Predicting the recombination potential of severe acute respiratory syndrome coronavirus 2 and Middle East respiratory syndrome coronavirus

Abstract

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) recently emerged to cause widespread infections in humans. SARS-CoV-2 infections have been reported in the Kingdom of Saudi Arabia, where Middle East respiratory syndrome coronavirus (MERS-CoV) causes seasonal outbreaks with a case fatality rate of ~37 %. Here we show that there exists a theoretical possibility of future recombination events between SARS-CoV-2 and MERS-CoV RNA. Through computational analyses, we have identified homologous genomic regions within the ORF1ab and S genes that could facilitate recombination, and have analysed co-expression patterns of the cellular receptors for SARS-CoV-2 and MERS-CoV, ACE2 and DPP4, respectively, to identify human anatomical sites that could facilitate co-infection. Furthermore, we have investigated the likely susceptibility of various animal species to MERS-CoV and SARS-CoV-2 infection by comparing known virus spike protein-receptor interacting residues. In conclusion, we suggest that a recombination between SARS-CoV-2 and MERS-CoV RNA is possible and urge public health laboratories in high-risk areas to develop diagnostic capability for the detection of recombined coronaviruses in patient samples.

Keywords: MERS-CoV; SARS-CoV-2; coronavirus; emergence; predictions; recombination.

Fig. 1.

Recombination potential of SARS-CoV-2 and MERS-CoV. (a) Pairwise alignment of the reference genomes of SARS-CoV-2 and MERS-CoV. The gene locations for both viral genomes are plotted on the x- and y-axes. Detected syntenic blocks are shown in the plot. The largest synteny block detected occurs between region 12 944–19 922 in SARS-CoV-2 and region 12909–19875 in MERS-CoV. (b) Structure of SARS-CoV-2 RNA-dependent RNA polymerase (PDB ID: 6M71 [26]), with the region of similarity with MERS-CoV from (a) highlighted in red. (c) Sliding window analysis of SARS-CoV-2 : MERS-CoV genome alignment, displaying the percentage identity of all 30 length nucleotide segments. The axis numbering corresponds to alignment position. (d) Alignments of three selected regions of sequence similarity between SARS-CoV-2 and MERS-CoV. Alignments from (c) were extended in the 5′ and 3′ direction if additional matching positions were present. Region 2 includes a 32 bp segment with only one mismatch (underlined). The axis numbering corresponds to the position within the SARS-CoV-2 genome.

Fig. 2.

Co-expression profile of angiotensin-converting enzyme 2 (ACE2) and dipeptidyl peptidase 4 (DPP4). (a) RNA-seq based expression profile of ACE2 and DPP4 across 54 tissues. Expression data are derived from the GTEx database [27]. ‘Small intestine – terminal ileum’ is highlighted as a key tissue of interest since ACE2 and DPP4 are co-expressed at high relative expression levels compared to other tissues. The centre line denotes per-sample median expression level. (b) Analysis of ACE2 and DPP4 co-expression in GEO dataset GSE75214, including microarray expression profiles of ileum samples from healthy individuals, and individuals with inflammatory bowel disease and Crohn’s disease [28]. Left: a heatmap of the top 100 ACE2 co-expressed genes. For the heatmap, the x-axis includes all samples in the microarray dataset, while the y-axis includes represented genes. Values in each cell represent gene expression levels. Horizontal lines on the right indicate ACE2 and DPP4. Right: ACE2/HNF4A and ACE2/DPP4 co-regulation. t.p.m., transcripts per million.

Fig. 3.

Evolutionary relationships of ACE2 and DPP4 proteins across various animal species. (a, b) Phylogenetic relationships of ACE2 and DPP4 orthologues across various animal species. Species that are known or suspected to be infected by either virus are indicated, namely cats, bats and the dromedary camel. Each tip in the tree represents a sequence from a different animal species, and the tips are annotated by coloured circles representing percentage identity to their human orthologue. Node labels represent bootstrap support as a fraction of a circle, where a greater proportion of the circle being black indicates a greater proportion of bootstrap support. The full list of sequences and their identities are available in Table S1. (c, d) Conservation of residues within the virus spike–receptor interfaces for ACE2 and DPP4 across animal species.