Peripheral nerve injury results in serious functional loss with paralysis, loss of sensation, and pain, with a prolonged course of recovery and only partial return of function. In recent years, there has been little improvement in the microsurgical techniques of repair of injured nerves, and the investigative efforts have focused on neurotrophic factors, peptides shown to be influential in the differentiation, function, and survival of neurons1. Delivery of factors to the site of injury as well as their dosage regimen have been major problems in this field of research.
In the study of central nervous system injury, it has been shown that the immune system may recognize certain injury-associated self-compounds as potentially destructive and answer with a protective antiself response mediated by a T-cell subpopulation that can recognize self-antigens2. This may be explained by the production of neurotrophic factors by the immune cells that support motor neuron survival before target reconnection occurs3,4.
A study of facial nerve injury examining severe combined immunodeficient (SCID) mice demonstrated a decrease in peripheral nerve survival in comparison with wild-type mice, and the investigators concluded that immune cells associated with acquired immunity play a role in regulating motor neuron survival after a peripheral nerve injury, specifically the CD4+ subgroup of T cells5,6. Reconstitution of SCID mice with wild-type splenocytes containing T and B cells restored the survival of facial motor neurons in these mice to the level of the wild-type controls.
An endogenous system may better modulate the milieu of neurotrophic factors and may achieve better nerve regeneration than any exogenous supply of specific factors.
Neuroprotective therapy is aimed at boosting the beneficial autoimmune response to injury-associated self-antigens. The development of neuroprotection has focused on antigens that would not induce autoimmune disease. The synthetic copolymer glatiramer acetate (Copolymer-1 or Copaxone; TEVA Neuroscience, Kansas City, Missouri) has been a successful example of this effort.
Glatiramer acetate is a synthetic amino acid polymer that is an approved drug for the treatment of relapsing multiple sclerosis and has been proved to be effective for decreasing the frequency of relapses in patients with relapsing multiple sclerosis7. It binds to the relevant major histocompatibility complex proteins and leads to the activation of T suppressor cells8. The main mechanism that has been reported is of a phenotype switch of Th1 cells to Th2 cells, resulting in a change in the profile of cytokine expression9-11. In the rat model of partial optic nerve crush, adoptive administration of glatiramer acetate-reactive T cells or vaccination with glatiramer acetate on the day of central nervous system injury had a marked preventive effect on the secondary degeneration of nerve fibers12.
Our hypothesis was that glatiramer acetate will be neuroprotective following peripheral nerve injury as well.
This hypothesis was examined in wild-type as well as nude-type rats. With the assumption that glatiramer acetate augments nerve regeneration by recruiting T-cell immunity, our hypothesis was that in the nude (T-cell-deficient) rat, the effect of glatiramer acetate will not be achieved or may be decreased.
Animals
Seventy-three female Harlan Sprague-Dawley rats (wild type; average weight, 225 g) (Harlan Laboratories, Madison, Wisconsin) and seventy-one female nude rats (NIHRNU-M) (Taconic, Germantown, New York) were divided into groups as depicted in Table I. This protocol was completed for a six-week period and for a three-week period following nerve injury. Each study period included two study groups and two control groups for wild-type as well as nude-type rats (with at least eight rats per group). The results were compared for each study period separately and for wild-type and nude-type rats separately.
All rats were provided with autoclaved pellets and water ad libitum. The rats were permitted one week to acclimate to their environment before being manipulated and used for the experiments; all experiments were performed when the rats were eight weeks of age. All rats were housed under a twelve-hour light/dark cycle in microisolator cages contained within a laminar flow system to maintain a specific pathogen-free environment. All experimental manipulations were performed under aseptic conditions and were completed in accordance with the National Institutes of Health guidelines on the care and use of laboratory animals for research purposes. The present study was approved by the institutional animal utilization study committee.
Induction of Nerve Injury
All rats were anesthetized with use of 56 mg/kg of Nembutal (sodium pentobarbital, 50 mg/mL), administered intraperitoneally. The surgical site was shaved and was washed with antiseptic solution prior to positioning for surgery, and local anesthetic (0.5 mL of 1% lidocaine hydrochloride) was injected subcutaneously at the surgical site. The sciatic nerve on the left side was approached with use of a semitendinosus-biceps femoris (long head) muscle-splitting incision. The nerve was dissected free of surrounding connective tissue. With use of a single pair of smooth microforceps, fully pressed for thirty seconds, a consistent crush injury was created. The location of the crush was marked with a single stitch of 9-0 nylon. The wound was closed with 4-0 sutures for the skin1.
