A Novel Model for Acute Peripheral Nerve Injury in the Horse and Evaluation of the Effect of Mesenchymal Stromal Cells Applied In Situ on Nerve Regeneration: A Preliminary Study

Transplantation of mesenchymal stromal cells (MSCs) to sites of experimentally created nerve injury in laboratory animals has shown promising results in restoring nerve function. This approach for nerve regeneration has not been reported in horses. In this study, we first evaluated the in vitro ability of equine bone marrow-derived MSCs (EBM-MSCs) to trans-differentiate into Schwann-like cells and subsequently tested the MSCs in vivo for their potential to regenerate a transected nerve after implantation. The EBM-MSCs from three equine donors were differentiated into SCLs for 7 days, in vitro, in the presence of specialized differentiation medium and evaluated for morphological characteristics, by using confocal microscopy, and for protein characteristics, by using selected Schwann cell markers (GFAP and S100b). The EBM-MSCs were then implanted into the fascia surrounding the ramus communicans of one fore limb of three healthy horses after a portion of this nerve was excised. The excised portion of the nerve was examined histologically at the time of transection, and stumps of the nerve were examined histologically at day 45 after transplantation. The EBM-MSCs from all donors demonstrated morphological and protein characteristics of those of Schwann cells 7 days after differentiation. Nerves implanted with EBM-MSCs after nerve transection did not show evidence of nerve regeneration at day 45. Examination of peripheral nerves collected 45 days after injury and stem cell treatment revealed no histological differences between nerves treated with MSCs and those treated with isotonic saline solution (controls). The optimal delivery of MSCs and the model suitable to study the efficacy of MSCs in nerve regeneration should be investigated.

inTrODUcTiOn Horses suffer injury to peripheral nerves from trauma, metabolic disease, toxins, genetic disorders, and degenerative and infectious diseases (1)(2)(3). Sequelae to nerve injury in horses frequently include poor performance, disability, or even death, which, in turn, result in profound financial and emotional burden. Schwann cells form the myelin sheath surrounding axons in the peripheral nervous system. This sheath is regularly segmented by the nodes of Ranvier, which help in transmitting the nerve impulse in a saltatory and, thus, extremely fast manner (4). Peripheral nerves are able to regenerate after injury due to the secretion of cytokines and neurotrophic factors emanating from the damaged cells (especially from Schwann cells) and to phagocytosis of cellular debris by local macrophages (5)(6)(7).
Residual nerve function depends on the magnitude and chronicity of damage to the nerve (8). The prognosis for recovery of a nerve is poorest when the nerve is transected, and its fibers and surrounding fascia (connective tissue) have been completely disrupted. This magnitude of damage is known as neurotmesis in Seddon's classification of peripheral nerve injury (9,10). Surgical techniques for repairing nerves involve apposing the stumps of the nerve with sutures or inserting a graft to bridge the gap between the stumps (5,6). Some of the most commonly used materials for fabricating grafts or scaffolds consist of vein, artery, nerve, silicone, collagen, laminin, gels made of platelet-rich plasma (PRP), and synthetic polymers (11,12). Regardless of the technique of tissue engineering used, clinical results are often disappointing (10,13).
Results of research using laboratory animals indicate that using mesenchymal stromal cells (MSCs) might be an alternative approach for repairing nervous tissue, including injury to the spinal cord and peripheral nerves (14)(15)(16). The neuroprotective effects of MSCs have been widely described and involve antiinflammatory, immunomodulatory, angiogenic, and nurturing mechanisms (17)(18)(19)(20). Furthermore, MSCs from bone marrow or adipose tissue are able to differentiate into Schwann-like cells (SLCs) after being chemically induced under specific conditions (16,21,22). SLCs transplanted into experimentally created nervous lesions of laboratory animals have enhanced repair of axons and myelin and improved sensory and motor functions, perhaps as a consequence of the secretion of specific neurotrophic factors that promote nerve repair (15,16,20).
There are no reports describing outcomes of horses treated for peripheral nerve injury with equine bone marrow-derived MSCs (EBM-MSCs) or any report that describes the ability of EBM-MSCs to differentiate into myelinated SLCs. These studies are necessary before EBM-MSCs can be used in veterinary medicine to enhance regeneration of damaged peripheral nerves. We have recently reported that EBM-MSCs are able to display morphological and protein characteristics of neural progenitors (23). In this preliminary study, we evaluated the ability of EBM-MSCs to differentiate into SLCs after chemical induction, in vitro. Additionally, we describe a new model of peripheral nerve injury in the horse and discuss the outcome after implanting undifferentiated EBM-MSCs into an experimentally created nerve lesion in three horses.

