采用抗体库优化法对抗肝癌单链抗体人源化改造的研究

发布时间：2018-04-11 16:51

本文选题：人源化 + 单链抗体　；参考：《暨南大学》2007年博士论文

【摘要】：目的：在已筛选获得1株高亲和力特异性抗肝癌单链抗体(scFv) DM的基础上,采用结合蛋白质同源模建技术、互补决定区(CDR)移植技术和噬菌体展示技术的抗体库优化法,同步实现人源化和优化非人源氨基酸残基,以获得特异性强、亲和力高、免疫原性低的人源化抗肝癌scFv,为肝癌的早期诊治提供新的方法与手段。方法：通过计算机辅助,采用同源模建技术模建抗肝癌scFv DM的三维结构；通过BLAST,从GenBank中选择与scFv DM的重链(VH)和轻链(VL)同源性最高的人源抗体序列为人源化模板；通过分子结构分析和序列分析,确定对抗原结合位点的构象有重要影响的骨架区(FR)残基,把难以判断回复突变残基的人、鼠源副本同时编入人源化抗体序列中；通过重叠延伸PCR技术合成人源化抗体全基因,利用噬菌体展示技术构建人源化组合抗体库；通过肝癌细胞筛选阳性克隆,ELISA方法测定各个克隆抗原结合活性；用硫氰酸盐洗脱法测定并比较亲本抗体和人源化scFv的相对亲和力。结果：抗肝癌scFvDM的三维结构示：六个CDR襻位于Fv段的N端,形成一个与抗原结合的口袋,口袋深16.1A。用Prostat验证,模建的三维结构的79.8%的Φ和Ψ分布于Ramachandran图的核心区域,模板1NQB的为83.2%；确定人抗体序列AAW67378和BAB18257分别作为VH和VL人源化模板,FR同源性分别为76.1%和71.6%；确定人、鼠VH和VL的FR序列分别有20个和23个不同残基,其中28个人源化残基、8个回复突变残基和7个难以确定残基；成功构建人源化组合抗体库,抗体库包含了由25不同重链和22不同轻链组成的128种人源化DM；经过筛淘及ELISA鉴定,获得HDM1、HDM2和HDM3三个阳性克隆,相对亲和力指数分别为：HDM1 1.6mol/L、HDM2 1.2mol/L。HDM3 2.2mol/L 结论：成功模建抗肝癌scFv DM的三维结构,模拟结构合理可靠；抗体库优化法是一个高效、简便的人源化方法；获得的3株阳性克隆HDM1、HDM2、HDM3与亲本抗体DM同样具有良好的抗原结合特异性和较高的亲和力,为进一步探索在肝癌早期诊疗的临床应用奠定了基础。
[Abstract]:Objective: on the basis of screening a high affinity specific single chain antibody (scFvDM) against hepatocellular carcinoma (HCC), an antibody library optimization method was used to optimize the antibody library by using the technique of protein homology modeling, complementary determinant region (CDR) transplantation and phage display.The humanization and optimization of non-human amino acid residues were carried out simultaneously to obtain the humanized anti-HCC scFv with strong specificity, high affinity and low immunogenicity, which provided a new method and means for the early diagnosis and treatment of HCC.Methods: the three-dimensional structure of anti-HCC scFv DM was constructed by computer aided modeling, and the human antibody sequence with the highest homology to scFv DM and light chain GenBank was selected as the humanized template by BLAST.Through molecular structure analysis and sequence analysis, the skeleton FR) residues which have important influence on the conformation of antigen-binding sites were determined.The whole humanized antibody gene was synthesized by overlapping extension PCR technique, the human combined antibody library was constructed by phage display technique, and the binding activity of each clone antigen was determined by Elisa method for screening positive clones of hepatoma cells.The relative affinity of parental antibody and humanized scFv was determined and compared by thiocyanate elution method.Results: the three-dimensional structure of anti-hepatoma scFvDM showed that six CDR loops were located at the N-terminal of the FV segment, forming an antigen-binding pocket with a depth of 16.1 A.By Prostat, 79.8% of 桅 and 蠄 were distributed in the core region of Ramachandran map and 83.2% of template 1NQB. The homology of human antibody sequence AAW67378 and BAB18257 as VH and VL humanized template FR were 76.1% and 71.6%, respectively.The FR sequences of mouse VH and VL have 20 and 23 different residues, of which 28 are humanized residues, 8 are reverse-mutated residues and 7 are difficult to determine the residues.The antibody library contains 128 kinds of humanized DMs composed of 25 different heavy chains and 22 different light chains. After screening and ELISA identification, three positive clones of HDM1, HDM2 and HDM3 were obtained, and the relative affinity index was 1.6 mol / L HDM2 1.2mol/L.HDM3 2.2mol/L, respectively.Conclusion: the three-dimensional structure of anti-hepatoma scFv DM was successfully constructed, and the simulated structure was reasonable and reliable, and the antibody library optimization method was an efficient and simple humanization method.The three positive clones, HDM1, HDM2HDM3, also had good antigen-binding specificity and high affinity with their parental antibody DM, which laid a foundation for further exploration of clinical application in the early diagnosis and treatment of liver cancer.
【学位授予单位】：暨南大学
【学位级别】：博士
【学位授予年份】：2007
【分类号】：R392

【参考文献】