Protective Effect of the Human Epineural Patch Application after Sciatic Nerve Crush Injury Followed by Nerve Transection and End-to-End Repair

The experimental procedures involving animals were conducted in strict accordance with protocols approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Illinois at Chicago, which holds accreditation from the American Association for the Accreditation of Laboratory Animal Care (AAALAC). All animals were provided humane care in accordance with the Principles of Laboratory Animal Care established by the National Society for Medical Research, as well as the Guide for the Care and Use of Laboratory Animal Resources. A total of 36 athymic nude rats (Crl:NIH-Foxn1^rnu, Charles River Laboratories, Wilmington, MA, USA), each weighing between 150 g and 250 g, were selected for this study. The animals were housed in two per hooded cages at room temperature, maintained on a 14/10 light–dark schedule, and provided with water and irradiated rodent feed ad libitum.

The human sciatic nerves used in this experimental study were provided by the Musculoskeletal Transplant Foundation (MTF; Edison, NJ, USA) to create hEP. Transported under sterile conditions on dry ice, the nerves were stored upon arrival in a (Glacier® Ultralow Temperature Freezer – Thermo Fisher Scientific, Waltham, MA, USA) set at −86°C. The sciatic nerves were defrosted on a heating pad (T/Pump®, Gaymar Industries, Orchard Park, NY, USA) with warm water circulating at 38°C. Next, the hEP patches were created based on our well-established protocol, where 3–4 cm long sciatic nerve segments were further dissected, and the epineurium was separated from the fascicles under 20× magnification of the microsurgical microscope (Wild M-691, Leica Microsystems, Wetzlar, Germany) (Siemionow et al. 2011, 2017a, b, 2019, 2022; Kozłowska et al. 2022; Strojny et al. 2023). After dissection, each hEP was thoroughly inspected for any inadvertent damage, and the damaged samples were excluded from the study. Finally, the created hEP was divided into 2 cm × 1 cm patches and was submerged in saline until application into the sciatic nerve repair site following the CTR injury.

The hAM used in this study was obtained from the MTF, our longstanding collaborator. Both the AmnioBand® Viable and lyophilized AmnioBand Membrane (MTF Biologics, Edison, NJ, USA; 2 cm × 4 cm; UPC: 840045712908) were processed following rigorous quality control protocols to ensure biological integrity. Transported under sterile, controlled conditions on dry ice to maintain low temperatures, the membranes were immediately stored upon arrival in a Glacier® Ultralow Temperature Freezer at −86°C to preserve its viability and structural properties until further use.

Rats were anesthetized using 5% isoflurane (Terrell Isoflurane, Piramal Critical Care Inc., Bethlehem, PA, USA) in an induction chamber and maintained at 1.5%–2.5% using the Classic T3™ SurgiVet® Vaporizer (Smiths Medical ASD Inc., St. Paul, MI, USA). Buprenorphine SR (1.2 mg/kg) was administered subcutaneously 15 min prior to the incision for analgesia. Each rat was positioned on the left side, with the right hind limb shaved and hair removed using Nair™ (Church & Dwight Co., Inc., Ewing, NJ, USA). The surgical site was then cleansed with 5% povidone-iodine (Betadine®, Purdue Pharma L.P., Stamford, CO, USA), and the rats were placed on a warm water circulation heating pad (T/Pump®, Gaymar Industries). All procedures were performed aseptically under 20×–40× magnification of an operating microscope (Wild 691, Leica Microsystems). To expose the sciatic nerve, a 3 cm incision was made at the right gluteal region. Next, as previously described (Gudemez et al. 2002; Siemionow et al. 2011, 2017a, b, 2019, 2022; Kozłowska et al. 2022; Strojny et al. 2023), a standardized crush injury was applied to the sciatic nerve with a 5-inch straight mosquito hemostat (Walter Lorenz®, Jacksonville, FL, USA) for 5 min. The crushed segment of the sciatic nerve was resected with microsurgical scissors, followed by the standard end-to-end repair using 10-0 nylon sutures (Ethicon Inc., Raritan, NJ, USA). Then, the nerve repair site was wrapped with either the hAM or hEP or was left without coverage as a control. Two additional 10-0 nylon sutures were used to secure the wrapping materials (Figure 1). The muscle layers were closed with 4-0 interrupted vicryl sutures, and the skin with 5-0 monocryl sutures (Ethicon). An antibiotic cream (Neosporin, Johnson & Johnson, USA – New Brunswick, NJ, USA) was applied to the incision site. Postoperative monitoring was conducted for 24 h to ensure normal recovery. Nerve samples were collected at both 6 weeks and 12 weeks after CTR for histological analysis.

