In an environment of high bacterial density, like the gut microbiome, horizontal gene transfer (HGT) between bacteria takes place. This exchange of genes is one of the mechanisms of evolution in bacteria, and is often how bacteria acquire antimicrobial resistance genes. One of these mechanisms of HGT is via plasmids, which play a role in evolution with consequences for health and disease. However, studying plasmids usually entails a time and labor intense process of growing bacterial cultures followed by the process of extracting plasmid DNA to sequencing. In addition, the method is limited to analyzing plasmids that are retained under laboratory conditions by culturable host bacteria. In this study, we explore the possibility of analyzing the evolution of plasmids in the human infant gut microbiome in a high-throughput manner using shotgun metagenomics data and the software tool metaplasmidSPAdes.
Daily samples from 12 infants from birth to one year of age were previously characterized by amplicon sequencing of the 16S rRNA V4 region. Here, monthly samples from the same sample set are sequenced with shotgun metagenomics and potential plasmid sequences are assembled using metaplasmidSPAdes. Filtering of the resulting scaffolds is based on the presence of a replication protein or a plasmid origin of replication, using the Pfam and DoriC databases respectively. To create bins in which all scaffolds are considered to originate from the same plasmid, scaffolds are pairwise aligned with the NUCmer tool of MUMmer. Pairwise total coverage, with at least 99.5% identity, of each scaffold by all other scaffolds is converted into a graph, with the nodes representing the individual scaffolds and edges a coverage of >90% of the scaffold length. Connected subgraphs form plasmid bins, represented by the scaffold that appears closest to the day of birth. Roughly, a third of the 244 bins are contained within the plasmid database PLSDB with a 99% identity using Mash screen. Given that most plasmids are still unknown, this is a good indication that most, if not all, bins contain plasmids or fragments of plasmids.
Scaffolds from two of the bins are found in five of the infants and scaffolds from about forty bins are found in two or three of the infants with the rest of the scaffolds in approximately 200 bins found in only one infant. Many of the bins contain scaffolds from several time points within the same individual, making it possible to follow the plasmid over time. Next, the binned scaffolds will be analyzed based on SNV-level similarity within and between infants, giving an indication of the dynamics of the most prominent plasmids during colonization of the infant gut. Further plans include samples with more time resolution and samples from the mothers around birth, which give the opportunity to compare plasmids and other mobile genetic elements between mother and infant pairs. Together, these results will give a first insight in the importance of plasmids in the establishment and adaptation of the gut microbiome of human infants.