2.1 SCAPs Culture

Human stem cells from apical papilla (SCAPs) were gifted by Dr. Anibal Diogenes at the University of Texas, Health Science Center, Department of Endodontics [18]. These cells were grown in alpha modified Minimum Essential Medium (A-MEM) containing 10% fetal bovine serum and 1% penicillin–streptomycin. The cells were kept at 37 °C in an incubator with 5% CO2. The medium was replaced every 2–3 days, and cells were subcultured when they reached 80% confluence.

2.2 Chemical Cocktail-Based SCAP Differentiation into Neural Lineages

At passage six, SCAPs were seeded onto 6-well culture plates (20,000 cells/cm2). Neural induction followed the protocol by Hu et al. [9] with some modifications, such as shortening the duration from 8 to 7 days. When SCAPs reached 60% confluence, the medium was changed to neural induction medium (NI; consisting of DMEM/F12: Neurobasal A [1:1] with 0.5% [v/v] N2, 1% [v/v] B27, 100 mM cAMP, 20 ng/mL basic fibroblast growth factor) supplemented with/without the chemical cocktail VCRFSGY (0.5 mM valproic acid, 3 µM CHIR99021, 1 µM Repsox, 10 µM forskolin, 10 µM SP600125, 5 µM GO6983, 5 µM Y-27632). The culture medium was refreshed on the third day.

2.3 Immunocytofluorescent of SCAP

Cells were fixed with 4% (v/v) paraformaldehyde in phosphate-buffered saline (PBS; pH 7.4) at room temperature for 20 min. Cells were then washed three times with PBS, permeabilized with PBS containing 0.5% (w/v) Triton X-100 for 5 min at 4 °C, and blocked in PBS containing 5% (w/v) bovine serum albumin (Sigma-Aldrich; Burlington, MA, USA) at room temperature for 1 h. The fixed samples were incubated with primary antibodies against β-Tubulin III (Tuj1; 1:500; cat. No. ab18207; Abcam, Cambridge, UK). The samples were washed in PBS and incubated with secondary Alexa Fluor 488-conjugated goat anti-rabbit (cat. No. ab150077; Abcam) for 1 h at room temperature in the dark, followed by washing with PBS. The cell nuclei were stained with DAPI for 5 min at room temperature and the samples were imaged using confocal microscopy (LSM710; Carl Zeiss AG, Jena, TH, Germany) at X200 magnification.

2.4 Cell Sheet Formation

The SCAPs were seeded at 20,000 cells/cm2 in a 35 mm UpCell dish (NunC, Rochester, NY, USA). After reaching confluence, the original culture medium was discarded and replaced with A-MEM, NI, and neural induction medium with small molecules (SM), respectively. After 3 days of induction, SM was replaced with only the NI for an additional 2 days.

On day 5, the first layer of the cell sheet was washed once in warm Phosphate-buffered saline (PBS), and 100 µL of the appropriate medium was added before placing a membrane onto the cell layer at room temperature to allow the cell sheet to attach to the membrane (30 min for A-MEM group, 90 min for NI and SM groups). After washing with PBS and adding 100 µL of the appropriate medium, the second cell sheet was placed on the membrane with the first cell sheet layer on top of the second cell sheet. The UpCell dish containing the two layers of cell sheets was placed back into the incubator (37 °C, 5% CO2) for 90 min. During this period, a third cell sheet was attached to a membrane in a similar manner to the first cell sheet at room temperature. After 90 min, 1 mL of the appropriate medium at 37 °C was added for 3 min to detach the two-layer cell sheet from the membrane. Next, 700 µL of the appropriate medium at 37 °C was added to the two-layer cell sheet. The third cell sheet was then added on top of the two-layer cell sheet and then placed back into the incubator for 90 min. After incubation, 2 mL of the appropriate medium was added to detach the three-layer cell sheet from the membrane.

