Chromatin Remodeling Drives Immune-Fibroblast Crosstalk in Heart Failure Pathogenesis

Abstract

Chronic inflammation and tissue fibrosis are common stress responses that worsen organ function, yet the molecular mechanisms governing their crosstalk are poorly understood. In diseased organs, stress-induced changes in gene expression fuel maladaptive cell state transitions and pathological interaction between diverse cellular compartments. Although chronic fibroblast activation worsens dysfunction of lung, liver, kidney, and heart, and exacerbates many cancers, the stress-sensing mechanisms initiating the transcriptional activation of fibroblasts are not well understood. Here, we show that conditional deletion of the transcription co-activator Brd4 in Cx3cr1-positive myeloid cells ameliorates heart failure and is associated with a dramatic reduction in fibroblast activation. Analysis of single-cell chromatin accessibility and BRD4 occupancy in vivo in Cx3cr1-positive cells identified a large enhancer proximal to Interleukin-1 beta (Il1b), and a series of CRISPR deletions revealed the precise stress-dependent regulatory element that controlled expression of Il1b in disease. Secreted IL1B functioned non-cell autonomously to activate a p65/RELA-dependent enhancer near the transcription factor MEOX1, resulting in a profibrotic response in human cardiac fibroblasts. In vivo, antibody-mediated IL1B neutralization prevented stress-induced expression of MEOX1, inhibited fibroblast activation, and improved cardiac function in heart failure. The elucidation of BRD4-dependent crosstalk between a specific immune cell subset and fibroblasts through IL1B provides new therapeutic strategies for heart disease and other disorders of chronic inflammation and maladaptive tissue remodeling.

Fig. 1.. Correlation of cardiac function and gene expression identifies a stress-activated Cx3cr1-expressing macrophage subpopulation that contributes to heart failure pathogenesis.

A. Schematic of experimental settings of Sham and Transverse Aortic Constriction (TAC) models in C57BL/6J mice with daily dose of the BET inhibitor JQ1 (50mg/kg). B. Left Ventricle (LV) ejection fraction quantified by echocardiography. Statistical significance is shown between TAC and TAC JQ1 at day 30 post-TAC. C. UMAP plot (scRNA-seq) of non-cardiomyocyte cells colored by cluster. D. Schematic of correlation analysis between LV ejection fraction and gene expression highlighting a negative or positive correlation. E. Number of strongly positively- and negatively-correlated genes (p-value < 1e⁻⁶, corresponding to correlation score < −5 and > +5) across the cell population composed by a minimum of 250 cells. F. Volcano plot showing correlation coefficients (refer to analysis depicted in Fig. 1D) and corresponding p-values for all the 17,519 genes captured in myeloid cells. Cx3cr1 and Il1b are highlighted in the regions with the most negative correlation coefficients (n=22 genes with score < −5). G. Expression Dot Plot of the 22 most-anticorrelated genes in Sham, TAC, and TAC JQ1. H,I. UMAP plot (scRNA-seq) of myeloid cells colored by cluster (H) and sample identity (I). J. Expression Violin Plots of Lyve1, Timd4, Ccr2, Cx3cr1, and Il1b across myeloid clusters. K. Expression Violin Plots of Cx3cr1 and Il1b across samples. L. Dot Plot indicating significance (−log10(p−val)) for indicated GO terms in genes upregulated in TAC versus TAC JQ1 (red) or upregulated in TAC JQ1 versus TAC (blue). M. Schematic of experimental settings for conditional Brd4 deletion in Cx3cr1^Pos cells. N. LV ejection fraction quantified by echocardiography in Cx3cr1^CreERT2 (CRE control) and Brd4^flox/flox and Cx3cr1^CreERT2;Brd4^flox/flox littermates. B,N Data are mean ± s.e.m. One-way ANOVA followed by Tukey post hoc test.

Fig. 2.. BRD4 in Cx3cr1-expressing macrophages triggers a proinflammatory transcriptional response in heart failure pathogenesis.

