The study design included in vitro and in vivo studies (Table 1, 2). All procedures described in this study were approved by the Ethics Committee of the Chinese PLA General Hospital (No. 2016-x9-07). All experiments were performed at the Beijing Key Laboratory of Regenerative Medicine in Orthopaedics, Chinese PLA General Hospital, Beijing, China.
In vitro study
1. Preparation of PRP
PRP was prepared by the method previously reported (Figure 1, step 1). A total of 8 mL whole blood of rat was collected via cardiac puncture and placed in a 10ml sterile tube containing 3.8 % w/v sodium citrate. The collected blood was immediately centrifuged at 400 g for 10 minutes. Following centrifugation, three layers were formed: the upper layer, that was acellular plasma; middle layer, containing abundant platelets; and bottom layer, with rich red blood cells. The supernatant and a small portion of the transition zone (buffy coat) were carefully transferred into another non-anticoagulant sterile tube and centrifuged at 800 g for 10 minutes, yielding approximately 1 ml of PRP. For quality testing, PRP sample and whole blood sample were sent to the laboratory with 200 μl each. The number of platelets, red blood cells, and white blood cells was counted. Further experiments were performed when the platelet concentration in PRP was 4 ~ 6 times of that in whole blood and the amount of erythrocytes and leukocytes was extremely low.
PRP was activated with a mixture of bovine thrombin (Sigma, T4648) and 10 % calcium chloride (Sigma, c1016) solution. After a few seconds, the PRP clot was formed, allowing the clot to retract for about 15 min to release enough growth factors. Next, the supernatant containing a large amount of growth factors was obtained by centrifugation at 2a800 g for 10 minutes. The supernatant was stored in a refrigerator at -80 °C until it was used to supplement the culture medium for further experiments.
2. Preparation and identification of active nerve microtissues
Nerve microtissues were harvested from the sciatic nerves of 3-day-old Sprague–Dawley (SD) rats (Vital River Laboratory Animal Technology Co., Ltd., Beijing, China; license number SCXK (Jing) 2016-0006) according to established method. Briefly, after the 15 rats were sacrificed and sterilized, the sciatic nerves were exposed and dissected through longitudinal surgical incisions behind the femurs on both sides. The epineuria of sciatic nerves were first removed and then the nerves were minced into 1 mm x1 mm fine particles (nerve microtissues). Nerve microtissues were cultured in DMEM/F-12 (Gibco, USA) containing 10% (v/v) fetal bovine serum (FBS; Gibco), 1 % penicillin–streptomycin solution (Gibco), 10 ng/ml heregulin-β1 (Sigma-Aldrich) and 2 mM forskolin (Sigma-Aldrich) in five six-well culture plates. Then, this was followed by incubation in an atmosphere of 5% CO2 at 37 °C for 1, 3, 5, 7, and 10 days. Culture medium were collected and stored in a refrigerator at -80 °C at each time point for further testing. The live nerve microtissues were identified by FDA/PI bichromatic fluorescence staining and the SCs derived from microtissues were identified by S-100 immunofluorescence staining at each time point. In addition, SCs in each well were enzymatically hydrolyzed with 300 ul 0.2 % (w/v) collagenase NB4 (Sigma-Aldrich) for 1 min. The mixture of DMEM/F-12 mixture containing 10% FBS was used to terminate enzymatic hydrolysis reaction, and then counted the SCs numbers.
3. Determination of neurotrophic factors secreted by nerve microtissues
The collected culture mediums at 1, 3, 5, 7, and 10 days were used for determining the concentration of the neurotrophic factors secreted by live nerve microtissues. The concentrations of nerve growth factor-β (NGF-β), vascular endothelial growth factor (VEGF), brain-derived neurotrophic factor (BDNF) and glial cell line-derived neurotrophic factor (GDNF) were determined using an NGF-β ELISA kit (Boster, EK0471), a VEGF ELISA kit (Boster, EK0540), a BDNF ELISA kit (Boster, 0308) and a GDNF ELISA kit (Boster, EK0363) as instructed by the manufacturer.
