Diabetes impairs wound healing by Dnmt1-dependent dysregulation of hematopoietic stem cells differentiation towards macrophages

Abstract

People with type 2 diabetes mellitus (T2DM) have a 25-fold higher risk of limb loss than non-diabetics due in large part to impaired wound healing. Here, we show that the impaired wound healing phenotype found in T2D mice is recapitulated in lethally irradiated wild type recipients, whose hematopoiesis is reconstituted with hematopoietic stem cells (HSCs) from T2D mice, indicating an HSC-autonomous mechanism. This impaired wound healing phenotype of T2D mice is due to a Nox-2-dependent increase in HSC oxidant stress that decreases microRNA let-7d-3p, which, in turn, directly upregulates Dnmt1, leading to the hypermethylation of Notch1, PU.1, and Klf4. This HSC-autonomous mechanism reduces the number of wound macrophages and skews their polarization towards M1 macrophages. These findings reveal a novel inflammatory mechanism by which a metabolic disorder induces an epigenetic mechanism in HSCs, which predetermines the gene expression of terminally differentiated inflammatory cells that controls their number and function.

Fig. 1

Morphometric analysis of impaired wound healing in Type 2 diabetic mice. a Wound closure rate measurement (n = 8, *p < 0.05 vs. WT). b Representative wound images. c–e Histological quantification of distance between epithelial tips/ends of panniculus carnosus and granulation tissue (n = 4, *p < 0.05 vs. WT). f Representative H&E staining wound images, magnification ×40. Scale bar, 100 μm. gt, granulation tissue. g, h Immunohistochemical (IHC) quantification of vascularization by CD144 and α-SMA staining. i Representative IHC staining images, magnification ×200. Scale bar, 100 μm. Results are expressed as means ± SEM. Two-tailed unpaired Student’s t test was used for a, c–e, g

Fig. 2

Type 2 diabetes reduces macrophage infiltration into wounds. a Quantification of monocytes concentration in bone marrow by flow cytometry (n = 6, *p < 0.05 vs. WT). b Schematic of flow cytometry gating. c Quantification of absolute number of monocytes in bone marrow (n = 6, *p < 0.05 vs. WT). d Quantification of macrophage concentration in the cutaneous wounds by flow cytometry (n = 6, *p < 0.05 vs. WT). e Schematic of flow cytometry gating. f–h Quantification of M1/M2 polarization in the cutaneous wounds by flow cytometry (n = 6, *p < 0.05 vs. WT). Results are expressed as means ± SEM. Two-tailed unpaired Student’s t test was used for a, c, d, f, g, h

Fig. 3

Type 2 diabetes impairs wound healing and monocytes/macrophages infiltration. a, b Wound closure rate measurement and representative images (n = 8, *p < 0.05 vs. WT HSC → WT). c–e Histological quantification of distance between epithelial tips/ends of panniculus carnosus and granulation tissue (n = 4, *p < 0.05 vs. WT HSC→WT). f Representative H&E staining wound images, magnification ×40. Scale bar, 100 μm. gt, granulation tissue. g Quantification of monocytes concentration in bone marrow by flow cytometry (n = 6, *p < 0.05 vs. WT HSC → WT). h Quantification of macrophage concentration in the cutaneous wounds by flow cytometry (n = 6, *p < 0.05 vs. WT HSC → WT). i–k Quantification of M1/M2 polarization in cutaneous wounds by flow cytometry. (n = 6, *p < 0.05 vs. WT HSC → WT). Results are expressed as means ± SEM. Two-tailed unpaired Student’s t test was used for a, c–e, g–k

