Metagenomic analysis of the South African human gut microbiome

Abstract

Understanding the ecology, evolution, structure and abiotic drivers of gut microbiota remains a major ecological endeavour. Previous studies have shown that several factors including diet, lifestyle and geography substantially control and modulate the human gut microbiome. The bulk of these studies have focused on phylogenetic analysis of bacterial communities with comparatively less centred on interpreting and understanding the role of other taxonomic groups including viruses, fungi, and functional dynamics of gut communities. This knowledge deficit is especially true for rural and urban African populations. Given the large degree of genetic variability and differences in diet among Africans, focused studies on these populations may reveal important insights regarding compositional trends and help to quantify the precise contributions of gut microbiota. To reduce this knowledge deficit, we established the first broad scale gut microbiome analysis of South African individuals. We assessed the structure and drivers of rural and urban gut mycobiota. We further assessed the influence that potential structural variations may impose on functional capacity in gut microbiomes, by undertaking an analysis of the resistomes of South African individuals. The participants in this study (n = 100) were balanced by geography and sex. The diversity of mycobiota in these geographically separated populations was characterized using amplicon sequencing of the Internal Transcribed Spacer (ITS) gene. We further assessed biomarker species specific to rural and urban cohorts. In addition to phyla which have previously been shown to constitute ubiquitous constituents of gut microbiota, Pichia were key constituents of the mycobiota. Our analysis revealed that geographic location was a major driver of gut mycobiota. Other factors such as smoking were also key determinants of gut mycobiota albeit to a lower extent, as explained by the small proportion of total variation. Linear discriminant and the linear discriminant analysis effect size analysis revealed several specific and distinct biomarkers in urban and rural samples, respectively. To elucidate the functional dynamics of gut microbiomes in African populations, a subset of samples (n =12) balanced by gender and geography were selected for shotgun metagenomic analysis. In addition, taxonomic and phylogenetic annotation, antimicrobial genes were also characterized. Taxonomic annotation and metagenome assembled binning revealed that several bacteria were dominant in the guts of South African individuals from a variety of phyla including Bacteroidetes, Firmicutes and Actinobacteria. However, no significant difference in relative abundance at phylum level were detected between the rural and urban metagenomes, suggesting that differences in diet and location may not result in clear differences in the gut microbiomes of these individuals. Interestingly, our analysis showed several antimicrobial resistance (AMR) genes associated with the resistance to tetracycline, macrolides, lincosamide, aminoglycosides, and vancomycin. These AMR profiles were geography-specific and correlated with the resistance profiles of some multi-AMR microbial species. Abiotic factors including age and gender were also shown to have a significant influence on AMR gene diversity in the gut microbiomes of South Africans. Taken together, our studies substantially broaden insights regarding the gut microbiomes of South African individuals. We reveal distinct fungal community structure and resistome in urban and rural South African individuals. These analyses demonstrate that the fungal mycobiota appears to harbour distinct lineages in contrast to those from individuals from the USA and other well studied locations. We show that geography is a key driver of rural and urban gut mycobiota and their functional attributes including AMR profiles. The insights generated from this study provide an important basis for understanding the phylogenetic patterns of rural and urban individuals. Moreover, the functional cues revealed through characterization of AMR profiles provide interesting fundamental questions for further studies.