Detection of adaptive immune receptor CDR3s in tumor proteomes and correlations with survival probabilities.

Abstract

e12597 Background: T cell receptors (TCRs) and immunoglobulins (IGs) play an important role in cancer. Specifically, the complementarity determining region-3 (CDR3), encoded at the adaptive immune receptor gene recombination site, is a crucial factor in binding of cancer antigens. There have been very few, if any, successful attempts to identify TCR or IG CDR3s in the tumor proteome. Proteomics-based detection of CDR3s in tumor samples may provide insight into potential biomarkers and better understanding of cancer antigen binding. Methods: Exome and RNA sequencing data from primary breast cancer samples were obtained from The Cancer Genome Atlas Breast Invasive Carcinoma (TCGA-BRCA) dataset. TCR and IG recombination reads were identified in the exome and RNA-seq files via a high-stringency search algorithm. These reads were translated into amino acid sequences, yielding productive CDR3 protein sequences. Then, mass spectrometry-based proteomics data from the TCGA-BRCA tumor samples were analyzed using the FragPipe proteomics pipeline. The abundance of proteins corresponding to the previously identified TCR and IG CDR3 sequences was quantified by the FragPipe algorithm. For each TCGA-BRCA case, mean CDR3 quantity (TCR and IG combined) was calculated. Cases were stratified into high- and low-abundance groups based on the cohort median. Overall survival (OS) probability differences were evaluated using Kaplan-Meier analysis and the log-rank test. Results: Proteins corresponding to CDR3 sequences identified in exome files were detected in mass spectrometry-based proteomics data, and their abundance was quantified. For exome-derived CDR3s, TCGA-BRCA cases represented by a mean CDR3 quantity above the 50th percentile had a higher OS probability than cases below the 50th percentile (n = 105, p = 0.049). Similarly, proteins corresponding to CDR3 sequences identified in RNA-seq files were detected and quantified in mass spectrometry data, and TCGA-BRCA cases with mean CDR3 quantity above the 50th percentile again exhibited higher OS probability than cases below (n = 105, p = 0.034). Conclusions: Adaptive immune receptor CRD3s were detected in the tumor proteome and, when present in higher quantity, were correlated with better OS probabilities. Future, more extensive proteomics analyses may offer better understanding of the potential of CDR3s as biomarkers and better understanding of CDR3 specificity for certain cancer epitopes.