Amniotic membrane hydrogel as novel injectable platform in combination with metformin for treatment of sciatic nerve injury

Abstract

Peripheral nerve tissue engineering is a field that uses cells, growth factors and biological scaffold material to provide a nutritional and physical support in the repair of nerve injuries. The specific properties of injectable human amniotic membrane-derived hydrogel including growth factors as well as anti-inflammatory and neuroprotective agents make it an ideal tool for nerve tissue repair, and metformin may also aid in nerve regeneration. The aim of this study was to investigate the effects of hydrogel derived from amniotic membrane (AM) along with metformin (MET) administration in the repair of sciatic nerve injury in male rats. We randomly divided 60 male rats into five groups. A control and four sciatic nerve compression groups including model; hydrogel; metformin and mix which received hydrogel and metformin. The recovery rate was assessed by Sciatic Functional Index (SFI), Static Sciatic Index (SSI) and von-frey test. Conduction velocity of the sciatic nerve was measured by Electrophysiological studies, and histological evaluations were performed 14 days after injury. SFI, SSI, latency time, remyelination rate and the expression of NF-200 and S-100β improved in hydrogel group. Response to mechanical stimulus, myelin density, axonal regeneration, and myelin sheath reconstruction improved in the mix group. The gastrocnemius muscle index was significantly reduced in the experimental groups while collagen fibers increased in these groups. These findings suggest that injection of hydrogel derived from decellularized amniotic membrane into the epineurium can be promoted reconstruction of peripheral nerve injury and improved functional nerve recovery. Also, metformin administration can reinforce the therapeutic effect of the hydrogel.

Introduction

Peripheral nerve injury can lead to different types of functional defeat or permanent impairment. It is believed that the formation of scar and neural fibrosis following nerve injury causes disruption of normal function, malformation, and hinder growth. ¹ Scar formation at the site of injury in a peripheral nerve can act as a mechanical barrier to axonal sprouting and block axonal repair, which can have negative effects on functional recovery. ² Various researchers have shown the positive impacts of scar-suppressing drugs such as triamcinolone acetonide, ³ methylprednisolone-acetate, cis-Hydroxyproline, ⁴ dexamethasone ⁵ , and aprotinin ⁶ on nerve damage, but few of them have achieved acceptable functional recovery and nerve repair remains a challenge in medicine.

Tissue engineering and regenerative medicine (TERM) are widely suggested as a suitable alternative to nerve repair, but several challenges remain before routine clinical application. ⁷ One of the most important challenges in tissue engineering is the development of three-dimensional (3D) scaffolds that can support the functional differentiation, survival, proliferation, and targeting of transplanted cells as well as sprouting of injured axons. Based on the origin, scaffolds can be classified into artificial and natural types. ⁸

Human amniotic membrane (hAM) has a long history as a medical biomaterial and has been widely used in various applications including neural tube construction, skin burns and dermatology, corneal recreation, and ocular surface repair.^9–12 It has a range of active factors, like polymorphic collagen, vascular endothelial growth factor, epidermal growth factor, stromal cell-derived factor, and transforming growth factor-β.^10,13,14 Some studies have demonstrated that biodegradable and amnion bridging channels can considerably improve nerve restoration and functional improvement.^15–17 Hydrogels derived from decellularized extracellular matrixes (dECMs) got from different organs or tissues are commonly applied for TERM applications because they are rich in growth factors and protein elements and have great biocompatibility.^18,19 Hydrogels are polymeric materials that can absorb large quantities of water or biological fluids, and maintain a distinct three-dimensional structure due to chemical or physical cross-linking of individual polymer chains.^20,21

The use of natural materials, particularly those derived from animals, presents challenges such as safety concerns, variability, and ethical issues, limiting their applicability in widespread clinical contexts. In this regard, we propose the amniotic membrane (AM) as a promising source of human proteins for developing a novel class of human-derived hydrogels. ²²