Fig 1.

Schematic illustration of the experimental design detailing the creation and application of the hEP as an innovative approach to enhance nerve regeneration following sciatic nerve crush injury, transection, and end-to-end repair. hEP, human epineural patch; hAM, human amniotic membrane.

Following the surgical procedure, each rat was kept in isolation with a protective collar to prevent self-inflicted wounds. After the initial 24-h period, the collar was removed, and the rats were returned to the original cages. Pain management was provided by administering buprenorphine (0.1 mg/kg) twice daily for the first 2 postoperative days. The surgical site was monitored daily for the first 14 days after surgery. Additionally, the overall wellbeing of the rats was assessed weekly by a board-certified veterinarian until study completion. Physical examinations of the animals assessed the signs of morbidity, changes in eating or drinking patterns, weight fluctuations, and mobility limitations. Additionally, rough hair or a hunched posture that could suggest that distress or pain was monitored closely.

The study included 36 athymic nude rats (Crl:NIH-Foxn1^rnu), which were randomly assigned to six experimental groups of 6 rats each, divided into the 6-week and 12-week follow-up studies (Table 1). Each group underwent a CTR procedure, including sciatic nerve crush injury, followed by nerve transection and standard end-to-end repair. In the 6-week study, Group 1 served as the control and did not receive protection at the nerve repair site; Group 2: hEP patch was applied to the nerve repair site; and Group 3: hAM patch was applied to the nerve repair site. In the 12-week study, Groups 4–6 underwent the same CTR procedures as Groups 1–3, and were evaluated over the 12-week follow-up period (Figure 2).

Fig 2.

Surgical technique of the CTR procedure and application of the created hEP and hAM patch. (a) Creation of the sciatic nerve crush injury. (b) Sciatic nerve crush site 0.5 mm long. (c) End-to-end repair after resection of the sciatic nerve crush site. (d) Sciatic nerve repair site wrapped with the hEP. (e) Sciatic nerve repair site wrapped with the hAM. hAM, human amniotic membrane; hEP, human epineural patch.

The animals were evaluated for functional recovery at specified time intervals: 1 week, 3 weeks, and 6 weeks in the 6-week study, and 1 week, 3 weeks, 6 weeks, 9 weeks, and 12 weeks in the 12-week study, following nerve CTR. Functional recovery assessments included the Toe-Spread test for motor recovery and the Pinprick test for sensory recovery. Samples were harvested at 6 weeks and 12 weeks post-surgery. The gastrocnemius muscle was harvested to assess muscle denervation atrophy using the standard gastrocnemius muscle index (GMI). Histomorphometric analysis of nerve samples harvested from the proximal, crushed, and distal sites quantified myelin thickness, axonal density, fiber diameter, and the percentage of myelinated nerve fibers.

The neuroregenerative potential of hEP was further evaluated through immunostaining, analyzing the expression of selected neurogenic (Glial fibrillary acidic protein [GFAP], Laminin B, nerve growth factor [NGF], S-100), angiogenic (vascular endothelial growth factor [VEGF], von Willebrand factor [vWF]), and immunogenic (human leukocyte antigen-DR [HLA-DR], human leukocyte antigen class I [HLA-I]) markers.

The motor nerve recovery of the right sciatic nerve of each rat was evaluated using the Toe-Spread test. With an uninjured limb, the rat extends and abducts the hind foot toes when it is held up by the tail. To access this, the Toe-Spread test was graded on a scale of 0–3 points, based on the following guidelines: no movement was given 0 points; any sign of toe movement was assigned 1 point; abduction of the toes was graded with 2 points; and full toe extension and abduction, representing a normal response, was assigned a score of 3 points. This test was administered at least three times per evaluation point for every rat within each group at the designated time marks. The Toe-Spread test was done alongside the Pinprick test.

