Preliminary study on the preparation of lyophilized acellular nerve scaffold complexes from rabbit sciatic nerves with human umbilical cord mesenchymal stem cells

Abstract

BACKGROUND

The gold standard of care for patients with severe peripheral nerve injury is autologous nerve grafting; however, autologous nerve grafts are usually limited for patients because of the limited number of autologous nerve sources and the loss of neurosensory sensation in the donor area, whereas allogeneic or xenografts are even more limited by immune rejection. Tissue-engineered peripheral nerve scaffolds, with the morphology and structure of natural nerves and complex biological signals, hold the most promise as ideal peripheral nerve “replacements”.

AIM

To prepare allogenic peripheral nerve scaffolds using a low-toxicity decellularization method, and use human umbilical cord mesenchymal stem cells (hUC-MSCs) as seed cells to cultivate scaffold-cell complexes for the repair of injured peripheral nerves.

METHODS

After obtaining sciatic nerves from New Zealand rabbits, an optimal acellular scaffold preparation scheme was established by mechanical separation, varying lyophilization cycles, and trypsin and DNase digestion at different times. The scaffolds were evaluated by hematoxylin and eosin (HE) and luxol fast blue (LFB) staining. The maximum load, durability, and elastic modulus of the acellular scaffolds were assessed using a universal material testing machine. The acellular scaffolds were implanted into the dorsal erector spinae muscle of SD rats and the scaffold degradation and systemic inflammatory reactions were observed at 3 days, 1 week, 3 weeks, and 6 weeks following surgery to determine the histocompatibility between xenografts. The effect of acellular scaffold extracts on fibroblast proliferation was assessed using an MTT assay to measure the cytotoxicity of the scaffold residual reagents. In addition, the umbilical cord from cesarean section fetuses was collected, and the Wharton’s jelly (WJ) was separated into culture cells and confirm the osteogenic and adipogenic differentiation of mesenchymal stem cells (MSCs) and hUC-MSCs. The cultured cells were induced to differentiate into Schwann cells by the antioxidant-growth factor induction method, and the differentiated cells and the myelinogenic properties were identified.

RESULTS

The experiments effectively decellularized the sciatic nerve of the New Zealand rabbits. After comparing the completed acellular scaffolds among the groups, the optimal decellularization preparation steps were established as follows: Mechanical separation of the epineurium, two cycles of lyophilization-rewarming, trypsin digestion for 5 hours, and DNase digestion for 10 hours. After HE staining, no residual nuclear components were evident on the scaffold, whereas the extracellular matrix remained intact. LFB staining showed a significant decrease in myelin sheath composition of the scaffold compared with that before preparation. Biomechanical testing revealed that the maximum tensile strength, elastic modulus, and durability of the acellular scaffold were reduced compared with normal peripheral nerves. Based on the histocompatibility test, the immune response of the recipient SD rats to the scaffold New Zealand rabbits began to decline3 weeks following surgery, and there was no significant rejection after 6 weeks. The MTT assay revealed that the acellular reagent extract had no obvious effects on cell proliferation. The cells were successfully isolated, cultured, and passaged from human umbilical cord WJ by MSC medium, and their ability to differentiate into Schwann-like cells was demonstrated by morphological and immunohistochemical identification. The differentiated cells could also myelinate in vitro.

CONCLUSION

The acellular peripheral nerve scaffold with complete cell removal and intact matrix may be prepared by combining lyophilization and enzyme digestion. The resulting scaffold exhibited good histocompatibility and low cytotoxicity. In addition, hUC-MSCs have the potential to differentiate into Schwann-like cells with myelinogenic ability following in vitro induction.

Key Words: Human umbilical cord mesenchymal stem cells; Peripheral nerve injury; Schwann cells; Acellular nerve scaffolds

Core Tip: The treatment of severe peripheral nerve injuries remains a clinical challenge, particularly in children. Autologous nerve grafts are the standard treatment for these severe neurologic deficits and the scarce number of autologous nerves and the loss of neurosensory function in the donor area are major obstacles, particularly in infants and young children. Allogeneic or xenografts are even more limited by immune rejection. Therefore, there is an urgent need for a peripheral nerve substitute that can bridge the two severed ends of the nerve, guide its axonal growth to avoid the formation of neuroma, and promote and guide the functional regeneration of the peripheral nerve. For patients with neonatal brachial plexus injury, human umbilical cord mesenchymal stem cells (hUC-MSCs) are an effective tool. In this study, we proposed to culture hUC-MSCs as seed cells on suitable decellularized scaffolds.

INTRODUCTION

The treatment of severe peripheral nerve injury (PNI) remains a clinical challenge, even for children. PNI in the brachial plexus region caused by birth trauma is the most common PNI (obstetrical palsy) in newborns. Autologous nerve transplantation is the standard treatment for these severe neurological deficits[1]. The number of precious autologous nerves and the loss of sensory function of the donor nerve are the primary constraints, especially for the transplantation of autologous nerves in infants and young children; however, nerve allograft or nerve xenotransplantation is limited by immune rejection. Currently, there is an urgent need for peripheral nerve substitutes that can bridge the two broken ends of nerves, guide the axial growth of nerve axons, avoid the formation of neuromas, and guide the functional regeneration of peripheral nerves.

