| Insulin-like growth factors (IGF I, IGFII) with 40% homology compared with the precurosor of insulin, play roles in cell proliferation. There are six members (IGFBP1～6) in the Insulin-like growth factor binding proteins (IGFBP) family. The IGF axis is comprised of the IGF, IGFBPs, IGF-receptor, insulin receptor and some specific protease against IGFBPs. Hydrolization of IGFBPs induced by these proteases could mediate the bioavailabiity of the IGF. In recent studies, new members were recruited into the axis, such as insulin receptor-related protein (IRR), several insulin-like growth factor binding protein related proteins (IGFBP-rp).Insulin-like growth factor binding protein 7 (IGFBP7, IGFBP-rp1), was screened from the cDNA substraction library of colonic adenocarcinoma-normal mucosa by suppression subtractive hybridization (SSH) in our laboratory in 1999, which was upregulated in the colorectal cancer. In our previous study work, we found that it might be a tumor suppress gene in colorectal cancer. Meanwhile we found positive correlation between IGFBP7 expression and the level of fasting serum glucose in patient who suffered CRC. And the concentration of IGFBP7 in peripheral blood was increased in type 2 diabetic patients who did not accept any therapy. So we think that IGFBP7 may be of great importance to CRC and type 2 diabetes mellitus. Compared with IGFBPs 1-6, the ratio of insulin:IGF binding for IGFBP-7 was approximately 500-fold higher than for IGFBPs 1-6 by densitometric analysis. These results highlight the hypothesis that IGFBP7 is relevant to insulin resistance, which was significant for studying the interaction between IGBFP7 and insulin. The high affinity of IGFBP7 for insulin was focused on by our research group and the laboratory of calculationeal chemistry in Zhejiang University. Our colleague in Department of Chemistry applied the dominance of homology modeling, docking analysis, interaction energy curve analysis and high-throughput molecular field modeling to predict the insulin binding domain in IGFBP7. By the simulation of the dynamic interaction process between IGFBP7 and insulin, they found the domain that could bind insulin directly. The binding sites of the amino acid residues are AA172-176, AA196-201, AA235-AA244. And we constructed expression vectors of the fragment proteins containing these domains, and purified the proteins expressed in E coli. The affinity of those proteins for insulin was examined by western blot, and C terminal of IGFBP7 molecules is the main domain that interaction with insulin directly, which exactly verified the prediction result.Based on the previous predictive results, we performed further precise prediction about the binding condition of these two molecules. In the simulative process of bonding and dissociation, we tried to find the key amino acid in C terminal of IGFBP7 that may bind to insulin. Then we also found that although the key amino acid was substituted by some amino acids, the space structure and the volume of side chain would not be changed obviously. The arginine (Arg198) and histidine (His200) were identified in the C terminal of IGFBP7 which may directly bind to insulin. Because when they were changed into isoleucine (Ile198) and phenylalanine (Phe200), the mutant type IGBFP7 would be disassociated from the complex formed by insulin and IGFBP7.According to this prediction, site-directed mutagenesis was performed for mutating the vector with the IGFBP7 DNA sequences. The three mutants of IGFBP7 including Arg198→Glu198(Ml),His200→Phe200(H200F),Arg198/His200→Ilel98/Phe200(M2) were constructed and expressed in E coli, and mutant proteins were purified by affinity chromatography. Then dot blot and pull down experiment were performed to qualitatively examine the affinity for insulin. The present results of our work verify the predictive outcomes, which indicated that Arg198 and His200 bind to insulin directly. Because the affinity for insulin of the mutant type M2 (Arg198/His200→Ilel98/Phe200) was decreased obviously. These findings not only certified the calculationeal models, but also highlight promising therapy target for insulin resistance; moreover, the further research into the interaction between IGFBP7 and insulin would bring perspective for interfereing the type 2 diabetes and metabolism syndrome.Subsequently, there are three conclusions from our prediction and verification data about the interaction between IGFBP7 and insulin:1. The binding site in molecules could be precisely predicted by the dominance of homology modeling, docking analysis, interaction energy curve analysis and high-throughput molecular field modeling.2. The proteins expressed in E coli and purified by AKTA system would not be changed greatly in molecular structure and the reactionogenicity.3. The Arg198 and His200 in the C terminal of the IGFBP7 might be the key binding sites that interacting with insulin because of the decreasing affinity for insulin of the mutant type M2 (Arg198/His200→Ile198/Phe200). |