4. Schwann cells were co-cultured with nerve microtissues and/or PRP supernatant
Microtissues and PRP supernatant were harvested following the procedures described above. Schwann cells were harvested as follows. Briefly, the sciatic nerves with the epineuria removed and were minced from 10 SD rats of 3-day-old and then enzymatically dissociated with 1 ml of 0.2 % (w/v) collagenase NB4 (Sigma-Aldrich) for about 10 min in a 37 ℃ incubator and stirred with a magnetic stirrer. After that, the mixture was centrifuged for 5 min with 1500 rpm/min and resuspended in SCs medium and transferred into a 25 cm2 cell culture flask that was incubated in an atmosphere of 5 % CO2 at 37 °C. The third passage (P3) cells were utilized in this study.
The SCs were placed in the lower chamber of the twelve-well transwell co-culture system. A total of 8×104 cells were seeded on the cell slide in each well, maintaining with 2 ml of serum DMEM/F-12 medium for 24 h and allowed to attach overnight. Subsequently, SCs were co-cultured with nerve microtissues and/or PRP supernatant. SCs medium was added to the upper chamber as control group. The nerve micro-tissues, PRP supernatant and the mixture of nerve microtissues and PRP supernatant served as corresponding three experiment groups were added to each upper compartment in quadruplicate. After co-cultured for 1, 3, 5 and 7 days, one of the wells in each group was enzymatically hydrolyzed with 200ul 0.2 % (w/v) collagenase NB4 (Sigma-Aldrich) for 1 min, and then DMEM/F-12 mixture containing 10 % FBS was added to terminate enzymatic hydrolysis reaction. Finally, the cell count was performed.
At the different time point, the 2 ml DMEM/F-12 medium and 200 μl of Cell Counting Kit-8 (CCK-8, Dojindo Laboratories, Shanghai) colorimetric assay reagent were added to each well to evaluate the effect of co-culture on SCs proliferation. The optical density (OD) value was measured at 450 nm by using a microplate reader (Epoch, Biotek, US). The absorbance was directly proportional to the number of living cells. Furthermore, the morphology and distribution of SCs in different groups were detected by S-100 immunofluorescence staining.
5. Dorsal root ganglions (DRGs) were co-cultured with microtissues and/or PRP supernatant
The DRGs were extracted from three 12-h-old SD rats. Briefly, the SD rats were firstly sacrificed and sterilized by soaking in 75% ethanol solution. The SD rat was placed in a prone position, then the skin was cut open along the midline of the back, and the blood was washed with DMEM/F-12 after the spine was completely removed. The spine was then divided into two halves with microscissors along the midsagittal segment under the microscope to expose the bilateral intervertebral foramen, and the DRG was removed completely. The epineurium of DRG was carefully stripped in the frozen DMEM/F-12 medium supplemented with 10 % FBS, and care was taken not to clamp the body of DRG. Finally, DRGs were inoculated on the cover glass of the lower chamber of the twelve-well transwell co-culture system, and an appropriate amount of DRG culture medium was added. The DRG medium, PRP supernatant, nerve microtissues and the mixture of nerve microtissues and PRP supernatant were added in the upper chamber of transwell co-culture system, respectively. They were placed and cultured in an incubator at 37 ℃with 5% CO2. After 5 days, the insert was carefully removed, and the DRGs were fixed with 4 % paraformaldehyde. The regenerative capacity of axons was detected by S-100 and NF-200 immunofluorescence staining. Five DRGs were randomly selected from each group and were integrally photographed. Each DRG image was divided into four quadrants, and the longest five axons in each quadrant were measured. Image Pro Plus 6.0 (IPP 6.0, Media Cybernetics, USA) image analysis software was used to calculate the average maximum axon length of each DRG.
6. Schwann cells migration study
In order to test the capacity of the PRP supernatant, neurotrophic factors secreted by nerve microtissues, and a combination of two to induce SCs migration, six-well transwell systems with 8 μm pores (Corning Costar, USA) were used in this study. A total of 1.5x104 SCs were added in each of the upper chamber. Serum medium (FBS group), serum medium with 500 μl PRP supernatant (PRP group), 5-day nerve microtissues medium 500 μl (Micro-T group), or medium with the latter two samples (Micro-T+PRP group) were added to the lower chambers. After incubation in a humidified atmosphere (37 ℃, 5% CO2) for 12 h, the upper surface of each membrane was cleaned with a cotton swab. The migrated SCs adhered to the underside of the membrane, which were fixed with 4% paraformaldehyde and stained by crystal violet solution. Five randomly selected visual fields (200 x magnification) were captured from each slide to calculate the number of migrated SCs using Image Pro Plus 6.0 software. The migration ratio of SCs refers to the ratio between the number of SCs migrated in each experimental group and the number of SCs migrated in FBS group.
