal a known protein-protein interaction domain, such as the PH domain, PTB domain, Src homology 2 domain, Src homology 3 domain, or WW domain (17)(18)(19)(20)(21). However, since it interacts with the phosphorylated regulatory loop of the insulin receptor ␤-subunit, we propose to designate it as the kinase regulatory loop binding (KRLB) domain. The binding of the KRLB domain of IRS-2 to the insulin receptor is independent of tyrosine phosphorylation sites in the COOH terminus and the NPEY motif in the juxtamembrane region (13). Since IRS-1 does not contain a KRLB domain (13,16), this domain may contribute to a unique signaling potential of IRS-2.
cDNA Constructs-Full-length mouse IRS-2 and IRS-2-(591-786) (IRS-2-KRLB) cDNAs were subcloned into the two-hybrid plasmid pAC-TII as reported previously (13). The subdomain of rat IRS-1 containing amino acids 494 -741 was subcloned into pACTII. This fragment was obtained by polymerase chain reaction using the following primers (5Ј to 3Ј): cgcggatccgcaccatggggacaagcccggcg and ccggaattcggggctgctggtgttggaatc. EcoRI and BamHI sites are underlined. The polymerase chain reaction product digested with EcoRI and BamHI were cloned into pACTII digested with EcoRI and BamHI. All IRS-1 and IRS-2 point mutants were generated by either site-directed mutagenesis of doublestranded DNA using the Transformer TM kit (CLONTECH, Palo Alto, CA) or the Quickchange TM site-directed mutagenesis kit (Stratagene, San Diego, CA).
Transformation of Yeast Strains and ␤-Galactosidase Assay-Plasmid DNA transformations were performed using the lithium acetate method of Gietz et al. (24). Cotransformants were selected on Trp Ϫ , Leu Ϫ plates. The transformants were tested for ␤-galactosidase activity by liquid culture assays using the substrate o-nitrophenyl-␤-D-galactopyranoside (ONPG) as described by Miller (25).
Construction of IRS-2 Fusion Proteins-GST fusion proteins were generated by subcloning a fragment of IRS-2-KRLB containing the amino acids 591-786, either wild type or mutated, into the vector pGEX-3X (Pharmacia Biotech Inc., Uppsala, Sweden). All GST fusion proteins were expressed in BL21 bacterial cells. Bacteria were lysed in presence of lysis buffer (20 mM Tris, pH 7.4, 1 M NaCl, 0.2 mM EDTA, 0.2 mM EGTA) containing protease inhibitors and 1 mg/ml lysozyme. Lysates were frozen twice in liquid nitrogen and sonicated twice for 30 s. After 20 min of centrifugation at 17,000 ϫ g, clarified lysates were incubated with glutathione-Sepharose beads (Pharmacia) for 1 h at 4°C. Then the glutathione-Sepharose pellets were washed twice with lysis buffer and concentrated using a column. The fusion proteins were eluted with elution buffer (100 mM HEPES, pH 8, 50 mM glutathione).