isolation, expansion, and characterization of eBM-Mscs
Bone marrow-derived MSCs, obtained from the sternum, previously characterized and cryopreserved from three equine donors, were assessed for their stem cell properties, as described (19). Briefly, all cells were confirmed to be mesenchymal stromal/ stem cells based on their colony-forming unit assays, MTS viability and proliferation assays, and the adipogenic, chondrogenic, and osteogenic differentiation patterns, as described in previous studies (24). Low-passage MSCs (P1 to P4) isolated from three equine donors were used in all in vitro experiments described below.
subjects Three healthy American Quarter Horse, crossbred mares, 9-13 years old, from the University of Tennessee's teaching herd were used in the study. All procedures were carried out according to a protocol approved by the Institutional Animal Care and Use Committee.

equine Model of Peripheral nerve injury
Horses were sedated with detomidine hydrochloride (0.01-0.02 mg/kg, IV) and butorphanol tartate (0.01-0.02 mg/kg, IV). The distal palmar aspect of both metacarpi was prepared for aseptic surgery, and 2 mL of 2% mepivacaine hydrochloride was deposited subcutaneously adjacent to the medial palmar nerve and adjacent to lateral palmar nerve proximal to the palpable ramus communicans lying palmar to the superficial digital flexor tendon. A scalpel blade was used to create a 15-mm long, cutaneous, longitudinal incision over the ramus communicans. Using a 6-mm diameter biopsy punch, the center of this anastomotic nerve, connecting the medial and lateral palmar nerves, was removed and placed in 10% formalin (Figure 1). The same procedure was performed on the contralateral fore limb.

Treatments and histological analysis of nerve Tissue
After excising a portion of the ramus communicans of one randomly selected fore limb, 10 × 10 6 allogeneic, undifferentiated MSCs (from a mare, whose MSCs had been previously characterized), suspended in 1 mL of sterile isotonic saline solution were instilled into the fascia surrounding the medial and lateral stumps of the ramus communicans (Figure 2). The same volume of sterile isotonic saline solution was injected around the stumps of the contralateral nerve (control). The cutaneous incision was closed with staples, and the distal portion of each fore limb was bandaged. The horses received phenylbutazone at the time of surgery (4.4 mg/kg, PO) and the day after surgery (2.2 mg/kg, PO). The limbs were bandaged until the staples were removed on day 14. Bandages were changed every third day.
The stumps of ramus communicans of each fore limb of each horse (n = 3) were harvested 45 days post stem cell therapy. To remove the stumps, the horses were sedated with xylazine hydrochloride (0.5 mg/kg, IV) and anesthetized with ketamine hydrochloride (2.2 mg/kg, IV). General anesthesia was maintained with isofluorane vaporized in oxygen and delivered through a semiclosed system into a cuffed endotracheal tube inserted orally into the trachea. The horses were placed in right lateral recumbency. A cutaneous, 2-cm, longitudinal incision was created over the medial and lateral aspects of the flexor tendons of both fore limbs to expose the medial and lateral stumps of the ramus communicans. The medial and lateral stumps of the nerve were transected close to the medial or lateral palmar nerve and placed in Carson's fixative until processed for histological examination. The same procedure was performed on the contralateral fore limb, and the incisions were closed with staples, and the distal portion of each fore limb was bandaged. All horses recovered uneventfully from anesthesia. The limbs were bandaged until the staples were removed on day 14. Bandages were changed every third day. Note the typical fibroblast-like morphology of the EBM-MSCs (control), the elongation of the SLCs into a spindle shape, the appearance of one or more cellular processes, and the growth of cells into a "whorl-like" pattern, at day 7. Scale bar = 100 μm. Five-micrometer thick sections of Carson's fixed, paraffinembedded nerve excised from the ramus communicans and from the stumps of the ramus communicans were examined by light microscopy.