Immunization
All injections were performed subcutaneously. The six-week protocol was performed first, and the emulsion, in a total volume of 0.2 mL, was injected into the animal's two hind footpads immediately after the crush injury. After review of the results, it seemed that, because of the rapid regenerative capability of the rat nerve, the six-week follow-up for the crush-injured nerves did not reveal the treatment effect. For this reason, the three-week protocol was performed; during this protocol, the progress was evaluated weekly with use of an additional measure, footprint analysis. In order to evaluate the footprints, footpad injections were not used for the three-week protocol, and the emulsion was subcutaneously injected into the base of the tail (resulting in the administration of the same total volume of 0.2 mL).
The rats in the study groups were injected with an emulsion containing glatiramer acetate with complete Freund's adjuvant (CFA; Becton Dickinson, Franklin Lakes, New Jersey). Groups B and F (Table I) were treated with an additional injection of glatiramer acetate (with incomplete Freund's adjuvant [IFA; Becton Dickinson]) forty-eight hours after the nerve injury. These groups were evaluated for a longer-lasting effect of glatiramer acetate. The dose of glatiramer acetate and complete Freund's adjuvant was based on previous reports13. Freund's adjuvant is an antigen solution emulsified in mineral oil, used as an immunopotentiator. It is composed of inactivated and dried mycobacteria. Incomplete Freund's adjuvant is devoid of the mycobacteria and is less potent yet less toxic than complete Freund's adjuvant.
The two control groups were injected with an emulsion containing complete Freund's adjuvant with normal saline solution or normal saline solution only.
Nerve-Regeneration Measures
Three or six weeks after the surgery, according to the protocol, both functional and histological evaluations were performed and the contralateral legs were used as the nonoperatively treated controls. The functional measurements that were obtained at three or six weeks included the tibialis anterior muscle response to sciatic nerve stimulation and the tibialis anterior muscle weight. In the groups of rats that were killed after three weeks, the functional evaluation included the additional measure of footprint analysis (specifically, toe spread factor, determined after seven, fourteen, and twenty-one days). The histological measurement at three or six weeks was an axon count in the sciatic nerve.
Functional Evaluation
Quantitative assessment of tibialis anterior muscle response to sciatic nerve stimulation:
Isometric tetanic contractile force measurements were made in situ with a technique adapted from that previously described by Zhao et al.14. After the induction of anesthesia (56 mg/kg of Nembutal), the hindlimb was immobilized on a frame and the foot was placed in a secure clamp to fix its position for the experimental analysis of muscle contraction force (response). An additional incision was made on the anterior aspect of the lower limb to expose the target of sciatic innervation, the tibialis anterior muscle. The distal tendon of the tibialis anterior muscle was divided at its insertion and mobilized up to the level of its origin, with care being taken to preserve the muscle's neurovascular supply. The transected tendinous part of the tibialis anterior muscle was secured to a metal hook, which was connected to a force transducer (model FT03; Grass Instrument, West Warwick, Rhode Island). The force transducer signal was sent to an amplifier whose output was recorded with use of LabTech Notebook software (Laboratory Technologies, Andover, Massachusetts) on a personal computer. The nerve was maintained at approximately 37°C with warm saline solution. The segment of the nerve proximal to the injury was dissected and was isolated proximally in preparation for stimulation. A silicone sheet was placed beneath the exposed nerve to isolate it from any conducting transudate on nearby tissues. An initial tension of approximately 330 mN was applied by stretching the muscle on the force transducer. This tension, corresponding to the normal length of the muscle, ensures that the contractions are isometric15. Platinum wire electrodes were used for electrical stimulation of the nerve proximal to the repair site. A stimulator (model S48; Grass Instrument) was used to deliver supramaximal stimuli (square wave impulses with a frequency of 100 Hz and duration of 0.6 msec). The maximal force of contraction, the tetanic force expressed in mN, was then recorded. A similar measurement was made for the nonoperatively treated, contralateral (control) side. The first muscular response was used to compare the injured side with the control side.
Tibialis anterior muscle weight:
After the animal was killed, both muscles were harvested and weighed and the difference between the injured and control legs was recorded.
Footprint analysis: toe spread factor:
For the groups of animals that were killed after three weeks, footprints were collected at three time points. Seven, fourteen, and twenty-one days after the injury to the nerve, footprints were collected and static toe spreading (for the first through fifth toes) was measured. This measurement has been found to be a good measure of functional recovery after sciatic nerve injury16. The analysis was performed by dipping the hind legs in ink and releasing the rats to walk over a large sheet (1 m × 20 cm) of white paper. The sheets were scanned, and measurements of toe spreading were made for three consecutive prints, which were averaged. A ratio was calculated as previously described16, with the injured side being compared with the control side.
