Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Oct 16;49(4):615-626.e6.
doi: 10.1016/j.immuni.2018.09.005.

The Nuclear Receptor PPARγ Controls Progressive Macrophage Polarization as a Ligand-Insensitive Epigenomic Ratchet of Transcriptional Memory

Free PMC article

The Nuclear Receptor PPARγ Controls Progressive Macrophage Polarization as a Ligand-Insensitive Epigenomic Ratchet of Transcriptional Memory

Bence Daniel et al. Immunity. .
Free PMC article


Macrophages polarize into distinct phenotypes in response to complex environmental cues. We found that the nuclear receptor PPARγ drove robust phenotypic changes in macrophages upon repeated stimulation with interleukin (IL)-4. The functions of PPARγ on macrophage polarization in this setting were independent of ligand binding. Ligand-insensitive PPARγ bound DNA and recruited the coactivator P300 and the architectural protein RAD21. This established a permissive chromatin environment that conferred transcriptional memory by facilitating the binding of the transcriptional regulator STAT6 and RNA polymerase II, leading to robust production of enhancer and mRNAs upon IL-4 re-stimulation. Ligand-insensitive PPARγ binding controlled the expression of an extracellular matrix remodeling-related gene network in macrophages. Expression of these genes increased during muscle regeneration in a mouse model of injury, and this increase coincided with the detection of IL-4 and PPARγ in the affected tissue. Thus, a predominantly ligand-insensitive PPARγ:RXR cistrome regulates progressive and/or reinforcing macrophage polarization.

Keywords: IFN-γ; IL-4; Nuclear receptor; PPARγ; coregulators; epigenomics; ligand-insensitive enhancers; macrophage polarization; muscle regeneration; progressive polarization; transcriptional memory.


