Palmitic acid alters enhancers/super-enhancers near inflammatory and efferocytosis-associated genes in human monocytes

Abstract

Free fatty acids like palmitic acid (PA) are elevated in obesity and diabetes and dysregulate monocyte and macrophage functions, contributing to enhanced inflammation in these cardiometabolic diseases. Epigenetic mechanisms regulating enhancer functions play key roles in inflammatory gene expression, but their role in PA-induced monocyte/macrophage dysfunction is unknown. We found that PA treatment altered the epigenetic landscape of enhancers and super-enhancers (SEs) in human monocytes. Integration with RNA-seq data revealed that PA-induced enhancers/SEs correlated with PA-increased expression of inflammatory and immune response genes, while PA-inhibited enhancers correlated with downregulation of phagocytosis and efferocytosis genes. These genes were similarly regulated in macrophages from mouse models of diabetes and accelerated atherosclerosis, human atherosclerosis, and infectious agents. PA-regulated enhancers/SEs harbored SNPs associated with diabetes, obesity, and body mass index indicating disease relevance. We verified increased chromatin interactions between PA-regulated enhancers/SEs and inflammatory gene promoters and reduced interactions at efferocytosis genes. PA-induced gene expression was reduced by inhibitors of BRD4, and NF-κB. PA treatment inhibited phagocytosis and efferocytosis in human macrophages. Together, our findings demonstrate that PA-induced enhancer dynamics at key monocyte/macrophage enhancers/SEs regulate inflammatory and immune genes and responses. Targeting these PA-regulated epigenetic changes could provide novel therapeutic opportunities for cardiometabolic disorders.

Keywords: diabetes; dyslipidemias; epigenetic mechanisms; free fatty acids (FFA); genomics; inflammation; macrophage; monocyte; obesity; phagocytosis.

Fig. 1

PA-induced gene expression is associated with genome-wide alterations in enhancer repertoire in human CD14⁺ monocytes. A: Experimental design used to identify PA-induced enhancers and super-enhancers in human CD14⁺ monocytes. RNA-sequencing and ChIP-seq were performed on two biological replicates in each condition. B and C: Bar graph showing pathways regulated by the upregulated (B) and downregulated (C) genes in human CD14⁺ monocytes treated with PA. Pathway analysis was performed using DAVID online tools. False discovery rate (FDR)-corrected P-values <0.05 were considered significant. D: Genomic distribution of H3K27ac peaks in control vehicle (BSA) and PA treated human CD14⁺ monocytes. E: Volcano plot of the differentially enriched H3K27ac peaks in PA treated human CD14⁺ monocytes versus control (BSA) (log2 fold change [log2FC] ≥1 and FDR ≤0.05). F: Heatmaps of read densities around the enhancer peaks (±5kb from the centre of the enhancer peaks). Red color represents increased and blue color represents decreased read densities. G, H: Graphs showing the read densities of H3K27ac enrichment around up and down regulated enhancers. Black color represents BSA and blue color PA treated human CD14⁺ monocytes.

Fig. 2

Enhancers regulated by PA are associated with increased expression of inflammatory genes, and reduced expression of genes associated with phagocytosis and efferocytosis. A, B: Bar graphs showing the enrichment of biological processes (BPs) in PA-upregulated (A) and PA-downregulated (B) enhancers. Differentially expressed enhancers were functionally annotated using Genomic Regions Enrichment of Annotations Tool (GREAT). Enhancer-associated genes were selected based on GREAT analysis (two nearest genes in ±1000kb from enhancer location). C: Box plot showing the relation between downregulated enhancers (Down E) and Upregulated enhancers (Up E) and associated gene expression. Expression values (log2 fold change) of the enhancer associated genes (±1000kb) were obtained from RNA-seq datasets of CD14⁺ monocytes treated with BSA and PA. Box plot shows the median and upper and lower quartile range. ∗∗∗∗P < 0.0001 as determined by unpaired t-tests. D: Heatmap of genes (shown on right) associated with enhancers in BSA and PA treated cells that have functions related to inflammatory response, phagocytosis, and apoptotic cell clearance. The genes were selected from biological processes (BPs) enriched from GREAT analysis (Inflammatory response, phagocytosis, and apoptotic cell clearance) shown in panels A, B. The heatmap was generated using RPKM values from RNA-seq data sets of CD14⁺ monocytes treated with BSA and PA. E: Enriched transcription factor motifs in upregulated and downregulated enhancers in PA-treated cells.

Fig. 3

PA-regulated enhancers harbor single-nucleotide polymorphisms (SNP)s associated with metabolic diseases. A: Workflow for identification of SNPs associated with type 2 diabetes, body mass index (BMI), and obesity that are located in PA-regulated enhancers and non-PA-regulated enhancers. B: The effect of PA-regulated enhancers(H3K27ac) overlapping with metabolic disease-associated SNPs on related gene expression (RNA). Data represents log2Fold changes in H3K27ac enrichment (ChIP-seq) and gene expression (RNA-seq) in CD14⁺ monocytes treated with PA versus BSA. C–H: RT-qPCR validation of PA-induced genes associated with enhancers harboring metabolic disease SNPs. THP1 monocytes were differentiated into macrophages and treated with BSA or PA for 24 h followed by RT-qPCR analysis of indicated genes. Error bars represent mean ± SD. ∗P < 0.05, ∗∗P < 0.01 and ∗∗∗∗P < 0.0001, as determined by unpaired t-tests in C–H (n = 4).

