CXCR2 mediated trafficking of neutrophils and neutrophil extracellular traps are required for myelin clearance after a peripheral nerve injury

Abstract

Neutrophils are a vital part of the innate immune system. Many of their functions eliminate bacteria & viruses, like neutrophil extracellular traps (NETs), which trap bacteria, enhancing macrophage phagocytosis. It was surprising when it was demonstrated that neutrophils are a part of Wallerian degeneration, a process that is essential for nerve regeneration after a nerve injury. It is not known what signals attract neutrophils into the nerve and how they aid Wallerian degeneration. Neutrophils accumulate in the distal nerve within one day after an injury and are found in the nerve from one to three days. We demonstrate that CXCR2 mediates the trafficking of neutrophils into the distal nerve, and without CXCR2 Wallerian degeneration, as indicated by luxol fast blue staining, was reduced seven days after a sciatic nerve crush or transection injury. NETs were detected in the distal nerve after a sciatic nerve transection. NET formation has been shown to require protein arginine deiminase 4 (PAD4), which citrullinates histone 3. Inhibiting PAD4 reduced NET formation significantly in the distal nerve at two days and myelin clearance at seven days indicating that NETs aid myelin clearance. These results demonstrate another function for NETs other than clearing pathogens. Neutrophils have been detected after injuries to the central nervous system and diseases in humans and animal models. Our results demonstrate neutrophils aid myelin clearance, suggesting a role for their presence in central nervous system injuries and diseases.

Keywords: Epineurium; Ly6B.2; Ly6G; Macrophage; Myelin; Neutrophil; Protein arginine deiminase 4 (PAD4); Wallerian degeneration.

Figure 1.. Neutrophils accumulate after injury at the site of injury and in the distal sciatic nerve.

Representative images of distal nerves from wildtype (WT) and Ccr2 KO mice 3 d after a sciatic nerve transection (A and B; scale bar 200 μm). Quantification of neutrophil accumulation of the whole distal nerve was performed by counting the number of Ly6B.2⁺ MPO⁺ cells (C) and Ly6G⁺MPO⁺ cells (D) in the distal nerve for both WT and Ccr2 KO at 1, 2, 3, 7 d after the injury. The accumulation at the injury site (E) and the distal portion of the nerve (F) was evaluated by measuring the percent area stained of Ly6B.2. for both WT and Ccr2 KO at 1, 2, and 3 d. Quantification of the percentage of MPO⁺ cells that are Ly6B.2 ⁺ (G), or Ly6G⁺ (H) in WT and Ccr2 KO mice 1, 2, 3, and 7 d after a sciatic nerve transection. Additional quantification of the percentage of Ly6B.2⁺ cells that are MPO⁺ (I) and Ly6G⁺cells that are MPO⁺ (J). Panel (K) is a representative image of a neutrophil labeled with Ly6B.2, MPO and DAPI. *p < 0.05. **p < 0.01. ***p < 0.001. ****p < 0.0001, # p < 0.05 compared between 7 d and earlier time points. ns indicates non-significant. n = 4-6 per group. One slide per animal was analyzed.

Figure 2.. Do neutrophils get phagocytosed by macrophages.

Representative immunofluorescent images from WT mice of macrophage (CD68⁺), neutrophils (Ly6G⁺), and DAPI⁺, 3 d after a sciatic nerve transection. Scale bar is 100 μm (A). B is a magnification of the image indicated with the white rectangle in panel A. Yellow arrowheads indicate cells that are both Ly6G⁺ and CD68⁺, while white arrowheads indicate Ly6G⁺ cells and grey arrowheads indicate CD68⁺ cells. Quantification of cells stained with Ly6G (C), CD68 (D), and both (E). Data is displayed as the mean ± standard error of the mean. *p < 0.05, **p < 0.01. n = 6 per group. One slide per animal was analyzed.

Figure 3.. Neutrophils are only infrequently associated with blood vessels in the distal injured nerve.