Partial Purification of Insulin Receptors-Insulin receptors were partially purified by chromatography on wheat germ agglutinin (WGA) (26). Briefly, RHIR (Rat 1 fibroblasts expressing 10 6 human insulin receptors/cell) were washed twice with PBS (137 mM NaCl, 2.7 mM KCl, 6.5 mM Na 2 HPO 4 , 1.5 mM KH 2 PO 4 , pH 7.4) and solubilized for 90 min at 4°C in 50 mM Hepes, 150 mM NaCl, 1% (v/v) Triton X-100, pH 7.6. The supernatant from an ultracentrifugation step (60 min, 100,000 ϫ g, 4°C) was applied to a WGA column, and insulin receptors were eluted with 0.3 M N-acetyl-D-glucosamine in 50 mM Hepes, 150 mM NaCl, 0.1% (v/v) Triton X-100, pH 7.6. Protease inhibitors were present throughout the procedure (20 M leupeptin, 1.25 mM bacitracin, 100 IU/ml aprotinin, and 1 mM phenylmethylsulfonyl fluoride. Insulin receptors were quantified by Scatchard analysis of 125 I-insulin binding. In Vitro Phosphorylation of the Insulin Receptor Peptide 1142-1156 -This peptide has been previously characterized (27) and corresponds to the sequence 1142-1156 (TRDIYETDYYRKGGK) of the human insulin proreceptor (28). 300 fmol of WGA-purified insulin receptors were incubated on WGA and were activated with insulin  (29). Proteins were resolved by SDS-PAGE (10%), and the gel was washed several times with H 2 O to remove SDS. Then phosphorylated GST-IRS-2-KRLB, wild type and mutated, detected after 15 h of autoradiography, were excised from the gel. The gel slices were minced, washed twice with 1 ml of H 2 CO 3 NH4, pH 8, and incubated overnight with the same buffer. The next day, the gel slices were subjected to tryptic digestion for a total of 24 h at 37°C in a solution containing 1 mM CaCl 2 and 50 g/ml trypsin. During the digestion procedure, trypsin (50 g/ml) was added twice after an 8-h interval. Then the buffer containing the tryptic peptides was recovered, and peptides were washed four times with 100 l of H 2 O and dried in a speed vacuum concentrator. Finally, the tryptic peptides were resuspended in NH 3 N/1000.
The peptides were spotted onto silica thin layer chromatography plates. The first dimension was electrophoresis for 3 h at 900 V in acidic buffer containing 30% (v/v) formic acid. After drying, the plates were placed in a chromatography tank containing a buffer of 37.5% isobutyl alcohol, 25% pyridine, and 7.5% acetic acid (v/v) for 10 h. Plates were dried and subjected to autoradiography.
Precipitation of Autophosphorylated Insulin Receptors by IRS-2 Fusion Proteins-0.5 g of GST-IRS-2-KRLB, mutated or not, was incubated with 40 l of glutathione-Sepharose for 1 h at 4°C. GST-IRS-2-KRLB was used either phosphorylated or not. When phosphorylated GST-IRS-2 was used, the fusion protein-containing pellets were incubated for 30 min with insulin (10 Ϫ7 M) and a phosphorylation mixture (30 M ATP, 4 mM MnCl 2 , 8 mM MgCl 2 ) in the presence of WGA-purified insulin receptors. Then the phosphorylated GST-IRS-2 pellets were washed five times with 0.5 M NaCl, 0.5% (v/v) Triton X-100 to remove the IR, while the unphosphorylated GST-IRS-2-KRLB pellets were washed twice with 30 mM Hepes, 30 mM NaCl, 0.1% (v/v) Triton X-100, pH 7.5. Concurrently, 2 pmol of WGA-purified insulin receptors were activated with insulin at 10 Ϫ7 M for 20 min in a final volume of 45 l. Thereafter, 5 l of phosphorylation mixture (300 M [␥-32 P]ATP (5 mCi/mmol), 40 mM MnCl 2 , 80 mM MgCl 2 ) were added. The reaction was stopped after 30 min by the addition of 500 l of buffer (50 mM HEPES, 150 mM NaCl, 100 mM NaF, 10 mM EDTA, and 10 mM Na 4 P 2 O 7 ), pH 7.5. Then 500 l of the activated receptors were added to the pellets of phosphorylated or unphosphorylated GST-IRS-2 and incubated for 4 h. The pellets were washed three times with 30 mM Hepes, 30 mM NaCl, 0.5% (v/v) Triton X-100, pH 7.5, and the samples were analyzed by SDS-PAGE and subjected to autoradiography. The experiment was then performed as described above.