Evaluation of the sensory nerve recovery was done using the Pinprick test. A pinching stimulus was applied to the skin of the rat’s right hind limb, extending from the toe to the knee joint by using Adson’s forceps (Walter Lorenz^®) until limb withdrawal and/or a vocal response to the stimulus occurred. The stimulus was applied at a specific pressure and amount of time while targeting the skin without harming the deeper tissues. Sensory recovery was graded from 0 to 3 in the following manner: a score of 0 indicated no sensation in the limb, a score of 1 represented a withdrawal response between the knee and the ankle, a score of 2 represented a withdrawal response between the ankle and the toes, and a score of 3 represented a withdrawal response of the toes themselves. This test was performed at least three times per evaluation point for every rat within each group at the designated time marks. Following the Pinprick test, rats were closely monitored for signs of ecchymosis, wounds, edema, or changes in skin color, none of which were observed.

Macroscopic evaluation of the sciatic nerve injury site was conducted following euthanasia of the rats at the designated 6-week and 12-week endpoints using SomnaSol™ (Henry Schein Inc., Melville, NY, USA). Following euthanasia, right sciatic nerve segments – protected with either no wrapping, hEP, or hAM – were carefully harvested for histomorphometric and immunofluorescence analyses. A 3 cm incision was made in the gluteal region of the right hind limb to expose the sciatic nerve and the applied patches. The harvested samples were then evaluated for adhesion to the surrounding tissues; signs of local inflammation; and the structural integrity, shape, and appearance of the patches. Additionally, assessments were made for the presence of fascicle-like structures within the hEP or hAM, signs of nerve atrophy distal to the hEP or hAM, and vascularization of the patches to determine the extent of healing and regeneration at the injury site.

The gastrocnemius muscles were harvested for macroscopic assessment, and muscle denervation atrophy was evaluated using the GMI at the 6-week and 12-week follow-ups. The wet weight of the gastrocnemius muscle was measured using an analytical balance Ohaus Precision Standard (Ohaus Corporation, USA – Parsippany, NJ, USA). The GMI was calculated as the ratio of the wet weight of the gastrocnemius muscle on the operated side to that of the contralateral gastrocnemius muscle on the unoperated side. Recovery from denervation atrophy was expressed as a percentage value, with a GMI of 100% indicating full recovery of the gastrocnemius muscle on the operated side.

The sciatic nerve samples from the proximal, crushed, and distal nerve segments with either No protection, hEP, or hAM application were excised, immersed in, and fixed with 2.5% glutaraldehyde. Post-fixation was performed using 4% aqueous osmium tetroxide according to the manufacturer’s protocol. For histological analysis, 1 µm thick cross-sections were stained with Toluidine Blue Solution (Thermo Fisher Scientific, Waltham, MA, USA). These sections were examined under a light microscope (Leica DM 5500B Automated Upright Microscope, RRID: SCR_020219, Leica Microsystems) using a high-magnification objective lens (100×) and captured with a digital camera (Leica DFC290 HD Color Digital FireWire Camera, Leica Microsystems). Images were analyzed using Image-Pro Plus (Media Cybernetics, Rockville, MD, USA). Six nonoverlapping fields were selected from each nerve section for quantification. Parameters such as myelin thickness (µm), axonal density (axons/µm²), fiber diameter (µm), and the percentage of myelinated nerve fibers (%) were measured using ImageJ software (Wayne Rasband, National Institutes of Health, Bethesda, MD, USA; RRID: SCR_003070), focusing exclusively on myelinated nerve fibers.