In vivo study
Seventy-five 4-month-old healthy, male and clean New Zealand white rabbits weighing 2.5 to 3.0 kg were provided by the Animal Breeding Centre of Long’ an, Beijing, China (licence no. SCXK [Jing] 2014-0003). The rabbits were housed individually in cages at room temperature with a 12-hour light/dark cycle. They were received food and water ad libitum. The rabbits were randomly divided into five groups with 15 rabbits in each group. The groups included (1) a Hollow group in which the defect of tibial nerve was repaired using a autogenous vein with saline infusion; (2) a PRP group, autogenous vein was filled with autogenous PRP to repair the transection injury of tibial nerve; (3) a nerve Micro-tissue (Micro-T) group, autogenous vein was filled with autogenous tibial nerve microtissues to repair the transection injury of tibial nerve; (4) a Micro-T+PRP group, autogenous vein was filled with the mixture of autogenous tibial nerve micro-tissues and PRP to repair the transection injury of tibial nerve; (5) an Autograft group, the defect of tibial nerve was repaired using an excised autogenous nerve graft from the same locale.
2. Autologous PRP preparation
Before surgery, 8 ml of whole blood was extracted from the central ear artery of rabbits in the PRP and Micro-T+PRP groups. The preparation procedures of PRP was the same in vitro study. The prepared PRP (1 ml) was stored in a -80 °C refrigerator for use during surgery . (Figure 1, step 1)
3. Surgery protocol
Rabbits were first weighed and then intramuscular injected with 3% pentobarbital sodium solution (1 ml / kg) for anesthsia. The right lower extremity of all animals was shaved and disinfected with iodine solution. All surgical procedures were performed under aseptic operating conditions by two surgeons. A 18-mm-long autogenous vein graft was harvested from the superficial subcutaneous vein at the surgical approach below the right femur. The blood clot inside the vein was removed by rinsing in saline solution. The vein was then stored temporarily in normal saline. Next, three major fascicles (tibial nerve, common peroneal nerve and sural nerve) of the right sciatic nerve were exposed by a gluteal muscle-splitting incision that was between the vastus lateralis and the biceps femoris muscles. A 12-mm-long tibial nerve was excised at the midthigh and a 12-mm-long gap was created. Since the vein was retracted after harvest, the vein that was excised is slightly longer than the length of the nerve defect. The distal and proximal stumps of the autologous vein were reversed and sutured to both ends of the tibial nerve, and 1 mm of each nerve stump was inserted into the vein graft. Every effort was made to avoid tension and keep correct rotational alignment throughout. (Figure 1, step 2)
The tibial nerve defect was repaired only by the autologous vein graft with saline infusion, avoiding the collapse of the vein wall.
The collected PRP was taken out from a -80 °C refrigerator and restored to room temperature. The 300 μl PRP was simultaneously injected into the vein with the activator (the mixture of 10 % calcium chloride (Sigma, c1016) solution and 1,000 units of bovine thrombin (Sigma, T4648)). The PRP was activated and formed a PRP gel in the vein.
After the excision of a 12-mm long tibial nerve, it was divided into three equal parts (4-mm/part), one of which was stripped of the epineurium and cut into nerve microtissues under sterile conditions. The minced nerve microtissues were distributed equally in the lumen of the vein graft.
The 4-mm length of microtissues stripped of the epimembrane was mixed with 300 μl PRP and injected into the vein simultaneously with the activator. The mixture was activated in the vein to form a gel that avoided leakage of microtissues.
The excised tibial nerve was reversed and sutured to both ends of the tibial nerve to repair the defect from the same locale.
After the animal model was established, the muscular layer and skin were sutured with 3-0 monofilament nylon and disinfected with iodine solution.
