Peripheral nerve injuries (PNIs), thought to be present in nearly 3% of all trauma patients, represent a major socioeconomic burden.1 If severe, with a long segmental gap, these injuries present a difficult clinical problem for peripheral nerve surgeons.
One particular issue with lower-extremity PNIs arises from a lack of autologous nerve material for repair, or an equally effective alternative. Patients with such injuries must undergo extensive sural nerve harvesting via additional incisions, which increases morbidity. Alternatives to nerve autografts, such as hollow guidance channels and decellularized nerve allografts, have been investigated extensively. Unfortunately, these are not well suited for large-diameter and/or long-segment nerve repairs, due to the lack of cellular support within the grafts.2,3
Second-generation axon guidance channels (AGCs) have recently been developed. A particular AGC with an internal lattice designed to replicate the longitudinal endoneurial tubes has demonstrated increased efficacy in nerve repair over hollow conduits.4 However, to our knowledge no studies have been performed to investigate combining these AGCs with cells or other growth-promoting substances.
One possible option for cellular supplementation is Schwann cell (SC) transplantation, due to the supportive role of these cells in the peripheral nervous system. SCs are integral to the regeneration and restoration of the function of injured nerves. In the last few decades, improvements in culture methods have enabled researchers to generate large populations of purified human cells from small samples.5–8 The use of these cells in nerve repair is supported by extensive preclinical data and, more recently, some clinical data. Our group has published data on the long-term follow-up in the first two cases in which autologous SCs were used in human PNIs.9,10
In the present study, we sought to investigate the ability of SCs to enhance axonal regeneration and functional recovery in a severe PNI with a gap beyond the “critical” gap length, previously observed as being > 10 mm in the rat sciatic nerve.11 Our hypothesis was that transplanting syngenetic rodent SCs into second-generation AGCs would lead to increased axon counts at 1 and 4 months after injury and improvements in functional metrics. We chose to use reversed autograft as a positive control, noting this to be an ideal option, albeit not clinically relevant.
Methods
Study Design
Eighty-six adult male Fischer rats (Envigo RMS, Inc.) with a mean weight of 361 ± 44 g were used in this study. All procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of our institution. All animals had ad libitum access to food and water and were housed according to the NIH Guide for the Care and Use of Laboratory Animals in the veterinary facilities at our institution. Animals were randomly assigned into three experimental groups (A–C) and one donor group for SC harvesting. Animals assigned to group A had sciatic nerve reversed autografts (positive control), group B received NeuraGen 3D nerve guides filled with green fluorescent protein (GFP)–labeled SCs, and group C received NeuraGen 3D nerve guides filled with serum only (negative control). Survival times for each group were 4 (30 rats) or 16 (30 rats) weeks postsurgery.
NeuraGen 3D Nerve Guide
The NeuraGen 3D nerve guide (Integra Lifesciences Corp.) is manufactured from a highly purified type 1 collagen derived from bovine deep flexor tendon. The collagen-glycosaminoglycan (GAG) matrix was prepared using a collagen and chondroitin-6-sulfate proteoglycan suspension.4,12 These nerve guides have an internal diameter of 1.5 mm and length of 15 mm.
SC Preparation
SCs were extracted from rat sciatic nerves and purified and expanded as previously described.5,8 Two weeks after harvest, once outward migration of fibroblasts occurred, explants were transferred onto new culture dishes, enzymatically dissociated, and then replated in DMEM/10% fetal bovine serum supplemented with 3 mitogens: bovine pituitary extract (2 mg/ml, Invitrogen), forskolin (0.8 mg/ml, Sigma), and heregulin (2.5 nM, Genentech).13 Cells were grown to confluency and passaged to new dishes 3 times. Their purity for grafting was found to be 95%–98%.13
Operative Procedure
Our surgical technique has been previously described.14–17 All animals had the right leg shaved prior to surgery. A posterior incision from the right iliac crest to the knee joint was made under aseptic conditions. The bicep femoris was reflected and the sciatic nerve beneath was dissected free from surrounding connective tissue. A 13-mm segment of nerve was removed via sharp transection distal to the notch and proximal to the bifurcation. Repair was performed according to group assignment (A, B, or C), using an operating microscope with 9-0 nylon sutures (Ethicon). Following surgery, the wound was closed in layers and skin was stapled. Staples were removed by postoperative day 10. Immediately after surgery, all animals were treated with buprenorphine for pain and gentamicin with ampicillin for infection prophylaxis. Apple bitter spray was used as prophylaxis for self-mutilation. Meloxicam droplets and Nylabone (Chewy, LLC) were added into all cages.
SC Loading
The conduit loading step occurred 2–4 hours prior to surgical repair. Our loading technique involved suspending SCs into a DMEM solution at a concentration of 100,000 cells/μl. A volume of 350 μl was then placed into a 1-ml syringe and the dry 15-mm conduit was added to the syringe. This volume of cell solution was chosen as the lowest volume, which would completely submerge the conduit. By application of negative pressure to the chamber with a syringe plunger, the cells were driven into the dry conduits (Fig. 1).