Hydrogels derived from human AM can stimulate intrinsic tissue repair without necessitating supplementation with exogenous bioactive molecules, such as growth factors or peptides. These hydrogels exhibit mechanical properties that support and guide axonal regeneration across injury sites, ultimately aiming to restore neuronal networks. Furthermore, the biochemical composition and biocompatibility of AM-derived hydrogels enhance cell adhesion, proliferation, and differentiation – key processes for effective neural recovery. Additionally, these hydrogels promote favorable in vivo tissue remodeling and regeneration, driven by extracellular matrix (ECM) degradation, which facilitates the recruitment of endogenous stem and progenitor cells. For clinical applications, thermosensitive injectable hydrogels with in situ gelling capabilities are particularly advantageous, as they can conform to irregular lesion geometries while minimizing tissue damage during administration. Thus, this approach leverages human-sourced materials without requiring new crosslinking strategies or the inclusion of growth factors and polypeptides compare to other hydrogels.^22,23

However, some appropriate potential of hAM-derived gels in the repair of nerve injuries is yet to be evaluated.

Metformin is an anti-hyperglycemia medication prescribed since the 1960s for the treatment of type 2 diabetes and other metabolic syndromes. ²⁴ The benefits of metformin extend beyond its ability to lower glucose levels. It has been shown to have various therapeutic benefits in central nervous system (CNS) disorders such as ischemic brain injury, Huntington’s disease, and Parkinson’s disease.^25–27 Recent studies have reported that MET also has anti-inflammatory and neuroprotective effects in central and peripheral nervous systems injuries.^28,29 These effects can help suppress scar formation in various contexts, such as CNS and spinal cord injury.

Therefore, we propose hAM-derived hydrogels as a means to bridge lesion cavities, serving as a temporary inductive niche that is fully degraded over time and subsequently replaced with anatomically appropriate and functional tissue, rather than forming scar tissue. Since, it is not clear whether co-treatment of nerve injuries with hAM-derived hydrogel and MET can repair nerve injury and improve functional recovery. Therefore, we aimed to evaluate the recovery of an animal model of injured sciatic nerve using injectable hAM-derived hydrogel as well as metformin.

Methods and materials

Amniotic membrane preparation

We obtained full-term placenta from healthy pregnant volunteer women; The placenta was collected following cesarean section in Afzalipour hospital, Kerman Iran. The amniotic membrane (AM) was separated from the underlying chorion after washing with sterile PBS containing 50 mg/ml penicillin and streptomycin. ³⁰ AM was frozen at −80°C for an hour × 3 times for half an hour, and finally kept in a nitrogen tank for 2 min. Frozen AM was divided into pieces and treated with 0.05% trypsin-EDTA in PBS, ³¹ followed by gentle mechanical scraping to produce decellularized amniotic membrane (dAM). After washing with deionized water, the dAM was lyophilized for 48 h and stored at −80°C until needed. Hoechst ³² and hematoxylin and eosin (H&E) ³³ were used to confirm successful decellularization. Masson’s trichrome staining ³³ and Alician blue staining ³⁴ were performed to evaluate the presence of collagen and glycosaminoglycans in dAM, respectively.

dAM-gel preparation and characterization

The lyophilized dAM was first powdered. Then, it was filtered using a 250 μM plastic mesh, and as a result, a suitable powder of dAM extra-cellular matrix was obtained. Ten milligram of powder was dissolved in 1 ml of 0.01 M HCl solution containing 1 mg/ml of pepsin. The solution was stirred at 4°C for 24 h, exposed to UV light to sterilize and was neutralized at 4°C with 1 M NaOH. The viscose dAM-pre-gel solution with a concentration of 5 mg/ml was used for experiments. The pre-gel solution was incubated at 37°C for 30 min to enable hydrogel formation. Once gelation was confirmed, the pre-gel was injected into the epineurium of the injured rats. ³⁵

Development of sciatic nerve injury model and interventional treatments

We used male Wistar rats (n = 12 per group, mean weight = 252 ± 15 g, approximately 8 weeks old) to induce sciatic nerve injury. Ketamine (80 mg/kg body weight) and xylazine (20 mg/kg body weight) ³⁶ were intraperitoneally injected to anesthetize animals. The posterior surface of the right thigh was incised, then the sciatic nerve trunk was exposed and compressed with constant pressure for 2 min using a smooth hemostat. ³⁷