4. Postoperative care
Postoperatively, all rabbits were provided free access to water and standard rabbit nutrients. Animals were examined twice daily for 10 days to check for wound healing and infection, and were recorded in the laboratory records. Postsurgical infection was controlled by injection of antibiotics (800,000 IU of penicillin daily) intramuscularly for 5 days. The animals were monitored by a specialist veterinarian under standard laboratory conditions during the 3-month postoperative period.
5. Nerve recovery assessment
5.1 Nerve function evaluation
The sciatic nerve function was evaluated at 12 weeks. The toe spreading score (graded from 1 to 4 points) and the modified Tarlov score (rated from 0 to 4 points) were used to evaluate nerve function. Higher scores indicated better nerve function recovery.
5.2 High-frequency ultrasound and contrast-enhanced ultrasonography (CEUS) examination of vein grafts
All of the ultrasound procedures were performed by a radiologist with nine years of ultrasound experience. All machine settings, such as depth, gain and focus, were kept constant during each measurement. The thickness of vein grafts was examined at 2 weeks and 12 weeks after operation and the perfusion of vein grafts was examined at 2 weeks after operation. In short, after the animals were anesthetized, the vein grafts were scanned longitudinally using a high-frequency ultrasonic equipment (Vevo 3100, Visualsonics, Canada) with a 14- to 28- MHz linear array probe (MX250). The probe was then rotated 90° to measure the thickness of the vein grafts in a cross section. The mean value of 3 measurements was taken for statistical analysis. (Figure 1, step 3)
Ultrasonic contrast agent is pure blood pool contrast agent, has a unique advantage for tissue microcirculation imaging. In this study, the CEUS examination, a method for detecting the regenerated microvessels and their perfusion in the vein grafts, was performed using a high-resolution ultrasound system (Mindray, Resona 7) at a low mechanical index (MI 0.05–0.07). This device was equipped with a linear array transducer (4–15 MHz). The procedures of CEUS examination was the same as previously reported[14b]. SonoVue (Bracco International, Milan, Italy), a sulfur-hexafluoride-filled microbubble contrast agent, which was encapsulated by a flexible phospholipid shell. It is very safe for use in animals or humans because adverse allergic reactions are rare, and it can be exhaled within 15 minutes after intravenous injection. At 2 weeks after the operation, SonoVue (mixed with 5 ml of saline) was injected into a peripheral ear vein at a bolus of 0.13 ml/kg, followed by a 2 ml saline flush for the CEUS examination. Dynamic ultrasound videos stored continuously for at least 60s were used to analyze the time to peak (TTP), peak intensity (PI) and area under the curve (AUC) of region of interest (ROI), which were all parameters reflecting blood perfusion in early postoperative vein grafts. The mean results of 3 repeated analyses of a vedio were presented.
5.3 Quantitative real-time RT-PCR (qRT-PCR)
At 2 weeks following the treatment, total RNA of vein/nerve grafts was isolated using RNA extraction solution (Servicebio, G3013). The FastStart Universal SYBR Green Master (Rox) (Servicebio, G3008) was used to assess VEGF gene expression. The glyceraldehyde-3-phosphate-dehydrogenase (GAPDH) served as reference household gene used to normalize the amount of mRNA. Then the sequence of the primers was selected and carefully checked as follows: NGF, forward 5'-CATCACTGTGGACCCCAAACTGT-3', reverse 5'-GTCCGTGGCTGTGGTCTTATCTC-3'; GAPDH, forward 5'- CTGGAGAAACCTGCCAAGTATG-3', reverse 5'- GGTGGAAGAATGGGAGTTGCT-3'. Relative changes in gene expression were calculated using the comparative Δ crossover threshold (CT) method.
5.4 Macroscopic evaluation of vein graft adherence
The nerve repair sites in animals from each group were evaluated at 12 weeks at the end of ultrasound follow-up with the animals under deep anesthesia. The assessment of vein graft adherence was performed blindly and followed an established numeric grade scheme. The degree of venous graft adherence was divided into three levels. No dissection or mild blunt dissection (grade I); some vigorous blunt dissection required (grade II); and sharp dissection required (grade III).
