Epigenetic regulation of the PGE2 pathway modulates macrophage phenotype in normal and pathologic wound repair
Frank M. Davis, … , Bethany B. Moore, Katherine A. Gallagher
Frank M. Davis, … , Bethany B. Moore, Katherine A. Gallagher
Published September 3, 2020
Citation Information: JCI Insight. 2020;5(17):e138443. https://doi.org/10.1172/jci.insight.138443.
View: Text | PDF
Research Article Endocrinology Inflammation

Epigenetic regulation of the PGE2 pathway modulates macrophage phenotype in normal and pathologic wound repair

Abstract

Macrophages are a primary immune cell involved in inflammation, and their cell plasticity allows for transition from an inflammatory to a reparative phenotype and is critical for normal tissue repair following injury. Evidence suggests that epigenetic alterations play a critical role in establishing macrophage phenotype and function during normal and pathologic wound repair. Here, we find in human and murine wound macrophages that cyclooxygenase 2/prostaglandin E2 (COX-2/PGE2) is elevated in diabetes and regulates downstream macrophage-mediated inflammation and host defense. Using single-cell RNA sequencing of human wound tissue, we identify increased NF-κB–mediated inflammation in diabetic wounds and show increased COX-2/PGE2 in diabetic macrophages. Further, we identify that COX-2/PGE2 production in wound macrophages requires epigenetic regulation of 2 key enzymes in the cytosolic phospholipase A2/COX-2/PGE2 (cPLA2/COX-2/PGE2) pathway. We demonstrate that TGF-β–induced miRNA29b increases COX-2/PGE2 production via inhibition of DNA methyltransferase 3b–mediated hypermethylation of the Cox-2 promoter. Further, we find mixed-lineage leukemia 1 (MLL1) upregulates cPLA2 expression and drives COX-2/PGE2. Inhibition of the COX-2/PGE2 pathway genetically (Cox2fl/fl Lyz2Cre+) or with a macrophage-specific nanotherapy targeting COX-2 in tissue macrophages reverses the inflammatory macrophage phenotype and improves diabetic tissue repair. Our results indicate the epigenetically regulated PGE2 pathway controls wound macrophage function, and cell-targeted manipulation of this pathway is feasible to improve diabetic wound repair.

Authors

Frank M. Davis, Lam C. Tsoi, Rachael Wasikowski, Aaron denDekker, Amrita Joshi, Carol Wilke, Hongping Deng, Sonya Wolf, Andrea Obi, Steven Huang, Allison C. Billi, Scott Robinson, Jay Lipinski, William J. Melvin, Christopher O. Audu, Stephan Weidinger, Steven L. Kunkel, Andrew Smith, Johann E. Gudjonsson, Bethany B. Moore, Katherine A. Gallagher

×

Figure 1

COX-2/PGE2 pathway is increased in murine and human diabetic monocytes and wound macrophages.

Options: View larger image (or click on image) Download as PowerPoint
COX-2/PGE2 pathway is increased in murine and human diabetic monocytes a...
(A) Peripheral blood (30 mL) was collected from patients with T2D and control subjects without diabetes and underwent red blood cell lysis followed by Ficoll separation. CD14+ monocytes were then positively selected by magnetic activated cell sorting (MACS), and PGE Synthase 1 gene expression was measured (n = 9 T2D and 4 nondiabetic, repeated 2 times in triplicate). (B) Human wound macrophages (CD68+) were sorted via MACS from patients with T2D and nonischemic controls, and PGE Synthase 1 expression was measured (n = 5, repeated 3 times in triplicate). (C) Wound monocyte/macrophages (MФs) (CD11b+[CD3–CD19–NK1.1–Ly6G–CD11b+]) were isolated on days 0, 1, 3, 5, and 7 postwounding and analyzed for secretion of PGE2 in DIO and control mice (n = 5/group, repeated 2 times). (D) Wound monocyte/macrophages (MФs) (CD11b+[CD3–CD19–NK1.1–Ly6G–CD11b+]) were isolated on day 5 postwounding and analyzed for secretion of PGE2 in db/db and control (db/+) mice (n = 3/group, repeated 2 times). (E) Wound cells were sorted from DIO mice and PGE2 levels were analyzed in myeloid cells (CD11b+[CD3–CD19–NK1.1–Ly6G–CD11b+]) and nonimmune cells (n = 3/group, repeated 2 times). (F) Wound monocyte/macrophages (MФs) (CD11b+[CD3–CD19–NK1.1–Ly6G–CD11b+]) from DIO and controls were isolated on day 5 postwounding and analyzed for Cox-2 expression. (n = 3/group, repeated in triplicate.) (G) Peripheral blood was collected from patients with T2D and control subjects without diabetes and underwent red blood cell lysis followed by Ficoll separation. CD14+ monocytes were then positively selected by MACS, and COX-2 gene expression was measured (n = 9 T2D and 4 nondiabetic, repeated 2 times in triplicate). (H) Wound monocyte/macrophages (MФs) (CD11b+[CD3–CD19–NK1.1–Ly6G–CD11b+]) from DIO and controls were isolated on day 5 postwounding and analyzed for cPLA2 expression. (n = 3/group, repeated in triplicate.) (I) Peripheral blood was collected from patients with T2D and control subjects without diabetes and underwent red blood cell lysis followed by Ficoll separation. CD14+ monocytes were then positively selected by MACs, and cPLA2 gene expression was measured (n = 9 T2D and 4 nondiabetic, repeated 2 times in triplicate). (J) ChIP analysis was performed for H3K4me3 at the NF-κB binding site on the cPLA2 promoter in DIO and control wound monocyte/macrophages (MФs) (CD11b+[CD3–CD19–NK1.1–Ly6G–CD11b+]) isolated on day 5 (n = 3/group, repeated 2 times in triplicate). (K) BMDMs and wound monocyte/macrophages (MФs) (CD11b+[CD3–CD19–NK1.1–Ly6G–CD11b+]) were isolated from Mll1fl/fl Lyz2Cre+ mice and littermate controls on day 5 postwounding. cPLA2 gene expression was quantified by quantitative PCR (qPCR; n = 3/group, repeated 2 times in triplicate). (L) Wound monocyte/macrophages (MФs) (CD11b+[CD3–CD19–NK1.1–Ly6G–CD11b+]) were isolated from Mll1fl/fl Lyz2Cre+ mice and littermate controls on day 5 postwounding. PGE gene expression was quantified by qPCR (n = 3/group, repeated 2 times in triplicate). *P < 0.05, **P < 0.01. Data are presented as the mean ± SEM. Data were first analyzed for normal distribution, and if data passed normality test, 2-tailed Student’s t test was used.