Figure 1.
Figure 1.. Macrophage polarization by IL-4 extends the PPARγ:RXR cistrome, exhibiting predominantly ligand-insensitive sites.
A; Read distribution plot of ATAC-seq, PPARγ and RXR ChIP-seq in non-polarized (CTR) and IL-4 polarized (IL-4) macrophages in a 1.5 kb window around the summit of the RXR peaks. Cluster I. represents constitutive RXR-bound genomic regions, while cluster II. shows de novo PPARγ:RXR sites. Enriched DR1 motif in Cluster II. (bottom). Three replicates were used to determine these clusters using DiffBind and differences were considered significant at p<0.05 using three replicates. B; Box plot representation of ATAC-seq, RXR and PPARγ read enrichments in the clusters defined. Reads from three replicates were merged and results were considered significant at p<0.0001 using paired t test. C; KEGG pathway analysis of RSG-induced genes identified by GRO-seq, dashed line represents —logl 0 (1.5) and used as a threshold. Heat maps depicting genes showing increased expression in the functional categories having p-values higher than the threshold (bottom). Fold changes are relative to IL-4+vehicle treated samples. D; Genome browser view of PPARγ:RXR peaks in the presence of IL-4. GRO-seq signals from IL-4-exposed cells (24 hours) followed by vehicle (veh) or RSG treatments (Ihour) on the indicated loci. E; Quantitative PCR (qPCR) measurement of the indicated genes. Data represent mean +/− SD of triplicate determinations.
Figure 2.
Figure 2.. Determination and characterization of the ligand-sensitive, -insensitive PPARγ cistrome.
A; Box plot depicting enhancer RNA expression (RPKM) in the presence of the indicated compounds determined by GRO-seq on PPARγ:RXR-bound loci. Numbers represent region count for enhancers (bottom) and annotated genes (Ann. Genes, right, in brackets) in each category. “veh” stands for solvent control. Reads from two biological replicates were merged and changes were considered significant at p<0.05 using two tailed paired t test. B, Crystal structure of the intact PPARγ:RXRa heterodimer on the indicated response element (Chandra et al., 2008). PPARγ is highlighted in red, RXRa in blue. DBD indicates DNA binding domain. Red arrows show the ACT sequence, located upstream to the PPARγ half site and its extensive interaction with the hinge region. DR1 motifs enriched for RSG-induced and IL-4-sensitive PPARγ:RXR enhancers are also shown (bottom). C; Histograms depicting read enrichments for ATAC-seq, P300 and RAD21 ChIP-seq around PPARγ summits at RSG- and IL-4-induced enhancers in IL-4 treated wild type (Pparg +/+) and Pparg −/− cells. Box plot panels show normalized read counts (Norm. R. C.) for each factor. Significant changes were identified by paired t test at p<0.05. Correlation analysis was performed between replicates (Figure S2D) and one representative experiment is shown. D; Genome browser view of PPARγ, ATAC- seq, P300 and RAD21 signals on the indicated loci. Overlay tracks are presented for ATAC-seq, P300 and RAD21 from IL-4-treated wild-type (+/+) and Pparg −/− cells.
Figure 3.
Figure 3.. Ligand-insensitive PPARγ facilitates STAT6 signaling.
A; Read enrichments (RPKM values from one representative experiment) for STAT6, RXR, P300 and RAD21 determined by ChIP-seq at IL-4-regulated enhancers in a time course experiment. B; ChIP-qPCR carried out for STAT6, RXR, PPARγ and RAD21 on the indicated loci. Experimental scheme is shown (bottom). C; Quantification of Arg1 gene expression (mRNA) using qPCR in wild-type (+/+) and Pparg −/− macrophages. First and second stimulation by IL-4 is indicated as 1st and 2nd. D; ChIP-qPCR for STAT6, RNAPII-pS2 and H3K27ac in wild-type and Pparg −/− cells upon 1st and 2nd IL-4 stimulation. E; 3C-qPCR experiments on the Arg1 locus. Interaction of the enhancer located close to the −198kb bait and the gene promoter is shown in untreated (CTR) and IL-4 stimulated wild-type cells (left). Interaction frequency of the bait and promoter is also presented in Pparg −/− cells. The experimental setup is the same as on panel B. F; Gene expression (mRNA) of the indicated genes in gain of function experiments using Pparg −/− macrophage cell lines expressing luciferase (LUC), wild-type PPARγ (Py) and ligand-insensitive, mutant PPARγ (Py E499Q). G; ChIP-qPCR for PPARγ, STAT6 and RAD21 on the Arg1 enhancer in the gain of function system. H; Genome browser view of PPARγ ChIP-seq in wild-type macrophages and ATAC-seq signals from gain of function experiments on the Arg1 locus. Overlay tracks are presented for ATAC-seq. I; Box plot showing ATAC-seq read enrichments from gain of function experiments on IL-4 induced enhancers from two replicates for each condition. Significant changes are determined by two tailed t tests at p<0.05 for all the panels (boxplots-paired, bar graphs- unpaired). Bargraphs present the mean +/− the SD of at least two biological replicates.
Figure 4.
Figure 4.. Ligand-insensitive PPARγ acts as an epigenomic ratchet, providing transcriptional memory on a coherent ECM-related gene set.
A; Experimental setup used to study transcriptional memory in wild type (Pparg +/+) and Pparg −/− macrophages. First stimulation (1st) and second stimulation (2nd) was performed and samples were collected for RNA-seq. B; Heat map representation of the genes (235) changing exclusively upon the 2nd IL-4 stimulation and dependent on the presence of PPARγ. Fold change > 2 and significant changes at p<0.05 using edgeR GLM (General Linear Model) are shown between wild-type and Pparg −/− from two replicates. Log2 fold change is presented. C; Genome browser snapshot of a select set of genes, showing upregulation only upon the 2nd IL-4 stimulation in a PPARγ- dependent manner (RNA-seq). RplpO is shown as a control. D; Enriched KEGG pathway analysis terms for PPARγ-dependent transcriptional memory, dashed line represents - log10 (1.5) and used as a threshold to focus on the most significant terms. E; Heat map representing the focal adhesion related gene set. Log2 fold change is presented. F; Genome browser view of the Collal locus with the indicated ChIP-seq and RNA-seq experiments. ChIP-seq experiments for RNAPII-pS2 was performed in the presence of the indicated nuclear receptor ligands. G; In vitro scratch assay using HREC (Human Retinal Endothelial Cells) cells. Wild-type (+/+) and Pparg −/− macrophages were stimulated with IL-4 for 24 hours or left untreated. Wash-out was performed and the cells were rested for 24 hours followed by 24 hours of IL-4 restimulation. HREC cells were incubated for 24 hours in the collected macrophage supernatants and wound closure was quantified. Percentage of re-epithelialization over untreated control (dashed line) is presented. Mean +/− SD of triplicate determinations are shown and changes were considered significant at p<0.05 using two tailed unpaired t test.

Comment in

  • Racheting Up Repair.
    Gatchalian J, Hargreaves DC. Gatchalian J, et al. Immunity. 2018 Oct 16;49(4):577-579. doi: 10.1016/j.immuni.2018.09.012. Immunity. 2018. PMID: 30332621

Similar articles

See all similar articles

Cited by 11 articles

See all "Cited by" articles

Publication types

MeSH terms