Platelet HMGB1 in Platelet-Rich Plasma Promotes Tendon Wound Healing

Platelet-rich plasma (PRP) is a widely used autologous treatment for tendon injuries in clinics, but clinical trials often produce conflicting results. Platelets (PLTs) are a major source of high mobility group box1 (HMGB1) that is gaining attention as a chemoattractant that can recruit stem cells to the wound area to enhance healing; however, the contribution of PLT HMGB1 in wounded tendon healing remains unexplored. This study investigated the effect of PLT HMGB1 within PRP to enhance healing in an acute patellar tendon injury model in PLT HMGB1 knockout (KO) mice and GFP mice. A window defect was created in the patellar tendons of both groups of mice, and wounds were treated with either saline, PRP isolated from PLT HMGB1 KO mice, or PRP isolated from GFP mice. Seven days post-treatment, animals were sacrificed and analyzed by gross inspection, histology, and immunostaining for characteristic signs of tendon healing and repair. Our results showed that in comparison to mice treated with PRP from PLT HMGB1-KO mice, wounds treated with PRP from GFP mice healed faster and exhibited a better organization in tendon structure. Mice treated with PRP from PLT HMGB1-KO mice produced tendon tissue with large premature wound areas and low cell densities. However, wounds of PLT HMGB1 KO mice showed better healing with PRP from HMGB1 KO mice compared to saline treatment. Moreover, wounds treated with PRP from GFP mice had increased extracellular HMGB1, decreased CD68, increased stem cell markers CD146 and CD73, and increased collagen III protein expression levels compared to those treated with PRP from PLT HMGB1 KO mice. Thus, PLT HMGB1 within PRP plays an important role in the healing of wounded tendon. Our findings also suggest that the efficacy of PRP treatment for tendon injuries in clinics may be affected by PLT HMGB1 within PRP preparations.


72
Tendon injuries to the Achilles and patellar tendons are prevalent in both occupational and 73 athletics populations. Overall, injured tendon healing is slow and yields an inferior quality of 74 tendon tissue that is prone to re-injury. Many therapeutic approaches including injection of 75 autologous platelet-rich plasma (PRP) have been devised to manage tendon injuries (1, 2). The  Although lacking nuclei, platelets are a rich source of high mobility group box-1 (HMGB1), a 85 highly conserved nuclear protein that is released by all cell types upon injury (9, 10). The 86 function of HMGB1 as an inflammatory molecule or chemoattractant is dependent upon its 87 redox state. When inside the cell, either in the nucleus or cytoplasm, HMGB1 is completely 88 reduced as fully reduced HMGB1 (frHMGB1). Once released to the extracellular matrix, 89 frHMGB1 is partially oxidized to disulfide HMGB1 (dsHMGB1) that is believed to initiate 90 inflammation (11). Research has suggested that during platelet activation, HMGB1 is 91 presented on the cell surface and released in significant amounts, with the current theory that 92 HMGB1 is fully reduced at this stage (12-15). Therefore, we hypothesized that platelet   109 Mice were anesthetized with isoflurane, and blood was drawn from the retro-orbital plexus into 110 anti-coagulated tubes. PRP was obtained by centrifugation at 500g for 10 min. Platelets were 111 first pelleted from PRP by centrifugation at 1,000g for 10 min, and were then resuspended in Isolated platelets were activated by adding 100 l of 10,000 U/ml bovine thrombin solution into 118 0.4 ml of 1 × 10 8 /ml platelet-ACD solution at room temperature for 30 min. The reaction mixture 119 was centrifuged at 1,000g for 10 min, and the supernatant was collected to determine the amount 120 of HMGB1 released from platelets using an HMGB1 ELISA kit according to the manufacturer's 121 protocol (Shino-Test Corporation, Tokyo, Japan). The pellet was re-suspended with 0.9% of 122 sodium chloride solution and reacted with a rabbit anti-mouse HMGB1 primary antibody for 3 123 hrs at room temperature (1:350, abcam, Cat. #ab18256), followed by a goat anti-rabbit secondary 124 antibody conjugated with Cy3 for 1hr at room temperature (1:500, Millipore, Cat. #AP132C).