Fig. 4

Nox-2-dependent increase in Dnmt1 expression. Quantification by flow cytometry of a DCF positive cells (n = 6, *p < 0.05 vs. WT; ^# p < 0.05 vs. db/db). b Monocytes concentration in bone marrow (n = 6, *p < 0.05 vs. WT; ^# p < 0.05 vs. db/db). c NADPH oxidase gene expression (n = 6, *p < 0.05 vs. WT). d Nox-2 and Nox-4 gene expression after knockdown of Nox-2 or Nox-4 (n = 6, *p < 0.05 vs. WT; ^# p < 0.05 vs. db/db). e Quantification of DCF positive cells after knockdown of Nox-2 or Nox-4 (n = 6, *p < 0.05 vs. WT; ^# p < 0.05 vs. db/db). f Dnmts gene expression (n = 6, *p < 0.05 vs. WT). g Dnmts expression after knockdown of Nox-2 or NAC antioxidant treatment (n = 6, *p < 0.05 vs. WT; ^# p < 0.05 vs. db/db). h Dnmts protein levels after knockdown of Nox-2 or NAC antioxidant treatment (n = 3, *p < 0.05 vs. WT; ^# p < 0.05 vs. db/db). Results are expressed as means ± SEM. One way ANOVA was used for a, b, d, e, g. Two-tailed unpaired Student’s t test was used for c, f

Fig. 5

Dnmt1 is the direct downstream target of microRNA let-7d-3p. a Relative expression of let-7d-3p in db/db HSCs (n = 6, *p < 0.05 vs. WT; ^# p < 0.05 vs. db/db). b Luciferase activity assay (n = 4, *p < 0.05 vs. control). Dnmt1 M1 represents mutant #1 and Dnmt1 M2 represents mutant #2. c Relative expression of let-7d-3p in WT HSCs following transfection with a let-7d-3p inhibitor (n = 6, *p < 0.05 vs. WT control). d Relative expression of Dnmt1 in WT HSCs following transfection of let-7d-3p inhibitor (n = 6, *p < 0.05 vs. WT control). e Relative expression of let-7d-3p in db/db HSCs following transfection with let-7d-3p mimic. f Relative expression of Dnmt1 in db/db HSCs following transfection of let-7d-3p mimic (n = 6, ^# p < 0.05 vs. db/db control). Results are expressed as means ± SEM. One way ANOVA was used for a. Two-tailed unpaired Student’s t test was used for b–f

Fig. 6

Knockdown of Dnmt1 in db/db HSCs increases wound closure rate. a qRT-PCR quantification of Dnmt1 expression (n = 6, *p < 0.05 vs. WT HSC, ^# p < 0.05 vs. db/db HSC). b Wound closure rate measurement (n = 8, *p < 0.05 vs. WT HSC → WT, ^# p < 0.05 vs. db/db HSC → WT). c–e Histological quantification of distance between epithelial tips/ends of panniculus carnosus and granulation tissue (n = 4, **p < 0.05 vs. W T HSC→ WT, ^# p < 0.05 vs. db/db HSC→ WT). f Quantification of macrophage concentration in the cutaneous wounds by flow cytometry (n = 6, *p < 0.05 vs. WT HSC → WT, ^# p < 0.05 vs. db/db HSC → WT). g–i Quantification of M1/M2 polarization in cutaneous wounds by flow cytometry. (n = 6, ^# p < 0.05 vs. db/db HSC → WT). Results are expressed as means ± SEM. One way ANOVA was used for a–f. Two-tailed unpaired Student’s t test was used for g–i

Fig. 7

Oxidant stress-dependent downregulation of Klf4, PU.1, and Notch1. Gene expression analysis following. a Nox-2 knockdown (n = 6. *p < 0.05 vs. WT; ^# p < 0.05 vs. db/db). b Dnmt1 knockdown (n = 6. *p < 0.05 vs. WT; ^# p < 0.05 vs. db/db). c Flow cytometry analysis of HSC-induced differentiation towards macrophages under in vitro conditions (n = 6, *p < 0.05 vs. WT; ^# p < 0.05 vs. db/db). d Representative images. e Flow cytometry analysis of M1/M2 macrophages (n = 6. *p < 0.05 vs. WT; ^# p < 0.05 vs. db/db). Results are expressed as means ± SEM. One way ANOVA was used for a–c, e

Fig. 8

T2DM upregulates Dnmt1 in db/db HSCs. a DNA methylation analysis by pyrosequencing (n = 6, *p < 0.05 vs. WT; ^# p < 0.05 vs. db/db). b ChIP-PCR analysis of histone modifications (n = 6. *p < 0.05 vs. WT; ^# p < 0.05 vs. db/db). Results are expressed as means ± SEM. One way ANOVA was used for a, b