Tyr 624 and Tyr 628 in IRS-2 Mediate Interaction of IRS-2 with
the Insulin Receptor-In the yeast two-hybrid analysis, IRS-2 interacts with the insulin receptor through the IH2 PTB and a region between residues 591 and 786 that we call the KRLB domain (13). The KRLB domain contains seven potential tyrosine phosphorylation sites at positions 594, 624, 628, 649, 671, 734, and 758. Alignment of IRS-1 and IRS-2 reveals that five of these residues are common to IRS-1, whereas Tyr 628 and Tyr 734 are unique to IRS-2 (3). To determine whether the unique tyrosine residues contribute to the interaction between the insulin receptor and the KRLB domain of IRS-2, each residue was replaced with phenylalanine, and the interaction of the mutant proteins with the cytoplasmic domain of the insulin receptor was tested in a yeast two-hybrid analysis. The KRLB domain of IRS-2 was fused to the Gal4 activation domain, whereas the catalytically active cytoplasmic portion of the insulin receptor (including the juxtamembrane region) was fused to the LexA DNA binding domain (LDBD); a IRS-2-IH2 PTB fusion protein was included as a positive control. The IH2 PTB and the KRLB domain interact with the insulin receptor independently during the yeast two-hybrid analysis (Fig. 1A). Furthermore, phenylalanine substitutions at Tyr 624 or Tyr 628 significantly reduced this interaction, whereas mutations at the other tyrosine residues had no effect (Fig. 1C). All of the mutant constructs interacted with a LDBD fused to N-and Cterminal Src homology 2 domains of p85 subunit of phosphatidylinositol 3-kinase, confirming that each one was expressed in the yeast cells (data not shown).
Alignment of IRS-1 and IRS-2 reveals that the KRLB domain of IRS-2 (IRS-2-(591-786)) is homologous to residues 544 -751 in IRS-1; Tyr 624 and Tyr 628 in IRS-2 reside in a PY 624 PEDY 628 motif that corresponds to a SY 578 PEEG 582 motif in IRS-1 (Fig.  2). To determine the important features for insulin receptor binding, several point mutations were prepared in this motif of the IRS1-(494 -741) region and the KRLB domain of IRS-2 and tested for binding to the insulin receptor in a yeast two-hybrid analysis. Substitution of Pro 623 in the KRLB domain with Ser 623 to create an IRS-1-like SYP motif significantly increased the binding to the insulin receptor (Fig. 3A); however, substitution of Ser 577 with proline to create an IRS-2-like PYP motif in IRS1-(494 -741) did not reconstitute insulin receptor binding (Fig. 3B). Furthermore, substitution of Gly 582 with Tyr 582 in IRS1-(494 -741) did not create an insulin receptor binding site (Fig. 3B). Thus, two tyrosine residues separated by three amino acids are essential for interaction of the KRLB domain with the insulin receptor, but this binding domain cannot be reconsti-tuted in IRS-1 by various point mutations. Thus, other determinants outside of the local sequence motif are also required.
The Role of Tyr 624 and Tyr 628 in the KRLB Domain-The binding properties of the KRLB domain were studied in vitro, using GST fusion proteins containing the wild-type sequence, or mutated at position 624, position 628, or both. Unphosphorylated GST fusion proteins containing the wild-type or the mutant KRLB domain were immobilized on glutathione-Sepharose and incubated with 32 P-labeled WGA-purified insulin receptors. After extensive washing, the phosphorylated insulin receptor associated with the wild-type GST-KRLB domain but not with GST alone (Fig. 4). By comparison, insulin receptor binding to the GST-KRLB domain with a mutation at position 624 was reduced by 60%; mutations at Tyr 628 or at both Tyr 624 and Tyr 628 decreased the association to 10 and 5%, respectively.
To evaluate the contribution of Tyr 624 and Tyr 628 in the interaction of the KRLB domain with the insulin receptor, we measured the coprecipitation of the 32 P-labeled insulin receptor by the GST-KRLB domain in the presence of the IRS-2 peptide comprising amino acids 623-633. Within a peptide concentration range of 10 -100 M, coprecipitation of IR was reduced in a dose-dependent manner (Fig. 5). At 100 M peptide, the amount of bound insulin receptor was reduced by 55%. A concurrent experiment carried out with a nonrelevant peptide at the same concentrations did not affect coprecipitation of IR (data not shown). Note that total inhibition of 32 P-labeled insulin receptor coprecipitation by GST-KRLB domain was not obtained, suggesting that additional determinants located outside of this amino acid sequence are required. This may be explained by the fact that the peptide does not adopt the same conformation as the intact domain.