Immunofluorescent staining was used to assess the expression of selected neurogenic (S-100, NGF, GFAP, Laminin B), angiogenic (vWF, VEGF), and immunogenic (MHC class I and II) markers within nerve samples. The paraffin sections of the crush and distal nerve segments were mounted onto glass slides and underwent deparaffinization with xylene, followed by subsequent rinsing in ethanol and distilled water. The slides were then rinsed with Tris Buffered Saline (TBS) Tween 0.1% (TBS, Agilent Technologies, Inc., USA – Santa Clara, CA, USA) for 2 min on an orbital shaker, fixed with cold acetone for 8 min, and watched three times with TBS-Tween 0.1% for 5 min each on an orbital shaker. Following blocking with Goat serum at 10% for 1 h at 4°C, the nerve samples were incubated at 4°C overnight in dark with antibody diluent (1:50) and the following primary antibodies: S-100 (S100 4C4.9, RRID: AB_795376, Invitrogen, Waltham, MA, USA), NGF (Anti-NGF beta, RRID: AB_10856084, Bioss Antibodies, Woburn, MA, USA), GFAP (anti-GFAP, ab68428, RRID: AB_1209224, Abcam, Cambridge, UK), Laminin B (Laminin, RRID: AB_2133633, Invitrogen, USA – Waltham, MA, USA), vWF (VWF, RRID: AB_10642840, Proteintech, Rosemont, IL, USA), VEGF (anti-VEGF VG1, RRID: AB_10001947, Novus Biologicals, Englewood, CO, USA), MHC class I (HLA class I APC, RRID: AB_1557426, Proteintech, USA – Rosemont, IL, USA), and MHC class II (BD Pharmingen^™ Purified HLA-DR, DP, DQ, RRID: AB_395939, BD Biosciences, Franklin Lakes, NJ, USA). The primary antibodies were selected and combined to perform co-staining. Next, the samples were rinsed three times with TBS for 5 min each on an Orbital shaker. The secondary antibodies (IgG Alexa Fluor^™ 488, RRID: AB_2534069, Invitrogen, USA – Waltham, MA, USA) (IgG Alexa Fluor^® 594, ab150132, RRID: AB_2810222, Abcam, UK – Cambridge, UK) were diluted in antibody diluent (1:500). The secondary antibodies application was performed in dark condition and followed by a 1-h incubation at 4°C in a humidity chamber. The slides were rinsed in TBS three times for 5 min each and mounted with Fluoromount-G^® (Southern Biotech, Birmingham, AL, USA). The slides were assessed using an upright confocal microscope (Leica TCS SP2 Upright Confocal Microscope, RRID SCR_020231, Leica Microsystems) with a digital camera (QImaging® Retiga-2000R Charge Coupled Device, QImaging, British Columbia, Canada) and ImagePro Plus (RRID SCR_016879) – Media Cybernetics, USA – Rockville, MD, USA software. Evaluation of immunofluorescence was scored based on signal intensity as follows: 0 = no staining, 1 = weak, 2 = moderate, and 3 = strong.

Statistical analyses were performed using GraphPad Prism software (RRID: SCR_002798, GraphPad Software, San Diego, CA, USA). One- or two-way ANOVA tests were conducted to assess statistical significance, followed by Tukey’s post hoc test for multiple group comparisons. Data from in vitro experiments, which included six randomly selected nerve samples, are presented as mean ± standard error of the mean (SEM). Statistical significance was defined as p < 0.05, with significance levels indicated by asterisks: ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, and ^****p < 0.0001.

At the 6-week follow-up, no significant motor response was observed at the 1-week time point across all groups, with the No protection, hEP, and hAM groups showing minimal or no measurable response (0.000 ± 0.000, p > 0.05). By 3 weeks, the hEP improved motor function compared with the No protection and hAM groups (0.778 ± 0.319, p < 0.05 vs. 0.222 ± 0.319 and 0.000 ± 0.000, respectively). At 6 weeks, the hEP group exhibited the highest motor response, significantly outperforming the No protection and hAM groups (2.165 ± 0.320, p < 0.05 vs. 1.167 ± 0.320 and 1.390 ± 0.320, respectively; Figure 3a).

Fig 3.

Functional assessment of sciatic nerve regeneration was done using Toe-Spread and Pinprick tests in the 6-week study at 1, 3, and 6-week study points. (a) Motor recovery (Toe-Spread test) revealed no significant response at 1 week, but the hEP group showed significant improvement by 3 weeks and the highest motor response at 6 weeks. (b) Sensory recovery (Pinprick test) showed significantly better outcomes in the hEP group at 3 weeks and 6 weeks compared with the No protection and hAM groups, with the hEP group demonstrating the highest response at 6 weeks. Data are presented as mean ± SEM, with significance indicated as *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. hAM, human amniotic membrane; hEP, human epineural patch; SEM, standard error of the mean.

