Application of genome-scale metabolic models for the prediction of drug targets for cancers and COVID-19

Abstract

Genome-scale metabolic models (GEM) play an important role in systems biology in understanding cell metabolism. GEM is a computer model that contains all biochemical reaction information involving metabolic genes in cells, and the activity of all biochemical reactions under various conditions can be predicted through optimization techniques. In particular, in the case of human metabolism, it becomes possible to build cell-specific GEMs by integrating omics data such as RNA-seq obtained from specific individuals or specific cells. In this thesis, we aim to predict drug targets in cancer-associated fibroblasts (CAFs) that can inhibit cancer proliferation by constructing cell-specific RNA-seq data-based GEMs, and further apply GEM to predict drug targets in COVID-19-infected patients that can inhibit the proliferation of coronavirus. CAFs are a major component of the tumor microenvironment (TME), which affects the progression, proliferation, and immune response of proximal cancer cells. Especially, interactions between CAFs and cancer cells via metabolites have profound effects on the metabolic reprogramming of cancer cells. To find out metabolites in CAF that can help control cancer growth, RNA-seq data was first used to build CAF-specific GEMs, and so-called metabolite-centered simulation was performed to evaluate the importance of each metabolite. In this study, a total of five CAF-specific GEMs were reconstructed, including breast cancer, colon cancer, skin cancer, lung cancer, and pancreatic cancer, and GEMs for normal fibroblasts derived from pancreas were also established to understand the common metabolic characteristics of CAF. Through a series of metabolite-based simulations, a total of 20 metabolites were predicted to be important for cancer suppression of CAF, and they will be subjected to 5 corresponding enzyme inhibitors for possible inhibition of cancer growth through CAFs. Meanwhile, in the study of coronavirus infections, patient-specific GEMs were constructed using RNA-seq data obtained from blood samples of 30 COVID-19-infected patients. Approaches and insights reaped from this study will be also helpful for predicting effective drug targets for other diseases that are associated with metabolism.