Tissue-engineered acellular nerve scaffolds are repair materials exhibiting natural nerve morphology, complex biological signals, and little immunogenicity obtained by the decellularization of homologous/heterogeneous peripheral nerves, which are the most promising nerve “substitutes”[2]. With the advent of tissue engineering, a variety of decellularization techniques have been applied to peripheral nerve scaffolds[3,4]; however, the use of detergents during the decellularization process increases the cytotoxicity of the scaffold itself and reduces the biomechanical properties of the material, which can affect the tissue reconstruction goals in vivo. Some researchers have completely abandoned detergents and simply used mechanical separation and enzymatic methods to obtain the ideal acellular scaffold[5,6].

Current studies suggest that Schwann cells (SCs) have important functions that other peripheral nerve-related cells do not in the process of peripheral nerve regeneration. A large number of experimental studies indicate that SC transplantation can enhance peripheral nerve regeneration[7]; however, the autologous source of SCs is very limited and undamaged peripheral nerves are often sacrificed to obtain culturable autologous SCs[8]. To replace the cultivation of SCs in tissue engineering, a pluripotent stem cell is needed that is easy to obtain, rapidly expands in vitro, and differentiates into SCs before implantation. Mesenchymal stem cells (MSCs) are the most widely used pluripotent stem cells for clinical applications and experimental studies, with a well-documented potential to differentiate into bone, fat, and cartilage[9]. Of these, bone marrow-derived MSCs are the most frequently used MSCs. They are derived from the bone marrow cavity with little trauma. They are abundant materials, easy to culture, and have the ability to differentiate into SCs under special conditions[10]. However, in the case of neonatal brachial plexus injury, there may be more appropriate pluripotent stem cells available. The Wharton’s jelly (WJ) in the newborn umbilical cord is rich in human umbilical cord MSCs (hUC-MSCs). hUC-MSCs are neonatal cells, but are located in vitro and are extremely convenient, non-invasive, and cell-rich[11]. hUC-MSCs represent a therapeutic option for neonatal brachial plexus injury. This study aimed to prepare a suitable acellular scaffold and co-culture of hUC-MSCs as seed cells for the scaffold.

MATERIALS AND METHODS

Experimental animals

All animal experiments were approved by the Association’s Animal Care and Use Committee and strictly adhered to animal protection guidelines. Twenty specific pathogen-free (SPF) 3-month-old New Zealand female rabbits and 30 SPF one-month-old SD rats were provided by the Animal Experimental Center of Fudan University. The animals were reared in an environment with an indoor temperature of 20 °C-25 °C and a relative humidity of 40%-50%. The experimental rabbits were fed by dedicated researchers with a quantitative diet and water intake each day. The animal experiment license number was SYXK (Shanghai) 2018.0003.

Acquisition of peripheral nerves

After an intraperitoneal injection of 30 mg/kg of pentobarbital sodium to anesthetize the New Zealand rabbits, an approximate 8 cm longitudinal incision was made in the middle of the bilateral posterior thigh. Blunt isolation of the posterior thigh muscles was then performed to find the deep sciatic nerve. The soft tissue around the sciatic nerve was freed, a nerve with a length of approximately 8 cm was selected and its far and near ends were tied together with silk thread. The nerve between the ligatures was cut with a sharp knife to remove it. After drying, the peripheral epineurium was mechanically separated. The 8 cm nerve was divided into two 4 cm long segments and stored at -80 °C. The nerve samples were transferred to a freeze-drying container pre-cooled to -56 °C, freeze-dried for 48 hours, rewarmed to 37 °C, and rinsed. After drying and removing the epineurium residue, the samples were stored at -80 °C.

Preparation of acellular scaffolds

The nerves were decellularized and the experimental samples were divided into groups 1, 2, 3, and 4, as well as groups A, B, C, and D based on their different lyophilization cycles (with 3 nerves in each group). Of these, groups 1-4 were lyophilized for one cycle and groups A-D were lyophilized for two cycles. Groups 1-4 and A-D were differentiated according to various enzyme treatment times. The specific steps were as follows: (1) Groups 1 and A were treated with serine protease inhibitors to stabilize the neural matrix components for 12 hours, pancreatic enzyme treatment for 5 hours to disrupts cell adhesion to the cell matrix, DNase treatment for 5 hours to destroy the nuclear components, and finally immersed in phosphate buffer saline (PBS) for 48 hours to remove the cellular components. Group 2 and group B were processed in the same manner, with the processing time being 12, 10, 5, and 48 hours, respectively. The treatment time for groups 3 and C were 12, 5, 10, and 48 hours, respectively, whereas that of groups 4 and D was 12, 10, 10, and 48 hours respectively. After drying the acellular scaffolds at a low temperature, 0.9% normal saline containing 5% penicillin (10000 U/mL)/streptomycin (10000 g/L) solution was added to prevent infection. After irradiation and disinfection with γ-rays (15K Gray), the samples were stored at -4 °C for later use.

