Proteome Microarray Based Autoantibody Repertoire Construction And Surgery-associated Biomarker Discovery For Lung Adenocarcinoma

Tumor associated autoantibodies are usually related to tumor progression and have great value for clinical applications.In addition,there are many natural autoantibodies in the blood of tumor patients as well.It is of great significance to study the structure and dynamic changes of autoantibody repertoire in tumor patients and to identify autoantibodies closely related to tumor progression,however,it is inadequate for this area duo to lack of effective tools.Lung cancer is the leading cause of death among all malignant tumors worldwide,and lung adenocarcinoma is the main pathological type.Based on high-throughput proteome microarray technology,this study comprehensively analyzed and compared the serum autoantibody profiles of lung adenocarcinoma patients and healthy people,established a random autoantibody model,and identified a number of surgery-associated autoantibody biomarkers for lung adenocarcinoma.Firstly,to dynamically analyze the constitution and changes of autoantibody repertoires of lung adenocarcinoma,we intensively collected longitudinal serum samples from 5 patients: serum samples were collected every 1 week or 2 weeks from the day before surgery to at least 3 months after surgery.In addition,15 sera from healthy people were collected as well.Based on a human proteome microarray with 20,240 recombinant proteins,we obtained the autoantibody repertoires of these samples and performed global analysis.Overall,there was no significant change among the longitudinal samples for each patient,indicating that the autoantibody repertoire is probably stability overtime,and even after surgery.While significant differences were observed for autoantibody repertoires from different individuals,indicating the heterogeneity.The autoantibodies from each individual can recognize thousands of proteins,of which hundreds are conserved.We also found that the isotype of autoantibodies is mainly Ig M.Secondly,we proposed a model for autoantibody repertoire,i.e.,the random autoantibody model.In addition to conserved antigens,there are a large number of protein antigens,and their antibodies appear in the population at a certain frequency,which is the major factor for the heterogeneity of autoantibody repertoires.We found that thousands of protein antigens in human body can elicit or cross-react with autoantibodies in an independent manner with different probabilities.Based on the autoantibody repertoires we obtained,we established a random autoantibody model.Using this model,we found that in the traditional case-control study for tumor-associated autoantigen discovery,due to the presence of a large number of physiological autoantibodies,some of them can randomly show up in the case group and lead to false positives.On this basis,we quantified false positives and false negatives and calculated the required sample size.Finally,a number of surgery-associated autoantibody biomarkers were identified.By comparing the changes in autoantibody response of each protein in preoperative and postoperative samples of each patient,we identified a number of autoantibodies with significantly reduced levels,we defined these antibodies as surgery-associated autoantibodies,which were further validated by intensively collected longitudinal sera.Finally,six surgery-associated autoantibodies were identified.Surgery-associated autoantibodies exist in all 5 lung adenocarcinoma patients,indicating high prevalence.However,for these identified surgeryassociated autoantibodies,there is very low overlap among the patients.On this basis,we propose an application plan for personalized recurrence monitoring of cancer based on surgery-associated autoantibodies.In summary,in this study,we profiled autoantibody repertoires of lung adenocarcinoma by longitudinal samples form 5 patients,analyzed its dynamic changes,and identified six surgery-associated autoantibody biomarkers.Based on the widespread presence of surgery-associated biomarkers,the clinical value for recurrence monitoring is preliminarily elucidated.A random autoantibody model was proposed to explain the significant differences in autoantibody repertoires among individuals,we believe this study can provide some guidance for the discovery of tumor-associated antigens.