Rapid and Multifaceted Characterization of the Effects of the Secreted EVome in Pancreatic Cancer: A Global Omics Approach for Screening Putative Biomarkers

Abstract

Background: Pancreatic cancer is one of the most difficult cancers to detect early and most patients die from complications arising due to distant organ metastases. The lack of bona fide biomarkers is one of the primary reasons for the late diagnosis of this cancer. Most researchers have focused on screening strategies to characterize biomarkers that show promise in preclinical models but eventually fail in a clinical scenario. Pancreatic cancer is a multifactorial disease and warrants a multi-omics approach to identify novel biomarkers for early diagnosis. Methods: In order to characterize the proteome, Extracellular vesicles (EVs) isolated from different in vitro conditions mimicking interactions between pancreatic cancer epithelial and stromal cells were isolated and subjected to high throughput mass spectrometry. Biological activity of the secreted EVome was analyzed by investigating changes in distant organ metastases and early changes in the microbiome. Putative biomarkers were selected and validated using a mouse-human hybrid TMA that was specifically generated for this study. As potential targets, immunohistochemical analysis of protein candidates (Kif5b, Sfpr2, Loxl2, and Mmp3) was performed on the TMA to assess their potential as candidates for early discovery of pancreatic cancer. Results: The EVome of epithelial and stromal cells co-cultured with each other is distinctly different from individual cells with distinct protein compositions. While these proteomes could not induce significant changes in Pre-Metastatic Niche (PMN) modulation and distant organ metastases, they did induce significant early changes in the microbiome, as indicated by changes in the α and β-diversities in experimental animal models. Novel biomarkers validated in this study, such as Kif5b and Sfrp2 show promise for early detection and investigation of pancreatic cancer. Conclusions: Multi-omics analyses of EVs from mimicking conditions of tumor stromal interactions resulted in the identification of Kif5b and Sfrp2 as bona fide biomarkers with biological activity which could prove to be immensely helpful in improving early detection and diagnosis of pancreatic cancer.