Representative images of distal nerves from WT sham injured (A) and distal injured sciatic nerve (B) 2 d after the nerve transection. Neutrophils were stained with Ly6B.2, blood vessels were stained with CD31, and cell nuclei with DAPI. Panel (C) is an enlarged area of panel B; the grey arrow indicates CD31 positive cells, white green arrows indicate Ly6B.2 positive cells, and yellow arrows indicate Ly6B.2 positive cells associated with CD31 positive cells. Images were quantified by counting the number of Ly6B.2+ cells that were associated with CD31⁺ cells in the sham injured nerve and the distal end of the injured nerve (D). Scale bar = 500 μm. Data is displayed as the mean ± standard error of the mean. One slide per animal was analyzed.

Figure 4.. Cxcr2 KO mice have a significant reduction in accumulation of neutrophils after sciatic nerve transection.

Two days after a sciatic nerve transection blood and spleens were collected from WT and Cxcr2 KO mice for flow cytometry analysis. Neutrophils were labeled as CD45⁺CD11b⁺Ly6G^high, monocytes were CD45⁺CD11b⁺Ly6G⁻Ly6C^High, and eosinophils were CD45⁺CD11b⁺Ly6G⁻SiglecF⁺. Panel A is the percentage of neutrophils in the blood, while D is the percentage in the spleen. Panel B and E are the percentages of monocytes in the blood and spleen, respectively. The eosinophils percentages in the blood (C) and spleen (F) were not significantly different between the genotypes. *p < 0.05. Representative images of distal nerves from litter mate WT (G) and Cxcr2 KO mice (H) 2 d after sciatic nerve injury. G’ and H’ are magnifications of the images indicated with the white rectangle. Yellow arrows indicated neutrophils. Scale bar = 200 μm. Neutrophils were stained with Ly6B.2 and myeloperoxidase (MPO), while DAPI stained nuclei. Quantification was performed using Ly6G counts (I) and the percent area stained for Ly6B.2 (J), Ly6G (K), and CD68 (L) of both sham injured and injured distal nerves for both genotypes. Data is displayed as the mean ± standard error of the mean. *p < 0.05. **p < 0.01 n = 5-7 per group. One slide per animal was analyzed.

Figure 5.. Cxcr2 KO mice have reduced myelin clearance 7 d after sciatic nerve.

The percent area stained was quantified for both genotypes (A). The percent area decreased compared to the sham is displayed in panel B. Representative images of luxol fast blue staining of sham and distal injured nerves from littermate WT (C, E) and Cxcr2 KO mice (D, F) 7 d after sciatic nerve transection. n = 7 per group. Scale bar = 100 μm. Three separate regions were analyzed and averaged to create an average value per nerve per animal. Panel G are representative images of Western blot bands for neurofilament-light and Sypro Ruby for sham nerves and injured nerves from both genotypes 7 d after a sciatic nerve transection, while Panel H is the quantification of the expression ratio, n = 3-4 per group. Representative images WT (I, K) and Cxcr2 KO (J, L) of myelin basic protein and CD 68. Panels M-O are the quantification of the MBP percent area stained, percent area changes over sham for MBP, and the percent area stained of CD68. n = 3-4 per group. Data is displayed as the mean, ± standard error of the mean. *p < 0.05, **p < 0.01, *** p< 0.001, ****p < 0.0001. Scale bar = 50 μm.

Figure 6.. Ccr2 KO mice have increased production of NETs.

Representative images of NETs: Citrullinated histone 3 (CitH3), neutrophils (Ly6B.2), and DAPI (A). The image was taken in the injury site. Scale bar is 20 μm. Quantification of the percentage of CitH3⁺Ly6B.2⁺ cells in WT and Ccr2 KO mice at 1, 2, 3 d after a sciatic nerve transection (B). Data is displayed as mean ± standard error of the mean. *p < 0.05. n = 4 per group. One slide per animal was analyzed. Yellow arrowheads indicate NET-positive cells, while the white arrowhead indicates a neutrophil, and the pink arrowhead indicates a cell that is Ly6B.2⁺CitH3⁺DAPI⁺ but does not have a NET or web like-structure outside of the cell.

Figure 7.. Inhibition of citrullination of histone 3 decreases neutrophils extracellular traps in distal nerve at 2 d and myelin clearance 7 d after sciatic nerve transection in wildtype mice.