After removing the membrane, 1.5 mL of medium was removed, leaving 0.5 mL of medium in the dish to allow the cell sheet to detach from the UpCell dish at room temperature (excessive medium will result in the cell sheet curling prematurely). The cell sheet was curled into a cylinder by using curved, smooth-surfaced micro forceps while tilting the dish at a 45-degree angle. The remaining medium was replaced with 500 µL of fresh, appropriate medium, and the cell sheet cylinder was placed back into the incubator for 1 day before surgery.

2.5 Animal Model of Severe Sciatic Nerve Injury

Sixty Male Sprague–Dawley rats (weighing approximately 250 g at the time of surgery) were housed in individual ventilated filtered cages (maximum of two rats per cage) under standard animal laboratory conditions with a controlled temperature of 25 °C, 12 h light/dark cycle, and food and water ad libitum. A 5 mm section of the left sciatic nerve was removed using diamond-coated micro scissors. A 7 mm long nerve guide (NeuraGen; Integra lifesciences, Princeton, NJ, USA) with an internal diameter of 1.5 mm was used to bridge the resulting gap. The rats were randomized into five experimental groups, with 12 rats in each group. Due to severe autotomy, the remaining rats in each group were as follows: empty nerve conduit (Empty) = 6, Reverse direction autograft (Reverse) = 4, cell sheet cultured in A-MEM (A-MEM) = 6, cell sheet cultured in NI (NI) = 7, and cell sheet cultured in neural induction with small-molecule cocktail (SM) = 5.

The nerve guide was split open horizontally with sharp micro scissors under sterile conditions in a culture hood. The cell sheet (trimmed to 5 mm using a surgical blade) or autografts soaked in saline were inserted into the nerve guide. After inserting the cell sheet or autograft into the middle of the nerve guide, the horizontal cut of the nerve guide was closed by suturing the opening with 8–0 nylon suture 1.5 mm away from the two ends. Next, 1 mm of the proximal and distal stumps of the sciatic nerve were each sutured at the lumen wall with 8–0 nylon suture and pulled into the nerve conduit. In the Reverse direction autograft group, the 5 mm excised nerve was inserted into the nerve guide in the opposite orientation to simulate a misaligned nerve recovery model. The muscle layer was sutured with 3–0 stitches (multifilament absorbable suture). The skin was clamped with 9 mm stainless steel reflex clips. All groups were treated daily with pharmaceutical-grade cyclosporine-A (10 mg/kg body weight) injected subcutaneously starting 24 h before surgery and fed paracetamol syrup (65 mg/kg body weight) until the end of the study.

2.6 Toe Spread Test

The toe spread test was used to evaluate motor recovery at week 12 post-surgery according to previous protocols with some modifications [19, 20]. Toe spreading was assessed by photographing the animal's paws inside a clear acrylic container. The camera was positioned beneath the transparent base of the container to capture images of the plantar surface of the paws.

A five-level scoring system was employed to analyze the toe spread functionality: a score of 0 denotes no observable toe spread; 1 denotes one toe spread out; 2 denotes two toes spread out; 3 denotes all four toes separated, but to a lesser degree than the unaffected side; and 4 denotes normal toe spread. A score of 0 is assigned if the assessment cannot be conducted, for example, due to severe toe flexion contracture. Additionally, the severity of toe contracture was evaluated using a three-level scoring system: a score of 0 denotes no contracture; 1 denotes mild contracture with visible toe spread; and 2 denotes severe contracture without any discernible toe spread.

2.7 Walking Track Analysis

The rats were tested in a modified elevated plus maze (75 cm long, 15 cm wide), elevated above the ground (75 cm). Two arms were guarded by a wall (5 cm high), and the other two arms had a shallow wall (3 cm high). Access to the closed arm was blocked by inserting two detachable walls (5 cm high) to form a straight, long path between the two open arms. The start of one side of the opened arm was blocked and covered at the center of the maze to form a black room. A piece of white paper (90 cm long, 15 cm wide) was placed on the bottom of the track. Chinese calligraphy ink was applied onto the plantar surface of the rat's hind feet with a brush, and the animal was allowed to walk down the track, leaving their hind footprints on the paper. Various measurements were obtained from the footprints, including: (1) the print length (PL), which is the distance from the heel to the third toe; (2) the toe spread (TS), which is the distance between the first and fifth toes; and (3) the intermediary toe spread (ITS), which is the distance between the second and fourth toes. These measurements were collected from both the experimental (E) and normal (N) sides. Rats were made to walk down the path until 3 complete gait movements without the animal pausing midway through the gait movements were recorded and used to calculate average SFI. The SFI was calculated as described by Bain et al. [21] according to the equation:

$$\mathrm= -38.3 \left(\frac}-}}}}\right)+109.5 \left(\frac}-}}}}\right)+13.3 \left(\frac}-}}}}\right)-8.8$$

The tracks were recorded at baseline before surgery and at weeks 2, 4, 6, 8, 10, and 12 after surgery.

2.8 Gastrocnemius Muscle Weight

Upon sacrificing the animals at week 12, the wet muscle weight of the gastrocnemius muscle on both the experimental and normal sides was measured using an electronic weighing machine. The wet weight ratios of experimental side in relation to the normal side were calculated to evaluate the functional recovery of sciatic nerve.

2.9 Immunofluorescent Analysis of the Nerve Conduit

Nerve conduits were harvested and fixed in 4% (v/v) paraformaldehyde in PBS (pH 7.4) overnight at 4 °C. The conduits were cryoprotected sequentially in 15% and 30% sucrose in PBS (pH 7.4) for 2 days and then snap-frozen in liquid nitrogen before storing at − 80 °C. Before cryosectioning into 10 μm sections, the nerve conduit was split in half. One half was sectioned longitudinally, while the other half was sectioned transversely. Transverse sections were separated into the mid-section of the nerve conduit (mid), the contact point between the cell sheet and rat sciatic nerve stump (joint), and the rat sciatic nerve itself (connecting).

Fixed tissues were rehydrated in PBS for 10 min on ice, permeabilized with PBS containing 0.3% (w/v) Triton X-100 for 10 min on ice, and then blocked in PBS containing 4% (w/v) bovine serum albumin (Sigma-Aldrich) and 0.3% (w/v) Triton X-100 for 1 h at room temperature. The fixed samples were incubated with primary antibodies against S100 calcium-binding protein B (S100B; 1:500; cat. no. ab52642; Abcam) and β-Tubulin III (Tuj1; 1:500; cat. no. NB100-1612; Novus, Centennial, CO, USA) overnight at 4 °C. The slides were washed in PBS and incubated with secondary Alexa Fluor 488-conjugated goat anti-chicken (cat. No. ab150173; Abcam) and Alexa Fluor 647-conjugated goat anti-rabbit (cat. No. #4414; Cell Signaling, Danvers, MA, USA) for 1 h at room temperature. Finally, nuclear staining was performed by adding DAPI mounting medium (cat. No. ab104139; Abcam). Images were captured at X100 magnification using a confocal microscope, LSM 900 Airyscan 2 (Carl Zeiss AG).

The total number of myelinated fibers on each transverse section of the mid, joint, and connecting sections was calculated by creating a trained segmentation model using ZEISS Arivis cloud’s (Carl Zeiss Microscopy GmbH, Jena, TH, Germany) deep learning AI segmentation, specifically to recognize nerve fibers encased within a myelin sheet. The trained model was then applied to the image analysis module of ZEN 3.8 (Carl Zeiss Microscopy GmbH) for image segmentation and analysis for accurate counting of myelinated fibers within each section.

2.10 Statistical Analysis

Data are presented as mean ± standard deviation (SD). Significant differences were assessed using one-way analysis of variance (ANOVA). Due to differences in the sample sizes across the groups in our study caused by severe autotomy, the Gabriel test was employed to evaluate the significance of the differences between groups. The correlation coefficients between the gastrocnemius muscle ratio and the total number of myelinated fibers in the connecting region were calculated using Pearson's correlation analysis. All statistical analyses were carried out using SPSS version 29.0 (Chicago, IL, USA).