SC loading of NeuraGen 3D conduits. Loading of SCs into the NeuraGen 3D conduits (upper) was performed by first submerging the dry 15-mm conduit (B) inside of a 1-ml syringe (for animal model) filled with DMEM-suspended SCs at a concentration of 100,000 cells/µl and then applying negative pressure with the syringe plunger (A). A homogenous distribution of GFP-labeled SCs (lower) was observed with this method of loading. For purposes of visualization, we demonstrate a longitudinal section of the human-sized conduit (30 mm) loaded (using 2.5 ml of solution in a 3-ml syringe) with GFP-labeled SCs. Upper panel copyright Roberto Suazo. Published with permission.
Electrophysiology
Animals assigned to the 16-week time point underwent electrophysiological testing 1 day prior to being killed. While the rat was under anesthesia, the sciatic nerve was stimulated percutaneously through a pair of monopolar needle electrodes at the sciatic notch with a single monophasic electrical pulse (20 μsec, supramaximal intensity). Electromyographic signals were recorded from the gastrocnemius (GM), tibialis anterior (TA), and plantar interossei (PI) muscles. The compound muscle action potentials (CMAPs) were amplified by ×100 or ×1000 (P511AC amplifiers, Grass) and band-pass filtered (3 Hz to 1 kHz). Digital sampling of the signals was made with a capacitance electronic disc (CED) recording system (CED1401 Micro3) at 20 kHz and fed into Signal software. The onset latency and the peak-to-peak amplitude of each CMAP (M-wave) were measured. Animal body temperature was maintained by means of a thermostatic heating pad. Control values were recorded from the intact left hind limb.
Autotomy Scores
Autotomy scores were measured on a weekly basis as a measure of the degree of automutilation behavior. The scoring scale, described by Wall et al.,18 ranges from 0 through 11. Injury of 2 or more nails was scored as 1 point, and injury of each distal half digit was scored as an additional point.
Muscle Weight
Animals were deeply anesthetized and perfused via cardiac ventricular puncture with 1× phosphate-buffered saline (PBS) followed by 4% paraformaldehyde (PFA) at 4 or 16 weeks after surgical transplantation, and sciatic nerves were explanted from the sciatic notch at the hip down to the level of the knee. The incision was then extended down to the paw and the GM harvested via section of the proximal and distal tendinous attachments. This was repeated on the uninjured leg. The GM from each leg was weighed dry on a precision balance.
Histological Analysis
One-micrometer plastic cross sections were stained with 1% toluidine blue (TB), 1% methylene blue, and 1% sodium borate solution. Sections were analyzed using an Olympus BX51 microscope at 100× magnification and Stereo Investigator software (version 2019.1.2, MBF Bioscience) with an Optical Fractionator. The cross-sectional area of the epineurium was estimated by point counting with a 20× dry objective. A sampling grid of 134 μm × 134 μm with a counting frame area of 30 μm × 30 μm was chosen. The total number of myelinated axons was extrapolated from the product of the area density and cross-sectional area. The randomly selected subset of counted axons was also used to determine axon diameter and myelin thickness using the Stereo Investigator measure line tool (MBF Bioscience). We calculated g-ratios using the inner and outer diameters of myelinated axons. Unbiased stereological analysis was completed by a single blinded investigator and averaged per animal and per group.
Immunohistochemistry
After overnight postfixation in 4% PFA, nerves designated for analysis with immunohistochemical (IHC) staining were transferred to a 30% sucrose solution for at least 48 hours. Nerves were then frozen and embedded in Shandon M-1 embedding matrix (Thermo Fisher Scientific) and frozen in preparation for cryosectioning. Nerves were cut into longitudinal, 20-μm-thick serial sections and placed onto charged slides using a cryostat (CM 1950, Leica Biosystems). Tissue sections were stored at −20°C until the IHC staining process. On the day of staining, sections were thawed to room temperature. Tissue was rinsed in 1× PBS and a fat pen was used around the edges of the slides. Nonspecific binding sites were then blocked using 0.3% Triton and 5% normal goat serum solution in PBS for 45 minutes. After blocking, tissue was incubated with primary antibody diluted in the same blocking solution as above at room temperature for 2 hours, followed by overnight at 4°C. Primary antibodies included chicken anti–p-zero (PZO, Aves Labs; 1:200) to label endogenous SCs and mouse anti-neurofilament (monoclonal RT97c, DSHB; 1:200) for axons. The following day slides were washed with PBS and then incubated with the secondary antibodies goat anti-chicken (AlexaFluor 555, Invitrogen; 1:500) and goat anti-mouse (AlexaFluor 647, Invitrogen; 1:500), respectively, for 2 hours at room temperature. Slides were again washed with PBS, counter-stained using the nuclear marker 4′,6-diamidino-2-phenylindole (DAPI, Thermo Fisher Scientific), rinsed 2 times for 5 minutes in PBS, and covered with a glass slip with fluorescence mounting medium (DAKO, Agilent Technologies). Slides were allowed to dry overnight in the dark prior to imaging. Imaging was performed with the Virtual Slide Microscope (VS120, Olympus) equipped with 20× (UPLSAPO, 0.75NA, Air) objectives and fitted with a Hamamatsu ORCA-Flash 4.0 camera, followed by analysis using ImageJ (version 1.8, NIH).
Statistical Analysis
All data are presented as means ± standard deviations. Statistical tests were performed using GraphPad Prism version 6.01. The data were first analyzed for outliers using the Grubbs test and then subjected to one-way ANOVA. Multiple comparisons between pairs of means were performed using the Tukey test. Statistical significance was considered as p < 0.05.
