Virus genomes have codon usage bias, which can be constrained by the host translational machinery. The effects of codon composition of a given mRNA on its own translation have been extensively reported in literature and is considered an important determinant of gene expression. However, how the codon usage of a viral genome affects the translation of host genes has been less explored. In fact, the virus replication requires a lot tRNA resource for the most demanded codons. Thus, the consumption of specific tRNAs for the virus replication could be an alternative manner, to control host protein synthesis and also to generate deleterious collateral effects. By integrating genomic, transcriptomic and proteomic data, we address whether the SARS-CoV-2 infection downregulates the most expressed genes in the host cell. Our analysis confirms this hypothesis. Extending this result to lung tissue we extract a list of 27 genes that could be affected by this mechanism of epistasis. Finally, we compile and discuss the functions and processes associated with these genes. We believe that our original hypothesis and analysis could be helpful for the challenging task of understanding the collateral effects of the COVID19.