Modulation of macrophages by a paeoniflorin-loaded hyaluronic acid-based hydrogel promotes diabetic wound healing

Abstract

The impaired wound healing in diabetes is a central concern of healthcare worldwide. However, current treatments often fail due to the complexity of diabetic wounds, and thus, emerging therapeutic approaches are needed. Macrophages, a prominent immune cell in the wound, play key roles in tissue repair and regeneration. Recent evidence has demonstrated that macrophages in diabetic wounds maintain a persistent proinflammatory phenotype that causes the failure of healing. Therefore, modulation of macrophages provides great promise for wound healing in diabetic patients. In this study, the potential of paeoniflorin (PF, a chemical compound derived from the herb Paeonia lactiflora) for the transition of macrophages from M1 (proinflammatory phenotype) to M2 (anti-inflammatory/prohealing phenotype) was confirmed using ex vivo and in vivo experimental approaches. A hydrogel based on high molecular weight hyaluronic acid (HA) was developed for local administration of PF in experimental diabetic mice with a full-thickness wound. The resultant formulation (HA-PF) was able to significantly promote cutaneous healing as compared to INTRASITE Gel (a commercial hydrogel wound dressing). This outcome was accompanied by the amelioration of inflammation, the improvement of angiogenesis, and re-epithelialization, and the deposition of collagen. Our study indicates the significant potential of HA-PF for clinical translation in diabetic wound healing.

Keywords: Adipic acid dihydrazide, ADH; Angiogenesis; Anti-inflammation; Hydrogel; Macrophage polarization; N-(3-Dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride, EDC.HCl; Regenerative medicine; arginase 1, Arg-1; bone marrow-derived macrophages, BMDMs; dimethyl sulfoxide, DMSO; fetal bovine serum, FBS; human umbilical vein endothelial cells, HUVECs; hyaluronic acid, HA; inducible nitric oxide synthase, iNOS; integrated optical density, IOD; interferon-γ, IFN-γ; interleukin-10, IL-10; interleukin-1β, IL-1β; lipopolysaccharide, LPS; macrophage colony-stimulating factor, M-CSF; paeoniflorin, PF; penicillin-streptomycin, P/S; phosphate-buffered saline, PBS; polyvinylidene difluoride, PVDF; scanning electron microscopy, SEM; signal transducer and activator of transcription, STAT; streptozocin, STZ; swelling ratio, SR; transforming growth factor-β, TGF-β; tumor necrosis factor-α, TNF-α; α-smooth muscle actin, α-SMA.

Graphical abstract

Fig. 1

Paeoniflorin modulates the phenotype and function of macrophages. (A) The chemical structure of paeoniflorin (PF). (B) The CD86⁺ and F4/80⁺ population (%) in LPS- and IFN-γ-stimulated BMDMs following treatment of DMSO and PF (10, 25 and 50 μM) (n = 3; ∗p < 0.05 and NS = no significance). (C) The CD206⁺ and F4/80⁺ population (%) in LPS- and IFN-γ-stimulated BMDMs following treatment of DMSO and PF (10, 25 and 50 μM) (n = 3; ∗p < 0.05 and NS = no significance). (D) The ratio of M2 cells (CD206⁺ and F4/80⁺) versus M1 cells (CD86⁺ and F4/80⁺) (n = 3; ∗p < 0.05 and NS = no significance). (E) The mRNA expression of cytokines in LPS- and IFN-γ-stimulated BMDMs following treatment of DMSO and PF (50 μM) (n = 3; ∗p < 0.05 and NS = no significance). (F) The release of cytokines in LPS- and IFN-γ-stimulated BMDMs following treatment of DMSO and PF (50 μM) (n = 3; ∗p < 0.05 and NS = no significance). (G) The activity of STAT signaling pathway in LPS- and IFN-γ-stimulated BMDMs following treatment of DMSO and PF (50 μM). The relative expression of the proteins of interest was quantified in Fig. S3.

Fig. 2

Preparation, physicochemical characterization, and in vivo performance of hyaluronic acid hydrogels. (A) The schematic for preparation of crosslinked HA. (B) SEM image of crosslinked HA with different molecular weights (Mw) (bar in the upper lane = 50 μm and bar in the lower lane = 10 μm). (C) The swelling rate of crosslinked HA with different Mw (n = 3). (D) The enzymatic hydrolysis resistance of crosslinked HA with different Mw (n = 3). (E) The release rate of BSA in crosslinked HA with different Mw (n = 3). (F) The images of crosslinked hydrogels (Mw = 1,800–2,200 kDa) at the concentration of 4%, 8% and 12%. (G) The healing rate (%) of mice with full-thickness wounds following treatment of crosslinked hydrogels (Mw = 1,800–2,200 kDa) at the concentration of 4%, 8% and 12% (n = 4; ∗p < 0.05 relative to untreated, 4% and 12%).

Fig. 3

Paeoniflorin-loaded hydrogel accelerates diabetic wound closure in mice. (A) The treatment schematic for STZ-induced diabetic mice with full-thickness wounds. (B) Images of incisional wounds after treatments of blank HA, INTRASITE gel, and HA-PF (containing 500 μM of PF). Healing rate (%) when compared with the wound area on day 0 (n = 8). The level of blood glucose in mice was monitored to ensure the formation of diabetes during the experiments (Fig. S4). (C) In H&E staining images (white bar = 20 μm), the blue, green and red arrows represent epidermal hyperplasia, proinflammatory cells, and new blood vessels, respectively (n = 6; ∗p < 0.05 and NS = no significance). (D) In Masson’s trichrome staining images (white bar = 20 μm) (red = keratin, muscle fibers or cytoplasm, blue = collagen), the collagen deposition was quantified (n = 6; ∗p < 0.05 and NS = no significance).

Fig. 4

HA-PF promotes diabetic wound healing via modulation of macrophages. (A) The population (%) of iNOS⁺ F4/80⁺ and Arg-1⁺ F4/80⁺ macrophages within the wound was determined using immunofluorescent assay (n = 4; ∗p < 0.05 and NS = no significance). (B) The mRNA expression of cytokines in the wound following treatment of blank HA, INTRASITE gel, and HA-PF (containing 500 μM of PF) (n = 4; ∗p < 0.05 and NS = no significance). (C) The release of cytokines in the wound following different treatments (n = 4; ∗p < 0.05). (D) The mRNA expression of CD31, VEGF, α-SMA and type I collagen in the wound following different treatments (n = 4; ∗p < 0.05 and NS = no significance). (E) The expression of CD31 in the wound was determined using immunofluorescent assay (n = 4; ∗p < 0.05 and NS = no significance). (F) The expression of α-SMA in the wound was determined using immunofluorescent assay (n = 4; ∗p < 0.05 and NS = no significance).

Fig. 5

Modulation of macrophages by a paeoniflorin-loaded hyaluronic acid-based hydrogel (HA-PF) promotes diabetic wound healing.