Results
SC Culture and Distribution
GFP expression was detected in 90% of the pool of transfected SCs. Conduits loaded with cultured SCs were fixed and sectioned longitudinally for imaging to confirm successful loading. GFP-labeled SCs were distributed homogenously throughout the conduit (Fig. 1).
IHC Observations
Longitudinal sections with immunostaining for neurofilament and SCs are displayed in Fig. 2. By 4 weeks, all three groups had regenerating axons growing into the distal segment of the implants, with notably less neurofilament signal in group C (Fig. 2A and B). A strong interaction between GFP-labeled SCs and neurofilaments was visible in all segments of group B at 4 and 16 weeks (Fig. 2C and D). A greater density of GFP signals was visible at 4 weeks, although both time points showed a good distribution of GFP signals throughout. On the other hand, the density of neurofilaments was visibly greater at 16 weeks.
Immunostained longitudinal sections showing extension of axonal neurofilaments and close interaction with GFP-labeled SCs. Regions of interest with dimensions of 1 mm2 at the midpoint and at each end (3 mm into the graft from both stumps) underwent thresholding and background subtraction using ImageJ (version 1.8) to highlight axonal neurofilament extension (anti-RT97c, red) at 4 (A) and 16 (B) weeks posttransplantation. A higher density of axonal neurofilaments (anti-RT97c, red) was visible in every segment of the NeuraGen 3D conduit with SCs at 16 (D) versus 4 (C) weeks posttransplantation. A close interaction between GFP-labeled SCs (green) and axonal neurofilaments was observed at both time points (bottom panels of C and D). Bars in A and B are 100 µm, top panels of C and D are 2 mm, and bottom panels of C and D are 100 µm.
Morphometric Analysis
Representative TB-stained axial semithin sections at 4 and 16 weeks are displayed in Fig. 3A and B. Graft diameters varied significantly between both time points (Fig. 3C). Conduits in groups B and C had similar outer diameters (2.11 ± 0.08 vs 2.99 ± 0.09 mm, respectively; p = 0.77) at 4 weeks. At 16 weeks, a significant reduction in diameter was observed in both groups (1.44 ± 0.66 vs 1.18 ± 0.45 mm, respectively). The mean values for all measurements are displayed in Table 1.
Supplementation of NeuraGen 3D conduits with autologous SCs enhances axon regeneration to a similar extent as within reversed autograft. Morphometric analysis was performed on plastic-embedded semithin cross sections stained with TB at 4 (A) and 16 (B) weeks posttransplantation. Cross sections of graft midpoints at 4 weeks are represented in the left column of A. The middle and right columns of A highlight myelinated axons within fascicles (arrows in bottom panels point to smaller fascicles) of each corresponding experimental group (magnification ×100). (B) TB-stained cross sections of 3 segments (proximal, middle, and distal) representative of each experimental group at 16 weeks after transplantation (magnification ×100). Significantly larger fascicles with a greater number of myelinated axons were visible in all groups at 16 weeks. Graft diameters were measured at the midpoint of the grafts at 4 and 16 weeks posttransplantation (C). Myelinated axon count (D), axon diameter (E), and myelin thickness (F) at 4 (left bar graphs) and 16 (right bar graphs) weeks posttransplantation were measured at magnification ×100 using stereological software. All bars are means ± SDs, and data were analyzed by one-way ANOVA using a Tukey multiple comparison test. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Bars in left column of A are 500 µm and the rest (A and B) are 10 µm.
Summary of morphometric analysis at 4 months
| Group A | Group B | Group C | |
|---|---|---|---|
| Proximal | |||
| Graft diameter (µm) | 1342 ± 257 | 1473 ± 584 | 1193 ± 496 |
| Myelinated axon count | 8070 ± 1273 | 7504 ± 1850 | 3739 ± 4278*† |
| Myelinated axon diameter (µm) | 3.4 ± 0.9 | 3.6 ± 1.1 | 2.3 ± 0.7*† |
| Myelin thickness (µm) | 0.81 ± 0.21 | 0.68 ± 0.20* | 0.56 ± 0.12* |
| g-ratio | 0.67 ± 0.06 | 0.72 ± 0.06* | 0.66 ± 0.08† |
| Midpoint | |||
| Graft diameter (µm) | 1210 ± 346 | 1438 ± 656 | 1176 ± 446 |
| Myelinated axon count | 7911 ± 1117 | 6468 ± 2462 | 1465 ± 1079*† |
| Myelinated axon diameter (µm) | 3.2 ± 1.1 | 3.6 ± 1.0 | 2.6 ± 0.9*† |
| Myelin thickness (µm) | 0.76 ± 0.22 | 0.73 ± 0.17 | 0.56 ± 0.13*† |
| g-ratio | 0.67 ± 0.07 | 0.70 ± 0.06* | 0.69 ± 0.06 |
| Distal | |||
| Graft diameter (µm) | 1326 ± 184 | 1542 ± 638 | 1351 ± 341 |
| Myelinated axon count | 9458 ± 1978 | 6455 ± 1425 | 2314 ± 2335*† |
| Myelinated axon diameter (µm) | 3.7 ± 1.6 | 4.0 ± 1.2 | 2.8 ± 0.7*† |
| Myelin thickness (µm) | 0.89 ± 0.40 | 0.88 ± 0.58 | 0.57 ± 0.13*† |
| g-ratio | 0.67 ± 0.06 | 0.70 ± 0.09* | 0.71 ± 0.07* |
Values are presented as mean ± SD. Group A, reversed autograft; group B, conduit + SCs; group C, conduit alone.
