Prediction of metabolites from cancer-associated fibroblasts that can have profound effects on cancer cells

Abstract

Cancer-associated fibroblasts (CAFs) are the most abundant component of tumor microenvironment (TME) that is an ensemble of non-cancerous cells surrounding cancer cells. In particular, CAFs are known to have tumor-assisting functions, which may promote active growth and survival of the interacting cancer cells through their metabolic reprogramming despite being in a harsh environment (e.g., hypoxia or low pH). In this regard, CAF-derived metabolites are expect to have profound effects on metabolism of the cancer cells. To identify novel CAF-derived metabolites that may heavily influence the cancer cells, CAF-specific genome-scale metabolic models (GEMs) were reconstructed and simulated. GEM is a computational model that can predict a metabolic state in a cell-specific manner by integrating omics data (e.g., RNA-seq data). In this study, a total of six GEMs were reconstructed by integrating RNA-seq data from one normal fibroblast, originating from pancreas, and five CAFs associated with breast, colon, skin, lung, and pancreatic tumors. Simulation of the GEMs predicted 20 metabolites that appeared to be heavily used by all the five CAF GEMs in comparison with the one normal fibroblast GEM. Such metabolites were expected to be significantly associated with establishment of TME that is favorable for the cancer cells’ growth. Some of these metabolites are associated with pyrimidine metabolism, fatty acid metabolism, and primary bile acid biosynthesis. Upon experimental validation, the predicted metabolites are expected to help better understand CAFs and TME in terms of metabolites, and also provide clues for identifying effective drug targets for cancer treatment.