resUlTs
No horses showed sign of discomfort or became lame after the surgical procedures, and none of the horses exhibited any sign of immunological response after allogenic EBM-MSCs were implanted in the nerve defects.

schwann cell Differentiation of eBM-Mscs In Vitro
The integrity of the nucleus and cytoplasm of the cells was examined using fluorescence microscopy. TO-PRO ® -3 stain and WGA, specific to the cell membrane, were used to demonstrate the nucleus and the cytoplasmic structure of the SLCs. Lowpassage EBM-MSCs demonstrated the potential to undergo trans-differentiation after being chemically induced for 7 days. Trans-differentiation was observed subjectively using morphological changes in cells exposed to the differentiation medium relative to the undifferentiated controls. The cells elongated until they displayed spindle-shaped morphology, accompanied by the appearance of one or two cell processes (Figure 3). These cells grew in a "whorl-like" pattern, a phenotype typical of Schwann cells. Cells undergoing differentiation displayed this morphological change at 4 days after chemical induction, and approximately 80% of the cells had acquired this morphologic change by 7 days. The undifferentiated controls maintained the typical fibroblastic appearance of a MSC throughout the 7-day period. Subjectively, no differences were observed in the phenotypic characteristics of SLCs, generated from EBM-MSCs, among equine donors.
The expression of S-100b and GFAP was confirmed by a combination of immunofluorescence (Figure 4) and immunoblot (Figure 5) analyses. The expression profiles were confirmed in SLCs differentiated from the EBM-MSCs of all equine donors. Interestingly, immunoblot analyses showed that the undifferentiated control cells expressed β3 and GFAP, but failed to express S-100b. Results suggest that β3 and GFAP may be neural progenitor markers, and their expression can be used as an indicator to demonstrate plasticity (i.e., the ability of equine MSCs to differentiate into other cellular lineages beyond that of the mesodermal lineages).

Peripheral nerve regeneration
The microscopic appearance of the nerves in all specimens transected with the punch biopsy on the first surgical procedure was within normal limits. On day 45, all stumps treated with isotonic saline solution and all stumps treated with MSCs had seromas characterized by cleft-like spaces rimmed by, or partially filled by, aggregates of fibrin overlain, bordered, or partially infiltrated by fibroblastic cells and macrophages (Figure 6). The isotonic salinetreated and MSC-treated stumps had formed post-traumatic, transectional neuromas characterized by haphazard streams, whorls, and fascicles of small vessels, fibroblasts, and Schwann cells (Figure 7). No localized or discrete population of MSCs or primitive cells (e.g., neural or Schwann-cell progenitors) was recognized.