A. Schematic of experimental settings for scRNA-seq in sorted CD45^Pos cells from hearts of Brd4^flox/flox (TAC) and Cx3cr1^CreERT2;Brd4^flox/flox (TAC Brd4-KO) littermates. B. UMAP plot (scRNA-seq) of monocytes/macrophages colored by cluster. C. Expression Violin Plots of Lyve1, Timd4, Ccr2, and Cx3cr1 across monocyte/macrophage clusters. D. Expression Feature Plots (scRNA-seq) of Lyve1, Timd4, Ccr2, and Cx3cr1 across monocytes and macrophages. E. UMAP plot (scRNA-seq) of monocytes/macrophages colored by sample identity. F. Dot Plot indicating significance (−log10(p−val)) for indicated GO terms across all monocyte/macrophage clusters. G. Expression Dot Plot of the 10 top gene markers in cluster 4 across all monocyte/macrophage clusters. H. Expression Dot Plots (scRNA-seq) of the 13 genes upregulated TAC versus Sham and downregulated in TAC-Brd4KO versus TAC in monocytes/macrophages across samples (LogFC>0.5 and FDR<0.5). I. Expression Violin Plots of Il1b across monocyte/macrophage clusters. J. Expression Feature Plot (scRNA-seq) of Il1b in monocytes and macrophages. K. Expression Violin Plot of Il1b in monocyte/macrophage across samples.

Fig. 3.. BRD4 in Cx3cr1-expressing cells functions non-cell autonomously to activate cardiac fibroblasts under stress.

A. Schematic of experimental settings for single nuclei RNA-seq and ATAC-seq from hearts of Brd4^flox/flox (TAC) and Cx3cr1^CreERT2;Brd4^flox/flox (TAC Brd4-KO) littermates. B,C. UMAP plot (snRNA-seq) of nuclei from cardiac tissue colored by cluster (B) and sample identity (C). D. Number of DE genes between TAC and TAC Brd4-KO across cell populations with at least 800 cells (LogFC>0.125; FDR<0.05). E,F. UMAP plot (snRNA-seq) of fibroblasts colored by cluster (E) and sample identity (F). G. Expression Feature Plot (snRNA-seq) of Postn and Meox1 in fibroblasts. H. Dot Plot indicating significance (−log10(p−val)) for indicated GO terms in genes upregulated in TAC versus TAC JQ1 (red) or upregulated in TAC JQ1 versus TAC (blue) in fibroblasts. I. UMAP plot (scATAC-seq) of all cardiac cells colored by cluster. J, K. UMAP plot (scATAC-seq) of fibroblasts colored by cluster (J) and sample identity (K). L. Chromatin accessibility gene score of Postn and Meox1 in fibroblasts. M. Sample distribution within clusters in fibroblasts. N. TF motif enrichment of fibroblast chromatin regions more accessible in TAC versus TAC Brd4-KO. Top 10 motifs are highlighted. O. Fibroblast chromatin accessibility in TAC and TAC Brd4-KO at the Postn and Meox1 locus. The signal at essential regulatory elements is highlighted in red together with Co-Accessibility with the respective gene promoter.

Fig. 4.. Chromatin accessibility and BRD4 occupancy in vivo identify the precise regulatory elements that control stress-dependent activation of Il1b.