Quality testing of the acellular scaffolds

DNA quantitative detection: DNA was extracted from each group of acellular scaffolds using a TIA Namp Genomic DNA kit for sample DNA quantification. The DNA was quantitated at 280 nm using a spectrophotometer (Thermo Spectronic, Biomate3, Rochester, NY, United States). The DNA was then analyzed using a nucleic acid/protein analyzer (Du800 BECKMAN, United States).

Integrity test: A hydroxyproline test kit was used to measure the collagen content of the scaffold. After hydrolyzing the samples, 1 mL of each sample was dripped onto a 96-well plate, placed in a microplate reader, and the absorbance of each sample was measured at 550 nm. The collagen content was calculated as follows: Collagen content = hydroxyproline content/13.4%.

Histological examination of acellular scaffolds

Hematoxylin and eosin (HE) and luxol fast blue (LFB) staining: Acellular scaffolds were immobilized with 10% paraformaldehyde, dehydrated, paraffin-embedded, and sliced, followed by histological observation. HE staining was performed and the structures, such as cells and myelin sheaths, were observed under a light microscope. The distribution of myelin sheaths was observed by LFB staining. Microscopically, the myelin sheaths of the peripheral nerve scaffolds were dark blue.

Biomechanical testing of lyophilized acellular peripheral nerve scaffolds

Rabbit sciatic nerve acellular scaffolds (7-10 cm) as well as 7-10 cm contralateral normal sciatic nerves (controls), were prepared and labeled individually according to the ptimal decellularization method (per 10 in each group). After installing the acellular scaffold directly onto the universal materials testing machine (Zwick, Germany), the acellular scaffold and approximately 10 mm of tissue at both ends of the nerve were fixed using the fixture on the testing machine. The sample was adjusted perpendicular to the ground using a plumb line. The tensile strength of the sample was adjusted to 1 Newton by the testing machine, and the sample diameter was measured by an electronic vernier caliper, based on which the cross-sectional area was calculated. Samples were sprayed with saline every 10 minutes throughout the test to ensure that they were moist. A test sample was assembled on the mechanical testing machine. After humidifying the acellular scaffold with normal saline, a length of 70 mm and a measuring length of 50 mm were taken. After presetting, the specimen was pulled apart at a loading rate of 10 mm/minute and the elastic modulus (MPa), displacement (mm), and maximum tensile load (N) were recorded.

Establishment of a histocompatibility detection model in SD rats

Twenty-four SD rats were divided into two groups: The first group was a sham operation group and the second group was an operation group, with 12 rats per group. The acellular scaffold was implanted into the right erector spinae with minimal injury. Pentobarbital sodium (30 mg/kg) was injected intraperitoneally for anesthesia. A median longitudinal incision was made along the spinous process notch of the spine, the skin and fascia were cut in turn, and bluntly separated into the erector spinae muscle layer. The right erector spinae was used as the insertion point. The acellular scaffold was fixed with silk threads and embedded into the muscle with minimal damage through round needle guidance. The sham operation group was prepared in the same manner as the operation group. After exposing the erector spinae muscle, the right erector spinae muscle was used as the insertion point. The needle was inserted, but no other foreign objects were implanted into the body. If the rats were infected after surgery, a compatibility testing model was re-established and the number of infected cases was recorded.

Detection of histocompatibility in SD rats

According to the time pattern for the post-transplantation rejection reactions in SD rats, three rats in each group were euthanized for local compatibility testing at 3, 7, 21, and 42 days following surgery. The rats in the operation group were dissected, the erector spinae muscle was removed, and fixed in a polyformaldehyde solution for preservation. Paraffin sectioning and HE staining were done to observe the degree of acellular scaffold degradation, inflammatory cell infiltration, and autologous cell migration. Systemic compatibility testing: After euthanizing the rats, three tubes (5 mL each) of blood were collected. After heparin anticoagulation, the samples were sent to the inspection center to test for three inflammatory indicators: Erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and IL-6.

Culture of rat fibroblasts

Fibroblasts from SD rats were used for cytotoxicity testing. After an intraperitoneal injection of 30 mg/kg of pentobarbital sodium, the rat tail was cut short and the tail skin was removed. The tail and coccyx were cut into a primary culture with a length of approximately 1 cm, placed onto the surface of a petri dish, and washed with DMEM three times. Subsequently, the tissue blocks were cut into 1 mm × 1 mm × 1 mm small pieces, with 2-3 mm spacing between the tissue blocks. After adding a high-sugar DMEM culture solution, the petri dish was incubated at 37 °C in a 5% CO₂ incubator. After the tissue blocks were attached to the wall as much as possible (after 4 hours), culture medium was carefully added at a volume that prevented the tissue blocks from floating, with the media slightly over the tissue blocks. The tissue blocks were cultured for 24 hours and the same culture medium was added appropriately according to the decrease in absorption of the culture medium, without allowing the tissue blocks float. The medium was replaced every 3 days. The fibroblasts underwent routine culture, fluid exchange, digestion, and passage. The third generation of fibroblasts was used for the cytotoxicity assay.