Histological Evaluation: Axon Count of Sciatic Nerve Tissue
Following nerve stimulation and after the animal was killed, sections measuring 0.5 cm in length were removed from the sciatic nerve, 0.5 cm proximal and 0.5 cm distal to the site of injury, and were immersed in Karnovsky fixative (for four hours at 25°C), were washed in 0.2-M cacodylate buffer (pH 7.4), were postfixed in 1.5% osmium tetroxide with 1.5% potassium ferricyanide followed by graded alcohol (30%, 50%, 70%, 95%, and 100%), and were embedded in EMBed 812 (Electron Microscopy Sciences, Fort Washington, Pennsylvania) for axon counting. Sections measuring 1 µm in thickness were prepared with use of a Reichert-Jung Ultracut E with a diamond knife (Reichert-Jung, Weiss, Austria). In all animals, the corresponding sections from the nonoperatively treated, contralateral (control) legs were harvested and processed in a similar manner. All sections were stained with toluidine blue prior to analysis with light microscopy.
The axon counts were performed blinded by a group of three investigators (J.C., N.M.N., S.L.) with use of an Eclipse E600 microscope (Nikon, Melville, New York) with a Retiga EX CCD camera (QImaging, Burnaby, British Columbia, Canada) and QCapture software (QImaging). Analysis was performed on a Macintosh personal computer (Apple, Cupertino, California) with use of the public-domain NIH Image program (developed at the United States National Institutes of Health and available on the Internet at http://rsb.info.nih.gov/nih-image/) for the calculation of area and point counting in order to express the results as a ratio of the density (axons/µm2) of the fibers of the experimental side to the control side. The ratio of axon density provides information relative to the contralateral, control nerve and thus is an animal-specific measure of regeneration valuable for between-group and among-group comparisons1.
Statistical Analysis
Standard two-sided tests with a significance level of a and a power of 1 - ß were used for power analysis using the validation data of previous studies1,17. The population sizes were computed with use of the following two-sided test, assuming equal sizes in both study populations: n = (s12 + s22) (Z1-a/2 + Z1-ß)2 (1/?2), where ? = µ2 - µ1, a = 0.05, ß = 0.2 (for 80% power with 95% confidence), Z1-a/2 = 1.96, and Z1-ß = 0.84. For axon counting, s1 (variance 1) = 0.448, s2 (variance 2) = 0.237, µ2 - µ1 = 0.56, and n = 6.4 for each study group. For muscle weights, s1 (variance 1) = 0.372, s2 (variance 2) = 0.186, µ2 - µ1 = 0.45, and n = 6.7 for each study group.
The differences in means were based on previous studies1,17. To provide an extra margin of safety, we elected to use a minimum of eight animals per group.
The hypothesis was tested for each dependent variable with use of analysis of variance, followed by a series of pairwise univariate comparisons using the Student t test to detect significant differences between groups. Multiple comparisons were accounted for with use of the Scheffe post hoc test for muscle contracture, weight, and axon count and the Fisher protected least significant difference test for the footprint analysis.
Source of Funding
There was no external funding source for the study.
One hundred and forty-six rats were included in the study. Two rats were killed because of malocclusion and consequent malnourishment (one rat) or because of autophagia of more than one toe proximal to the distal interphalangeal joint of the injured limb (one rat) according to the criteria of the institutional ethical committee.
The results are summarized in tables in the Appendix.
Functional Evaluation
Tibialis Anterior Muscle Response to Sciatic Nerve Stimulation
Significantly greater muscle responses were measured after three weeks in the group of wild-type rats treated with glatiramer acetate in comparison with all other groups, including the double dose of glatiramer acetate, complete Freund's adjuvant, and normal saline solution (control:injured limb ratio, 0.05 mN compared with 0.52, 0.53, and 0.51, respectively; p < 0.05). No differences were found between the groups of wild-type rats after six weeks or nude-type rats after three or six weeks (at the 95% confidence level).
Tibialis Anterior Muscle Weight
No difference was found in muscle weights between the groups of wild or nude-type rats (at the 95% confidence level).
Footprint Analysis: Toe Spread Factor
A significant decrease in toe spread was found after twenty-one days in the wild-type group treated with two doses of glatiramer acetate in comparison with the complete Freund's adjuvant group (ratio, -0.083 compared with 0.044; p < 0.01). Although the differences were not significant, the group that received two doses of glatiramer acetate was found to have decreased toe spread after twenty-one days in comparison with the group that received one dose of glatiramer acetate group and the group that received normal saline solution. No differences were found between the groups of nude-type rats (at the 95% confidence level).