Study design and experimental groups

Animals were randomly assigned into 5 groups of 12 animals. A control group received no intervention while a model group underwent sciatic nerve compression without any other intervention. The animals in the hydrogel group, metformin group and hydrogel + metformin group (Mix group) received hydrogel in the epineurium of the injured sciatic nerve, received 200 mg/kg metformin intraperitoneally ³⁸ or both (Mix group), respectively. Following sciatic nerve injury, various treatments were performed on the animals. In the hydrogel group, the pre-gel solution was carefully injected into epineurium. In the MTF group 200 mg/kg metformin was daily injected intraperitoneally. In the Mix group both MTF and hydrogel was administered as described above. In the model, hydrogel and Mix groups the injury site was washed with normal saline, the muscles were returned back to their position, and the skin was sutured with 4-0 nylon sutures. After that the surgical site was washed with an antiseptic solution to reduce the risk of infection. Once the surgery was finished, the rats were supervised during their recovery. They were warmed up using a heat lamp to maintain their body temperature and were allowed to drink and eat freely. This post-operative care was performed for the well-being and successful recovery of the animals.

Sciatic functional index (SFI) calculation

Before injury and 2, 9, and 14 days after the sciatic nerve injury the sciatic functional index was calculated using walking track and footprint analysis. The hind paws of animals were dyed with ink and permitted to walk on a narrow track covered by a white paper.

Following factors were used for SFI calculation

• Print Length (PL): The length between the heel and the middle toe.

• Toe Spread (TS): The length between the first toe and the last toe.

• Intermediate Toe Spread (ITS): The length between the second toe and the fourth toe.

• E: experimental.

• N: normal.

The SFI was calculated by the following equation ²⁸ :

$$\begin{array}{c}\mathrm{SFI}=\left[-38.3\left(\mathrm{EPL-NPL}\right)/\mathrm{NPL}\right)]\\ +[109.5\left(\mathrm{ETS-NTS}\right)/\mathrm{NTS})]\\ +[13.3\left(\mathrm{EITS-NITS}\right)/\mathrm{NITS})]-8.8\end{array}$$

In the given formula, the experimental side is represented by the variable E, while the normal side is represented by the variable N. The value of 0 for E represents normal function, while the value of −100 represents complete disorder.

Static sciatic index (SSI) calculation

This is a simple and quick method for evaluating peripheral nerves functional recovery in rats with accurate precise outcomes. Static parameters are used for the SSI evaluation and does not consider the print length parameter (PL). The formula for calculating the SSI is ³⁹ :

$$\mathrm{SSI}=\left(108.44\times \mathrm{TSF}\right)+\left(31.85\times \mathrm{ITSF}\right)-5.49$$

Where:

• TSF represents the static toe spread parameter, calculated as (ETS−NTS)/NTS

• ITSF represents the static intermediate toe spread parameter, calculated as (EIT−NIT)/NIT

A normal SSI score is 0, while a score of −100 represents complete impairment.

Pain behavioral quantification

Mechanical sensitivity to non-painful stimuli of the hind paws was measured by applying von Frey filaments of 2, 4, 6, 8, 10, 15, 26, 60 g. Rats were put on a metal mesh plate and in a plastic enclosure with a height of 25 cm and dimensions of 20 cm × 18 cm. Thirty minutes later, rat gets used to the recent conditions and exploration of the new environment was stopped. Each filament was applied to the experimental hind paw three times for about 2–3 s with an interval of 5 s. If the animal withdrew its foot from the stimulus two out of three times when the filament was applied (positive response), that filament was considered as the mechanical threshold of pain, and otherwise, after 5 min, the stimulus was applied with a higher intensity. The lowest stimulation intensity in grams that could produce two positive responses out of a total of three stimulations was considered as a positive response. Answers that were due to movement and other natural behaviors of the animal were not considered. If the rat did not react to 26 g filament, it was considered as reaction threshold. ⁴⁰

Electrophysiological evaluation

Electrophysiological studies were performed 14 days post injury (dpi) to investigate the latency time (in ms) and compound muscle action potential (in mv). To prevent animals’ discomfort, the animals were given an i.p. injection of ketamine/xylazine for light anesthesia, then the backs of the rats was shaved. During the experiment, the body temperature of the animals was monitored by a probe placed in the anal canal. A bipolar stimulating electrode (1.5 V) was positioned in the proximal area of the grafts and a second electrode was placed on the gastrocnemius muscle (frequency of 20 Hz). The elicited action potential, in response to the impulse, was reported and then the latency and amplitude were calculated. ⁴¹

