Fig 1.
GA2 segregate into five different subtypes.
A. Schematic of the GA2 locus of B. caccae CL03T12C61 as example. Segments highlighted in gray represent the conserved regions made into a concatemer used for the analyses shown in panels B and C. B. Average global DNA percent identity between the conserved concatemers of different GA2 subtypes. C. Maximum likelihood phylogenetic tree of the GA2 regions showing the separation into 5 subtypes. Only some strains are shown, with the numbers at the end of each branch indicating the total number of isolate genomes, out of the 1434 Bacteroidales genomes queried that contain that subtype of GA2. Some isolates from non-human sources are highlighted in colored font.
Fig 2.
Abundance of GA1 and GA2 loci in different Bacteroidales species.
Shown is the fraction of strains within each species harboring each type of GA. *Three or less sequenced isolates. Strains with no species call were not included. Shaded boxes highlight the species with high GA2 prevalence.
Fig 3.
Screen for intra-community GA2 spread in the Comstock lab strain collection.
A. Schematic of a GA2 T6SS locus. Small arrows indicate screening primer annealing sites and bars indicate the corresponding PCR product. Color scheme of genes is as described in Fig 1. B. PCR screen of the isolates from the three communities indicated, using primers for the GA2 conserved region. Strains bolded in green were selected for whole genome sequencing. C. Strains sequenced in this study and whether they contain GA1 or GA2 T6SS loci. x and y in GA1 column indicate the regions are distinct.
Fig 4.
Intra-community spread is common for GA1 and GA2 loci.
A. Number of communities per dataset, classified by GA1 and GA2 loci presence/absence. Numbers in parentheses under each dataset name indicate total number of volunteers analyzed in that collection. *Includes CL08 and CL14 which were screened by PCR as described in the text. B. Percentage of communities in the CGR and BIOML collections that show the indicated event, out of the total communities in that dataset that have at least one strain with the indicated GA type: no transfer, single-species spread (present in two species), multi-species spread (present in three species or more). C. Communities with a single-species intra-community transfer (highlighted in light grey) or multi-species spread (highlighted in dark grey). Letters in parentheses indicate type of GA2. Numbers listed separately in an individual box indicate the T6SS regions are different (<99.99% DNA sequence identity and/or divergent variable regions). **Screened by PCR and sequencing of variable region, as described in the text. Species determined by 16S rRNA gene sequence. ‡The two relevant strains from donor AF15 and AF16 are nearly clonal and appear to have both a GA2b and GA2c (see details in text).
Fig 5.
Sequence specificity of the GA1 and GA2 integrases in Bacteroides and Parabacteroides genomes.
A. GA1 recombinase and integration site of the B. fragilis YCH46 ICE (ICE spans BF2788 to BF2921). The insertion of the GA1 ICE in the B. fragilis YCH46 genome relative to the 638R genome is shown. B. The recombinase and integration site of the GA2(a,b,c) showing the GA2b region from B. uniformis BIOML-A2 as an example (GAP40_16280 to GAP40_16650 and GAP40_07160 to GAP40_07320). This ICE integrates into the tRNAPhe gene. C, D. The recombinases and integration sites of two different GA2d ICE. Panel C shows the GA2d region from B. uniformis BIOML-A77 (GAS23_04785 to GAS23_05290) with the IntGA2d integrase. Shown below are the proposed recognition sequences deduced from BIOML-A77 and BIOML-A84 comparison, the sequence reported for B. plebeius DSM 17135 (a PUL ICE), and the proposed consensus sequence recognized by IntGA2d. This integrase recognizes the sequence shown in tRNALys genes. Panel D shows the GA2d ICE from B. uniformis AF39-16AC (DW039_02125 to DW039_02640) which harbors the IntGA2ch recombinase which integrates at a tRNAPro gene sequence.
Fig 6.
Schematic of genomes that have a GA1 and a GA2, where one of the regions is disrupted.
Corresponding genomic coordinates for each region shown in S6 Table are included. Regions of the GA1 ICE outside of the T6SS locus are shown as hatched green boxes and those outside the GA2 T6SS locus are indicated by hatched blue boxes. All GA2 regions shown are of the GA2b subtype except where indicated. Grey hatched boxes indicate gaps in contigs or contigs that were joined manually. Color legend for the T6SS genes as specified in Fig 1. Strains with bolded names have disrupted T6SS regions, whereas non-bolded correspond to the same ICE carrying a different cargo. Insertions are indicated by filled triangles, deletions by arrows. Boxed regions indicate the bounds of a transposable element insertion or recombination event that replaced the cargo.
Fig 7.
Distribution of the T6SS GA loci and subtypes in gut metagenomes from different human populations.
The percentage of metagenomes with each GA1 and GA2 subtype of the total metagenomes for each population is shown by bar graph. The size of each population sampled is designated by the size of the circle. No GA loci were detected in the Peru metagenomic dataset. The exact numbers for each metagenomic dataset are reported in S7 Table as well as the health status of each individual in the various cohorts and the abundance of Bacteroides spp. and Prevotella spp in the metagenomic sample. Map data copyrighted OpenStreetMap contributors and available from https://www.openstreetmap.org.