NEW YORK (GenomeWeb) – By analyzing more than 9,000 metagenomes from human microbiome samples, an international team of researchers has uncovered more than 150,000 microbial genomes, many of which represent species that have never before been described.
Much of the diversity of the human microbiome remains unexplored, particularly among body sites other than the gut and among non-Westernized populations. Researchers led by Nicola Segata at the University of Trento in Italy reconstructed microbial genomes from more than 9,000 metagenomes from various human populations, geographies, lifestyles, and ages. As they reported in Cell today, they uncovered nearly 5,000 species-level genome bins, 77 percent of which were not present in public databases.
"We genetically characterized and catalogued a large number of bacteria and archaea that are part of the human microbiome, but remained so far unexplored, uncharacterized, and undescribed," Segata said in a statement. "We also observed that many of these microbes tend to be only rarely identified in Westernized populations, most probably as a indirect consequence of the complex industrialization processes."
The researchers used a large-scale metagenomic assembly approach to reconstruct bacterial and archaeal genomes from 9,316 metagenomes, generated by short-read sequencing. The metagenomes came from 46 datasets and included samples from different populations, body sites, and host ages. From these, the researchers reconstructed 154,723 genomes.
This, they noted, more than doubles the about 150,000 microbial genomes that were previously publicly available.
Using an all-versus-all genetic distance quantification and clustering approach, they organized these 154,723 genomes into about 5,000 species-level genome bins (SGBs). About 4,930 SGBs were from 22 known phyla and 345 could not be assigned a taxonomic family.
Still, 77 percent of the SGBs represented species without any publicly available genome, the researchers noted. These unknown SGBs include, on average, nine reconstructed genomes. Most of these unknown SGBs represent rare human-associated microbes — about half of the unknown SGBs contained one reconstructed genome — though the researchers noted that some are more prevalent.
Having these unknown SGBs increased mappability of metagenomic reads, the researchers noted. The read mappability for stool samples increased by 29 percent to 87.5 percent, and for the oral cavity by 26 percent to 82 percent. In particular, they increased the mappabilty of the gut microbiomes from non-Westernized populations, which now reached 83 percent.
The researchers dubbed one of the most prevalent unknown SGBs "Candidatus Cibiobacter qucibialis." Through a phylogenetic analysis, they placed this candidate species between Faecalibacterium and Ruminococcus, both key members of the gut microbiome. Within this phylogenetic analysis, Ca. Cibiobacter qucibialis genomes from non-Westernized populations clustered together in a monophyletic subtree.
Functional profiling of Ca. Cibiobacter qucibialis found differences between the Westernized and non-Westernized strains. The non-Westernized strains, for instance, included the full tryptophan metabolism operon, while the Westernized strains did not. This, they noted, could be a sign of divergent evolution.
Even among well-studied microbiome bacteria, like Bacteroides, the researchers uncovered additional intra-species diversity.
But the addition of gut microbiomes from rural, non-Westernized populations in Madagascar uncovered a number of unknown SGBs, including in the Firmicutes and Actinobacteria phyla, as well as in other phyla that are not always associated with humans.
Functional annotation of these SGBs also uncovered some differences in what modules were enriched, though the researchers noted that the same functions were often encoded in both non-Westernized and Westernized microbiomes, but sometimes by different enzymes and pathways.
This, they said, suggests there are numerous ways the gut microbiome adapts to the diversity of its human hosts.