Complex virome in feces from Amerindian children in isolated Amazonian villages

Abstract

The number of viruses circulating in small isolated human populations may be reduced by viral extinctions and rare introductions. Here we used viral metagenomics to characterize the eukaryotic virome in feces from healthy children from a large urban center and from three Amerindian villages with minimal outside contact. Numerous human enteric viruses, mainly from the Picornaviridae and Caliciviridae families, were sequenced from each of the sites. Multiple children from the same villages shed closely related viruses reflecting frequent transmission clusters. Feces of isolated villagers also contained multiple viral genomes of unknown cellular origin from the Picornavirales order and CRESS-DNA group and higher levels of nematode and protozoan DNA. Despite cultural and geographic isolation, the diversity of enteric human viruses was therefore not reduced in these Amazonian villages. Frequent viral introductions and/or increased susceptibility to enteric infections may account for the complex fecal virome of Amerindian children in isolated villages.

Fig. 1

Virus distribution by children in four sites. Each circle represents the log 10 transformed number of virusgenome/contig matching viral reads per million total reads

Fig. 2

Phylogenetic analysis of P1 amino acid sequences from members of the family Picornaviridae. The analysis was carried with maximum likelihood method and 100 bootstrap replicates. The viral sequences assembled in this study for which the P1 region was available are in bold and colored according to the location (Urban center A in yellow; village B in blue; village C in magenta; and village D in green). Bootstrap values higher than 0.7 (70%) are shown. Site of sample collection is also reflected in the taxa name (containing A, B, C, or D) for each specific location. GenBank accession numbers for study sequences are listed in Supplementary Data 1

Fig. 3

Phylogenetic analysis of P1 amino acid sequences from members of the family Caliciviridae. The analysis was carried with maximum likelihood method and 1000 bootstrap replicates. The viral sequences assembled in this study for which the P1 region was available are in bold and colored according to the location (village B in blue; village C in magenta; and village D in green). Bootstrap values higher than 0.7 (70%) are shown. Site of sample collection is also reflected in the taxa name (containing A, B, C, or D) for each specific location. GenBank accession numbers for study sequences are listed in Supplementary Data 1

Fig. 4

Phylogenetic analysis of 459 amino acid region of RdRp from unassigned members of the Picornavirales order. The analysis was carried with maximum likelihood method and 1000 bootstrap replicates. The viruses sequences assembled in this study for which the RdRp region was available are in bold and colored according to the location (village B in blue; village C in magenta; and village D in green). Only bootstrap values higher than 0.7 (70%) are shown. Site of sample collection is also reflected in the taxa name (containing A, B, C, or D) for each specific location. GenBank accession numbers for study sequences are listed in Supplementary Data 1