Cytotoxicity assay

The prepared acellular peripheral nerve scaffolds were immersed in a DMEM culture solution to prepare acellular scaffold extracts. The cultured third-generation SD rat fibroblasts were prepared at a concentration of 10⁹/L, suspended, and inoculated into a 96-well plate. DMEM culture solution was added to the culture and incubated for 24 hours in a 5% CO₂ cincubator at 37 °C. Once cell adhesion was observed, the remaining culture medium was discarded, and two different concentrations of lyophilized acellular scaffold extracts (50% and 100%) were added. DMEM (200 µL) was added to the negative control group. COne culture plate was removed from the incubator on day 0, 1, 3, 5, 7, and 9 followed by incubation with MTT solution (20 µL/well) for an additional 4 hours. After removing the culture solution, 150 µL of DMSO was added and shaken for 10 minutes. The absorbance value was measured with a microplate reader at 490 nm. The results were calculated by the six-level toxicity grading method for the relative proliferation and toxicity of fibroblasts in different acellular scaffold extracts.

Isolation and culture of hUC-MSCs

Healthy umbilical cords were obtained from full-term neonates born in the Obstetric Department of our hospital with the informed consent of their parents. Fresh umbilical cords of full-term neonates were obtained and sterilized. Blood was removed and placed in DMEM/F12 culture dishes to remove arteriovenous vessels. WJ was the remaining tissue after removal of the umbilical artery. The separated WJ tissue was sheared to a particle size of approximately 1 cm³, collagenase preheated to 37 °C was added to the tissue block, and incubated for 18 hours. After the tissue block was washed with PBS, 0.25% trypsin was added and digested for 5 minutes. The digested tissue block was placed in a 1/6 ratio in a petri dish and a special culture medium for MSCs containing 2 mmol/L L-glutamine and 1% double antibody (100 U/mL penicillin and 100 U/mL streptomycin) was added. After washing with culture medium several times, the umbilical cords were cut into 1 mm³ tissue blocks, inoculated in a petri dish at a diameter of 15 cm in an appropriate density, 5.0 mL of DMEM/F12 (containing 10% fetal bovine serum) was added, and the samples were place in a 37 °C humidified incubator containing 5% CO₂. The next day, 10.0 mL of the same culture medium was added and incubation continued for 3-5 days before the medium was changed. After 10-14 days, the tissue blocks were removed. The cells were digested with 0.25% trypsin-EDTA upon 80% confluence, collected, and passed to a third generation at a ratio of 1: 3. The cells were then digested with 0.25% trypsin-EDTA and collected for re-inoculation at a density of about 10⁷/L.

Differentiation ability of hUC-MSCs

Determination of the ability of hUC-MSCs to differentiate into osteoblasts: HUC-MSCs were subcultured for three generations and trypsinized (0.25%) to prepare a cell suspension. hUC-MSCs were re-inoculated at a density of 1 × 10⁵/cm² and cultured in a trypsin zed special MSC medium, which was changed every three days. The degree of cell fusion was observed using an inverted phase contrast microscope. When hUC-MSCs grew to near confluence, an osteogenic induction solution was added, and the cells were stained with alkaline phosphatase on the 7^th day after continuous culture. On the 14^th day, the mineralized crystals in the culture dish were stained with alizarin red and observed using an inverted phase contrast microscope. Alizarin red staining: After discarding the culture medium, the contents remaining in the petri dish were fixed with paraformaldehyde for 30 minutes, rinsed with PBS, stained with alizarin red for 8 minutes, and rinsed sequentially with distilled water and 0.5% acetic acid. Following dehydration with gradient ethanol, the samples were observed under a microscope.

Identification of the ability of hUC-MSCs to differentiate into adipocytes: The degree of cell fusion was the same as that of ‘Determination of the ability of hUC-MSCs to differentiate into osteoblasts’. When hUC-MSCs were nearly confluent, an adipocyte induction solution was added. The cells were stained with oil red O on the 14^th day after continuous culture and observed under an inverted phase contrast microscope. Oil red O staining: After removing the culture medium, the cells were fixed with paraformaldehyde for 30 minutes, rinsed with PBS, and stained with oil red O solution for 20 minutes. Color separation was achieved with 60% ethanol followed by slow rinsing with distilled water, and mounting with gum.

Induction into SCs and identification after differentiation

Induction into SCs: Third-generation hUC-MSCs were seeded into a 6-well plate at a density of 5 × 10⁵/cm² to prepare round coverslips. Special MSC culture medium containing 1 mmol/L β-mercaptoethanol was added. Once the cells were confluent, a pre-induction solution (DMEM medium, 10% FBS, and 35 ng/mL of all-trans retinoic acid) was added for 72 hours. After washing with 0.01 mol/L of PBS three times, an induction solution (MSC medium, 5 mmol/L Forskolin, 10 ng/mL bFGF, 5 ng/mL PDGF, and 200 ng/mL of heregulin) was added. The medium was replaced every three days and the cells were identified on day 7.

Identification after differentiation: After 7 days of induction and culture, the cells were placed under an inverted phase contrast microscope to evaluate their growth status and morphological changes. The induced MSCs were fixed with 4% paraformaldehyde for 30 minutes and washed three times with PBS. To eliminate endogenous peroxidase in cells, a peroxidase-blocking solution was added for 20 minutes followed by incubation with goat serum working solution. Three monoclonal antibodies (MBP, GFAP, and S100) were added separately and incubated overnight at -4 °C FITC and TRITC secondary antibodies were added, respectively, and incubated for 2 hours at 37 °C. The cells were observed by fluorescence microscopy and the ratio of the number of cells positive for MBP, GFAP, and S100 to the total number of cells was calculated.

