Therapeutic Targeting of Neutrophil Extracellular Traps Improves Primary and Secondary Intention Wound Healing in Mice

Abstract

Background: Neutrophils are the first responders in wound healing after injury that mediate pro- and anti-inflammatory activities i.a. through the formation of extracellular traps (NETs). However, excessive NETs presence in wound tissue can cause local hyperinflammation and -coagulation resulting in delayed wound healing. To improve wound healing, we aimed to examine the role of NETs and DNase1 on primary and secondary wound healing.

Methods: The study included 93 C57BL/6 mice, with 3 different genotypes: wildtype, Pad4-, and DNase1-Knockout (KO). Pad4-KO mice show limited NETs formation, while DNase1-KO mice cannot disintegrate them. All 3 genotypes were included in (1) a laparotomy group and (2) a thermal injury group. Animals in both groups either received DNase1 or a vehicle i.p. post wound induction and wound assessment and euthanasia were conducted. Laparotomy and burn scars were assessed using the stony brook scar evaluation scale and modified Yeong scale respectively. Tissue was analyzed histologically using H&E staining. Ly6g, Collagen I and III, SMA, and Fibrinogen were visualized and neutrophils activation (NE, MPO) and NETs (H3cit) formation assessed.

Results: All animals survived with no complications. DNase1 treatment led to a significantly improved scar appearance in both groups, which was also seen in Pad4-KO mice. In the laparotomy group DNase1 improved collagen deposition and fibrin concentration was significantly reduced by DNase1 treatment. Markers of neutrophil activation were significantly reduced in the treatment and Pad4-KO group. In the thermal injury group wound closure time was significantly reduced after DNase1 treatment and in the Pad4-KO group. Even though inflammation remained high in the thermal injury model over time, neutrophil activation and NETs formation were significantly reduced by DNase1 treatment compared to controls.

Discussion: Primary and secondary intention wound healing is improved by targeting NETs through DNase1 treatment or genetic KO, as assessed by wound closure time and scar appearances. Additionally, wound stability was not affected by DNASE treatment. The results suggest that overall wound healing is accelerated and DNase1 appears to be a promising option to reduce scar formation; which should be evaluated in humans.

Keywords: DNases; burns; neutrophil extracellular traps; scars; wound healing.

Figure 1

Therapeutic targeting of NETs improves primary intention wound healing. In all animals a 2.5 cm long laparotomy was performed. For the wildtype and knockout mice two timepoints (72 h, 21 days) were performed. Controls n=6, DNase1 n=6, DNase1-KO n=5, Pad4-KO n=6. (A, B) Animals that received DNase1 had significantly superior scar scores than controls. In mice that were unable to produce NETs (Pad4-KO) a similar effect was found. (C–E) Animals with DNase1 treatment or animals without NETs showed a significantly faster switch from collagen 3 to 1 and better collagen alignment indicating a faster maturation of the scar. (F) As previously reported DNase1 reduced Fibrin levels in the scar. (G) SMA was not affected by DNase1 treatment or in animals without NETs. Data shown as Mean ± SD. Comparison was performed always in comparison with controls. Statistics: mixed-effect model with Geisser-Greenhouse correction as well as Dunnett’s multiple comparison test. *DNase1 vs. controls. ^#PAD4-KO vs. controls.

Figure 2

Therapeutic targeting of NETs formation results in decreased neutrophil activation and NETs formation in a model of primary intention wound healing. (A) Ly6G a marker of granulocytes was not affected by genetic alterations of NETs formation or DNase1 treatment. (B–D) Neutrophil activation and NETs formation was significantly reduced by DNase1 treatment or genetic knockout. (E) Representative immunofluorescence images. DNase1 treatment vastly reduced neutrophil activation (MPO) and NETs formation (H3cit). Data shown as Mean ± SD. Comparison was performed always in comparison with controls. Statistics: mixed-effect model with Geisser-Greenhouse correction as well as Dunnett’s multiple comparison test. *DNase1 vs. controls. ^#PAD4-KO vs. controls.

Figure 3

Therapeutic targeting of NETs improves secondary intention wound healing. In all animals a 1.5 cm² thermal injury was induced. For the wildtype mice four timepoints (72 h, 7, 14, and 28 days) were performed and two (7, 28 days) for the knockouts. Controls n=5-6, DNase1 n=6, DNase1-KO n=5–6, Pad4-KO n=5–6. (A–C) Animals that received DNase1 or with limited NETs formation (Pad4-KO) had significantly superior scar scores than controls. (D–H) Animals with DNase1 treatment or without NETs showed a significantly faster wound closure and switch from collagen 3 to 1. Additionally, an improved collagen alignment after DNase1 treatment was found, indicating a faster maturation of the scar. (E) H&E staining of the thermal injury on day 7 showing the left wound border. Arrows indicate the reepithelization tongue progressing significantly faster in animals with DNase treatment. S indicates scab which is made up of necrotic tissue and found is less-optimal wounds. M indicates muscle layer and F adipose tissue. In control mice the subcutaneous fat and muscle layer was lost. However, in mice that were treated with DNase1 both layers were not affected possibly suggesting a secondary injury induced by extracellular traps. (F) Parallel collagen fiber formation was observed predominantly during early stages and persisted significantly in control and DNase1-KO scars until day 28; whereas basket wave-like texture reminiscent of normal skin was more evident in animals with DNase treatment or with limited NETs formation (Pad4-KO). Moreover, in controls and DNase-KO mice immature fibers (Collagen III; green vs. Collagen I, red) remained dominant. (I) SMA was only affected partially by DNase1. Data shown as Mean ± SD. Comparison was performed always in comparison with controls. Statistics: mixed-effect model with Geisser-Greenhouse correction as well as Dunnett’s multiple comparison test. *DNase1 vs. controls. ^#PAD4-KO vs. controls.

Figure 4

Therapeutic targeting of NETs reduces neutrophil activation and NETs formation in a model of secondary intention wound healing. (A) Comparable Ly6G levels indicate that granulocytes were not affected by DNase1 after thermal injury. However, in PAD4-KO mice a significant difference was found after 28 days, possibly indicating a loop effect of NETs and neutrophils. (B, C) Neutrophil activation remained high but was significantly lower in animals that received DNase1 or with limited NETs formation (Pad4-KO). (D) NETs formation can be significantly reduced by DNase1. (E) Representative immunofluorescence images. It appears that in burn wound neutrophils remain activated and produce NETs even after 4 weeks. DNase1 treatment is able to reduce this effect to some extent. Data shown as Mean ± SD. Comparison was performed always in comparison with controls. Statistics: mixed-effect model with Geisser-Greenhouse correction as well as Dunnett’s multiple comparison test. *DNase1 vs. controls. ^#PAD4-KO vs. controls.