Compared with group A, p < 0.05.
Compared with group B, p < 0.05.
Myelinated axons were analyzed at ×100 magnification. At 4 weeks, group A had a significantly greater number of myelinated axons at the midpoint (5.86 ± 0.80 × 103 axons) than both conduit groups (p < 0.0001) (Fig. 3D). Although not statistically significant, group B had a greater number of axons on average than group C (226 ± 157 vs 28 ± 6 axons, p = 0.82). Group B also had visibly larger fascicles with a greater number of myelinated axons than group C (Fig. 3A). At 16 weeks, the myelinated axon counts in group B were statistically similar to those in group A (p > 0.05) and significantly higher than those in group C in all segments (p < 0.05). Despite a trend toward lower total axon counts at successively distant segments, this finding was not statistically significant, indicating that axons were able to grow into the conduit and elongate all the way through to the distal nerve stump.
Axon diameter and myelin thickness were also analyzed (Fig. 3E and F). At 4 weeks, the axon diameters in group B (3.4 ± 1.0 μm) were significantly larger than those in group A (2.7 ± 0.6 μm, p < 0.0001). Although diameters in group C were on average larger (2.9 ± 0.9 μm) than those in group A, this finding was not statistically significant (p = 0.058). At 16 weeks, in groups A and B the diameters were similar in all segments (p > 0.05) and were significantly larger than those in group C (p < 0.05). Myelin thickness also varied significantly between the groups at both time points, with group B having thicker myelination than group C in nearly all segments (Fig. 3F). The g-ratios, an index of axonal myelination, were significantly larger among axons in group B compared with those in group A in all 3 segments at 16 weeks (p < 0.05).
In summary, conduits loaded with SCs transmitted a larger aliquot of myelinated axons and axons of greater maturity, comparable to autografts and superior to those delivered by non–SC-loaded conduits.
Electrophysiology
Onset latencies and CMAP amplitudes measured at 16 weeks are shown in Fig. 4 and reported in Table 2. With increasing distance from the stimulation site, mean onset latencies progressively increased in all experimental groups, differing significantly from the contralateral control at the TA and PI. No statistically significant difference was observed between groups A and B at the GM (p = 0.9989) or TA (p = 0.0684). At the longest distance (PI), group B latencies had longer onset than group A latencies (p = 0.0030) but remained significantly shorter than group C latencies (p = 0.024). At all three muscles, onset latencies for group C were significantly longer than those for the other groups (p < 0.0001).
Supplementation of NeuraGen 3D conduits with autologous SCs improves nerve conduction recovery. The sciatic nerve in both injured and uninjured limbs (positive control) was stimulated percutaneously at 16 weeks posttransplantation with monopolar needle electrodes at the sciatic notch, and CMAPs were measured at successively distant muscle groups (GM, TA, and PI). Representative CMAP waves in the uninjured limb are displayed (A). Onset latencies (B) and CMAP amplitudes (C) are measured at the GM, TA, and PI. Linear regression analysis shows a statistical correlation between increasing myelinated axon counts and both faster conduction velocity (top left; R = 0.84, p = 0.0002) and larger CMAP amplitudes (top right; R = 0.73, p = 0.003) (D). Axonal diameter (bottom left; R = 0.78, p = 0.001) and myelin thickness (bottom right; R = 0.73, p = 0.003) were also correlated with faster conduction velocities. All bars are means ± SDs, and data were analyzed by one-way ANOVA using the Tukey multiple comparison test. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Figure is available in color online only.
Summary of nerve conduction studies
| Contralateral | Group A | Group B | Group C | |
|---|---|---|---|---|
| Onset latency (msec) | ||||
| GM | 1.3 ± 0.1 | 1.5 ± 0.2 | 1.6 ± 0.1 | 3.9 ± 1.1*†‡ |
| TA | 1.4 ± 0.1 | 2.1 ± 0.4* | 2.8 ± 0.6* | 4.9 ± 1.2*†‡ |
| PI | 2.4 ± 0.2 | 3.4 ± 0.5* | 5.0 ± 1.5*† | 6.2 ± 1.3*†‡ |
| Amplitude (mV) | ||||
| GM | 66.7 + 5.4 | 24.2 ± 6.8* | 23.4 ± 6.1* | 8.8 ± 5.9*†‡ |
| TA | 67.2 ± 10.2 | 18.0 ± 7.3* | 17.2 ± 6.6* | 5.1 ± 4.2*†‡ |
| PI | 4.40 ± 1.3 | 0.7 ± 0.8 | 0.8 ± 0.5 | 0.8 ± 0.6 |
| Conduction velocity (m/sec) | ||||
| GM | 40.1 ± 4.0 | 33.3 ± 3.7* | 32.7 ± 2.3* | 15.3 ± 5.7*†‡ |
| TA | 39.6 ± 3.8 | 26.6 ± 4.3* | 20.5 ± 4.5*† | 12.2 ± 3.0*†‡ |
| PI | 36.7 ± 2.6 | 27.4 ± 4.5* | 19.6 ± 5.3* | 16.0 ± 4.8*† |
Values are presented as mean ± SD.
Compared with contralateral control, p < 0.05.
Compared with group A, p < 0.05.
Compared with group B, p < 0.05.