In vitro myelination capacity assessment of cells after induction

MPZ expression: When intracellular cAMP increases to a certain level, true SCs can differentiate into myelin-forming cells and express myelin protein 0 (MPZ, or p0). Therefore, based on the published method[12], the level of cAMP in the induced cells was measured by the FSK method, and the expression of MPZ was assessed by immunofluorescence. The treated cells were observed under a fluorescence microscope to determined whether there were any cells with positive MPZ expression.

Co-culture with PC12 cells: When PC12 cells were co-cultured with true SCs, which can induce and accelerate PC12 cell differentiation and form myelin sheath structures with PC12 neurites[13]. Therefore, we used PC12 to assess the myelination capacity of SC-like cells after induction and used transmission electron microscopy to observe and document the formation of myelin.

Statistical analyses

The data were analyzed using SPSS 26.0 software. A two-tailed Student’s t-test or one-way analysis followed by Tukey’s multiple comparisons post hoc test were used to determine statistical significance. All continuous variables for normality were tested and expressed as the mean ± SD. A P value of P < 0.05 (two-tailed) was considered statistically significant.

RESULTS

Selection of the optimal decellularization protocol

DNA content and collagen content were used to detect the degree of cell elution, and HE staining was used to observe the integrity of the scaffold. The decellularization method we tried remains a decellularization protocol that combines instrumental isolation, lyophilization and enzymatic methods and is detergent-free. After different cycles of lyophilization and application of different reagent times, we obtained eight scaffolds with different degrees of decellularization and different degrees of integrity. Figure 1 shows the residual DNA content in samples from the 8 groups following the various decellularization protocols (Group 1: 0.5560 ± 0.02, Group 2: 0.3552 ± 0.03, Group 3: 0.4913 ± 0.02, Group 4: 0.3406 ± 0.02, Group A: 0.2705 ± 0.02, Group B: 0.0525 ± 0.01, Group C: 0.0376 ± 0.01, Group D: 0.0358 ± 0.01), indicating a statistically decreased DNA content in each group compared with that in the control group (1.4294 ± 0.04). The decrease was the most significant in groups B-D, whereas no significant inter-group difference was observed in the DNA content between groups C and D (P = 0.783). Overall, the collagen content in the acellular scaffold decreased with extension trypsin and DNase digestion time (Group 1: 49.65 ± 1.62, Group 2: 52.62 ± 1.26, Group 3: 42.07 ± 2.73, Group 4: 49.06 ± 2.37, Group A: 50.24 ± 0.99, Group B: 42.42 ± 4.36, Group C: 49.20 ± 2.91, Group D: 45.25 ± 2.88). However, the collagen content of groups 1, 2, 4, A, and C showed no significant difference compared with the control group. Thus, for the preparation plans, we selected the method with the most complete degree of decellularization and preservation of the scaffold. Through this comparison, we proceeded with the decellularization method of group C.

Figure 1 Comparison of DNA and collagen content among the various groups. A: DNA content in each group; B: Collagen content in each group. ^aP < 0.05 vs control group; n = 3.

Histological observation of acellular scaffolds

After HE staining, we found that the structure of the acellular peripheral nerve scaffold exhibited a regular tubular structural arrangement (Figure 2A). In the longitudinal section, the structure of the peripheral nerve was almost identical to that of normal peripheral nerves. After depleting the SCs and the cytoplasmic components of the neuron cells, the structure of the peripheral nerve neurolemma was nearly intact. At low or high magnification, little staining of the nuclei was observed (Figure 2B).

Figure 2 Histological observation of acellular scaffolds . A: Hematoxylin and eosin (HE) staining shows the peripheral nerve fibers enveloped by the perineurium and a large amount of nuclear components may be seen in the peripheral nerves; B: After decellularization, the matrix structure of the HE-stained nerve fibers is preserved; however, the cell structure is no longer visible; C: Luxol fast blue (LFB) staining of peripheral nerve tissue reveals significant blue staining of myelin sheath tissue; D: LFB staining shows decreased staining of decellularized nerve tissue and myelin sheaths.

The peripheral nerve myelin sheath was visible and enveloped around the central nerve fibers by LFB staining (Figure 2C). After decellularization, myelin staining of the peripheral nerves was attenuated, but the pore structure of the central nerve fibers was preserved (Figure 2D). These data indicate that after mechanical separation of the outer nerve membrane, lyophilization and reheating (2 cycles), five-hour pancreatic digestion, and ten-hour DNase digestion, a tissue-engineered peripheral nerve scaffold with complete acellular and extracellular matrix preservation was prepared.

