Establishment Of Array-ELISA Assay Towards Multiplex Detection Of Gastric Cancer-associated Biomarkers

The sandwich ELISA-based multiplex immunoassays (Array-ELISA) are gaining the spotlight and prospect since the ongoing discovery and validation work of novel protein biomarker candidates keep stimulating the desire for large-scale clinical sample processing as well as in-parallel quantification. The essence of this technique is how to customize the immunoassay towards biased analytes.However, the systematic study on the development of Array-ELISA assays toward novel biomarker candidates from scratch is still lacking due to limited resources of protein and antibody. In this study, we comprehensively focus on the technical problems during the customization process in an attempt to develop an Array-ELISA assay towards the gastric cancer-associated biomarker candidates (GCABC), and preliminarily explore the individual or combinatorial performance of GCABC upon the sera samples.Based on the GCABC found in our laboratory and reported by other groups, we collected dozens of them as the analytes. Large-scale recombinant protein expression and antibody production were conducted for generation of high-quality antigens and antibodies pertaining to different analytes. We further screened antibody pairs in large-scale in order to get as many antibody pairs to individual analytes as possible. With the available antibody pairs and cognate antigens, we then developed the Array-ELISA assays toward different GCABC with multiple optimizations and evaluation to guarantee multiplexing performance. At last we adopted the custom Array-ELISA assays to the sera screening in order to explore the potential predictive performance of GCABC.We obtained 43 recombinant proteins toward 27 analytes,72 monoclonal antibodies (mAbs) toward 26 analytes and 23 polyclonal antibodies (pAbs) toward 23 analytes during the large-scale protein and antibody preparation process, and 18 capture mAb vs. detection pAb antibody pairs toward 15 analytes,18 capture pAb vs. detection mAb antibody pairs toward 15 analytes and 37 capture mAb vs. detection mAb antibody pairs toward 10 analytes during the antibody pair screening. Qualified antibody pairs, which are the basic requirement for the multiplex immunoassay development, are often well-known with minimal background interference. We thus focused on exploring the causes for high background interference during the antibody pairing process aiming at improvement of pairing performance as well as screening efficiency. We discovered that the high background interference was irrelevant of the antibody purification procedure as well as the cross-species recognition between secondary antibody and primary antibody, but was tightly associated with the ascites preparation. We thus for the first time proposed that the heterophilic antibodies derived from animal were one of the major causes of high background interference. By using the non-immune mouse sera blocking coupled with the antigen-mediated affinity chromatography treatment, the high background interference was effectively reduced during the pairing process with ascites-derived mAbs.Through systematic optimizations such as printing, reaction, non-specific binding and cross-reactivity, we successfully developed the multiplex quantitative Array-ELISA assay toward 6 GCABC. The variation of coefficient was less than 5%, the detection limit less than 10 pg/mL and the dynamic range at 4 degrees of magnitude. After applying the Array-ELISA assay on qualitative screening toward 70 cases of sera, we found the combination of MMP11, CCN1, PGC and PRDX6 was capable of predicting gastric cancer with both sensitivity and specificity higher than 90%. We further conducted the quantitative Array-ELISA assay based on 5 GCABC (MMP11, GKN1, REG4, CK8, and OPN) toward 80 cases of sera. The 5-GCABC panel could predict the gastric cancer with 67% of sensitivity and 80% of specificity. The preliminary screening results based on the GCABC-oriented Array-ELISA assays thus indicate the potential application value in clinic. With deeper optimizations and large-scale sera screening, it is very promising to establish robust prediction models and conventional clinical assays toward gastric cancer upon the multiplex assay system.