Establishment And Preliminary Application Of Antibody Variable Region Stabiliy Evaluation Model

Recently, engineering antibody was development quickly in the pharmaceuticalmarket in the world, and has become a pillar product of one of the world’s biologicalengineering pharmacy. Unlike other biological drugs, antibody needs to reform beforeit becomes drug. However, in the process of antibody reform, the stability change ofthe antibody becomes bottleneck and restricts the antibody development. A series ofresearch results showed that the antibody variable region is the key factor to affect theantibody stability. So, the antibody variable region stability and optimization havebecome the important step for the recombinant antibody research and development.With the fast development of structural biology, bioinformatics and computer science,the protein-protein interaction was studied using the chemical thermodynamicparameters (such as enthalpy, entropy, free energy etc.) based on the protein structuralinformation. Furthermore, the interface character and the complex stability ofprotein-protein interaction were analyzed theoretically.In the present work, using the established antibody variable region sequence andstructure database, according to the following methods, including sequence alignment,structure modeling and optimization, distance geometry, intermolecular hydrogenbond forming theory, and binding energy calculation, the key factor influencing theantibody variable region stability was determined theoretically. The results providethe theoretical model to reconstruct the antibody stability. The research work wasshown as following in detail:To construct the informational database of antibody sequence andstructure—Based on the bioinformatics technology, the mouse and human antibody variable region sequence information were collected from network database (such asKabat, IMGT, PDB and SwissProt and so on). With the help of computerprogramming technology (i.e. Java language), the translation between nucleic acidsequence and protein amino acid sequence was realized. According to the judgmentmodel of Kabat, chothia and AbM, the framework region (i.e. FR) andcomplementarity-determining region (i.e. CDR) of the antibody were classified andannotated. And then, the antibody variable region sequence database was built. Usingthe multiple alignment ClustalW method, the human antibody variable regionsequences were analyzed based on cluster analysis method. The amino acid positionuseful frequency of human antibody variable region was determined. The results werelaid the foundation for the antibody stability and humanization.Using the antibody variable region sequence database and the protein data bank(i.e. PDB), the mouse and human antibody variable region structure information werecollected. Based on the computer guided molecular homology modeling method, the3-D structures of the antibody variable region were optimized. And then, the rationalstable conformations of the antibody variable region, heavy chain and light chainvariable region were stored in the local structural informational database of theantibody variable region. Furthermore, the topological structure of the antibodyvariable region was drawn and stored in the database.To determine the key factor influencing the stability of the antibodyvariable region—Using the antibody variable region sequence and structure database,based on the intermolecular hydrogen bond formation theory, and free energy ofreaction theory, the dynamic structure and energy characteristics between the heavyand light chain variable region of the antibody were studied. Furthermore，with thehelp of the principle of statistics, nonlinear fitting and regression analysis, themathematical model was set up. According to the simulation and statistics analysis,there is linear relationship between the binding energy and the number of theintermolecular hydrogen bonding, van der Waals interaction of the heavy and light chain of the antibody; there is polynomial correlation between the binding energy andthe physicochemical properties of the antibody.Using the amino acids position frequency of and established model analysis, thehumanized anti-ricin antibody, which could not obtain the stable engineering cell line,was evaluated and optimized. The sequence and3-D structure of the anti-ricinantibody were analyzed to design a novel mutant. And then, according to thetheoretical results mentioned above, the stable engineering cell line of the humanizedanti-ricin antibody was obtained in the experiment.The self structure of the antibody variable region (conformation,physicochemical properties) possesses great influence on its stability. The antibodystability can be improved with the antibody structure optimization.