Chromatin accessibility analysis identifies the transcription factor ETV5 as a suppressor of adipose tissue macrophage activation in obesity

Obesity induces a chronic low-grade inflammatory state that contributes to insulin resistance, diabetes, and metabolic syndrome. Adipose tissue expansion causes adipocyte death, hypoxia, and metabolic stress which drive immune cell infiltration and activation. Adipose tissue macrophages (ATMs) are abundant and major sources of inflammatory mediators such as IL-6 and TNF-α that interfere with insulin signaling. Monocytes recruited via MCP-1/CCR2 pathways differentiate into ATMs in lipid-rich adipose environments where free fatty acids (FFAs), particularly saturated fatty acids like palmitate, activate TLR4 signaling, PKCs, ER stress, and ROS generation, triggering inflammatory responses. While ATMs in obesity overexpress inflammatory genes (e.g., IL6, Nos2), emerging evidence describes a metabolically activated macrophage (MMe) phenotype induced by high glucose, insulin, and palmitate. Although several epigenetic and transcriptional programs in macrophage activation are known, the transcription factors and regulatory elements governing ATM activation during obesity remain incompletely defined. This study profiles chromatin accessibility of monocytes and ATMs from lean and obese mice to identify regulatory elements and transcription factors underlying activation, highlighting ETS translocation variant 5 (ETV5) as a candidate regulator.

Prior work links inflammation to insulin resistance and metabolic disease, with ATMs as key producers of cytokines impairing insulin signaling. Monocyte recruitment to adipose tissue during obesity is mediated by MCP-1/CCR2 and FFAs activating TLR4, PKCs, ER stress, and ROS. ATMs in obesity can adopt a metabolically activated (MMe) phenotype distinct from classical M1/M2 states, induced by glucose, insulin, and palmitate. Epigenetic remodeling underlies macrophage phenotypic changes, and regulatory elements responsive to polarization cues have been mapped, though less is known for ATM differentiation/activation. Transcription factors such as NF-κB, AP-1, HIFs, STATs, and PPARs have been implicated in macrophage activation. ETV5 has been associated with insulin exocytosis, hepatic lipid metabolism via PPARγ, and human obesity GWAS signals, but its role in macrophages was previously unclear.

Animal model: Male C57BL/6 mice (6 weeks) were fed either chow diet (CD; 3.8 kcal/g, 15% protein, 75% carbohydrate, 10% fat) or high-fat diet (HFD; 5 kcal/g, 20% protein, 23% carbohydrate, 57% fat) for 12–14 weeks. Glucose tolerance tests (GTT; 2 g/kg after 16 h fast) and insulin tolerance tests (ITT; 0.5 U/kg after 4 h fast) assessed metabolic status. Cell isolation and flow cytometry: Epididymal/visceral WAT was digested (collagenase II) to isolate stromal vascular fraction; RBC lysis performed. Peripheral blood mononuclear cells were isolated by density gradient. Cells were blocked and stained (CD45.2, F4/80, Ly6C, CD11c, Ly6G, CD11b). Sorting of CD11b+Ly6Chigh monocytes and CD11b+F4/80+ ATMs used FACSAria III. Data acquired on BD LSRFortessa and analyzed with FlowJo. ATAC-seq: Prepared libraries from sorted monocytes and ATMs; quality assessed by TSS enrichment and fragment length distribution. Peaks called with MACS2; counts quantified by BedTools and normalized with DESeq. Peaks with mean count > 8, FC > 1.5, p < 0.05 were compared between CD and HFD. Differential regions between monocytes and macrophages in CD and HFD were contrasted; regions with LogFCHFD − LogFCCD > 2×log2(1.5) selected, yielding 9,461 differential peaks. Unsupervised clustering (Cluster 3.0) defined seven clusters (C1–C7). Functional enrichment of cis-regulatory regions was analyzed with GREAT. Motif enrichment and TF footprinting assessed TF occupancy. BMDM culture and polarization: Bone marrow cells differentiated to BMDMs using M-CSF or 15% L929-conditioned media for 7 days. M1 induced with LPS (100 ng/mL) + IFN-γ (20 ng/mL). M2 induced with IL-4 (10 ng/mL) + IL-13 (10 ng/mL). MMe induced with palmitate (0.4 mM), insulin (10 nM), and glucose (30 mM) for 24 h. Gene perturbation: Lentiviral shRNA or overexpression vectors (PSIN-GFP, pLKO) were produced in 293T cells and used to infect RAW264.7 macrophages; selection with puromycin. siRNAs targeting Etv5 were transfected into RAW264.7 or BMDMs using TurboFect. Knockdown/overexpression validated by RT-qPCR and Western blot. Target sequences are listed (Table 1). Molecular assays: RNA isolated by TRIzol or RNeasy; cDNA synthesized and qPCR performed; normalization to GAPDH/β-actin/Hprt (ΔΔCt). Western blot probed ETV5 and β-actin. IL-6 protein in supernatants measured by ELISA. RNA-seq: Libraries prepared with Illumina TruSeq; differential expression identified with limma; GO enrichment via clusterProfiler; GSEA performed. Upregulated genes (log2FC > 2) used for GO analysis. Single-cell RNA-seq analysis: Analyzed mouse iWAT stromal vascular fraction data (GEO GSE154047) using Seurat; filtered cells/genes; PCA and tSNE; focused on myeloid clusters identified by Lyz2 expression. Statistics: Two-tailed unpaired Student’s t-test for comparisons; body weight change and GTT analyzed by two-way ANOVA with repeated measures and Holm–Sidak post hoc; ITT by Mann–Whitney U. Shapiro–Wilk and Levene tests assessed distribution and variance. Results from at least two independent experiments; mean ± SD reported.