DiscUssiOn
Peripheral nerves can be injured by chemical, thermal, or mechanical trauma (25). Direct trauma, metabolic disease, such as hypoparathyroidism, electrolyte imbalance, and equine motor neuron disease, and idiopathic disease, such as damage to the left recurrent laryngeal nerve (recurrent laryngeal neuropathy), appear to be the nerve injuries most commonly reported in the horse. Treating horses for damage to a peripheral nerve, by administering one or more anti-inflammatory drugs and physical rehabilitation, is often unrewarding. Repair of transected nerve fibers of human beings is enhanced by closing the gap by apposing the nerve endings and their surrounding fascia with sutures or by inserting an autograft (11). The use of autografts in horses, however, might not be practical and could result in unwanted sequelae (26). Techniques investigated to repair damaged nerves using regenerative therapies include transplantation of MSCs, alone or in combination with bioengineered materials, for the purpose of providing proper environmental conditions for survival and proliferation of cells that can aid nerve repair, particularly Schwann cells (20,27). Based on experimental studies in laboratory animals, the transplantation of MSCs after peripheral nerve injury speeds regain of motor and sensory nerve functions (15,20,28,29). Additionally, studies have demonstrated that MSCs are capable of extra-mesodermal differentiation, including differentiation into cells from neural lineage. Transplantation of MSCs differentiated into SLCs on to injured nerves has revealed promising results on nerve function and morphology (14,15,30).
We chemically induced low-passage, bone marrow-derived MSCs from young and middle-aged horses to determine their plasticity into SLCs. Morphological changes, determined by phase-contrast and fluorescence microscopic examinations, became evident by 7 days after chemical induction in all horses. Differentiated cells were elongated, had an oval-shaped cytoplasm, and had formed one or multiple cellular processes. These cells appeared to multiply to form patches, whereas undifferentiated MSCs multiplied to form flat, even layers.
Western blot and immunofluorescence analyses revealed the expression of the Schwann cell markers S-100b and GFAP in SLCs. Undifferentiated MSCs also expressed β3 and GFAP, as previously reported by our group (23), but not S-100b. Expression of β3 and GFAP suggests that EBM-MSCs are capable of differentiating into cells of extra-mesodermal lineages (31).
The in vitro portion of this study relied mainly on morphological analysis and Schwann-cell protein markers to describe the events occurring during differentiation of EBM-MSCs into SLCs. Our results were similar to those experiments performed using bone marrow-derived MSCs from rats and humans, in which cells were able to differentiate and express markers typical of Schwann cells (7,21,30,32).
The in vivo portion of this study used a model for acute peripheral nerve injury in the horse, which has not previously been reported. As opposed to other models of peripheral nerve injury, transecting the ramus communicans does not cause sensory or functional impairment in the horse. This is an important welfare matter. Additionally, we evaluated the effects on speed and pattern of nerve regeneration after transplantation with allogeneic MSCs. Transplanting MSCs that had differentiated into SLCs was not possible due to poor viability of the differentiated cells (25% viable) after the cells were detached from the tissue culture flasks.
A method to optimize collection of equine SLCs is necessary. Furthermore, no histological differences were observed between nerve stumps treated with MSCs and those treated with isotonic saline solution (controls).
For this specific study, we had previously characterized the stemness of the MSCs from all our equine donors. Horses with various diseases presented to our institution have been treated with allogeneic MSCs previously characterized and stored. We have subjectively perceived benefits from transplanting these allogeneic cells into damaged tendons and ligaments. Our study was aimed at differentiating allogeneic MSCs into SLCs for treating horses for nerve injuries because the ultimate aim of our research is to have cells readily available to use for regenerating damaged nerves, because to regenerate nerves, expeditious treatment after injury is critical for success. Allogeneic MSCs can be used as an "off the shelf " product soon after injury, whereas culturing and expanding autogenous MSCs requires several days.
Results were inconclusive using our model for acute peripheral nerve injury in the horse, perhaps because of the low number of horses used and perhaps because of other factors, such as surgical technique, anatomical region of the experimentally injured nerve (the distal aspect of the limb of the horse has less  vasculature than other areas in the body), method of cellular delivery, and labeling and tracking of cells. A portion of the ramus communicans may, perhaps, be excised with less trauma to adjacent tissue with the horse anesthetized than with the horse standing. With better exposure, the cells could, perhaps, be injected directly into the nerve stumps. Similarly, suturing the incision in the fascia and subcutaneous tissue may prevent the formation of seromas.
To the best of our knowledge, this is the first report describing the morphological features and protein expression changes of EBM-MSCs into SLCs. In this study, we validated the crossreactivity of rat Schwann cell-specific antibodies with protein samples from horses. Further studies exploring the viability, homing, and functionality of MSCs and SLCs after transplantation into horses with peripheral nerve injuries are warranted. Similarly, a practical and affordable method for delivering these cells is necessary. aUThOr cOnTriBUTiOns CCV, JS, and MD designed and conceived the work. CV and MD acquired and analyzed all data. CCV and MD carried out all in vitro experiments. CCV and JS carried out all in vivo experiments and acquired the nerve tissues. RD analyzed the nerve tissues. CCV, JS, and MD drafted the manuscript, and all authors revised it and approved the last version for publication. All authors agree to be accountable for all aspects of the work ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

acKnOWleDgMenTs
The authors thank Nancy Nielsen and Lisa Amelse for their technical support. This research study was funded by the Center of Excellence UTIA.