Table 1.
Methylation motifs in B. subvibrioides.
Fig 1.
Genes misregulated in ccrM mutant compared to wild-type in B. subvibrioides.
A) Workflow showing the cutoffs used for defining misregulated genes in ccrM mutant. Using >2-fold and P<0.01 as cut offs, 129 genes were found misregulated compared to WT out of 3393 total genes in B. subvibrioides. Out of 129 misregulated genes, 56 of them had at least one GANTC site in their promoter suggesting potential direct regulation by CcrM. B) List showing genes downregulated (left) and upregulated (right) in the ccrM::pNPTS139 strain along with COG functional category. For both left and right, Column 1 shows the heat map of the magnitude of fold change in log2 scale. Column 2 shows if those genes have GANTC site within their promoter (grey—GANTC site present, white—GANTC site absent). Genes were clustered by COG functional category (Column 3).
Fig 2.
Common genes misregulated in ccrM mutant in B. subvibrioides and C. crescentus.
A) Workflow showing the cutoffs used for defining misregulated genes in ccrM mutant in both organisms. Using >2-fold and P<0.01 as cut offs, 129 genes and 152 genes were found misregulated in ccrM in B. subvibrioides and C. crescentus respectively. Only 4 genes were found in common. B) Concentric circle diagram showing common genes using different parameters. C. crescentus CcrM regulates 152 genes and B. subvibrioides has orthologs to 89 of those genes. Only 29 of the B. subvibrioides orthologs have GANTC sites in their promoter regions. Of those 29, only 2 genes showed significant transcriptional changes in a ccrM mutant strain (highlighted in orange). C) List of 89 B. subvibrioides orthologs to C. crescentus ccrM regulated genes sorted by COG functional category. Column 1 and 2 shows the heat map of the magnitude of fold change in log2 scale in the C. crescentus ccrM strain (data obtained from [23]) and B. subvibrioides ccrM::pNPTS139 strain respectively. Columns 3 and 4 show if those genes have GANTC site within their promoter in C. crescentus and B. subvibrioides respectively (grey—GANTC site present, white—GANTC site absent). Genes were clustered by COG functional category (Column 5). Orthologs that met the transcriptional change cutoffs are highlighted in green; orthologs that met the transcriptional change cutoffs and have a GANTC site are highlighted in orange.
Fig 3.
Genes directly regulated by GcrA in B. subvibrioides.
A) Workflow showing the cutoffs used for defining the GcrA regulon. Using transcriptional change cutoffs of >2-fold and P<0.01, 131 genes were characterized as misregulated in the gcrA mutant. Of those 131 genes, GcrA peaks (obtained from ChIP-seq) were detected in the promoter regions of 78 genes. B) MEME analysis of promoters activated by GcrA that had only one GANTC site within the promoter region in B. subvibrioides. In total, 18 genes were activated by GcrA with only one GANTC site within the promoter region in B. subvibrioides. MEME analysis showed no preference for any bases in an extended GANTC motif beyond a slight preference for C before GANTC site. C) Venn diagram showing common genes of CcrM/GcrA regulons in B. subvibrioides. There were 56 genes directly regulated by CcrM (>2-fold, P<0.01 and GANTC site in promoter), and 45 genes directly regulated by GcrA with methylation sites (>2-fold, P<0.01 and GcrA peak with GANTC site in promoter). Only 19 genes were found in common between them (highlighted in orange in Fig 3D). D) List showing all misregulated genes with GcrA peaks (with or without GANTC site) that were downregulated (left) and upregulated (right) in the ΔgcrA strain sorted by COG functional category. For both left and right, Column 1 shows the heat map of the magnitude of fold change in log2 scale. Column 2 shows if GANTC site is also present in the GcrA peak (grey—GANTC site present, white—GANTC site absent. Genes were clustered by COG functional category (Column 3).
Fig 4.
Common genes regulated by GcrA in B. subvibrioides and C. crescentus.
A) Workflow showing the cutoffs used for defining common genes belonging to GcrA regulon in B. subvibrioides and C. crescentus. Out of 204 GcrA regulated genes in C. crescentus, 147 orthologs found in B. subvibrioides. Only few genes were in common despite using different cutoffs to define GcrA regulon in B. subvibrioides. B) Concentric circle diagram showing common genes using different cutoffs. B. subvibrioides has 147 orthologs to the 204 GcrA targets in C. crescentus. Of those 147 genes, only 48 had detectable GcrA peaks (obtained from ChIP-seq data) with GANTC sites. Only 24 of those 48 genes had transcriptional changes meeting a P<0.01 cutoff in the gcrA mutant (all highlighted genes in Fig 4C), and only 12 of those met the >1.75-fold change transcriptional cutoff (highlighted blue in Fig 4C). C) List showing all 147 B. subvibrioides genes orthologous to the 204 members of the published C. crescentus GcrA regulon, sorted by COG functional category. For both left and right, Column 1 and 2 is the heat map showing the magnitude of fold change in log2 scale in C. crescentus gcrA strain (data obtained from [19]) and B. subvibrioides gcrA strain respectively. Column 3 shows genes that met P<0.01 criteria or not in B. subvibrioides (grey—P<0.01 is met, white—P<0.01 is not met). Column 4 shows if those genes have GcrA peaks with GANTC sites within their promoter in B. subvibrioides (grey—GcrA peak with GANTC site present, white—GcrA peak with GANTC site absent). Genes were clustered by COG functional category (Column 5). Orthologs with GcrA peaks containing a GANTC site are shown in the left and orthologs without GcrA peaks containing a GANTC site are shown in the right. Orthologs with GcrA peak and P<0.01 are highlighted (blue and gray). Orthologs with GcrA peak, P<0.01 and >1.75-fold change are shown in highlighted blue.
Fig 5.
Heatmap showing top 50 ranked highly conserved ortholog groups for Caulobacteraceae and Hyphomonadaceae families.
Each column designates a species, following the order dictated by a phylogenetic tree inferred from a multiple sequence alignment of GcrA homologs. Rows correspond to identified orthologous groups. For each cell, the cyan-yellow scale coloring indicates the posterior probability of regulation of the ortholog in that species (cyan blue—1, yellow—0). White cells indicate absence of the ortholog in that particular species. The left ancillary columns indicate, using the same color scale, the number of orthologs in each ortholog group (lowest value 20 out of 23) and the inter-species average of best extended GANTC instance score , which has been used to rank the ortholog groups. Both values are shown normalized to the (0,1) range. The right ancillary columns indicate the two primary functional categories for the COGs assigned to each ortholog group, the description and identifier of which is shown adjacent. Highlighted descriptions denote ortholog groups also present in Fig 6.
Fig 6.
Heatmap showing top 50 ranked highly-conserved ortholog groups for representatives of each Alphaproteobacteria order.
Each column designates a species, following the order dictated by a reference cladogram adapted from [44], placing Emcibacter congregatus and Rhizomicrobium palustre following their reported phylogeny in [45] and [46] respectively. Rows correspond to identified orthologous groups. For each cell, the cyan-yellow scale coloring indicates the posterior probability of regulation of the ortholog in that species (cyan blue—1, yellow—0). White cells indicate absence of the ortholog in that particular species. The left ancillary columns indicate, using the same color scale, the number of orthologs in each ortholog group (lowest value 11 out of 13) and the inter-species average of best extended GANTC instance score , which has been used to rank the ortholog groups. Both values are shown normalized to the (0,1) range. The right ancillary columns indicate the two primary functional categories for the COGs assigned to each ortholog group, the description and identifier of which is shown adjacent. Highlighted descriptions denote ortholog groups also present in Fig 5.