CMAP amplitudes are displayed in Fig. 4C. At the two most proximal muscles, groups A and B displayed similar amplitudes that were also significantly larger than those of group C (p < 0.0001). No significant differences were observed between any of the groups at the PI. None of the groups achieved values similar to those of the contralateral control after 16 weeks.
Conduction velocities were calculated using the distance between stimulation and recording sites and the onset latencies (Table 2). Mean conduction velocities were similar between groups A and B at the GM (p = 0.613) but differed significantly at the TA (26.6 ± 4.3 vs 20.5 ± 4.5 m/sec, p = 0.0061) and PI (27.4 ± 4.5 vs 19.6 ± 5.3 m/sec, p = 0.0003). At all three muscle sites, conduction velocities for group C were significantly slower than those for the other groups (p < 0.0001).
A linear regression analysis comparing morphological data with electrophysiological results was performed (Fig. 4D). Faster conduction velocities and larger amplitudes correlated with increasing axon counts (p = 0.0002 and p = 0.003, respectively). Faster conduction velocities also correlated with larger axon diameter and thicker myelination (p = 0.001 and p = 0.003, respectively).
Muscle Atrophy
Dry GM weights are displayed in Fig. 5. Groups A and B had percent muscle recoveries of 55.6% ± 4.9% and 52.4% ± 4.6%, respectively, and did not differ significantly. Muscles in group C recovered to a significantly lesser extent (33.3% ± 10.4%, p < 0.0001). Although muscle weights of all groups remained significantly smaller than their contralateral counterparts, adding SCs to the conduits demonstrated a significant benefit in reducing muscle atrophy, presumably on the basis of improved reinnervation.
Supplementation of NeuraGen 3D conduits with autologous SCs reduces muscle atrophy. The dry weight of the GM from the injured limb is represented as a percentage of the contralateral limb. Bars are means ± SDs, and data were analyzed by one-way ANOVA using the Tukey multiple comparison test. ****p < 0.0001.
Behavioral Analysis
Most animals recovered well from the surgery. One animal in group A died immediately after the procedure. No wound infections developed postoperatively. A complete foot drop paresis of the injured limb was noted in all animals. The trend in autotomy scores for the 16-week animals is demonstrated in Fig. 6.
Self-mutilation behavior. Autotomy scores for self-mutilation at 2, 4, and 6 weeks posttransplantation were not statistically significantly different between any of the experimental groups. Bars are means ± SDs, and data were analyzed by one-way ANOVA using the Tukey multiple comparison test.
Discussion
Manufactured AGCs present a promising alternative strategy for the repair of PNIs, allowing regenerating nerve fibers to extend toward their target in a noncollapsible conduit. AGCs can be tailored in diameter and length to fit the cut nerve ends without the use of an autologous graft.19 Importantly, AGCs provide the opportunity to incorporate neurotrophic factors (NTFs), extracellular matrix components, and support cells that enhance peripheral nerve regeneration.14,17,20–23 We have previously described and reported a cell-based therapeutic strategy that uses synthetic conduits in conjunction with cultured autologous SCs to repair long-segment PNIs.7,9,10,14–17,20,23,24
There has been 3 decades of focused research developing and refining SCs toward adjunctive therapy for nerve repair. The major milestones include the isolation of SCs from a number of adult mammalian sources,5,8 the survival of autologous human SCs (ahSCs) posttransplantation in a peripheral nerve environment,17 the expansion of primary and mitogen ahSCs to promote regeneration in a guidance channel,14 the ability of ahSCs to divide in response to heregulin and cAMP analogs while retaining their functional abilities,7,23 the absence of tumor formation despite extensive stimulation and ahSC proliferation,25 the importance of SCs in supporting long-distance regeneration,10,13,24 the observation that primate SCs enhance peripheral nerve regeneration in two injury models,15,16 and finally, the recent exciting safety and efficacy findings when ahSCs were used to supplement sural nerve graft repair after devastating long-segment human sciatic nerve injury.9,10
Prior studies have described a critical gap length in which nerve regeneration does not occur within conventional AGCs without the addition of exogenous growth factors and/or a small segment of nerve tissue.11,26–30 In rat sciatic nerve, this distance is > 10 mm. Matrix-filled conduits have demonstrated superior improvement in structural and electrophysiological properties in the repair of long-segment sciatic nerve injuries compared with hollow conduits.4,31,32 Although hollow conduits have significant advantages in preventing permeation of fibroblasts and enabling buildup of NTFs for regeneration, hollow conduits may contribute to incomplete reinnervation from dispersal of axonal fibers within the hollow tube.33 Therefore, hollow conduits have been suggested for small nerve gaps and matrix-filled conduits for larger gaps.34
In the present study, we used a 13-mm gap as part of our injury model to test regenerative capacity beyond the critical gap length of a novel AGC (NeuraGen 3D) with a porous collagen-GAG inner matrix supplemented with SCs. The inner matrix provides a synthetic mimic of the SC extracellular matrix and is axially aligned to allow for SCs to distribute along the path of growing axons, emulating the Bungner bands seen in native peripheral nerves. Upon loading of conduits, we observed an effective and robust distribution of GFP-labeled SCs. At 4 and 16 weeks, exogenous SCs closely interacted with regenerating axons throughout all segments of the NeuraGen 3D conduits, suggesting that exogenously implanted SCs not only are capable of long-term survival within these conduits but also continue to play a functional role past the acute phase of recovery in axonal regeneration across a long-segment injury.