Representative images of Citrullinated histone 3 (CitH3), neutrophils (Ly6G), and DAPI staining 2 d after a sciatic nerve transection in WT mice without (A and A’) or with (B and B’) the inhibitor protein-arginine deiminase type 4 (PAD4). The orange arrowhead indicates a Ly6G⁺ neutrophil, the yellow arrowheads indicate NET formation, while the white arrowhead indicates a Ly6G⁺CitH3⁺ neutrophil without the NET structure. The PAD4 inhibitor was administered daily via direct injection into the distal nerve after the injury. Scale bar is 20 μm. Quantification of neutrophils, neutrophil extracellular traps (NETs), and CD68 (macrophages) are in panels C, D, E, and F. ***p < 0.001. ****p < 0.0001. n = 7 per group. One slide per animal was analyzed. Representative images of LFB staining of sham and distal nerves from wildtype (WT) treated with either vehicle (G, I) or inhibitor of citrullination of histone 3 (H, J) 7 d post sciatic nerve injury (DPI). The PAD4 inhibitor was administered daily via direct injection into the distal nerve after the injury. Quantification of luxol fast blue (LFB) staining in WT mice sham injured or injured nerves with either vehicle treatment or inhibitor (K). The percent change over sham is displayed in panel L. Data is displayed as the mean ± standard error of the mean. ****p < 0.0001. Scale bar = 100 μm. n = 5 per group. Three separate regions were analyzed and averaged to create an average value per nerve per animal.

Figure 8.. Myelin clearance was not affected by MPO inhibitor.

The percent area stained was quantified for sham and injured nerves for both genotypes (A). The percent area change over sham is displayed in panel B. Representative images of luxol fast blue staining of sham and distal injured nerves from WT mice 7 d after a sciatic nerve transection and treatment with the MPO inhibitor (D, F) and vehicle (C, E). Data is displayed as mean ± standard error of the mean. Scale bar = 100 μm. ****p < 0.0001. n = 9 per group. Three separate regions were analyzed and averaged to create an average value per nerve per animal.

Figure 9.. Accumulation of neutrophils is not different between the crush or transection injuries in the WT or Cxcr2 KO mice.

Representative images of Ly6G and CD68 staining from distal nerves of littermate WT (A, C) and Cxcr2 KO mice (B, D) 2 d after sciatic nerve crush or transection. Scale bar = 100 μm. Neutrophils were stained with Ly6G, macrophages were stained with CD68, and nuclei with DAPI. Cell counts were used to quantify the number of cells per mm² for Ly6G (E), and CD68 (F) for both injuries for both genotypes. Representative images of Ly6G and Collagen Type 1 staining from distal nerves of littermate wildtype (WT) (G, I) and Cxcr2 KO mice (H, J) 2 d after sciatic nerve crush or transection. G’ is a section of A to show the collagen type 1 is staining of the epineurium between the branches of the sciatic nerve with a large aggregation of neutrophils in the parenchyma and touching the epineurium. Scale bar = 200 μm. Neutrophils were stained with an anti-Ly6G antibody, and the epineurium was stained with anti-collagen type 1 antibody. The percentage of neutrophils (Ly6G⁺DAPI⁺cells) associated with epineurium staining is displayed in panel K. The percentage of cells inside the parenchyma is displayed in panel L. Data is displayed as the mean, ± standard error of the mean. n = 5-6 per group. One slide per animal was analyzed. The dotted line indicates the crush site.

Figure 10.. Ccr2 KO mice undergo normal Wallerian degeneration after a crush injury, while Cxcr2 KO mice have reduced myelin clearance after a crush.

The quantification of luxol fast blue (LFB) staining (A, B). Representative images of LFB staining of WT (C, E), and Ccr2 KO mice (D, F) uninjured or sciatic crush injured nerves 7 d after the injury. Data is displayed as mean ± standard error of the mean. Scale bar = 100 μm. ns indicates not significantly different, ****p < 0.0001. n = 5 animals per group. The percent area stained was quantified for both genotypes and injuries (G). The percent area change over sham is displayed in panel (H). Representative images of luxol fast blue staining of sham and distal injured nerves from littermate WT (I, K) and Cxcr2 KO mice (J, L) 7 d after sciatic nerve crush. Data is displayed as the mean, ± standard error of the mean. Scale bar = 100 μm. ****p < 0.0001. n = 9 animals per group. Three separate regions were analyzed and averaged to create an average value per nerve per animal.