At the 12-week follow-up, all groups showed minimal or no measurable response at the 1-week and 3-week time points, with no significant differences between No protection, hEP, and hAM (0.000 ± 0.000 for all groups, p > 0.05). By 6 weeks, hEP exhibited a significantly higher motor response compared with No Protection (1.777 ± 0.262 vs. 0.833 ± 0.262, p < 0.01), though the difference between hEP and hAM did not reach statistical significance (1.777 ± 0.262 vs. 1.167 ± 0.262, p > 0.05). At 9 weeks, hEP continued to show superior motor recovery compared with No Protection (2.278 ± 0.262 vs. 1.112 ± 0.262, p < 0.0001), while the difference between hEP and hAM remained nonsignificant (2.278 ± 0.262 vs. 1.722 ± 0.262, p > 0.05). By 12 weeks, hEP demonstrated the highest level of motor function recovery (2.500 ± 0.262), significantly outperforming No Protection (1.667 ± 0.262, p < 0.01), while no significant difference was observed between hEP and hAM (2.057 ± 0.262, p > 0.05) (Figure 4a).

Fig 4.

Functional assessment of sciatic nerve regeneration was done using Toe-Spread and Pinprick tests in the 12-week study at 1, 3, 6, and 12-week study points. (a) Motor recovery (Toe-Spread test) showed minimal response across all groups at 3 weeks. By 6 weeks, the hEP group maintained significantly higher response levels than the No protection group. By 9 weeks, the hEP group outperformed both groups, though statistical differences were observed for No protection. At 12 weeks, the hEP group achieved the highest recovery, significantly outperforming the No protection group. (b) Sensory recovery (Pinprick test) revealed that the hEP group had significantly higher responses at 6 weeks and 9 weeks than the other groups, with comparable recovery observed across all groups by 12 weeks. Data are mean ± SEM, with significance marked as *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. hAM, human amniotic membrane; hEP, human epineural patch; SEM, standard error of the mean.

At the 6-week follow-up, no differences were observed between the No protection, hEP, and hAM groups. By 3 weeks, the hEP group showed a significantly better recovery compared with the hAM group (1.667 ± 0.201 vs. 1.112 ± 0.201, p < 0.05), but no significant difference was found between hEP and No protection (p > 0.05). At 6 weeks, the hEP group demonstrated the highest recovery, significantly outperforming both the No protection (2.945 ± 0.201 vs. 2.167 ± 0.201, p < 0.01) and hAM groups (2.945 ± 0.201 vs. 2.332 ± 0.201, p < 0.05; Figure 3b).

At the 12-week follow-up, the hEP group continued to show superior recovery outcomes. At 6 weeks, the hEP group demonstrated significantly better recovery compared with No protection (2.555 ± 0.207 vs. 1.723 ± 0.207, p < 0.001) and hAM (2.555 ± 0.207 vs. 2.000 ± 0.207, p < 0.05). By 9 weeks, the hEP group continued to show superior recovery, significantly outperforming No protection (3.000 ± 0.207 vs. 2.500 ± 0.207, p < 0.05). At the 12-week mark, recovery outcomes were comparable across all groups, with no statistically significant differences (3.000 ± 0.207 for all, p > 0.05) (Figure 4b).

During nerve sample harvesting at 6-week and 12-week study end-points, the evaluation of the repair site after the CTR procedure revealed no signs of adhesion, fibrosis, or local inflammatory response around the hEP application. Remarkably, fascicle-like structures were visible at the site of the CTR procedure. Specifically, at the application site, the hEP demonstrated preserved structural integrity and morphology, displaying adequate vascularization. In contrast, the hAM was difficult to detect at the site of application during the CTR procedure upon macroscopic evaluation at both the 6-week and 12-week following nerve repair.

At the 6-week follow-up, the hEP group had higher GMI values compared with the hAM and No protection groups, though these differences were not statistically significant (Figure 5a).

Fig 5.

Evaluation of muscle denervation atrophy by GMI. (a) At 6 weeks, no significant differences in the GMI were observed between groups. (b) In contrast, at 12 weeks, the hEP group exhibited a significantly higher GMI compared with the No protection group. Data are presented as mean ± SEM. Statistical significance was determined using a one-way ANOVA test for group comparisons, with significance levels as follows: *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. GMI, gastrocnemius muscle index; hAM, human amniotic membrane; hEP, human epineural patch; SEM, standard error of the mean.

