Robust Axonal Regeneration in a Mouse Vascularized Composite Allotransplant Model Undergoing Delayed Tissue Rejection

Abstract

Background: Nerve regeneration in vascularized composite allotransplantation (VCA) is not well understood. Allogeneic transplant models experience complete loss of nerve tissue and axonal regeneration without immunosuppressive therapy. The purpose of this study was to determine the impact of incomplete immunosuppression on nerve regeneration. Methods: In this study, transgenic mice (4 groups in total) with endogenous fluorescent protein expression in axons (Thy1-YFP) and Schwann cells (S100-GFP) were used to evaluate axonal regeneration and Schwann cell (SC) migration in orthotopic-limb VCA models with incomplete immunosuppression using Tacrolimus (FK506). Survival and complication rates were assessed to determine the extent of tissue rejection. Nerve regeneration was assessed using serial imaging of axonal progression and SC migration and viability. Histomorphometry quantified the extent of axonal regeneration. Results: Incomplete immunosuppression with FK506 resulted in delayed rejection of skin, muscle, tendon, and bone in the transplanted limb. In contrast, the nerve demonstrated robust axonal regeneration and SC viability based on strong fluorescent protein expression by SCs and axons in transgenic donors and recipients. Total myelinated axon numbers measured at 8 weeks were comparable in all VCA groups and not statistically different from the syngeneic donor control group. Conclusions: Our data suggest that nerve and SCs are much weaker antigens compared with skin, muscle, tendon, and bone in VCA. To our knowledge, this study is the first to prove the weak antigenicity of nerve tissue in the orthotopic VCA mouse model.

Introduction

Vascularized composite allotransplantation (VCA) promises the potential for significant quality of life improvement for individuals who have experienced major tissue loss or amputations. With advances in immunosuppressive therapies, additional improvements to nerve regeneration and functionality of these transplanted parts have become increasingly relevant. Nerve regeneration and axonal reinnervation of graft musculature are critical for functional recovery in transplanted tissue.

Schwann cells (SCs) play an important role in peripheral nerve regeneration by providing a positive regenerative environment, including releasing neurotrophic factors and producing the microstructure for guiding axonal progression. Multiple studies have demonstrated that SC migration precedes axonal regeneration.^6,13 Furthermore, in a nerve allograft model, episodes of rejection stimulate host SC migration into the allografts from both proximal and distal host nerve ends to repopulate the allografts.¹³ The host SCs replace lost donor SCs and support continued axonal regeneration. This migration is the key to long-term nerve function. Thus, the repopulation of donor nerve with host SCs in a VCA is hypothesized to be vital to protecting its long-term functional utility.^2,15 Although an acute rejection episode does not alter nerve regenerative capacity or function, a late rejection episode diminishes function possibly due to loss of SCs and resulting conduction block.¹⁵ Therefore, a greater understanding of SC migration and axonal regeneration in response to immune rejection is paramount to improve functional recovery and long-term stability.

Previously, it was demonstrated that the use of immunosuppression prevented hindlimb VCA rejection and facilitated robust axonal regeneration. However, as expected, lack of immunosuppression resulted in complete rejection of all tissues including nerve.¹⁴ In the current study, we used a sub-immunosuppressive dose of Tacrolimus (FK506) to induce selective and delayed rejection of the VCA. This incomplete immunosuppressive drug regimen allowed us to consider the antigenicity of nerve and its regenerative capacity under these conditions.

Material and Methods

Animal Care

All housing, care, and surgical procedures were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals, and the experimental protocols were approved by our university’s Animal Studies Committee. Animals were housed in a central animal care facility with access to standard rodent chow and water ad libitum. All animals were evaluated daily for signs of rejection or distress. The criteria used to diagnose rejection have been previously detailed.^14,15 Briefly, these signs include erythema, progressive edema, and exudation, followed by eschar formation, necrosis, mummification, and slough.

Experimental Design

A total of 34 mouse orthotopic limb transplantations were performed in this study (Table 1). All mice used in the studies were obtained from Jackson Laboratory (Bar Harbor, Maine), and transplants were performed in 2- to 3-month-old mice weighing 20 to 25 g. Animal surgeries were performed sequentially from groups 1 to 4. For group 1 (syngeneic transplant and positive control), Thy1-YFP transgenic mice (background of C57BL/6 mice [H-2k^b]) were used as recipient and C57BL/6 as donor. In group 2 (allogeneic transplant), Thy1-YFP mice received donor Balb/c (H-2K^d) hindlimb transplants. In groups 3 and 4 (semi-allogeneic transplants), S100-GFP transgenic mice were used in combination with C57BL/6 as recipient and donor, respectively. Due to the mixed background of S100-GFP transgenic mice (B6D2F1/J [H-2k^b/d] from a mating of C57BL/6 [H-2k^b] and DBA/2J [H-2k^d]), a semi-allogeneic transplant was constructed. In the study, a negative control group (allogeneic transplant without FK506 treatment) was not used as our previous data demonstrated transplant rejection signs by 5 days postoperatively and resulting nerve tissue was completely rejected leaving essentially no tissue to analyze.¹⁴

Table 1.

Study Design.

