Cardioviruses are genetically diverse and cause common enteric infections in South Asian children

Abstract

Cardioviruses cause enteric infections in mice and rats which when disseminated have been associated with myocarditis, type 1 diabetes, encephalitis, and multiple sclerosis-like symptoms. Cardioviruses have also been detected at lower frequencies in other mammals. The Cardiovirus genus within the Picornaviridae family is currently made up of two viral species, Theilovirus and Encephalomyocarditis virus. Until recently, only a single strain of cardioviruses (Vilyuisk virus within the Theilovirus species) associated with a geographically restricted and prevalent encephalitis-like condition had been reported to occur in humans. A second theilovirus-related cardiovirus (Saffold virus [SAFV]) was reported in 2007 and subsequently found in respiratory secretions from children with respiratory problems and in stools of both healthy and diarrheic children. Using viral metagenomics, we identified RNA fragments related to SAFV in the stools of Pakistani and Afghani children with nonpolio acute flaccid paralysis (AFP). We sequenced three near-full-length genomes, showing the presence of divergent strains of SAFV and preliminary evidence of a distant recombination event between the ancestors of the Theiler-like viruses of rats and those of human SAFV. Further VP1 sequencing showed the presence of five new SAFV genotypes, doubling the reported genetic diversity of human and animal theiloviruses combined. Both AFP patients and healthy children in Pakistan were found to be excreting SAFV at high frequencies of 9 and 12%, respectively. Further studies are needed to examine the roles of these highly common and diverse SAFV genotypes in nonpolio AFP and other human diseases.

FIG. 1.

Phylogenetic analysis of cardioviruses based on full-length genomes (A) and on amino acid sequences of P1 (B), P2 (C), and P3 (D) regions of polyproteins. Bootstrap values from 1,000 replicates are shown for the nodes.

FIG. 2.

Recombination analyses. (A) Sliding-window SimPlot graph generated by using the sequence of the TLV of rats as a query against all other cardiovirus and EMCV sequences. (B) Bootscan analysis of the sequence of the TLV of rats in comparison to those of SAFV, TMEV-GDVII, and EMCV. (C) Sliding-window SimPlot graph generated by using the SAFV-Pak5152 sequence as a query against all other available SAFV sequences. (D) Bootscan analysis of the sequence of SAFV-Pak5152 in comparison to other available SAFV sequences. The virus closest to TLV of rats is rat theilovirus 1. The nucleotide position is given on the x axis, and the percent nucleotide sequence similarity is given on the y axis. The positions relative to viral genes are shown by genome diagrams above the plots and scans.

FIG. 3.

Pairwise amino acid identity levels (in percentages) of available human cardiovirus VP1 sequences.

FIG. 4.

Phylogenetic analysis of the VP1 proteins of human and animal theiloviruses and EMCVs.

FIG. 5.

Alignments of SAFV proteins with homologues from other theiloviruses and EMCVs. (A) Alignments of the L proteins showing different domains with conserved amino acid motifs (highlighted). The Y and T residues in the acidic domain thought to be phosphorylated in EMCVs are labeled with a star. (B) Alignments of potential L* proteins encoded by SAFV with those of other theiloviruses. SAFV proteins are depicted as being initiated by an ACG (Thr) codon and terminated by premature stop codons. The carboxy-terminus 72 aa of the other theilovirus L* proteins are not shown. (C) Alignments of the EF loops (in VP2 proteins) of theiloviruses and EMCVs. (D) Alignments of the CD loops (in VP1 proteins) of theiloviruses and EMCVs. Columns highlighted in black show absolute amino acid conservation, while those highlighted in gray show amino acids with highly similar properties. Terminal residue positions of the aligned L protein and EF and CD loop structures according to those in a reference strain (TMEV-DA) are used.