Gastrocnemius muscle wet weight and Masson’s trichrome staining

The study involved the following steps and measurements:

• 1. Muscle harvesting and wet weight recording:

•  ● Gastrocnemius muscles (n = 12) of hind limbs of rats were harvested bilaterally

•  ● The wet weight of the muscles was measured by a digital balance

• 2. Wet weight improvement ratio calculation ²⁸ :

•  ● Wet weight recovery rate was assessed using the following formula:

Wet weight improvement ratio= (Muscle wet weight on the experimental Side/ Muscle wet weight on the normal side) ×100

• 3. Muscle Fixation and Sectioning:

•  ● 10% formaldehyde was used to fix the harvested muscles, followed by dehydration, embedding in paraffin, and finally sectioning with a microtome at 5 µm thickness.

•  ● The 5 µm slices were stained with Masson’s trichrome ⁴² and the slides were examined under a light microscope with ×400 magnification.

•  ● Image-Pro Plus 6.0 software was used for evaluating muscle fiber regions

Immunohistochemistry

Grafts representing all experimental requirements were harvested on dpi 14. The sciatic nerve was then cross-sectioned 5 mm below and above the injury site. Paraffin embedded sciatic nerve tissues were sectioned at 5 µm thickness and immunostaining was performed with anti-NF-200 (Santa cruz, 1:100) and anti-S100 (Abcam, 1:100) primary antibodies at 4°C overnight, followed by incubation with DB detection kit - rabbit/mouse dual, HRP/DAB (Biotechnology) in order to visualize the nerve fibers and myelin sheet, respectively. ⁴¹ Expression of these proteins were quantitated by image analyzer Photoshop Software CS6.

Statistical analysis

The results were expressed as means ± SD. Statistical analyses were performed using GraphPad Prism (version 8.4.3) software. Statistical differences between groups were analyzed by one-way analysis of variance (ANOVA) for parametric data followed by Tukey’s Post-hoc test for comparisons between groups. For nonparametric assessments, statistical analysis was performed by Kruskal-Wallis followed by the Mann-Whitney U test.

Results

Body weight measurement

Body weight was taken before surgery and before sacrificing the animals. The average BW of animals at the beginning of the study did not show any significant difference in various groups and was equal to 254 ± 20 g (Figure 1(a)). Also, at the end of the study, the average BW of animals in different groups was comparable between the groups (Figure 1(b)).

Figure 1.

Body weight index: (a) initial body weight and (b) final body weight.

SFI outcomes

Footprint evaluation was performed to assess the functional recovery after sciatic nerve compression at dpi 2, 9, and 14. The average SFI at dpi 2 and 9 did not show significant difference between different experimental groups, while SFI in the experimental groups indicated a significant difference (p < 0.001) compared to the control group (Figure 2(a) and (b)). At dpi 14 the average SFI in the hydrogel groups increased significantly compared to the model group (p < 0.001; Figure 2(c)). These data confirm the beneficial effects of the hydrogel in improving motor function in animals.

Figure 2.

Sciatic Functional Index (SFI). Effects of concurrent administration of metformin and amniotic membrane derived hydrogel on sciatic functional index at days 2(a), 9(b), and 14(c) after sciatic nerve compression injury. Data were expressed as mean ± SD (n = 12), ***p ˂ 0.001 versus control group. ^#p < 0.05 versus model group.

SSI outcomes

The changes in the SSI on days 2, 9, and 14 were similar to the changes observed in the SFI, indicating that compression of the sciatic nerve for 2 min can cause deficits in animal motor function. The average SSI on day 2 and day 9 after injury did not demonstrate a significant difference between various experimental groups, while SSI in the experimental groups indicated a significant difference compared to the control group (p < 0.001; Figure 3(a) and (b)). The same pattern was correct for day 14, at which time SSI was significantly lower the control group (p < 0.001), but SSI in the hydrogel group was insignificantly higher in the model group (Figure 3(c)).

Figure 3.

Static Sciatic Index (SSI). Effects of co-administration of metformin and amniotic membrane derived hydrogel on static sciatic index at days 2 (a), 9 (b), and 14 (c) after sciatic compression injury in rats. Data were expressed as mean ± SD (n = 12), ***p ˂ 0.001 versus the control group.