The Role of Tyr 624 and Tyr 628 in the KRLB Domain during in Vitro Tyrosine Phosphorylation-To determine whether Tyr 624 or Tyr 628 is phosphorylated by the insulin receptor, the IRS-2 peptide 623-633 was used as a substrate. This peptide was not phosphorylated by the activated insulin receptor regardless of the concentrations tested (Fig. 6). However, a peptide based on the amino acid sequence of the insulin receptor kinase regulatory loop (IR 1142-1156) used commonly as an in vitro substrate was phosphorylated normally (Fig. 6). These results suggest that Tyr 624 and Tyr 628 in the IRS-2 peptide 623-633

FIG. 1. Quantitative analysis of the interaction between the insulin receptor and IRS-2-KRLB mutants.
Measurement of the ␤-galactosidase activity in transformed yeast is shown. The yeast reporter strain L40 was cotransformed with a plasmid encoding the LDBD-IR␤ in combination with a plasmid encoding different constructs of GAD-IRS-2. Transformants were isolated on selective plates. Activation of the LexA-LacZ reporter gene was monitored by measuring ␤-galactosidase activity in cell lysates using ONPG as substrate. The activities are expressed in Miller's units (25) and are the average Ϯ S.E. of values obtained with samples prepared from three independent transformants. A, cotransformations of LDBD-IR␤ in combination with a plasmid encoding different domains of GAD-IRS-2. As a negative control, the GAD-IRS-2-KRLB was coexpressed with the LDBD-lamin hybrid. B, localization of tyrosine residues in IRS-2-KRLB. C, cotransformations of LDBD-IR␤ in combination with a plasmid encoding GAD-IRS-2-KRLB mutants. All IRS-2-KRLB mutants have the indicated tyrosine residues mutated to phenylalanine. WT corresponds to GAD-IRS-2-KRLB wild type. are not phosphorylated by the insulin receptor. Alternatively, it is possible that these residues are not phosphorylated in a small peptide but are phosphorylated in the intact domain, especially if conformational determinants are crucial for the interaction with the receptor.
Therefore, we performed phosphopeptide maps of GST-IRS-2-KRLB, wild type or mutated on both Tyr 624 and Tyr 628 . The two fusion proteins, mutated or not, were phosphorylated by the insulin receptor in the presence of [␥-32 P]ATP. After digestion with trypsin, they were separated by two-dimensional chromatography. A negative control experiment performed with GST alone showed that GST is not phosphorylated by the insulin receptor (data not shown). Therefore, the phosphopep-tides obtained by tryptic digestion of the fusion proteins can be considered to be generated from the KRLB domain.
Analysis of autoradiograms revealed several separated phosphopeptides; a set of major peptides and a set of minor peptides localized on the two autoradiographs. This indicates that trypsin had not totally digested the proteins and that the KRLB domain of IRS-2 contains several tyrosine residues that can be phosphorylated by the insulin receptor. As shown in Fig. 7, four peptides were present only in the IRS-2-KRLB wild type autoradiogram, whereas they were absent in the Tyr 624  To determine the role of Tyr 624 and Tyr 628 for phosphorylation of the KRLB domain, the wild-type or mutated GST fusion proteins were incubated with [␥-32 P]ATP and WGA-purified insulin receptors. The proteins were separated by SDS-PAGE, and the phosphorylation of the insulin receptor ␤-subunit and the GST-KRLB domain fusion proteins was monitored by autoradiography. The wild-type KRLB domain was phosphorylated by the insulin receptor, whereas its phosphorylation was reduced to 50, 20, and 10% by mutations at Tyr 624 , Tyr 628 , and both positions, respectively (Fig. 8). Thus, Tyr 624 and Tyr 628 are involved in the interaction between the IR and the KRLB domain of IRS-2, including tyrosine phosphorylation, and Tyr 628 seems to be more important than Tyr 624 in this process.