- ATAC-seq revealed distinct chromatin accessibility landscapes in peripheral blood monocytes and ATMs from lean (CD) vs obese (HFD) mice; PCA clustered samples by cell type and condition. - Identified 9,461 differential chromatin-accessible regions, grouped into seven clusters (C1–C7). Clusters linked to immune activation, myeloid differentiation, chemotaxis, cytokine-mediated signaling, NF-κB signaling, and IL-6 production. - Monocyte activation (C1) showed enriched motifs/footprints for SP1, NFYA, NFKB1, and KLF5; many elements were proximal to TSS. - Monocyte-to-ATM differentiation (C4, C5) enriched for JUN and BATF, as well as negative regulators BACH2 and ATF3. - ATM activation (C6, C7) enriched for JUN, ATF3, FOS, and particularly ETV5; GREAT linked ETV5-associated elements to IL-6 production. - ETV5 expression was significantly decreased in ATMs from HFD-fed mice and in metabolically activated macrophages (MMe) in vitro; ETV5 was not altered under M1 or M2 polarization. - Single-cell RNA-seq of iWAT SVF identified macrophage subsets: Etv5 was high in C1qa macrophages (alternative markers Retnla, Selenop), but low in Itgax-high (activated) macrophages in vivo. - Palmitate, but not glucose or insulin, downregulated ETV5 protein in RAW264.7 cells and BMDMs; combinations including palmitate drove the effect. - Etv5 knockdown in RAW264.7 cells upregulated 2,017 genes and downregulated 2,525 genes; upregulated genes included Il6, Plin2 (MMe marker), and Nfkbiz (IκB-ζ). - GO and GSEA showed enrichment of ER stress response, inflammatory response (p < 0.001), and positive regulation of IL-6 production (p = 0.044) upon Etv5 knockdown; Il6 ranked among top genes in these gene sets. - Functional assays: Etv5 knockdown increased Il6 mRNA in BMDMs and RAW264.7; ETV5 overexpression decreased Il6. Under MMe induction, Etv5 silencing elevated Il6, while overexpression suppressed it. - Palmitate alone increased IL6 mRNA and IL-6 protein secretion; adding glucose/insulin did not further enhance secretion. Etv5 knockdown (siRNA or shRNA) increased IL-6 secretion by ELISA. - Data suggest ETV5 suppresses ER stress and inflammatory signaling (ERK/NF-κB/NFKBIZ axis), thereby limiting IL-6 production and ATM activation.

This study addressed how transcriptional regulatory mechanisms and chromatin accessibility changes underlie monocyte and ATM activation in obesity. ATAC-seq defined regulatory element clusters and TF networks spanning monocyte activation, differentiation to ATMs, and ATM activation. The enrichment of ETV5 motifs in ATM-activation clusters, its downregulation in ATMs from obese mice and in palmitate-treated macrophages, and the induction of inflammatory/IL-6 pathways upon Etv5 knockdown together identify ETV5 as a suppressor of ATM activation. ETV5 appears to restrict ER stress and inflammatory signaling leading to IL-6 production, positioning it as a key negative regulator in the MMe context. These findings expand the repertoire of macrophage regulatory TFs beyond NF-κB, AP-1, HIFs, STATs, and PPARs, linking ETV5 to control of chronic inflammation in obese adipose tissue. The data also align with prior observations of ETV5 roles in metabolism (insulin secretion, hepatic fatty acid metabolism), suggesting tissue-specific functions: in macrophages, ETV5 suppresses inflammatory outputs without altering Pparg expression under tested conditions. The scRNA-seq analysis corroborates an in vivo reduction of Etv5 in activated Itgax-high ATMs. Mechanistically, increased Nfkbiz and ER stress signatures after Etv5 loss indicate that ETV5 may attenuate ERK/NF-κB/IκB-ζ–mediated IL-6 transcription. Overall, delineating these chromatin and TF networks provides insight into obesity-associated ATM activation and suggests therapeutic modulation of ETV5 to mitigate adipose inflammation and insulin resistance.

Chromatin accessibility profiling of monocytes and ATMs from lean and obese mice uncovered regulatory element clusters and TF networks governing activation states. ETV5 emerged as a critical negative regulator of ATM activation: it is downregulated by palmitate, suppresses ER stress and inflammatory signaling, and limits IL-6 expression and secretion. Loss of ETV5 promotes an MMe-like inflammatory phenotype. These findings propose ETV5 as a potential therapeutic target to reduce obesity-related adipose inflammation and insulin resistance. Future work should define the direct genomic targets and co-factors of ETV5 in macrophages, clarify its interaction with ERK/NF-κB/IκB-ζ pathways, and dissect cell-state heterogeneity and dynamics using single-cell multi-omics and in vivo models.

The study primarily uses mouse models and in vitro macrophage systems, which may limit direct translation to humans. Mechanistic details of how ETV5 modulates ER stress and ERK/NF-κB/NFKBIZ signaling remain to be elucidated. Monocytes and ATMs are heterogeneous; while scRNA-seq provided some insights, comprehensive single-cell analyses of chromatin accessibility and regulatory dynamics during monocyte activation, differentiation to ATMs, and ATM activation were not performed and are needed.