5.5 Electrophysiological recovery evaluation
A fully functional electromyography (EGM) machine (Keypoint, Medtronic) was used for the electrophysiological evaluation. At 12 weeks, after macroscopic evaluation of vein graft, electrophysiological tests were performed to assess the regeneration of myelinated nerve fibers. After a moderate dose of anaesthetic was administered to the rabbit, the tibial nerve operating area in the right thigh, the tibial nerve in the healthy side and the bilateral triceps surae muscles were exposed. The two stimulation electrodes (6 Hz, 5 mA) were placed in order at the proximal and distal ends of the vein grafts, and the monopolar recording electrode was placed on the belly of the triceps surae muscle of the same side to induce and record the compound muscle action potentials (CMAPs). The CMAPs of the normal tibial nerve also need to be detected and recorded. The amplitude and latency of bilateral CMAPs were recorded 5 times for each animal, and the analysis results were expressed as the ratio of the injury side to the normal side. (Figure 1, step 4).
5.6 Histological evaluation of regenerated nerves at early stage
Animals were sacrificed by overdose injection of a sodium pentobarbital at 4 weeks after vein transplantation. The implanted vein grafts were isolated from the surrounding tissue and divided into three parts after the integral excise. The middle and distal transversal segment of grafts was rapidly fixed in 4% paraformaldehyde for 2 hours and then embedded in paraffin. The 4 µm thick transversal sections were obtained H&E staining and immunofluorescence staining.
Commercial H&E staining kit (G1120, Solarbio) was used for H&E staining, followed by a series of routine dehydration, transparency and sealing steps. As for immunofluorescence staining, the sections of each vein graft were blocked at room temperature with 10 % goat serum for 1 hour after washing three times (5 min/time) each time with PBS. Rabbit anti-S100 antibody (1:200, bs-2015R, Bioss) and mouse anti-Neurofilament 200 antibody (1:200, N5389, Sigma) were, respectively, applied as the primary antibodies and incubated in a humidified chamber overnight at 4 ℃. Next morning, the remaining liquid was removed from the sections and washed three times in PBS. Then, the sections were incubated with goat anti-rabbit IgG H&L (Alexa Fluor 488, 1:200, ab150077, Abcam) and goat anti-mouse IgG H&L (Alexa Fluor 594, 1:200, ab150116, Abcam) secondary antibody in the dark for 2 hours at room temperature. After washing with PBS for three times (5 min/time), the nuclei of SCs were counterstained with DAPI (1:200). A fluorescence microscope (200× magnification, Nikon Eclipse C1, Japan) and IPP 6.0 software were used for capturing images (5 random fields were selected from each slice) and analyzing the mean density of regenerated axons of each group at 4 weeks, respectively.
5.7 Morphometrical assessment of regenerated nerves
The semithin sections and ultrathin sections were acquired to evaluate the morphology of regenerative nerves (fiber diameter, and myelin sheath thickness). At 12 weeks, animals in each group were randomly sacrificed and 5 mm length of regenerative nerves at the end of distal vein grafts were removed. The excised nerves were rapidly fixed in precooled 2.5 % (w/v) glutaraldehyde for 3 hours, then in 1 % (w/v) osmic acid solution for 1 hour, finally washed, dehydrated through a series of grades of ethanol solutions, and embedded in epoxy resins. The nerve segments were cut into 1.5 μm-thick semithin slices with a glass knife and 70 nm-thick ultrathin slices with a diamond knife.
The semithin sections were stained with Toluidine Blue solution (1% in sodium borate, G3663, Solarbio), and then five fields at 400x magnification were randomly selected for each animal. The mean density of myelinated nerve fibers was counted by IPP 6.0 software. Furthermore, the morphology of myelin sheath, diameter of fiber, and myelin sheath thickness were observed in ultrathin sections and analysed by IPP 6.0 software.
6. Triceps surae muscle recovery assessment
6.1 Multimodal ultrasound evaluation
6.1.1 High-frequency ultrasound evaluation
At 12 weeks, high-frequency ultrasound equipment with a 21- to 44- MHz linear array probe (MX4000, Vevo 3100, Visualsonics, Canada) was used for assessing echogenicity and thickness difference (measured in maximal cross-sectional area, CSA) of innervated targeted muscle (triceps surae muscle) in each group. The mean value of 3 measurements was taken for statistical analysis.