The Interaction between the IR and KRLB Domain Decreases When IRS-2 Is Phosphorylated-Next we examined whether phosphorylation of the KRLB domain modifies the in vitro binding to the insulin receptor. Immobilized GST-KRLB domain was phosphorylated or not phosphorylated by activated WGA-purified IR and then extensively washed to remove the receptor; no 32 P-labeled insulin receptor was retained after stringent washes (Fig. 9A, lane 3). Thereafter, the immobilized KRLB domain was incubated with fresh samples of WGApurified insulin receptor autophosphorylated in the presence of [␥-32 P]ATP in separate incubations. GST alone did not coprecipitate the 32 P-labeled insulin receptor (lanes 1 and 2). However, the unphosphorylated GST-KRLB domain bound strongly to the 32 P-labeled insulin receptor (lane 4). In contrast, the binding of the phosphorylated GST-KRLB domain to the insulin receptor was reduced by 80% (p Ͻ 5 ϫ 10 Ϫ4 ) (lane 5 and Fig. 9B).
To ascertain that this result is not due to contamination of the WGA preparation by proteases and/or by kinases different from the insulin receptor, we performed experiments using immunopurified insulin receptors. We obtained similar results, indicating that the difference between the phosphorylated and the nonphosphorylated forms of GST-KRLB is likely to be due to insulin receptor-induced phosphorylation (data not shown).
The autophosphorylation sites in the regulatory loop of the IR kinase (Tyr 1146 , Tyr 1150 , and Tyr 1151 ) are required for the interaction between the IR and KRLB domain. To address this issue in more depth, we measured the coprecipitation of the IR by the GST-KRLB domain in the presence of a peptide containing Tyr 1146 , Tyr 1150 , and Tyr 1151 . Different concentrations of peptide corresponding to the sequence 1142-1156, phosphorylated or not, were incubated with the GST-KRLB domain immobilized on glutathione-Sepharose. At all concentrations tested, the nonphosphorylated peptide had no effect on receptor binding to the GST-KRLB domain (Fig. 10). In contrast, when the peptide was phosphorylated, competition between the insulin receptor and the peptide was observed, and the amount of insulin receptor bound to the GST-KRLB domain was reduced in a dose-dependent manner (Fig. 10B). These results suggest that the KRLB domain binds to the phosphorylated regulatory loop of the insulin receptor.

DISCUSSION
In addition to the IH2 PTB in IRS-1 and IRS-2 that engages the phosphorylated NPEY motif in the insulin receptor juxtamembrane region, we identified a second region in IRS-2 between residues 591 and 786 that binds to the insulin receptor (13). Recently, He et al. (16) reported similar results. Since this novel region interacts with the regulatory loop of the insulin receptor, we call it the kinase regulatory loop binding, or KRLB, domain. The KRLB domain in IRS-2 provides a unique mechanism of interaction between IRS-2 and the insulin receptor, which is absent from IRS-1. Insulin-stimulated phosphorylation of the regulatory loop plays an important role in activation of the insulin receptor kinase (30). Mutations in this region decrease the kinase activity of the receptor in vitro and in vivo and reduce the strength of the insulin biological responses (31,32). The finding that the regulatory loop may also play an important role for IRS-2 binding functionally couples the activation step to substrate selection. The absence of the KRLB domain from IRS-1 may facilitate the formation of the flexible complex, which is more easily phosphorylated by the insulin receptor. In contrast, by engaging IRS-2 through both the IH2 PTB domain and the KRLB domain, the degrees of freedom in the complex may be reduced and restrict access to certain tyrosine phosphorylation sites. This difference between IRS-1 and IRS-2 is likely to generate differences in the tyrosine phosphorylation pattern of these substrates. Hence, this mechanism could result in different signaling potentials for IRS-1 and IRS-2.
Our experiments suggest that Tyr 624 and Tyr 628 in the KRLB domain are essential for binding to the insulin receptor. Substitution of these residues individually or together with phenylalanine inhibited insulin receptor binding, whereas mutations at the other tyrosine residues in this domain did not affect the interaction. Since the unphosphorylated peptide inhibits the interaction between the KRLB domain and the insulin receptor, we conclude that Tyr 624 and Tyr 628 function in an unphosphorylated state to promote the interaction.
