Extensive horizontal gene transfer in cheese-associated bacteria
- University of California, San Diego, United States
- Tufts University, United States
- Jacobs School of Engineering, University of California, San Diego, United States
Figures
Figure 1 with 2 supplements
Extensive horizontal gene transfer in the cheese microbiome.
(A) All HGT events in analyzed cheese-associated bacteria. Connection thickness is scaled to number of shared protein coding sequences. Maximum likelihood tree based on 16S RNA alignment using …
Figure 1—figure supplement 1
Schematic of software pipeline to identify HGT.
(1) Sequenced genomes are annotated with IMG/ER and downloaded in Genbank format. (2) All annotated genes in all genomes are used to assemble a BLAST database using BLAST+ command-line tools. (3) …
Figure 1—figure supplement 2
Same as Figure 1A with branch labels.
All HGT events in analyzed cheese-associated bacteria. Connection thickness is scaled to # of shared protein coding sequences. Phylogenetic tree based on 16S RNA alignment using Ribosomal Database …
Figure 2 with 1 supplement
HGT Groups in Actinobacteria, Firmicutes, and γ-Proteobacteria groups.
(A) The 31 largest HGT groups that contain predominantly Actinobacteria. The areas of circles are scaled to log2(n), where n is the total number of nucleotides in that group for each species. The …
Figure 2—figure supplement 1
Group A: Expected clustering: contiguous genes in multiple species are in a single group.
Although island 6 (i6) lacks one gene present in i1 and i4, (possibly because of a transposon insertion), it is still considered related. Group B: Ambiguous grouping: islands 2 and 3 from species 1 …
Figure 3
Structure of RUSTI islands.
(A) At-scale schematics for genomic context of HGT Group 1 (top) for B. linens JB5 and G. arilaitensis JB182 and alignment of RUSTI operon (bottom). Nucleotide position values (top) refer to contigs …
Figure 4
Presence of RUSTI in cheese metagenomes.
Genes in ActinoRUSTI (G. arilaitensis JB182) and ProteoRUSTI (V. casei JB196) regions were compared with 32 assembled metagenomes from the US and Europe. Filled CDS represents a positive (>97% …
Figure 5
Mobility of RUSTI.
(A) Schematic for PCR primer design - see Materials and methods for details. (B) PCR testing for the presence of RUSTI and for the excision of the ICE in an overnight culture of G. arilaitensis …
Additional files
-
Supplementary file 1
Genome information.
(a) Genome statistics for newly sequenced genomes, determined by IMG/ER. Gene IDs refer to IMG bioproject or RefSeq Accession. Genomes from Almeida et. al. do not yet have accession numbers. (b) Pairwise species comparison summary total. Protein coding sequences (column ‘Shared CDS’) and nucleotides (in base-pairs - column ‘Shared nt’) determined to be horizontally transferred for every pair of species that were compared by the HGT detection pipeline. Also shows calculated ANI and 16S similarity in % (column ssu - see Materials and methods for method of determining 16S similarity). Species pairs that have ANI > 0.89 were not compared and are not shown. (c) HGT identification parameters. Different parameters for minimum length of gene match for HGT, maximum % ANI identity for related species, and maximum distance between genes in an island were compared. Number of positive HGT hits identified when varying the minimum protein coding gene length. Number of HGT groups constructed when varying the maximum separation between hits that are classified as belonging to the same group. Number of nucleotides or number of protein coding sequences in HGT regions by 16S similarity. Note - There are no results below 500 as 500 bp is the minimum length for protein coding sequences in this analysis. (d) Full group annotations. All protein coding sequences identified as HGT, sorted by group # (ranked by total nucleotide content), species and genome location within species. Certain functional annotations are identified by color (e.g. orange for iron) based on text in annotation. Locus tags and contig IDs beginning with lower case letters were assigned by kvasir, and do not correspond to any published database. (e) Group summary statistics. Summary statistics for each HGT group. (f) Highly conserved genes in Brevibacterium species. Protein coding sequences from Group 29, as well as selected highly conserved genes from Brevibacterium antiquum CNRZ918 were compared with other Brevibacterium strains by BLAST. B. linens 947.7 has substantially lower nucleotide identity for the four genes found in Group 29 than other B. linens strains, despite similar nt distance for other highly conserved genes. This suggests that Group 29 is a true example of HGT between CNRZ918 and other B. linens strains, rather than a false positive. (g) RUSTI gene expression during competition. Gene expression data from RNA seq analysis for genes in JB182 RUSTI. Related to Figure 3B (h) TCBD hits for transporters in RUSTI. Representative CDS of Actino- and ProteoRUSTI from G. arilaitensis JB182 and V. casei JB196, respectively, were compared with the Transporter Classification Database (TCDB). (i) RefSeq BLAST Actino- and ProteoRUSTI from G. arilaitensis JB182 and V. casei JB196, respectively, as well as the consensus sequence for StaphRUSTI (see Figures 3 and 4) were compared with the NCBI RefSeq database using BLAST.
- https://doi.org/10.7554/eLife.22144.011