The von Frey sensory evaluation

In order to check mechanical sensitivity to non-painful stimuli (mechanical allodynia index), von Frey filaments (2, 4, 6, 8, 10, 15, 26, and 60 g) were used on days 7 and 13 after surgery, and the results were presented in Figure 4(a) and (b). The index of mechanical allodynia increased significantly in the model group compared to the control group (p < 0.001). This index was not significantly different in hydrogel and metformin groups compared to the model group; while, it was significantly higher in the mix group compared with the model group.

Figure 4.

Von frey test. Investigating the mechanical sensitivity to non-painful stimuli on days 7 (a) and 13 (b). Data were expressed as mean ± SD (n = 12), **p < 0.01 and ***p ˂ 0.001 versus the control group. ^#p < 0.05 and ^##p < 0.01 versus model group.

Electrophysiological results

At day 14, latency time and compound muscle action potentials (CMAP) were determined in each group in order to investigate the function of the sciatic nerve and the posterior compartment of leg. Latency index values showed that there was a significant difference between the experimental groups compared to the control group (p < 0.001). Figure 5(a)show that in the model group, the latency time to the electrical stimulus was significantly increased compared to the control group (p < 0.01). In the hydrogel group (p < 0.05) and the mix group the latency index was significantly (p < 0.05) reduced compared to the model group. There was no significant difference between the hydrogel and mix group. Treatment with metformin had no significant effect on the latency time (Figure 5(a)).

Figure 5.

Electrophysiological evaluations at day 14. (a) Comparison of mean latency time. (b) Comparison of mean Compound Muscle Action Potential (CMAP)/nerve conduction velocity. Data were expressed as mean ± SD (n = 12), **p < 0.01 and ***p ˂ 0.001 versus control the group. #p < 0.05 versus model group.

Comparing the average CMAP among the studied groups with ANOVA test showed that there is a significant (p < 0.001) difference between the experimental groups compared to the control group (Figure 5(b)). As shown in Figure 5(b), CMAP in the model group was significantly (p < 0.001) reduced compared to the control group. Although different therapeutic interventions increased CMAP in the injured sciatic nerve groups, these increases were not statistically significant.

Gastrocnemius muscle wet weight and Masson’s trichrome staining

The gastrocnemius muscle recovery from atrophy was evaluated by calculating the wet weight of the gastrocnemius muscle on the affected side to the intact side 14 days after the start of the experiments (Figure 6(a)). Compared to the control group, the muscle mass ratio in the model, hydrogel, metformin, and mix groups was significantly lower (p ˂ 0.001), however, among the treated animals the weight of gastrocnemius muscle in the hydrogel group was insignificantly higher than other treated animals (Figure 6(b)). Masson’s trichrome staining was performed to reveal the collagen content of the samples (Figure 7(a)).

Figure 6.

Gastrocnemius Muscle Index. The gastrocnemius muscles of both sides (operated R and unoperated L) are excised (a) and weighed (b) at day 14. Scale bar = 10 mm. Data were expressed as mean ± SD (n = 12), ***p ˂ 0.001 versus the control group.

Figure 7.

Histological evaluation of the gastrocnemius at day 14. (a) Masson’s trichrome staining of transverse sections from control and experimental groups. Arrows were pointed to collagen fibers. Scale bar = 50 μM. (b) Collagen fibers density. Data were expressed as mean ± SD (n = 12), ***p ˂ 0.001 versus the control group.

The quantitative results of transverse sections of gastrocnemius muscle stained with Masson’s trichrome indicated that the collagen fiber area was significantly increased in the experimental groups compared with the control group (p < 0.001); While there was no significant difference between the collagen content in the hydrogel, metformin and mix groups compared with the model group (Figure 7(b)).

Histological evaluations

Luxol Fast Blue (LFB) staining results indicated remyelination at the injury site (Figure 8(a)). Quantitative evaluations showed that myelin density in the experimental groups has decreased significantly compared to the control group (p < 0.001). Myelin regeneration in the hydrogel (p < 0.05) and mix (p < 0.001) groups was significantly higher than the model group (Figure 8(b)).

Figure 8.

Histological evaluation of the sciatic nerve at day 14. (a) Longitudinal sections from control and experimental groups’ injured nerve region were stained by Luxol Fast Blue. Scale bar = 50 μm (b) Remyelination rate. Data were expressed as mean ± SD (n = 12), ***p ˂ 0.001 versus the control group. ^#p ˂ 0.05 and ^###p < 0.001 versus model group.

