SARS-CoV-2 Codon Usage Bias Downregulates Host Expressed Genes With Similar Codon Usage

Abstract

Severe acute respiratory syndrome has spread quickly throughout the world and was declared a pandemic by the World Health Organization (WHO). The pathogenic agent is a new coronavirus (SARS-CoV-2) that infects pulmonary cells with great effectiveness. In this study we focus on the codon composition for the viral protein synthesis and its relationship with the protein synthesis of the host. Our analysis reveals that SARS-CoV-2 preferred codons have poor representation of G or C nucleotides in the third position, a characteristic which could result in an unbalance in the tRNAs pools of the infected cells with serious implications in host protein synthesis. By integrating this observation with proteomic data from infected cells, we observe a reduced translation rate of host proteins associated with highly expressed genes and that they share the codon usage bias of the virus. The functional analysis of these genes suggests that this mechanism of epistasis can contribute to understanding how this virus evades the immune response and the etiology of some deleterious collateral effect as a result of the viral replication. In this manner, our finding contributes to the understanding of the SARS-CoV-2 pathogeny and could be useful for the design of a vaccine based on the live attenuated strategy.

Keywords: SARS-CoV-2; codon optimality; codon usage bias; pathogeny; translational control; vaccine design.

Figure 1

SARS-CoV-2 ORFome have lower content of GC3. (A) Bar chart comparing the G and C nucleotides on the third nucleotide (GC3) at every codon in the ORFeome of SARS-CoV and SARS-CoV-2. (B) Bar chart showing the proteins level of nine viral proteins at 10 h post-infection (data obtained from Bojkova et al., 2020). The GC3 content for each viral ORF is showed in the corresponding bar.

Figure 2

The translational rate of genes with high CCorr in SARS-CoV-2-infected cells is lower than the mock-infected cells. Scatter plot showing negative correlation between translational rate of the 100 most expressed proteins in SARS-CoV-2 infected CACO-2 cells vs. virus codon usage.

Figure 3

High codon correlation between SARS-CoV-2 and lung cells is associated with a translational rate decay over specific transcripts. (A) Frequency distribution of CCorr on the 100 highest expressed genes in lung cells and CACO-2 cells. (B) Comparison between the CCorr frequency distributions obtained from the 100 highest expressed genes in lung and arterial tissues. (C) The tissues of the lungs and arteries share most of the highly expressed genes. (D) Histogram of the translational rate of 100 highest expressed genes in lung cells. Transcripts with high and low codon correlation (CCorr) with SARS-CoV-2 are highlighted.

Figure 4

SARS-CoV-2 codon usage may have an impact on host translational machinery. GO enrichment analysis performed over genes of interest. The most enriched biological processes and molecular functions are highlighted. Both size and color shade of bars represent the level of significance.