Fig. 4

PA-regulated super-enhancers (SE)s are enriched in phagocytosis and other key monocyte phenotypes. A: Graph showing the differentially enriched SEs in PA-treated human CD14⁺ monocytes. B: Box plot showing the association of PA-regulated SEs and gene expression. SEs-associated genes were selected based on GREAT analysis (two nearest genes in ±1000kb from SEs). Box plot shows the median and upper and lower quartile range (n = 2). C, D: Gene ontology biological process analysis of upregulated (C) and downregulated (D) super-enhancers associated genes. E, F: Integration of ChIP-seq (H3K27ac), RNA-seq and publicly available promoter capture Hi-C data shows the upregulated SE around IRAK2 (E), downregulated SE around MERTK gene (F); Vertical orange boxes (E1-E4)- H3K27ac enrichment peaks; Red color loops: chromatin interactions. G–L: qPCR results of the chromosome conformation capture (3C) analysis of enhancer-promoter interactions between SEs and indicated gene promoters in THP1-macrophages treated with BSA or PA for 24 h. (G–L). P1, P2, P3 and P4 represent different primer sets. Error bars represent mean ± SD. ∗P < 0.05, ∗∗P < 0.01 and ∗∗∗∗P < 0.0001, as determined by unpaired t-tests in G–L (n = 3). E1-E4, refer to enhancers shown in panels E, F.

Fig. 5

Effects of PA on gene expression in THP-1 macrophages and the effects of inhibitors of BRD4 or NF-kB. A–I: Bar graphs showing the RT-qPCR validation of indicated up-regulated genes (A–E) and downregulated genes (F–I) in THP1-macrophages (TMC) treated with PA (200 μM, 24 h) versus control (BSA) treated TMC. J–M: Validation of PA-regulated SEs by H3K27ac ChIP-qPCR in TMC. TMC were treated with BSA or PA (200 μM) for 24 h. ChIP-assays were performed using H3K27ac specific antibody or IgG control. ChIP-DNA were used for qPCR analysis using indicated SE specific primers for indicated genes. N–P: Gene expression analysis of indicated genes in THP1 macrophages treated with BRD4 inhibitor (JQ1) for 2 h followed by 24 h of PA. Q–S: Gene expression analysis of indicated genes in THP1 macrophages pretreated with NF-κB inhibitor BAY 11–7082 (BAY) for 1 h followed by PA treatment for 24 h. Error bars represent mean ± SD. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001 and ∗∗∗∗P < 0.0001 as determined by unpaired t-tests A-I (n = 3) or Ordinary one-way ANOVA with Sidak's multiple comparison test for J–S (J–M, n = 3, N-S, n = 6).

Fig. 6

PA inhibits phagocytosis and efferocytosis in human macrophages. A, B: Effect of PA treatment on phagocytosis in macrophages. Representative images showing phagocytosis of E. coli bioparticles (green) (A) and quantification of fluorescence intensity (B). THP1-macrophages were treated with BSA or PA for 24 h followed by incubation with fluorescently labeled E. coli bioparticles (green). Fluorescence intensity was determined using a 96-well plate reader. Error bars represent mean ± SD. Images were acquired with the fluorescence microscope using a 20X objective (Keyence). C, D: Representative images showing inhibition of efferocytosis in THP1-macrophages treated with PA (200 μM, 24h) versus BSA treated cells. THP1-macrophages treated with BSA or PA (24 h) were incubated with fluorescently labeled apoptotic Jurkat cells (green). Images were taken using a confocal microscope (Zeiss LSM700) with 40x/1.2NA with water immersion objective from n = 4 replicates (C) and quantified using CellProfiler software (D). Box plot on the right shows quantification of efferocytosis expressed as percent of efferocytosis (Cells with efferocytosis X 100/total cells) (D). ∗∗∗∗P < 0.0001 as determined by unpaired t-tests for B and D (n = 4–5 images).

Fig. 7

Inflammation-associated genes are upregulated and phagocytosis/efferocytosis genes are downregulated in macrophages in vivo in atherosclerosis and diabetes. A, B: Expression of PA-regulated enhancer-associated genes in bone marrow-derived macrophages from mice with accelerated atherosclerosis. Bar graphs show upregulated genes (A) and downregulated genes (B). C: Expression of inflammatory and efferocytosis genes in macrophage clusters 1–3 from human atherosclerotic plaques Data from GSE155512. were plotted using PlaqView, an open-source single-cell portal for cardiovascular research. D: Inflammatory response genes increased in human macrophages with the M1 phenotype, and MERTK expression downregulated in the M1 phenotype compared to the M2 phenotype. Heatmap showing the Gene expression analysis (FPKMs) from a published study GSE55536) of inflammatory genes TNF, IL1B, IRAK2, RIPK2, and IL6 and efferocytosis gene MERTK in human macrophages with M1 and M2 phenotypes. E–I: qRT-PCR analysis of indicated genes in peritoneal macrophages from type 2 diabetic db/db or control db/+ mice. ∗P < 0.05 and ∗∗P < 0.01 as determined by unpaired t test for D–H (n = 5).