Immunohistochemistry results

In order to investigate the process of nerve repair and its related mechanisms, 14 days after sciatic nerve injury, the expression of S100 and NF200 proteins in tissue sections of sciatic nerve was evaluated by immunohistochemistry. As shown in Figure 9(a) and (b) the expression of S100 marker in the model group was significantly reduced compared to the control group (p < 0.001). While compared to the model group, the expression of S100 in all therapeutic groups (hydrogel, metformin and mix) had increased significantly (p < 0.001), and there was no significant difference between the experimental groups in this index. Nevertheless, expression of S100 in the experimental groups was still significantly lower than the control group (p < 0.001).

Figure 9.

Immunohistochemistry staining with anti-S100 at day 14. (a) Longitudinal sections of control and experimental groups’ injured nerve region were stained by the anti-S100 antibody. Scale bar = 50 μm (b) Schwann cell migration rate/ Myelin sheath regeneration rate. Data were expressed as mean ± SD (n = 12), ***p ˂ 0.001 versus the control group. ^###p ˂ 0.001 versus model group.

The expression results of NF200 marker are presented in Figure 10(a) and (b). Damage to the sciatic nerve caused a significant decrease in the expression of NF200 in the model group compared with the control group (p < 0.001). However, NF200 expression significantly increased in the mix and hydrogel groups compared to the model group (P < 0.05). Also, there was no significant difference between the hydrogel and mix groups (p = 0.78).

Figure 10.

Immunohistochemistry staining with anti-NF200 at day 14. (a) Longitudinal sections of control and experimental groups’ injured nerve region were stained by the anti-NF200 antibody. Scale bar = 50 μm (b) Axonal regeneration rate. Data were expressed as mean ± SD (n = 12), ***p ˂ 0.001 versus the control group. ^#p ˂ 0.05 and ^##p < 0.01 versus model group.

Discussion

Successful interaction between the axonal connections of peripheral nervous system is essential to maintain the unity of the nervous system. Following damages to peripheral nerves, which is the commonest form of trauma, disruption of axonal connections is the main defect, ³⁷ which causes disability and must be treated with surgery, drugs and new methods including the use of stem cells and scaffolds. However, none of the mentioned methods guarantee the outcomes of the treatment and future studies are needed to pave the way. To address this issue, we injected amniotic membrane derived-scaffolds along with intraperitoneal metformin administration, and after 14 days, a significant improvement in the function of the damaged sciatic nerve of rats was observed. To obtain a suitable injectable hAM scaffold, amniotic cells must be removed without damaging other membrane components, including collagen and proteoglycans, which are active in the gelation process. We removed the amniotic cells by exposing the amniotic membrane to trypsin and EDTA,^33,43 and the presence of macromolecules in hAM scaffolds was confirmed by Masson’s trichrome and allicin blue staining. To create sciatic nerve injury, we modified the method presented by Liu et al., ³⁷ and applied a constant pressure on the rat’s sciatic nerve for 2 min to cause moderate damage to the sciatic nerve. Preliminary tests (data not shown) showed 2 min compression results in complete blockade of nerve impulses (confirmed by SFI and SSI assessment 2 days after compression, and Von Fery sensation evaluation 7 days after compression and histological evaluations) which was recovered partially after 14 days. Other compression times (1 –3 min) and repeated compressions were not helpful. This experiment further corroborates prior findings indicating that AM exhibits excellent biocompatibility.^14,44