126
Mouse tendon wound healing model 127 The effect of PLT HMGB1 on tendon wound healing was tested with a window defect created in 128 each patellar tendon (PT) of PLT HMGB1-KO mice and GFP mice using a 1 mm diameter 129 biopsy punch. Wounded PLT HMGB1-KO mice and GFP mice were divided into three groups 130 (3 mice/group). Group 1 mice were treated with 10 l of saline (Saline), group 2 mice were 131 treated with 8 l of PLT HMGB1-KO PRP and 2 l of bovine thrombin for PRP activation (KO-132 PRP), and group 3 mice were treated with lox 8 l of GFP-PRP and 2 l of bovine thrombin 133 (GFP-PRP). All mice were sacrificed at 7 days post-injury, and patellar tendons were harvested.

134
The effect of PLT HMGB1 on wounded tendon healing and cell migration was assessed by 135 histological analysis.

137
Histochemical staining on mouse tendon tissue sections 138 Tendon tissue sections were fixed with 4% paraformaldehyde for 20 min at room temperature, 139 and then washed three times with PBS. Slides were stained with H&E at room temperature 6 140 according to the standard protocols, washed with water 3 times, and dehydrated through 15%, 141 30%, 50%, 75%, 95% alcohol, and absolute alcohol for five minutes each. Finally, slides were 142 treated with xylene and mounted with resinous mounting medium. The staining results were 143 observed and imaged on a microscope (Nikon eclipse, TE2000-U).

145
Immunostaining on mouse tendon tissue sections 146 For immunostaining, the patellar tendons were dissected from the mice and were immediately    All results were obtained from 6 tendons (three mice) from each group and presented as the 187 mean ± SD. The statistical analyses were performed with an unpaired student t-test. When P < 188 0.05, the two groups in comparison were considered significantly different. 193 First, the expression of PLT HMGB1 was assessed for both transgenic lines, specifically Pf4-Cre 194 Hmgb1 fl/fl mice, referred to as PLT HMGB1-KO mice, and GFP mice. Platelets were isolated 195 and stained as described in the methods. Immunostaining showed that PLT HMGB1 was 196 decreased within PLT HMGB1-KO mice (Fig. 1A, B) compared to GFP mice (Fig. 1C, D).   206 Patellar tendon (PT) wounds of mice from each group were treated individually with PRP 207 generated from either PLT HMGB1-KO mice or from GFP mice. A saline treatment was used as compared to GFP mice (Fig. 2E, F). PLT HMGB1-KO mice treated with GFP-PRP (Fig. 2C) 214 exhibited slightly better healing compared to treatment with KO-PRP (Fig. 2B). However, PLT 215 HMGB1-KO mice treated with KO-PRP ( Fig. 2B) had better healing than a saline treatment 216 ( Fig. 2A, red arrow). These results indicated that PLT HMGB1 within PRP preparations is 217 required for promoting tendon wound healing by PRP.

298
Similar results were found with CD73 levels in wounded tendons (Fig. 7). Few CD73 + cells can 299 be seen within saline-treated tendons of PLT HMGB1-KO mice (Fig. 7A, B), compared to saline treated GFP mice (Fig. 7G, H). Overall, both PLT HMGB1-KO PRP and GFP-PRP were able to 301 increase the level of CD73 + cells but to different degrees. GFP-PRP treatment increased CD73 + 302 cells in both mouse lines (Fig. 7E, F, 7K, L), surpassing the effect of PLT HMGB1-KO PRP on 303 CD73 levels (Fig. 7C, D, 7I, J). Semi-quantification of CD73 staining supports these results 304 (Fig. 7M), with GFP-PRP treatment producing elevated CD73 + cells in both transgenic lines 305 with the highest levels in the treated tendons of GFP mice. Taken together, these results suggest 306 that HMGB1 ablation in PLTs can negatively affect tendon wound healing by decreasing stem 307 cell recruitment.  313 Collagen type III (Col III) has an important role in the healing process of tendon (24).