Axonal morphology differed significantly between groups. At 4 weeks, larger and denser fascicles filled with myelinated axons were evenly distributed throughout the SC-filled conduits, while acellular conduits had few and small mini-fascicles scattered near the outer edges. Most axons in the acellular conduits had very little to no myelination, although diameters did not differ from axons in the SC-filled conduits. At 16 weeks, the mini-fascicles were more pronounced in both groups, but SC-filled conduits maintained a significantly greater number of fascicles with a higher density of myelinated axons throughout. The axons in SC-filled conduits also had larger diameters with thicker myelination. In all morphological measurements, SC-filled conduits performed statistically similarly to reversed autografts.
These results extend our previously published findings on using a hollow first-generation NeuraGen channel to repair a 13-mm gap injury.13 In the prior work, we showed improvement in regeneration of myelinated axons with the addition of SCs. Comparing data between our previous results and the present study, we find that second-generation, matrix-embedded conduits allowed for superior regeneration as compared with the first-generation hollow conduit. In this study, the addition of SCs into matrix-embedded conduits produced the highest degree of axonal regeneration and myelination of all groups in both time points, save for reverse autograft. These experiments combined demonstrate that SC-supplemented collagen-GAG conduits provide a better regenerative environment, producing a greater number of myelinated axons with larger diameters and thicker myelination than hollow and acellular channels.
It should also be recognized that the use of a reverse autograft creates an artificially high regenerative bar to reach in the development of novel nerve prostheses. In the clinical setting of sciatic nerve repair with a large gap, the surgeon must suture multiple small sural nerve grafts to reconstruct the nerve. The insufficiency of such autograft material serves as a major impetus for this approach.9,10,35 More importantly, the multiple layered sural nerve grafts are in aggregate structurally very dissimilar to a “reversed” sciatic nerve autograft—that is, exactly the same size as the host nerve, given that there are few branch points in the midthigh. Ideally, we could consider using multiple sural nerve autografts as the positive control in the rat, but these grafts are too small and too short to consider in this model.
From the nerve conduction studies, we observed improved onset latencies at all distances from the stimulation site with SC-supplement conduits. By 16 weeks, onset latencies at the GM in the SC-filled group recovered to values similar to those of autografts and contralateral controls. Conduction velocities also improved significantly at all 3 sites, results not dissimilar to those of reversed autografts. Similar to previous studies that have observed correlations between nerve conduction and axonal morphology,36–39 we also observed a linear relationship between increasing numbers of myelinated axons and greater conduction velocities and amplitude. Increasing axon diameter and thicker myelination were also correlated with greater conduction velocities.
Muscle weight has also been used to evaluate the success of reinnervation. The GM is innervated by the posterior tibial branch of the sciatic nerve and, with complete sciatic nerve transection, will atrophy. The extent at which it regains its mass is therefore proportional to the amount of muscle reinnervation.40–42 In the present study, at 16 weeks the extent of muscle mass recovery of the SC-filled group was similar to that of the reversed autograft (52% vs 56%, respectively) and significantly better than that of the empty conduit (33%). This provides further support that collagen-GAG channels supplemented with autologous SCs enhance both structural and functional recovery in long-segment PNIs beyond the critical gap length.
Last, we observed a significant reduction in autotomy scores in all groups compared to the groups in our previous study with the first-generation NeuraGen channels.13 We believe that a more aggressive approach to prevent self-mutilation, by using apple bitter spray, meloxicam droplets, and Nylabone, was mostly responsible for the reduction in autotomy scores.
There are a few limitations in the present study. One such limitation involved a gait analysis that was performed on the 16-week animals, carried out at 2-week intervals on the CatWalk XT system (version 10.0, Noldus Information Technology). Due to foot drop paresis and plantar flexion contractures in the animals, we were unable to obtain any meaningful data (Supplemental Fig. 1). This has been observed in various peripheral nerve studies and may be associated with an imbalance in the reinnervation of distal muscles.32,43,44 Another limitation is that we were unable to process the proximal and distal segments of the 4-week grafts for histological analysis and therefore were unable to compare these with first-generation conduits at those segments. However, histological data at the midpoints were successfully collected, and we believed these data were the most important. Additionally, due to the extreme length of the repair, nerve tissue proximal and distal to the suture lines was not consistently obtained and thus not studied.
The data in this study provide further evidence and opportunity to develop a cell-based therapy combining two FDA-approved products—a 3D collagen matrix nerve conduit (NeuraGen 3D) and a cell product (ahSC) currently determined to be an investigational new drug (IND 14856). Further research combining ahSCs with NeuraGen conduits, including cell concentration and the feasibility of filling longer and larger nerve guides, will be crucial translational steps to ultimately repairing long-segment nerve gaps.
Conclusions
A critical nerve gap (13 mm) in the rat sciatic nerve was successfully repaired with SC-filled NeuraGen 3D conduits based on histological, electrophysiological, and functional endpoints. Significant improvements in the total number of myelinated axons, axon diameter, and myelin thickness throughout the SC-filled conduits allowed for improvements in nerve conduction and a subsequent decrease in muscle atrophy compared to an acellular nerve matrix. The degree of regeneration was noninferior to that of the reversed sciatic nerve graft, the positive control in this investigation.
Acknowledgments
This work was supported by the National Institute of Neurological Disorders and Stroke (R21NS111334-01).
