Prediction of metabolic drug targets that restore the biological status of a human host cell infected with respiratory viruses

Abstract

Viruses infect human host cells to replicate themselves, thereby replicating their genes and structural proteins. Such viral infection consequently transforms metabolic phenotype of an infected host cell among many possible phenotypic changes incurred. Thus, restoring the metabolic phenotype of the infected host cell can be an effective means to control and even treat viruses. In this regard, we undertook a systems biology approach to identify effective metabolic drug targets in a human lung adenocarcinoma cell line, A549, as a host cell against seven representative respiratory viruses. In this study, we focused on multiple respiratory viruses, including SARS-CoV-2, as they have become notable threats to the mankind. In our systems biology approach, genome-scale metabolic models (GEMs) were first reconstructed for the A549 cell line infected with each of the seven strains of respiratory viruses. GEMs are a type of computational model that contains information on entire metabolic genes and biochemical reactions for a target cell, and can be simulated to predict the cell’s metabolic phenotype under a given condition. Importantly, a cell-specific GEM can be built through an established integration method if omics data, often RNA-seq data, are available. Each cell line-specific GEMs were subsequently subjected to the metabolic simulation method called robust Metabolic Transformation Algorithm (rMTA) to identify effective metabolic drug targets, some of which were predicted to be effective against multiple strains of the respiratory viruses examined in this study. These predicted treatment targets were shown to restore metabolic flux distribution of the infected cell line towards the non-infected states. This study showcases a novel systems biology approach that allows effective metabolic drug targets in a data-driven manner.