Use of proximity ligation shotgun metagenomics to investigate the dynamics of plasmids and bacteriophages in the gut microbiome following fecal microbiota transplantation

Abstract

Proximity ligation shotgun metagenomics facilitate the analysis of the relationships between mobile genetic elements, such as plasmids and bacteriophages, and their specific bacterial hosts. We applied this technique to investigate the changes in the fecal microbiome of patients receiving fecal microbiota transplantation (FMT) for recurrent Clostridioides difficile infections (rCDI). FMT was associated with successful engraftment of donor bacteria along with their associated bacteriophages. While fecal microbial diversity increased in all patients, the extent of specific bacterial taxa engraftment varied among individual patients. Interestingly, some donor bacteriophages remained closely linked to their original bacterial hosts, while others expanded their associations across different bacterial taxa. Notably, FMT partially reduced the content of vancomycin resistance and extended-spectrum beta-lactamase genes in the fecal microbiome of rCDI patients.

Keywords: Clostridioides difficile; Proximity ligation shotgun metagenomics; antimicrobial resistance genes; fecal microbiota transplantation.

Figure 1.

Proximity guided metagenomics links mobile elements to hosts. The graphic shows the major steps involved in the biochemical methodology. (A) The sample contains contact microbial cells along with mobile elements (plasmids and active phages) in close proximity to the host chromosome. (B) First, proximity ligation crosslinks neighboring dsDNA within intact cells using formaldehyde. Next (arrow), cells are lysed and dsDNA is digested with restriction enzymes. After that (arrow), the crosslinked molecules are end-joined and paired-ended libraries are created. (C) Alignment of paired end reads links assembled contigs of mobile elements and host microorganisms.

Figure 2.

Diversity of bacteria and phages across the FMT cohort. Alpha-diversity of (A) Bacteria and (B) phages in pre-FMT (time point 0) and post-FMT (time points 1–4 weeks) samples compared to the donor (dotted blue line). Beta-diversity comparisons for bacteria (C) and phages (D) between patient samples at each time point and the corresponding pre-FMT (purple) and donor (green) samples.

Figure 3.

Engraftment of bacteria and phages. Beta-diversity comparisons between post-FMT samples for (A) bacteria and (B) phages. (C) Taxonomic distribution of engrafted bOtus and vOtus. (D) Log2 ratio of relative abundances, post-FMT vs. donor, observed for engrafted bOtus and vOtus.

Figure 4.

Graph representation of donor phage engraftment. The number of interactions detected between donor phage (vOtus, hexagons) and bacteria (mOtus,circles) is indicated by the edge weight. The order level assignment of bacteria is indicated by the color key. The number of patients vOtus and bOtus detected is indicated by the node size. Specific interactions discussed in the text are indicated by a, b, c, and d.

Figure 5.

Abundance and distribution of clinically relevant AMR genes. The heatmap (A) depicts the sum of abundances for bacteria harboring individual AMR genes for each patient where a pre-FMT plus 2-week and 4-week post-FMT samples were available, corresponding to the three rows depicted for each patent indicated on the left axis. Color intensity indicates the sum of bacteria TPM values in each sample (log10 transformed). “*” indicates three clinical failures. AMR genes are listed individually, or where appropriate combined such as with erm and tet genes. A full list of genes and aggregations is available in Supplementary Table S3. (B) Genomic sources for each observed AMR gene. (C) Taxonomy of bacteria hosts for each AMR gene.