The yeast two-hybrid analysis reveals that the interaction between the KRLB domain and the insulin receptor requires a functional kinase domain and all three phosphorylation sites (Tyr 1146 , Tyr 1150 , and Tyr 1151 ) in the regulatory loop of the insulin receptor (13). In this case, a phosphopeptide corresponding to the amino acid sequence in the regulatory loop inhibits the binding of the KRLB domain to the insulin receptor, whereas the unphosphorylated peptide does not interfere. Together with the yeast two-hybrid analysis, these results suggest that the Tris-phosphorylated regulatory loop constitutes an interaction site for the KRLB domain of IRS-2. Although the peptide competition experiments favor this hypothesis, it is possible that the phosphorylation of the regulatory loop exposes another region that is masked in the unstimulated receptor.
From our results, we propose the following model. The IRS-2 PTB domain would be the primary anchor of IRS-2 to the insulin receptor; the KRLB domain of IRS-2, including the PY 624 PEDY 628 motif, would stabilize this binding to the insulin receptor, which then phosphorylates tyrosines of this domain. When we compared the amount of insulin receptor coprecipitated by the GST-KRLB phosphorylated or not phosphorylated, we found that phosphorylation of IRS-2-KRLB decreased the interaction with the insulin receptor. Hence, it would appear that the binding between the insulin receptor and the KRLB domain of IRS-2 results in tyrosine phosphorylation of the KRLB domain, and this leads to decreased binding of IRS-2 to the insulin receptor. It is tempting to think that such a mechanism might be involved in releasing IRS-2 from the receptor.
Although IRS-1 contains significant amino acid sequence homology to the KRLB domain of IRS-2, it does not bind to the regulatory loop of the insulin receptor (13). The PY 624 PEDY 628 motif in IRS-2 that contains these residues is similar but not completely conserved in IRS-1 (SY 578 PEEG 582 ). Comparison of the two sequences shows that they contain three different residues, two of which (proline 623 and tyrosine 628) are differently charged. However, mutations converting this motif in IRS-1 to resemble the motif in IRS-2 did not create a KRLB domain in IRS-1. In particular, substitutions of Gly 582 with Tyr did not create the binding site, although in vitro binding experiments suggest that Tyr 628 is the predominant functional residue in the KRLB domain. Moreover, mutation of Pro 623 to FIG. 10. Competition between the IR and a phosphorylated or nonphosphorylated IR peptide containing Tyr 1146 , Tyr 1150 , and Tyr 1151 . The GST-IRS-2-(591-786) containing the KRLB domain of IRS-2 (0.5 g/sample) was incubated with 32 P-labeled IR (300 fmol/ sample) in the absence or in the presence of a peptide containing amino acids 1142-1156. The phosphorylated peptide or the nonphosphorylated one was added at different concentrations (shown in M). IR coprecipitated with the GST fused to IRS-2-KRLB was separated by SDS-PAGE and autoradiographed (A). IR-associated 32 P radioactivity was quantified (B). Results are expressed as a percentage of IR coprecipitated in the absence of peptide and are the mean Ϯ S.E. of three different experiments. *, a value of two different experiments.
Ser 623 in IRS-2 enhances rather than diminishes the binding of the KRLB domain to the insulin receptor. This result indicates that the particular structure induced by the prolines surrounding Tyr 624 is not a major determinant for the IRS-2 interaction with the insulin receptor. This suggests that the Tyr 624 and Tyr 628 in the KRLB domain mediate the insulin receptor binding, but other regions of the KRLB domain that are absent from IRS-1 must create the binding pocket.
In conclusion, in addition to the pleckstrin homology domain and the phosphotyrosine binding domain in IRS-1 and IRS-2, IRS-2 has the region called the KRLB domain comprising amino acids 591-786 for interacting with the insulin receptor. Knowing that this region is absent in IRS-1 and that the IRS-molecules are substrates for the kinases associated to receptors for cytokines and GH (33)(34)(35)(36)(37)(38)(39)(40), it is tempting to think that these differences between IRS-1 and IRS-2 might contribute to a unique regulation by the different receptor and nonreceptor tyrosine kinases.