Mechanical properties of acellular scaffolds

To examine the mechanical properties of acellular scaffolds, we performed measurements on durability, elastic modulus, and maximum tension of the acellular scaffolds. The results indicated that the maximum tensile strength of the sciatic nerve in normal rabbits reached an average of 9.007 ± 3.07 N, whereas that of the acellular rabbit nerve scaffold decreased significantly (current average maximum tensile strength: 6.005 ± 1.82 N, P < 0.05, Figure 3A). The elastic modulus of the control group was 16.75 ± 4.91 MPa compared with 10.91 ± 3.31 MPa for the acellular scaffold. Therefore, the tensile strength that the acellular scaffold can withstand was significantly smaller compared with that in the normal nerve tissue and its elastic modulus was also lower (Figure 3B).

Figure 3 Mechanical properties of the acellular scaffolds . A: Comparison of the maximum load between acellular scaffolds and peripheral nerves; B: Comparison of elastic modulus between acellular scaffolds and peripheral nerves; C: Stress durability test of acellular scaffolds and peripheral nerves; D: Stress-time diagram of peripheral nerves; E: Stress-time relationship of acellular scaffolds. ^aP < 0.0001 (control vs acellular nerve scaffolds); n = 10.

A total of ten scaffolds and nerves were tested. The stress required for maintaining 2% of the sample length during stretching in the control group did not show a significant decrease after removing the viscoelastic effect in the first cycle, which indicated good durability of the sciatic nerve in the control group. However, two samples in the acellular scaffold group experienced breakage during the 7^th and 11^th cycles. The stress of the remaining eight samples continued to decrease for five cycles while maintaining a 2% tensile length before it began to plateau (Figure 3C). Figures 2E and 3D shows the test cycles for sample No. 3 in the control group and sample No. 2 in the acellular scaffold group, respectively.

Histocompatibility testing

To test the histocompatibility of acellular scaffolds, we compatibilized the scaffolds with SD rat tissues and tested their compatibility on the 3^rd postoperative day. A large number of inflammatory cells began to appear around the scaffold, including those surrounding the scaffold and spreading toward the middle. The overall shape of the scaffold could still be recognized; however, the pore structure associated with the original scaffold could not be identified (Figure 4A). On the 7^th day following surgery, the number of inflammatory cells around the scaffold was further increased and the scaffold was further enveloped by the surrounding inflammatory cells, with some degradation compared with that on the 3^rd day. In areas with relatively few inflammatory cells, the pore structure of the scaffold could be distinguished (Figure 4B). Three weeks after the operation, the inflammatory cells around the scaffold began to fade. Although the scaffold was degraded to a certain extent, its continuous morphology was maintained and the proliferation of connective scar tissue was observed around the scaffold (Figure 4C). Six weeks following the surgery, the inflammatory reaction further subsided and the shape of the scaffold was preserved. Neovascularization was observed at the scaffold–muscle junction (Figure 4D).

Figure 4 Hematoxylin and eosin staining after acellular scaffold implantation. A: 3 days after surgery, a large number of inflammatory cells began to appear around the scaffold, including those surrounding the scaffold and spreading toward the middle; B: 7 days postoperatively, the number of inflammatory cells around the scaffold was further increased and the scaffold was further enveloped by the surrounding inflammatory cells; C: 3 weeks after surgery, the inflammatory cells around the scaffold began to fade; D: 6 weeks postoperatively, the inflammatory reaction further subsided and the shape of the scaffold was preserved.

Systemic compatibility testing

In this step of the study, we will compare the mechanical properties of this scaffold with those of nerve scaffolds synthesized from other materials to fully evaluate the advantages and disadvantages of lyophilized decellularized scaffolds. Autologous grafts of nerves do not undergo rejection, allogeneic grafts need to be transplanted either by comparing antigens or after treatment to remove antigens, whereas inter-allogeneic grafts are more likely to undergo rejection, and the grafts generally do not survive in the recipients without special treatment. Thus, we tested systemic compatibility. SD rats in the sham operation group and the acellular scaffold group showed increased CRP levels 3 days after surgery, whereas the levels in the sham operation group returned to normal 1 week following surgery. In contrast, CRP levels in the acellular scaffold group increased 7 days after surgery, but did not exhibit a downward trend until 3 weeks after surgery, and essentially returned to normal levels 6 weeks following surgery (Figure 5A).

Figure 5 Systemic compatibility testing. A: Changes in postoperative C-reactive protein levels; B: Postoperative erythrocyte sedimentation rate changes; C: Changes in postoperative IL-6 Levels; D: MTT detection of cytotoxicity. n = 10. CRP: C-reactive protein; ESR; Erythrocyte sedimentation rate.

Postoperatively, the ESR in the sham operation group did not change, but fluctuated slightly from 3 days to 1 week after surgery. However, ESR in the acellular scaffold group began to rise continuously one week after surgery, but did not show a downward trend until the last test (Figure 5B).

There were no significant postoperative changes in IL-6 levels in the sham operation group. In the acellular scaffold group, IL-6 began to increase at 1 week after surgery, peaked at three weeks, and began to decline at 6 weeks following surgery (Figure 5C).

The A490 values measured on days 0-9 are shown in Figure 5D. Except for a decrease in the cell proliferation rate after 1 day of cultivation with 100% concentration of the extract, the toxicity of the extract in each group at other time points did not inhibit cell growth compared with the negative control group. On the 9^th day of cytotoxicity testing, both groups of cells performed well at different concentrations of nerve scaffold extracts, with positive proliferation. The results indicated that there was no significant difference in cell proliferation following treatment with 50% extract compared with the control group (P > 0.05).