Our data successfully demonstrated that therapeutic intervention with hydrogel was more effective than other interventions for sciatic nerve regeneration, considering the improvement of motor and sensory function. Sciatic functional and static indices including latency time, myelin regeneration rate and the expression of NF-200 and S100 improved in this group, in which the hydrogel was injected into the epineurium of the injury site. In addition, in the mix group, metformin administration increased the effectiveness of the hydrogel in response to mechanical stimuli, myelin density, axonal regeneration and Schwann cell migration to the injured area. The CMAP parameter in electrophysiological evaluations showed that there was no significant difference between the experimental groups, but this index considerably improved in the treated animals compared to the model group. Compression of sciatic nerve for 2 min led to a significant decrease in the weight of the gastrocnemius muscle as well as an increase in the amount of collagen fibers detected by Masson’s trichrome staining. Following the disconnection of muscle and nerve fibers, muscle function is impaired, resulting in muscle weight loss and deposition of collagen fibers. None of treatments on the injured sciatic nerve could improve the condition except in the hydrogel group that a minor nonsignificant improvement was detected. As shown by other clinical and animal studies, neuromuscular junctions are difficult to establish. Whether longer treatment times lead to functional improvement was not demonstrated in our research and requires further studies. However, by other methods and longer time points some improvement in the gastrocnemius mass has been reported (Zhou et al., 28 Bai et al., ⁴⁵ and Wolfe et al. ⁴⁶ ). Unlike motor recovery, mechanical sensitivity to non-painful stimuli which was examined by Von Frey test, increased significantly, so that it was comparable with the control group using hydrogel and metformin (Mix group). In agreement with our findings, in a recent study Zhou et al. ²⁸ demonstrated that metformin induces M2 polarization through AMPK/PGC-1α/PPAR-γ pathway to improve peripheral nerve regeneration. NF200 marker expression as a primary sensory neuron that give rise to myelinated axons was strongly expressed in the hydrogel and mix groups compared to the model group. In addition, the S100 marker as a Schwann cell marker was significantly expressed in the experimental groups compared to the model group, and no significant difference was observed between the hydrogel, metformin and mix groups. LFB staining also revealed a significant improvement in axonal remyelination in the hydrogel and mix groups which is consistent with the Wu et al. ⁴⁷ study, which demonstrated that metformin accelerates myelin debris clearance and improves myelin preservation in 4 weeks. Also, our findings are in agreement with Lemke et al. ⁴⁸ who reported a significant improvement after 12 weeks of using amniotic membrane at the site of sciatic nerve injury.

We also showed that SFI and SSI indices were significantly improved in the hydrogel group, and insignificantly in the mix group. Our results are in agreement with Mohammad et al., ⁴⁹ Zhang et al., ⁵⁰ Sadraie et al., ⁵¹ and Meng et al. ¹⁵ who used cell-free amniotic membrane to regenerate the sciatic nerve following nerve dissection.

Our findings in the electrophysiological evaluations of the animals 14 days after sciatic nerve compression indicated that the latency index was significantly decreased in the hydrogel and the mix groups, while the CMAP improvement was not significant in the treatment groups. Sadraie et al. ⁵¹ reported that these two parameters improved in the decellular amniotic membrane-treated animals 8 weeks after injury. While, Zhang et al. ⁵⁰ showed that AM wrapping significantly improved the amplitude of nerve impulses 16 weeks after surgery. In addition, Liu et al. revealed that i.p. injection of metformin for 4 weeks improved amplitude due to its protective role against neurotoxins and rescuing neurons from neurodegeneration.

The body weight of the animals did not change significantly at the beginning of the experiments and 14 days after the interventions, indicating that compression of the sciatic nerve does not lead to a change in body weight. So, the weight of the gastrocnemius muscle on the affected side compared to the body weight have not influenced from the body weight gain. However, Zhou et al. ²⁸ reported an increase in the gastrocnemius muscle mass after 14-day treatment of animals with metformin, which disagree with our data.

Conclusion

Our findings showed that a hydrogel prepared from human decellularized amniotic membrane is capable of improving sciatic nerve injury when injected into the epineurium of the affected nerve. A daily administration of metformin could improve some features of nerve regeneration when used in combination with hAM hydrogel. Although previous studies have demonstrated that intraperitoneal administration of metformin and the application of hydrogel at the injury site individually improve sciatic nerve function following injury, in this study, the simultaneous administration of hydrogel and intraperitoneal metformin did not enhance the therapeutic effects of these factors when used alone. It is possible that localized delivery of metformin through a hydrogel at the injury site, with controlled local release, could potentiate the therapeutic effects of both hydrogel and metformin individually. Since the hAM hydrogel can be easily harvested with less expenses, the lyophilized hAM hydrogel can be used in human studies without fear of the risk of xenograft materials. However, the appropriate time of delivery according to the time of injury, the amount of hydrogel delivered, and the route of delivery of the hydrogel, also administration of metformin loaded hydrogel to the injured peripheral nerve should be investigated before any human use.