Fig 1.
Changes of epigenetic histone modifications in Y. enterocolitica infected primary human macrophages.
A, Experimental design. CD14+ monocytes were isolated and differentiated into macrophages by cultivation with 20% human serum for 6 days. Macrophages from ≥ two independent donors were mock infected or infected with avirulent Y. enterocolitica strain WAC, wild type strain WA314 or the single Yop-mutant strains WA314ΔYopM and WA314ΔYopP with multiplicity-of-infection (MOI) of 100 for 6 h. Samples were subjected to ChIP-seq for histone modifications H3K27ac, H3K4me3 and RNA-seq. H3K4me1 and H3K27me3 ChIP-seq was performed only for mock, WAC and WA314 infected macrophages. For all replicates used in the analysis see S17 Table. B, Bar plot showing proportions of dynamic and constant regions of ChIP-seq data obtained from mock-, WAC- and WA314 infected macrophages as described in (A). Dynamic regions were defined as MACS (H3K4me3, H3K27ac) or SICER (H3K4me1, H3K27me3) peaks overlapping significantly (≥ 2-fold change, adjusted P-value ≤ 0.05) up- or down-regulated differentially enriched regions from diffReps in the three pairwise comparisons WAC vs mock, WA314 vs mock and WAC vs WA314. C, Bar plot showing distribution of dynamic regions from (B) at gene promoters and enhancers. D, Heatmap showing Spearman correlation (cor.) of H3K27ac tag density at all H3K27ac peaks from mock-, WAC- and WA314-infected samples and publicly available H3K27ac ChIP-seq data of naive* and LPS* treated macrophages. Low to high correlation is indicated by blue-white-red colour scale. E, Number of differentially enriched regions (diffReps ≥ 2-fold change, adjusted P-value ≤ 0.05) for the indicated histone marks. Total (unique) refers to all unique regions in each row or column after merging overlapping regions. F, G, Venn diagrams showing overlaps of differentially enriched regions from diffReps depicting suppression of WAC induced upregulation of H3K4me3 marks by WA314 (F) and suppression of WAC induced up-regulation of H3K27ac marks by WA314 or prevention of WAC induced down-regulation of H3K27ac marks by WA314 (G). H, Bar plot showing the number of genes associated with promoters, enhancers or both for the indicated histone marks associated with dynamic regions in (E) or when taking all regions together (total).
Fig 2.
Dynamic chromatin modifications at promoters and enhancers in Y. enterocolitica infected macrophages.
A, Heatmap showing clustering of all H3K4me3 (module P1) and H3K27ac (module P2) differentially enriched regions at promoters in mock-, WAC- and WA314 infected human macrophages. P1 contains dynamic H3K4me3 regions and P2 contains dynamic H3K27ac regions at largely constant H3K4me3 regions. H3K4me3 and H3K27ac tag densities are shown in both modules. The identified clusters (colour coded bars on the left side) were grouped in classes P1a-P2d. B, Boxplots of H3K4me3 and H3K27ac tag counts for the classes in (A). Profiles describing relation of histone mark levels between mock, WAC and WA314 are indicated on top. Data are representative of at least two independent experiments. C, Peak tracks of H3K4me3 (red) and H3K27ac (blue) ChIP-seq tag densities at promoter regions of the genes OASL from Suppression profile classes P1a, ACOT7 from Prevention profile class P1b, ZNF513 from Down profile class P2b and SLC30A3 from Up profile class P2c. D, Heatmap showing clustering of H3K27ac differentially enriched regions at enhancers of mock-, WAC- and WA314 infected macrophages. H3K4me1 tag counts are shown for the associated regions. Clustering analysis yielded 6 clusters (color coded, left side) which were assembled into classes E1-4. E, Boxplot of H3K27ac tag counts of classes E1-E4 in (D). Profiles describing relation of histone mark levels between mock, WAC and WA314 are indicated on top. Data are representative of two independent experiments. F, Heatmap showing H3K4me1 and H3K27ac tag counts at latent enhancers induced by WAC vs mock. For (A), (D) and (F) rows are genomic regions from -10 to +10 kb around the centre of the analyzed regions. “n” indicates number of regions.