Above all, the lyophilized decellularized scaffolds in this experiment showed excellent anti-degradation properties and histocompatibility: the recipients were eventually able to tolerate the implantation of the scaffolds, the local reactions in the recipients after xenografting were not dramatic, the inflammatory reactions began to diminish at 6 weeks after the procedure, and the scaffolds ultimately retained a relatively intact continuity.

Culture and identification of hUC-MSCs

To further construct seed cells, we cultured hUC-MSCs. After six days of culture in the special MSC culture medium, the growth of round-like cells was observed under an inverted phase contrast microscope. After approximately one week of culture, the cells were adhering to the wall (Figure 6A). After passaging, no significant differences were observed in the cells microscopically. After 14 days of cultivation, the cells had fused into clusters were arranged in a spindle shape, with no significant differences in cell morphology (Figure 6B). Isolated and cultured hUC-MSCs were identified, and positive results of hUC-MSC surface markers CD29 and CD44 and a negative result of CD34 were obtained (Figure 6C), indicating successful isolation of hUC-MSCs. After continued culture of the hUC-MSCs with osteogenic induction solution, the morphology of the cells showed polygonal and oval changes. After 2 weeks of cultivation, obvious extracellular mineral crystals were observed following alizarin red staining (Figure 6D). This indicates that hUC-MSCs have the ability to differentiate into osteoblasts. After hUC-MSCs were continuously cultured with an adipogenic induction solution, the morphology of the cells was altered and lipid droplets with varying refractive indices were evident in the culture medium. After culturing for two weeks, oil red O staining (Figure 6E) revealed lipid substances in the cytoplasm. This indicates that hUC-MSCs have the ability to differentiate into adipocytes.

Figure 6 Culture and identification of human umbilical cord mesenchymal stem cells. A: Morphological observation of human umbilical cord mesenchymal stem cells (hUC-MSCs) after 3 days of culture; B: Morphological observation of hUC-MSCs cultured for one week; C: Flow cytometry for measuring the levels of hUC-MSC surface markers CD44, CD29, and CD34; D: Alizarin red staining; E: Oil red O staining. APC: Adipocyte precursor cell.

Identification of Schwann-like cells after induced differentiation

After induction, we observed that the cell morphology began to change approximately 24 hours after the addition of the inducer, and gradually became slender from the initial spindle shape (Figure 7A) to form a long-shuttle shape similar to SCs on the 7^th day (Figure 7B). Immunofluorescence revealed positive expression of MBP, GFAP, and S100 (78.4%, 92.9%, and 77.6%, respectively) in differentiated SC-like cells (Figure 7C). These results indicate that hUC-MSCs differentiate into SCs after induction culture.

Figure 7 Identification of Schwan-like cells after induced differentiation . A: Cell morphology after 3 days of induced differentiation and culture; B: Cell morphology after 7 days of induced differentiation and culture; C: Immunohistochemical fluorescence identification of differentiated cells.

In vitro myelination capacity of cells after induction

We confirmed whether hUC-MSC-differentiated SC-like cells have myelin-forming capacity. The medium for inducing differentiated cells was replaced with an MPZ induction medium to enhance intracellular cAMP levels. After seven days of culture, some SC-like cells exhibited a positive immune response to p0 (Figure 8A) and some MPZ-positive cells showed changes, such as irregular shape and swelling, which may be the result of cell membrane expansion caused by the secretion of myelin-like substances. These results suggest that SC-like cells differentiated from hUC-MSCs have a myelination ability (Schwann-like cells).

Figure 8 In vitro assessment of myelination capacity of the cells after induction. A: MPZ expression in the induced cells; B: PC12 cells; C: Morphology of PC12 cells co-cultured with differentiated cells; D: Layered myelin structure suspected to be formed under scanning electron microscopy.

When subculturing PC12 cells separately, only some stretching processes in PC12 cells were observed by microscopy (Figure 8B). Next, we co-cultured PC12 and SC-like cells in the SC medium for 14 days. PC12 rapidly differentiated into neuron-like cells after the addition of SC-like cells, whereas PC12 with obvious neurites were observed (Figure 8C). After 14 days of co-culture, transmission electron microscopy revealed that the SC-like cells formed a layered myelin structure with PC12 neurites (Figure 8D). The results also indicated that SC-like cells differentiated from hUC-MSCs have a myelination ability. Therefore, we found that when differentiated hUC-MSCs (Schwann-like cells) were co-cultured with PC12 cells, the Schwann-like cells could induce PC12 cells to rapidly differentiate and form myelin structures with PC12 cell neural protrusions.

DISCUSSION

Thus far, various methods of decellularization have been developed to make acellular scaffolds for peripheral nerves in the field of tissue engineering. Sondell et al[3] used allogenic acellular peripheral nerves as transplants for the first time. They used Triton X-100 and SDC to decellularize the sciatic nerves of SD rats. Compared with Sondell’s method, the detergent Triton-200 can better preserve the extracellular matrix of the scaffold[4]. Considering that residual detergents are cytotoxic and can penetrate cells, researchers have tried to reduce or remove such detergents using hyperosmotic methods for cell washing[14,15] and apoptosis for cell removal[16,17].