At the 12-week follow-up, the hEP group showed a significant improvement with a GMI value of 0.955 ± 0.014, which was notably higher than the no protective wrapping group’s GMI of 0.857 ± 0.030 (p < 0.05). Additionally, the hEP group had higher GMI values compared with the hAM group, which had a GMI of 0.903 ± 0.015; however, this difference was not statistically significant (p > 0.05) (Figure 5b).

The efficacy of hEP was evaluated by assessing myelin thickness, fiber diameter, percentage of myelinated fibers, and axonal density at 6 weeks (Figure 6) and 12 weeks (Figure 7) following nerve CTR injury.

Fig 6.

Histological assessment of the proximal, crushed, and distal parts of the injured segment at 6 weeks after sciatic nerve crush injury followed by nerve transection and end-to-end repair. (a) Proximal myelin thickness, (b) Crushed myelin thickness, (c) Distal myelin thickness, (d) Proximal fiber diameter, (e) Crushed fiber diameter, (f) Distal fiber diameter, (g) Percentage of proximal myelinated fibers, (h) Percentage of crushed myelinated fibers, (i) Percentage of distal myelinated fibers, (j) Proximal axonal density, (k) Crushed axonal density, and (l) Distal axonal density. The data are presented as mean ± SEM and statistical significance was determined using a one-way ANOVA test for group comparisons, with significance levels denoted as *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. hAM, human amniotic membrane; hEP, human epineural patch; SEM, standard error of the mean.

Fig 7.

Histological assessment of the proximal, crushed, and distal parts of the injured segment at 12 weeks after sciatic nerve crush injury followed by nerve transection and end-to-end repair. (a) Proximal myelin thickness, (b) Crushed myelin thickness, (c) Distal myelin thickness, (d) Proximal fiber diameter, (e) Crushed fiber diameter, (f) Distal fiber diameter, (g) Percentage of proximal myelinated fibers, (h) Percentage of crushed myelinated fibers, (i) Percentage of distal myelinated fibers, (j) Proximal axonal density, (k) Crushed axonal density, and (l) Distal axonal density. The data are presented as mean ± SEM and statistical significance was determined using a one-way ANOVA test for group comparisons, with significance levels denoted as *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. hAM, human amniotic membrane; hEP, human epineural patch; SEM, standard error of the mean.

To evaluate the neuroregenerative potential of the hEP, selected neurogenic (GFAP, Laminin B, NGF, S-100), angiogenic (VEGF, vWF) and immunogenic (HLA-DR, HLA-I) markers were assessed following the nerve CTR injury model (Figures 8 and 9).

Fig 8.

Expression evaluation of selected neurogenic, angiogenic, and immunogenic markers assessed at the crush injury sites and distal ends by immunofluorescence staining at 6 weeks following sciatic nerve crush injury, transection, and repair. (a) GFAP, (b) Laminin B, (c) NGF, (d) S-100, (e) VEGF, (f) vWF, (g) HLA-DR, and (h) HLA-I. Data were presented as mean ± SEM. A two-way ANOVA test for group comparison was used to define statistical significance at *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. GFAP, glial fibrillary acidic protein; HLA-DR, human leukocyte antigen-DR; HLA-I, human leukocyte antigen class I; NGF, nerve growth factor; SEM, standard error of the mean; VEGF, vascular endothelial growth factor; vWF, von Willebrand factor.

Fig 9.

Expression evaluation of selected neurogenic, angiogenic, and immunogenic markers assessed at the crush injury sites and distal ends by immunofluorescence staining at 12 weeks following sciatic nerve crush injury, transection, and repair. (a) GFAP, (b) Laminin B, (c) NGF, (d) S-100, (e) VEGF, (f) vWF, (g) HLA-DR, and (h) HLA-I. Data were presented as mean ± SEM. A two-way ANOVA test for group comparison was used to define statistical significance at *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. GFAP, glial fibrillary acidic protein; HLA-DR, human leukocyte antigen-DR; HLA-I, human leukocyte antigen class I; NGF, nerve growth factor; SEM, standard error of the mean; VEGF, vascular endothelial growth factor; vWF, von Willebrand factor.