SEREX技术分析筛选乳腺癌自身抗体标志物的研究

SEREX screening autoantibodies for breast cancer detection

目的:自身抗体是目前最受关注的一类肿瘤标志物候选分子,有助于提高乳腺癌检测的敏感性和特异性。鉴于单一一个乳腺癌文库的丰度有限,而且乳腺癌和肺癌发生来源相近,本研究从肺癌文库筛选乳腺癌特异性自身抗体标志物。方法:重组cDNA表达文库血清学分析(SEREX)联合生物淘洗富集技术从肺癌cDNA T7噬菌体文库中筛选乳腺癌相关蛋白,测序鉴定得到噬菌体表达蛋白信息。免疫印迹和酶联免疫吸附(ELISA)分析每一个开放阅读框架内的噬菌体表达蛋白的自身抗体水平。逻辑回归模型和留一交叉验证评估诊断价值。结果:筛选得到2个开放阅读框架内蛋白序列信息。通过免疫印迹和ELISA分析,这2个噬菌体表达蛋白能显著区分出病人和正常人。逻辑回归模型联合检测的敏感性达到82.4%,特异性达到88.2%;留一交叉验证两者联合的预测值敏感性和特异性分别达到78.5%和82.1%。结论:SEREX是有效的肿瘤自身抗体标志物筛选技术,从肺癌文库筛选乳腺癌特性抗原的方法是可行的,同时HSPA4L和RINT-1蛋白的血清自身抗体联合对乳腺癌的检测准确率要好于其中任何一种标志物。

Objective：Autoantibodies induced by tumor-associated proteins may be used in a serum profile and improve the sensitivity and specificity of breast cancer detection.In view of the limited information of a breast cancer cDNA library and the similar tumorigenesis of breast cancer and lung cancer,this study aimed to screen and analyze breast cancer autoantibody markers from a lung cancer cDNA library.Methods：Combinations of serological analysis of recombinant cDNA expression library （SEREX） and biopanning enrichment techniques were used to screen tumor-associated proteins from a lung cancer cDNA T7 phage library with sera from normal and breast cancer patients.After biopanning,the enrichment of tumor-associated proteins was assayed using immunochemical detection.Some phage clones which showed obviously different from breast cancer pattents and normal serum samples were collected for PCR and BLAST sequencing analysis.Unique and ORF phage-expressed proteins were then used to develop phage protein immunochemical detection and enzyme-linked immunosorbent assays （ELISA） to measure corresponding antibodies level in 62 breast cancer patients and 62 normal serum samples.Logistic regression model and leave-one-out validation were used to assess the value of autoantibodies of single marker and combined markers.Results：We harvested 43 putative breast tumor-associated phage clones from biopanned lung cancer cDNA T7 phage library.After PCR,the productions of ten clones which have significant different electrophoresis bands were sequenced.Eight proteins were blast from NCBI.Antibodies to HSPA4L and RINT-1 were measured by ELISA and results showed that two of the phage clones had statistical significance in discriminating patients from normals.Measurements of the two predictive phage proteins were combined in a logistic regression model that achieved 82.4％ sensitivity and 88.2％ specificity in prediction of sample status,whereas leave-one-out validation achieved 78.5％ sensitivity and 82.1％ specificity among 62 patients and 62 control samples.Conclusion：SEREX was a effective method for screening tumor-associated autoantibodies,and it was feasible to screen breast cancer-associated markers from a lung caner cDNA library.Serum autoantibodies to HSPA4L and RINT-1 proteins showed superior accuracy for breast cancer detection than each one of them.Further refinements will need to verify the accuracy from other types of cancer.