Dr. Levi receives a teaching honorarium from the American Association of Neurological Surgeons and grant support from the Department of Defense and the National Institutes of Health (NIH-NINDS).
The authors have no financial interests in the manufacture or sale of the NeuraGen tube, and the company’s (Integra LifeSciences Corporation) support of the research is limited to donation of the tubes for the project at no cost.
The authors wish to thank Vania Almeida, Dr. Simon J. Archibald, Dr. Marcia Boulina, Ms. Adriana Brooks, Dr. Susana Cerqueira, Mr. Ramon German, Dr. Aisha Khan, Dr. Alexander E. Marcillo, Mrs. Yelena Pressman, Ms. Risset Silvera, and Mr. Roberto Suazo for their invaluable technical assistance.
Disclosures
The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.
Author Contributions
Conception and design: Levi, Burks, Haggerty, Midha. Acquisition of data: Burks, Diaz, Haggerty, de la Oliva. Analysis and interpretation of data: Levi, Burks, Diaz, Haggerty, de la Oliva. Drafting the article: Levi, Burks, Diaz, de la Oliva. Critically revising the article: Levi, Burks, Diaz, Midha. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Levi. Statistical analysis: Diaz, Haggerty, de la Oliva. Administrative/technical/material support: Levi, Burks, Diaz. Study supervision: Levi, Burks.
Supplemental Information
Online-Only Content
Supplemental material is available with the online version of the article.
Supplemental Figure 1. https://thejns.org/doi/suppl/10.3171/2020.8.JNS202349.
References
- 1↑
Noble J, Munro CA, Prasad VS, Midha R. Analysis of upper and lower extremity peripheral nerve injuries in a population of patients with multiple injuries. J Trauma. 1998;45(1):116–122.
- 2↑
Berrocal YA, Almeida VW, Levi AD. Limitations of nerve repair of segmental defects using acellular conduits. J Neurosurg. 2013;119(3):733–738.
- 3↑
Moore AM, Kasukurthi R, Magill CK, et al. Limitations of conduits in peripheral nerve repairs. Hand (N Y). 2009;4(2):180–186.
- 4↑
Lee JY, Giusti G, Friedrich PF, et al. The effect of collagen nerve conduits filled with collagen-glycosaminoglycan matrix on peripheral motor nerve regeneration in a rat model. J Bone Joint Surg Am. 2012;94(22):2084–2091.
- 5↑
Morrissey TK, Kleitman N, Bunge RP. Isolation and functional characterization of Schwann cells derived from adult peripheral nerve. J Neurosci. 1991;11(8):2433–2442.
- 6
Morrissey TK, Kleitman N, Bunge RP. Human Schwann cells in vitro. II. Myelination of sensory axons following extensive purification and heregulin-induced expansion. J Neurobiol. 1995;28(2):190–201.
- 7↑
Levi AD, Bunge RP, Lofgren JA, et al. The influence of heregulins on human Schwann cell proliferation. J Neurosci. 1995;15(2):1329–1340.
- 8↑
Levi AD. Characterization of the technique involved in isolating Schwann cells from adult human peripheral nerve. J Neurosci Methods. 1996;68(1):21–26.
- 9↑
Gersey ZC, Burks SS, Anderson KD, et al. First human experience with autologous Schwann cells to supplement sciatic nerve repair: report of 2 cases with long-term follow-up. Neurosurg Focus. 2017;42(3):E2.
- 10↑
Levi AD, Burks SS, Anderson KD, et al. The use of autologous Schwann cells to supplement sciatic nerve repair with a large gap: first in human experience. Cell Transplant. 2016;25(7):1395–1403.
- 11↑
Lundborg G, Dahlin LB, Danielsen N, et al. Nerve regeneration across an extended gap: a neurobiological view of nerve repair and the possible involvement of neuronotrophic factors. J Hand Surg Am. 1982;7(6):580–587.
- 12↑
Yannas IV, Lee E, Orgill DP, et al. Synthesis and characterization of a model extracellular matrix that induces partial regeneration of adult mammalian skin. Proc Natl Acad Sci U S A. 1989;86(3):933–937.
- 13↑
Berrocal YA, Almeida VW, Gupta R, Levi AD. Transplantation of Schwann cells in a collagen tube for the repair of large, segmental peripheral nerve defects in rats. J Neurosurg. 2013;119(3):720–732.
- 14↑
Levi AD, Bunge RP. Studies of myelin formation after transplantation of human Schwann cells into the severe combined immunodeficient mouse. Exp Neurol. 1994;130(1):41–52.
- 15↑
Calancie B, Madsen PW, Wood P, et al. A guidance channel seeded with autologous Schwann cells for repair of cauda equina injury in a primate model. J Spinal Cord Med. 2009;32(4):379–388.
- 16↑
Levi AD, Sonntag VK, Dickman C, et al. The role of cultured Schwann cell grafts in the repair of gaps within the peripheral nervous system of primates. Exp Neurol. 1997;143(1):25–36.
- 17↑
Levi AD, Guénard V, Aebischer P, Bunge RP. The functional characteristics of Schwann cells cultured from human peripheral nerve after transplantation into a gap within the rat sciatic nerve. J Neurosci. 1994;14(3 Pt 1):1309–1319.
- 18↑
Wall PD, Devor M, Inbal R, et al. Autotomy following peripheral nerve lesions: experimental anaesthesia dolorosa. Pain. 1979;7(2):103–111.