6.1.2 Shear wave elastography (SWE) and Angio PlaneWave UltrasenSitive (AngioPLUS) imaging evaluation
At 12 weeks, the stiffness and the microvascular flow of the triceps surae muscle were detected and measured with the Aixplorer system (Supersonic Imagine, Aix-en-Provence, array transducer Super Linear L15-4, France) equipped with SWE and AngioPLUS tecniques. Different from other ultrasound techniques, the images displayed by SWE and AngioPLUS tecniques can be diaplayed simultaneously on the screen, so that the relationship between the degree of the target muscle fiber tissue hyperplasia, that is, the stiffness variation and the changes of microvascular flow density in the muscle can be obtained.
Double image mode was adopted for detection. The ROI area (a circular area of 10 mm in diameter) was selected at the mid-belly of the triceps surae muscle in a longitudinal plane of ultrasound to measure the stiffness and microvascular flow density of the muscle at the same time. The image was frozen when the map was stable in 3-5 s. The mean Young’s modulus values were averaged for three measurements. The interval between every two measurements should be at least 5 s. In addition, the blood flow frequency spectrum was used to confirm that the images presented were microvessels but not artifacts. The value of Young’s modulus represents the stiffness. The amount of microvascular flow density (total area of microvascular flow/ ROI area) was analyzed by IPP 6.0 software, and the mean value of three measurements were taken in each case for the statistical analysis. The monochrome pattern of the microvascular muscle was also displayed. Finally, the correlation analysis was performed between Young’s modulus values and microvascular flow density.
6.1.2 CEUS evaluation
CEUS examination of targeted muscles was performed at 12 weeks after the operation, and the examination method was similar to that of vein grafts. Briefly, first placed the probe at the maximal CSA of the triceps surae muscle on transverse imaging plane, then rotated the probe (4–15 MHz) at a 90° to obtain a longitudinal imaging plane for CEUS examination. The contrast agent (0.23 ml/kg) was then administered as a bolus injection through the peripheral ear vein. As soon as the contrast agent injection began, digital cine loops were stored for at least 60 s. Two ROIs of the same size were used for blood perfusion analysis in all animals. The video of each animal was analyzed three times and averaged for final statistical analysis.
6.1.3 Macroscopic evaluation
At 12 weeks, after multimodality ultrasound examination, the animals were sacrificed in each group. Bilateral triceps surae muscle were removed and weighed immediately. The wet weight recovery rate of the muscle was the ratio of the wet weight of the muscle from operative side to that of the normal side.
6.1.4 Morphological evaluation
At 12 weeks after surgery, the H&E and Masson’s trichrome staining were performed for detecting the morphology of the muscle. In other words, H&E staining was used to observe adipocytes infiltration and nuclear distribution in muscle, and Masson’s trichrome staining was used to evaluate collagen proliferation in muscle tissue and atrophy recovery of muscle fibers. The triceps surae muscles in each group were harvested and fixed in 4 % paraformaldehyde for 2 hours and then embedded in paraffin. The 4 µm thick transversal sections were obtained. The method of H&E staining was the same as that of nerve tissue, and the Masson’s trichrome staining was performed by using a modified Masson’s trichrome stain kit (G1345, Solarbio). Five randomly selected fields in each slide at 200x magnification were captured to measure the average area of muscle fibers and positive area percentage of collagen with IPP 6.0 software. The mean value of each slide was used for the final statistical analysis.
7. Statistical analysis
Statistical analyses were performed using Statistical Program for Social Sciences (SPSS) software (version 22.0) and GraphPad 8.0 (Graphpad Software, Inc. San Diego, CA, Unit). The Kolmogorov-Smirnov test was used for a normal distribution test of the data. If the data was normally distributed and the variance was uniform, the differences between two groups were compared by Student’s t-test and the multiple comparisons were tested by one-way ANOVA analysis. Tukey’s multiple comparison post hoc test was applied when P>0.05 in the test of homogeneity of variances; otherwise, Dunnett’s T3 post hoc test was applied. Pearson’s correlation analysis was performed to analyze the correlation between the two variables. Statistically significant was defined as P<0.05 between groups. According to the post hoc power analysis, a power of the main indicators of each group was >80% at 12 weeks after the operation, with a significance level of 0.05, indicating that no additional animals were needed.