Fig 3.
Association between histone modifications and gene expression in Y. enterocolitica infected macrophages.
A, Heatmap from clustering of all DEGs from mock, WAC and WA314 comparisons. Clustering identified 4 major classes R1-R4 (color coded, right side). Profiles describing relation of expression levels between mock, WAC and WA314 are indicated on the left. In the heatmap, gene rlog counts are row-scaled (row Z-score). B, C, Venn diagrams of DEG overlaps between WAC vs mock up and WAC vs WA314 up (B) and WAC vs mock down and WAC vs WA314 down (C). Overlaps show the number of genes whose up-regulation by WAC is suppressed by WA314 (B) and whose down-regulation by WAC is prevented by WA314 (C). D, Barplot showing percentage of genes from ChIP-seq profiles with promoter or enhancer modifications overlapping with corresponding RNA-seq profiles from DEG analysis. E, Heatmap presentation showing relative overlap of genes in promoter classes P1a-P2d and enhancer classes E1-4. Light to dark color scale indicates low to high overlap. F, Barplot showing percentage of genes with coordinated promoter and enhancer changes from ChIP-seq profiles in (E) overlapping with corresponding RNA-seq profiles. G, Bar plot showing fraction and number (numbers in bars) of genes from RNA-seq profiles overlapping with genes from corresponding ChIP-seq profiles at promoters or enhancers. H, Bar plot showing enriched GO and KEGG (hsa prefix) terms for genes overlapping in RNA-seq and ChIP-seq as in (G). Bars represent log10 transformed P-values correlating with significance of enrichment. Pos: positive, reg: regulation, pol: polymerase, prom: promoter. I-K, Heatmaps of row-scaled (row Z-score) RNA-seq rlog gene counts for genes with RNA-seq and ChIP-seq promoter or enhancer overlaps from pathways in (H). L, Representative enriched transcription factor motifs in promoter and enhancer regions of genes with RNA-seq and ChIP-seq (promoter or enhancer) overlaps as in (G).
Fig 4.
Epigenetic and transcriptional regulation of Rho GTPase pathway genes in Y. enterocolitica infected macrophages.
A, Number of Rho GTPase pathway genes associated with changed regions in promoter (P1-b, P2a-d) or enhancer (E1-4) classes. “Unique” refers to total unique genes or regions (overlapping regions were merged together) when counting together all promoter and enhancer classes. B, Heatmap of all DEGs from Rho GTPase pathway belonging to classes R1-R4 (color coded, right side) in Fig 3A. Association with profiles is shown on the left side. “n” refers to the number of genes. Gene rlog counts were row-scaled (row Z-score). C, Bar plot showing number of DEGs belonging to Rho GTPase pathway from RNA-seq profiles in (B) overlapping with genes with corresponding promoter (P1a-b, P2a-d classes) and/ or enhancer (E1-4 classes) changes from the same profiles. D, Bar plot showing Rho GTPase specificity of active GAPs [90] for genes with RNA-seq and ChIP-seq overlaps in (C). Fraction and number (in bars) for all active GAPs and for GAPs in the Suppression and Prevention profiles are shown. E, F, Heatmaps of RNA-seq rlog counts for GAP (E) and Rho GTPase genes (F) with RNA-seq and ChIP-seq overlaps in (C). Associated classes are indicated on the left and specificity of GAPs for Rho GTPases is indicated on the right. Gene rlog counts were row-scaled (row Z-score). G,Table showing DEGs (from Fig 3A and S6 Table) known to regulate ubiquitination and/ or proteosomal degradation of Rho GTPases. H, Schematic representation of Rho GTPase cycle showing reprogrammed Rho GTPase pathway genes divided into GAPs, GEFs, Rho GTPases and effectors, and assigned to profiles.
Fig 5.
Role of YopP and YopM in the epigenetic and transcriptional reprogramming of macrophages by Y. enterocolitica.