- 19↑
Tyner TR, Parks N, Faria S, et al. Effects of collagen nerve guide on neuroma formation and neuropathic pain in a rat model. Am J Surg. 2007;193(1):e1–e6.
- 20↑
Guénard V, Kleitman N, Morrissey TK, et al. Syngeneic Schwann cells derived from adult nerves seeded in semipermeable guidance channels enhance peripheral nerve regeneration. J Neurosci. 1992;12(9):3310–3320.
- 21
Guénard V, Xu XM, Bunge MB. The use of Schwann cell transplantation to foster central nervous system repair. Semin Neurosci. 1993;5(6):401–411.
- 22
Rodríguez FJ, Verdú E, Ceballos D, Navarro X. Nerve guides seeded with autologous Schwann cells improve nerve regeneration. Exp Neurol. 2000;161(2):571–584.
- 23↑
Morrissey TK, Levi AD, Nuijens A, et al. Axon-induced mitogenesis of human Schwann cells involves heregulin and p185erbB2. Proc Natl Acad Sci U S A. 1995;92(5):1431–1435.
- 24↑
Hood B, Levene HB, Levi AD. Transplantation of autologous Schwann cells for the repair of segmental peripheral nerve defects. Neurosurg Focus. 2009;26(2):E4.
- 25↑
Emery E, Li X, Brunschwig JP, et al. Assessment of the malignant potential of mitogen stimulated human Schwann cells. J Peripher Nerv Syst. 1999;4(2):107–116.
- 26
Francel PC, Francel TJ, Mackinnon SE, Hertl C. Enhancing nerve regeneration across a silicone tube conduit by using interposed short-segment nerve grafts. J Neurosurg. 1997;87(6):887–892.
- 27
Hess JR, Brenner MJ, Fox IK, et al. Use of cold-preserved allografts seeded with autologous Schwann cells in the treatment of a long-gap peripheral nerve injury. Plast Reconstr Surg. 2007;119(1):246–259.
- 28
Lee AC, Yu VM, Lowe JB III, et al. Controlled release of nerve growth factor enhances sciatic nerve regeneration. Exp Neurol. 2003;184(1):295–303.
- 29
Lundborg G, Dahlin LB, Danielsen N, et al. Nerve regeneration in silicone chambers: influence of gap length and of distal stump components. Exp Neurol. 1982;76(2):361–375.
- 30
Lundborg G, Gelberman RH, Longo FM, et al. In vivo regeneration of cut nerves encased in silicone tubes: growth across a six-millimeter gap. J Neuropathol Exp Neurol. 1982;41(4):412–422.
- 31↑
Chamberlain LJ, Yannas IV, Hsu HP, et al. Collagen-GAG substrate enhances the quality of nerve regeneration through collagen tubes up to level of autograft. Exp Neurol. 1998;154(2):315–329.
- 32↑
Chamberlain LJ, Yannas IV, Hsu HP, et al. Near-terminus axonal structure and function following rat sciatic nerve regeneration through a collagen-GAG matrix in a ten-millimeter gap. J Neurosci Res. 2000;60(5):666–677.
- 33↑
Carvalho CR, Oliveira JM, Reis RL. Modern trends for peripheral nerve repair and regeneration: beyond the hollow nerve guidance conduit. Front Bioeng Biotechnol. 2019;7:337.
- 34↑
Du J, Chen H, Qing L, et al. Biomimetic neural scaffolds: a crucial step towards optimal peripheral nerve regeneration. Biomater Sci. 2018;6(6):1299–1311.
- 35↑
Burks SS, Levi DJ, Hayes S, Levi AD. Challenges in sciatic nerve repair: anatomical considerations. J Neurosurg. 2014;121(1):210–218.
- 36
Cragg BG, Thomas PK. The conduction velocity of regenerated peripheral nerve fibres. J Physiol. 1964;171:164–175.
- 37
Dellon AL, Mackinnon SE. Selection of the appropriate parameter to measure neural regeneration. Ann Plast Surg. 1989;23(3):197–202.
- 38
Boyd IA, Kalu KU. Scaling factor relating conduction velocity and diameter for myelinated afferent nerve fibres in the cat hind limb. J Physiol. 1979;289:277–297.
- 39
Hursh JB. Conduction velocity and diameter of nerve fibers. Am J Physiol. 1939;127(1):131–139.
- 40
Menderes A, Yilmaz M, Vayvada H, et al. Effects of nerve growth factor on the neurotization of denervated muscles. Ann Plast Surg. 2002;48(4):415–422.
- 41
Fitton AR, Berry MS, McGregor AD. Preservation of denervated muscle form and function by clenbuterol in a rat model of peripheral nerve injury. J Hand Surg Br. 2001;26(4):335–346.
- 42
Day CS, Riano F, Tomaino MM, et al. Growth factor may decrease muscle atrophy secondary to denervation. J Reconstr Microsurg. 2001;17(1):51–57.
- 43↑
Chang A, Yannas IV, Perutz S, et al. Electrophysiological study of recovery of peripheral nerves regenerated by a collagen-glycosaminoglycan copolymer matrix. In: Gebelein CG, Dunn RL, eds. Progress in Biomedical Polymers. Springer; 1990:107–120.
- 44↑
Hare GM, Evans PJ, Mackinnon SE, et al. Walking track analysis: a long-term assessment of peripheral nerve recovery. Plast Reconstr Surg. 1992;89(2):251–258.