In the present study, we used a decellularization scheme that combines instrument separation, lyophilization, and an enzyme method without detergents. After various cycles of lyophilization and application of reagents for different times, we obtained eight scaffolds with varying degrees of decellularization and integrity. After the identification of DNA and collagen on the scaffold, we found the most suitable peripheral nerve decellularization method among the eight protocols. Lyophilization-rewarming of peripheral nerves can effectively destroy the cell membrane structure of myelin and neuronal cells, whereas DNase and pancreatic enzymes can effectively act following cell membrane destruction and has a similar role as detergents. In addition, the successful preparation of peripheral nerve scaffolds demonstrated the convenience of lyophilization and decellularization for organ preparation. For identification after scaffold preparation, we were satisfied to find no staining of the myelin sheath, neurons, and other cell nuclei in the acellular scaffold by HE staining. In addition, a porous structure remained between the perineurium and the endoneurium, which is conducive to nerve repair, neurotrophic factor transmission, and cell excreta transport. LFB staining revealed that after various steps of decellularization, the myelin tissue that was originally filled with nerves and the SCs that comprised the myelin sheath were significantly reduced. This indicates that the myelin tissue of the acellular scaffold was completely washed, which meets the criteria for clinical use of peripheral neural scaffolds.

The performance evaluation of acellular scaffolds includes many aspects, including their biomechanical properties, pore size, degradation performance, histocompatibility, and cell tropism[18]. A good tissue substitute must meet most of the above requirements to achieve better substitution effects. We tested the mechanical properties of the acellular scaffolds, which were decreased compared with normal nerves in terms of maximum load, elastic modulus, and durability, which is obviously related to the loss of extracellular matrix, cellular components, and factors secreted by cells during the process of manufacturing acellular scaffolds. We believe that mechanical separation of the epineurium before scaffold preparation greatly reduces the maximum load on the nerve. Moreover, the lyophilization process will destroy the fiber connections between the endoneurium and the perineurium connective tissue and the subsequent application of enzymes further damages the protein connections between them. Therefore, it is possible to obtain acellular scaffolds with reduced mechanical properties. In addition, there is no rejection reaction during nerve autotransplantation. Allogeneic transplantation requires antigen comparison or antigen removal before transplantation, whereas xenotransplantation is associated with a large rejection reaction, and generally the graft does not survive in the recipient without special treatment[19]. Tissue engineering reduces the content of cells from primary immunogenic sources by acellular decellularization to reduce post-transplant rejection. The anti-degradation performance and histocompatibility of the lyophilized acellular scaffold in the present study were excellent. The results indicated that the receptor was ultimately able to tolerate the implantation of the scaffold and the local response of the receptor after xenotransplantation was not severe. The inflammatory response began to weaken six weeks after surgery and the scaffold ultimately retained relatively good continuity.

Following the successful fabrication of the acellular scaffold, we considered the repair of the nerve. SCs are constituent cells of the myelinated nerve fiber myelin sheath, which are widely present in the peripheral nervous system of vertebrates[20]. They function by secreting neurotrophic factors and peripheral nerve matrixes; however, SCs have few sources and are difficult to culture. Moreover, as mature cells, their proliferative potential is not as active as stem cells[21]. As seed cells, SCs do not meet the needs for clinical treatment. In contrast, WJ from the umbilical cords of newborns is rich in hUC-MSCs. Although they are cells from newborns, they are located outside of the body, which are extremely convenient to obtain non-invasively and are rich in cell content. Compared with bone marrow MSCs, hUC-MSCs are more primitive and have a greater differentiation potential. After subculturing hUC-MSC for three generations, the morphology of umbilical cord-derived cells tended to stabilize. We used an antioxidant-induced growth factor co-induction scheme for induction. After induction and culture, the morphology of hUC-MSCs began to change. The cells began to lengthen to form a long-shuttle shape and grew into a synapse-like structure, compounding the morphological structure of the SCs. Fluorescence staining revealed that most of the cells were differentiated in the direction of SCs. In addition, Schwann-like cells from hUC-MSCs are very similar to true SCs and the experimental Schwann-like cells expressed MPZ by enhancing intracellular cAMP levels. We found that Schwann-like cells when co-cultured with PC12 cells induced rapid differentiation of PC12 cells to form myelin structures with PC12 neurites. These results provide an experimental basis for the use of hUC-MSCs as seed cells for peripheral nerve scaffolds in the future.

CONCLUSION

In summary, by combining lyophilization and enzymatic digestion, we successfully prepared acellular peripheral nerve scaffolds with complete cell depletion, intact matrix preservation, good histocompatibility between different species, and extremely low cytotoxicity. In addition, hUC-MSCs can differentiate into Schwann-like cells with myelinogenic ability after in vitro induction. However, there were some limitations to this study. It was the first attempt to prepare peripheral nerve scaffolds. Although seed cells more suitable for neonates were found, neonatal animal models or models with neonate-specific PNI were not used. In follow-up experiments, it will be necessary to prioritize the production of neonatal and animal models of obstetrical palsy before further research on treatment.