A. Schematic of experimental settings for scATAC-seq in sorted CD45^Pos cells from hearts of Brd4^flox/flox (TAC) and Cx3cr1^CreERT2;Brd4^flox/flox (TAC Brd4-KO) littermates. B,C. UMAP plot (scATAC-seq) of monocytes/macrophages colored by cluster (B) and sample identity (C). D. Chromatin accessibility gene score of Timd4, Lyve1, Ccr2, Cx3cr1, and Il1b in monocytes/macrophages. E. Dot Plot indicating significance (−log10(p−val)) for indicated GO terms in genes driving monocyte/macrophage clusters 3 and 4. F. TF motif enrichment of monocyte/macrophage chromatin regions in cluster 4 (cluster 3 was used as background). Top 5 motifs are highlighted. G. Chromatin Accessibility changes between TAC and Sham (x axis) and TAC and TAC Brd4-KO (y axis) for the 749 super enhancers identified in monocytes/macrophages. Dotted red lines at +5 and −5 in both axes are indicated. Genes proximal to candidate enhancers are highlighted in top right quadrant. H. Targeting strategy to generate Brd4^flag/flag animals (top) and schematic of experimental settings for anti-FLAG CUT&RUN performed on Cx3cr1^Pos cells from Brd4^flag/flag Sham and TAC mice (bottom). I. Il1b locus showing from top to bottom: scATAC-seq mean normalized accessibility across monocytes/macrophages in Sham, TAC and TAC Brd4-KO; anti-FLAG CUT&RUN signal in Cx3cr1^Pos cells from Brd4^flag/flag Sham and TAC mice; RELA/p65 ChIP-seq on BMDMs. 7 regions (Peak1 to 7) of high chromatin accessibility are highlighted in red. J. Il1b expression by qPCR in Unstimulated or LPS (1ng/ml) treated RAW 264-7 macrophages in CRISPR WT and Peak1 to Peak7 KO clones. Significance is shown between the 2 WT clones and Il-1b Peak5-KO and Il-1b Peak6-KO clones. K. Transcriptional activity by luciferase assay of Il-1b Peak5 and Peak6 enhancer regions in the presence of BRD4 and p65/RELA. L.UMAP plot (scATAC-seq) of human cardiac myeloid cells colored by cluster. M. Chromatin accessibility gene score of CX3CR1 and IL1B in human cardiac myeloid cells. N. Human IL1B locus showing scATAC-seq mean normalized accessibility across human cardiac myeloid cells in controls (green) and patients with myocardial infarction (red). Red boxes with the same height of the y axis are placed on peak 5 and 6. J,K, Data are mean ± s.e.m. Two-way (J) and One-way (K) ANOVA followed by Tukey post hoc test.

Fig. 5.. IL1B neutralization improves cardiac function in heart failure and prevents fibroblast activation.

A. Schematic representation of experiment (left): medium from unstimulated or LPS-treated BMDM is treated with either IgG isotype control or anti-IL1B antibodies, then the conditioned medium is added onto the iPS-CFs and collagen contractility over 120 hours is measured (right). B. Bulk RNA-seq in iPS-CFs. LogFC between indicated conditions is shown. Dotted red lines at +2 and −2 in both axes are indicated. MEOX1 gene in the top-right quadrant is highlighted. C. Human MEOX1 locus showing H3K27Ac and p65/RELA in unstimulated iPS-CFs. The syntenic region of the mouse Meox1 Peak9/10 regulatory element is highlighted. D. Transcriptional activity by luciferase assay of MEOX1 human enhancer region (highlighted in Fig. 5C) in the presence of BRD4 and p65/RELA. E. MEOX1 expression by qPCR in Unstimulated iPS-CFs or treated with IL1B, TGF-B, or TGFB+IL1B with control or p65/RELA-targeting siRNAs. F. Schematic of experimental settings in TAC C57BL/6J mice treated with either 500ug of IgG isotype control and anti-IL1B antibodies injected intraperitoneally every 3 days. Non-CM cells were processed at day 30 for scRNA-seq. G. LV ejection fraction quantified by echocardiography. Statistical significance is shown between IgG and Anti-IL1B group 30 days post-TAC. H,I. UMAP plot (scRNA-seq) of fibroblasts colored by cluster (H) and sample identity (I). J. Expression Violin Plots of Postn and Meox1 across fibroblasts clusters. K. Sample distribution within clusters in fibroblasts. L. Expression Violin Plot of Postn and Meox1 in fibroblasts across samples. M. Working model depicting the molecular mechanisms regulating the crosstalk between Cx3cr1-positive cells and activated fibroblasts through IL1B and MEOX1. A,D,E Data are mean ± s.e.m. Two-way (A) and One-way (D,E) ANOVA followed by Tukey post hoc test.