Histological Evaluation
Sciatic Nerve Axon Count (Figs. 1-A Through 1-D)
After three weeks, a significantly higher number of axons was counted distal to the injury in the group of wild-type rats that received one dose of glatiramer acetate in comparison with the group that received normal saline solution (control:injured limb ratio, -0.07 compared with 0.29; p < 0.05). After six weeks, a significantly higher number of axons was counted proximal to the injury in nude-type rats receiving glatiramer acetate in comparison with the normal saline solution group (control:injured limb ratio, -0.32 compared with 0.05; p < 0.05).
Differences were not found between the wild-type rats after six weeks or between nude-type rats after three weeks (at the 95% confidence level).
Glatiramer acetate caused accelerated nerve regeneration in the wild-type rats when measured three weeks after the injury. This finding was expressed both histologically (on the basis of the axon count) and functionally (on the basis of the tibialis muscle response). This effect was not seen in the nude-type rats, although there was an increase in the axon count in the nude glatiramer acetate group after six weeks. Furthermore, the increase was found in the wild-type rats after three weeks, distal to the site of injury, and in the nude type at six weeks, proximal to the site of injury. Following crush injury, it is known that the regeneration of rat sciatic nerve is rapid and that no measurable differences should be expected after four weeks18,19. The regeneration of peripheral nerves may be seen to advance with time20, which may explain these results. At the time of intervention in the present study, the accelerated growth was found to advance distal to the injury in the wild-type rats (at three weeks). Only at a later time (at six weeks), and to a lesser extent (proximal to the injury), was it seen in the nude-type rats.
The effect of nerve growth factors is known to be dose-dependent21-24. The dose response to glatiramer acetate has not been described in the past, although it has been suggested that its effect should be limited for optimum neuroprotection25. In the present study, we found that the additional dose of glatiramer acetate, forty-eight hours after the injury, did not result in improved regeneration compared with the control groups. This finding may have been due either to the dose of glatiramer acetate or to the timing of its administration13,26. It is possible that the additional late effect (at forty-eight hours) of glatiramer acetate or its elevated dose had a negative effect on the dynamics of the immune response, or the growth factor milieu, at the site of injury (as seen with the footprint analysis). Specifically, brain-derived neurotrophic factor (BDNF) has been known to have a bimodal effect on peripheral nerve regeneration24. Several explanations have been suggested: (1) differential activity of BDNF motor-neuronal receptors (trkB facilitatory receptor as compared with p75 inhibitory receptor), (2) downregulation of trkB activity with prolonged exposure to BDNF, (3) generation of free radicals with higher levels of BDNF, and (4) p75 receptor-mediated decrease in supportive Schwann cell activity24. BDNF has been implicated as having a major role in the mechanism of action of glatiramer acetate9, which may explain the results of the present study.
Although we saw a positive effect on nerve regeneration with glatiramer acetate both histologically and functionally, other functional measures failed to demonstrate improvement in the recovery from injury. Several factors may explain this finding. Some measures may be less sensitive than others, the regeneration may have been too rapid to evaluate differences in this model, or the dose or treatment regimen may not have been the optimum one. It is also known that there is limited correlation between the different outcome measures, which depict different aspects of the recovery process18. Finally, it should be noted that the crush injury model is a more difficult test of therapeutic benefit in comparison with the nerve transection injury. The crush injury requires a shorter evaluation period, and the recovery is more complete and less variable than is the case with transection18.
An increase in axon count in comparison with the control limb was noted at six weeks in most groups. This increase represents the robust capability of the rat peripheral nerve to regenerate18. This increase was significantly higher in the nude-type rats treated with glatiramer acetate after six weeks. This finding can not be explained by the proposed mechanism of T-cell manipulation by glatiramer acetate. It may be explained by other mechanisms of glatiramer acetate action described in the literature. The late and less pronounced regeneration may be explained by the minor role of two other suggested mechanisms: (1) production of neurotrophic factors by cells other than T cells9,25; or (2) regulation of free radical production as shown in multiple sclerosis models27,28.
We found a beneficial effect of the antigen glatiramer acetate on the regeneration of the injured peripheral nerve. This was limited by the addition of another, later dose of glatiramer acetate and was significantly decreased in nude-type rats undergoing similar intervention. Additional investigation is needed to evaluate the relationship between dosage and regimen.