A, Principal component analysis of H3K27ac tag counts in classes P1a-P2d and E1-4 in two replicates of macrophages infected with the indicated strains. B, Number of statistically significant (≥ 2-fold change, adjusted P-value ≤ 0.05) DEGs and differentially enriched H3K4me3 and H3K27ac regions for WA314ΔYopM vs WA314 and WA314ΔYopP vs WA314. C, D, Percentage of WA314 effect on expression of Rho GTPase pathway (C) or Inflammatory response (D) genes that is caused by YopP. Genes with associated promoter or enhancer changes from the same profile and significant change for RNA-seq (DESeq2) and ChIP-seq (diffReps) between WAC vs WA314 are shown. The percentage value was calculated from the ratio of fold change (FC) between WA314ΔYopP vs WA314 and WA314 vs WAC for Suppression (C, D) and Prevention (C) profiles. E, Peak tracks of ChIP-seq tag densities showing H3K4me3 and H3K27ac changes associated with IL2RA gene, whose suppression of gene expression by WA314 is produced in a YopP-dependent manner (D). H3K4me3 and H3K27ac densities at the IL2RA promoter (green shade) and H3K27ac densities at the enhancers (purple shade) are not affected by YopP (compare effects of WA314 and WA314ΔYopP). F, Dot plots of H3K4me3 and H3K27ac ChIP-qPCR signals at indicated genes in macrophages not infected (mock) or infected with indicated strains. WA314ΔYopP infections occurred in the absence or presence of MAPK inhibitors (inh). The ChIP-qPCR signal was expressed as relative (rel.) enrichment (enr.) vs mock. Lines represent mean of 3 biological replicates (dots with different shapes). SOCS3: Suppressor of cytokine signaling 3, IDO1: Indoleamine 2,3-Dioxygenase 1, PTGS2: Prostaglandin-Endoperoxide Synthase 2. G, Western blot analysis (top) and quantification (bottom) of ABI1 levels in primary human macrophages infected with WA314, WA314ΔYopP and WA314ΔYopP+inh (MAPK inhibitors) at an MOI of 100 for 6 h. GAPDH was used as a loading control and for normalization of ABI1 signal. Western blot is representative of three independent experiments shown in quantification. H, Immunofluorescence staining of primary human macrophages infected with WA314, WA314ΔYopP+inh (MAPK inhibitors) or WA314ΔYopP with MOI of 100 for 6 h. Cells were stained with Alexa568 phalloidin (red) to visualize actin and antibodies for ABI1 (cyan), vinculin (blue) and ZO-1 (green). Fourth column shows an overlay of actin and ABI1 (top) and actin, vinculin and ZO-1 (bottom) for WA314ΔYopP infection. Data are representative of two independent experiments. I, Quantification of actin (left) and ABI1 (right) signal intensity at cell junctions from immunofluorescence of primary human macrophages infected with WA314, WA314ΔYopP+inh (MAPK inhibitors) or WA314 YopP with MOI of 100 for 6 h. Data are representative of two independent experiments. J, Immunofluorescence staining of primary human macrophages infected with WA314, WA314ΔYopP+C646 (p300 inhibitor) or WA314ΔYopP with MOI of 100 for 6 h (left). Cells were stained with Alexa568 phalloidin (red) to visualize actin and antibody for ABI1 (cyan). Quantification of ABI1 signal at cell cell contacts is shown on the right. Data are representative of two independent experiments. In (I) and (J) bars show mean and error bars represent standard deviation. ****: P-adjusted < 0.0001 by unpaired Wilcoxon test.
Fig 6.
Summary schematic showing the main effects of Y. enterocolitica on macrophage chromatin in gene expression.
Plasmid-encoded virulence factors suppress upregulation (Suppression profile) and prevent downregulation (Prevention profile) of PAMP induced H3K27ac and H3K4me3 changes at gene promoters and/ or enhancers and consequently the expression of associated genes. Suppression profile is associated with genes from Inflammatory response and Type I IFN signaling, whereas Prevention profile is associated with genes from Metabolic pathways and Rho GTPase signaling. Representative genes belonging to each pathway are shown. Enhancer activity is indicated with arched arrows. Increased promoter activity is indicated with more mRNAs.