Estimating molecular preservation of the intestinal microbiome via metagenomic analyses of latrine sediments from two medieval cities

Abstract

Ancient latrine sediments, which contain the concentrated collective biological waste of past whole human communities, have the potential to be excellent proxies for human gastrointestinal health on the population level. A rich body of literature explores their use to detect the presence of gut-associated eukaryotic parasites through microscopy, immunoassays and genetics. Despite this interest, a lack of studies have explored the whole genetic content of ancient latrine sediments through consideration not only of gut-associated parasites, but also of core community gut microbiome signals that remain from the group that used the latrine. Here, we present a metagenomic analysis of bulk sediment from medieval latrines in Riga (Latvia) and Jerusalem. Our analyses reveal survival of microbial DNA representative of intestinal flora as well as numerous parasites. These data are compared against parasite taxon identifications obtained via microscopy and ELISA techniques. Together, these findings provide a first glimpse into the rich prokaryotic and eukaryotic intestinal flora of pre-industrial agricultural populations, which may give a better context for interpreting the health of modern microbiomes. This article is part of the theme issue 'Insights into health and disease from ancient biomolecules'.

Keywords: aDNA; archaeology; metagenomics; microbiome; parasitology.

Figure 1.

Map indicating locations of the two archaeological sites considered in this analysis. Precise locations of the latrines are shown with bordered red rectangles.

Figure 2.

α-Diversity across all sequenced libraries. (a) Species richness (count of total number of species) from the metagenomic profile of each library. The bar chart is stacked by taxonomic group. Values can be found in electronic supplementary material, table S1. (b) Simpson's diversity indices across all libraries for each site, partitioned by taxonomic group. Simpson's index considers evenness in its formula. Indices closer to 0 have lower diversity, and indices closer to 1 have higher diversity (see formula in Methods). Values can be found in electronic supplementary material, table S2.

Figure 3.

Bacterial and archaeal source analyses. (a) SourceTracker2 [38] was used to estimate contributions from five model sources: hunter–gatherer human gut [39,40], industrial human gut [39], sewage/wastewater [41] and soil/sediment [–44]. Contribution was estimated using combined libraries for each site with reads mapping to the HG19 human reference genome removed. Human reads were also removed from the model source metagenomes. Estimations are based on reads summarized to the genus level within the Bacteria and Archaea nodes. (b) Donut plots depicting the proportion of reads summarized to bacterial and archaeal species-level nodes with at least 100 reads, grouped based on typical isolate source according to a PubMed literature survey using taxon name as the search term.

Figure 4.

Eukaryotic parasite taxa with read alignments from latrine samples. (a) The taxon names to the right of the heatmap correspond to eukaryotic parasite species. All taxon nodes listed here have at least one assigned read aligning to associated sequences from the MALT results of the Jerusalem and Riga combined libraries (with reads aligning to the HG19 human reference genome removed). All taxa in this figure had assigned reads in at least one of the combined libraries, and the assigned read count (not normalized) can be found in the respective cell in the heatmap. The colour of each cell indicates the HOPS criteria passed by the alignments to that taxon. ‘No pass’—no criteria were passed, level 1—edit distance criterion was passed, level 2—damage criterion was passed, level 3—edit distance of damaged reads criterion was passed. The red and blue microscope symbols indicate the taxon was identified by microscopy in samples from Jerusalem and Riga, respectively. The red and blue antibody symbols indicate the taxon was previously identified by ELISA assay. (b) Radial trees representing the relationships between taxa identified in the latrine libraries, spanning multiple levels of taxonomic classification. Tabular read count results for the latrine samples can be found in electronic supplementary material, table S12, and results for negative controls can be found in electronic supplementary material, table S13.

Figure 5.

Cytosine-to-thymine mismatch plots calculated from direct alignments to H. sapiens, A. lumbricoides and D. latus genomes. Sequencing reads from the combined libraries were aligned using BWA (Burrows-Wheeler Aligner) and mismatch frequencies for each alignment were calculated by DamageProfiler as implemented in EAGER [57]. As the sequencing reads were extracted from latrine sediment, all alignments likely represent more than one individual of each species.