Group	Donor (background)	Recipient (background)	Treatment
Group 1 (syngeneic)	C57BL/6 (H-2Kb)	Thy1-YFP (C57BL/6, H-2Kb)	No FK506
Group 2 (allogeneic)	Balb/c (H-2Kd)	Thy1-YFP (C57BL/6, H-2Kb)	FK506 (2 mg/kg, daily)
Group 3 (semi-allogeneic)	C57BL/6 (H-2Kb)	S100-GFP (B6D2F1/J, H-2Kb/d)	FK506 (2 mg/kg, daily)
Group 4 (semi-allogeneic)	S100-GFP (B6D2F1/J, H-2Kb/d)	C57BL/6 (H-2Kb)	FK506 (2 mg/kg, daily)

Subcutaneous delivery of Tacrolimus (FK506, LOT-ATH-114; LC Laboratories, Woburn, Massachusetts) was performed daily for immunosuppression from the time of surgery until the end point of the study at a dose of 2 mg/kg for the allogeneic (group 2) and semi-allogeneic (groups 3 and 4) mice. All mice in the study underwent serial live imaging of their sciatic nerve of the transplanted limb at 2, 4, 6, and 8 weeks post transplantation for monitoring axonal progression in groups 1 and 2 and SC migration in groups 3 and 4. After 8 weeks, nerve, skin, and muscle tissues were harvested for histologic assessment. An end point of 8 weeks post limb transplantation was chosen based on a pilot study in which syngeneic limb transplants demonstrated muscle force production comparable with uninjured mice (data not shown).

Surgical Technique

Anesthesia was administered via subcutaneous injection of ketamine (75 mg/kg) and medetomidine (1.0 mg/kg). Right hind limbs of donor and recipient animals, including bone, muscle, femoral vessels, sciatic nerve, saphenous nerve, and skin, were amputated at the mid-femoral level as discussed in our previous rat model¹⁵ and as Zhang et al described.¹⁶ Donor hind limbs were orthotopically transplanted onto recipients. Osteosynthesis was achieved using a 25-gauge needle (about 8- to 10-mm length) as an intramedullary rod. Musculature was re-approximated by groups and secured using 6-0 silk suture (Sharpoint, PBN Medicals, Denmark). Donor femoral artery and vein were microsurgically anastomosed end-to-end to recipient femoral artery and vein using 11-0 nylon suture on a 50-µm needle (Sharpoint, Reading, Pennsylvania). The anastomoses were performed nearly “underwater,” with the vessels submerged in heparinized saline (50 U/mL), as previously described,^12,14 for better filling and visualization of the lumen. Following reperfusion, host and donor sciatic and saphenous nerves were coapted with epineurial neurorrhaphy using 11-0 nylon suture. Skin incisions were then closed with 6-0 silk suture. All surgery was performed under magnification (6×-25×) as necessary. Postoperatively, animals were maintained on a warming pad overnight for observation prior to returning to the animal care facility. No dressings, immobilization, or systemic anticoagulation was used.

Zhang et al considered hypovolemic shock a major reason for animal death after limb transplantation.¹⁶ Therefore, two bolus injections of 0.3 mL saline were subcutaneously injected in recipients of groups 1 and 2 before transplant or vascular clamp release. In groups 3 and 4, one bolus of 0.3 mL Lactated Ringer’s solution was injected before clamp release (Table 2).

Table 2.

Summary of Surgical Survival and Complications.

Group	No. of VCA procedures	No. of death post surgery	No. of rejection	No. of autophagia	No. of failed anastomosis	No. for live imaging and histology
Group 1 (syngeneic)	15	6 (40%)	0	3 (killed 2)	1 (vein, killed)	6
Group 2 (allogeneic)	7	3 (42.9%)	3	1 (killed)	0	3
Group 3 (semi-allogeneic)	6	1 (16.7%)	3	2	1 (artery, killed)	4
Group 4 (semi-allogeneic)	6	1 (16.7%)	2	1 (killed)	1 (vein, killed)	3

Serial Live Imaging

All recipient mice underwent serial in vivo live imaging at 2, 4, 6, and 8 weeks post limb transplantation. For the live imaging, animals were anesthetized, prepped, and the sciatic nerves were exposed. Under a MVX10 fluorescence dissecting macroscope (Olympus, Center Valley, Pennsylvania), images were captured with a cooled CCD camera controlled by MetaMorph Software, version 7.1 (Universal Imaging Corporation, Downingtown, Pennsylvania). After imaging, the skin was closed with 6-0 silk suture. Animals were reversed from anesthesia via subcutaneous injection of atipamezole HCl (2 mg/kg; Orion Pharma, Espoo, Finland) and recovered on a circulating water heat pad before they returned to the animal facility. All imaging data were measured and analyzed with NIH ImageJ software (v.1.46r).

Axonal progression and SC migration were determined by measuring the “intensity ratio” of fluorescent expression from the imaged nerve. The intensity was calculated by measuring the mean pixel values for 3 areas spaced approximately 2 to 3 mm apart spanning about 15 mm within the donor and recipient sciatic nerve (Figure 2a). The 3 areas were averaged to represent the donor and recipient intensity values. The values were compared as a ratio to indicate increasing or decreasing presence of axons and SCs based on fluorescent intensity.

Figure 2.

VCA models demonstrated robust axonal regeneration despite strong signs of transplant rejection.

Note. Serial images of nerve taken from transplantation models demonstrate axonal regeneration into nerve in both syngeneic models (a) and allogeneic models (b). The right panels of Figure 2a depict how fluorescent intensity was measured at 3 locations in both the donor and recipient nerves. “P” indicates proximal and “D” distal in the complex of sciatic nerve. Red arrows indicate suture line in the complex. Yellow line indicates scale bar. (c) The intensity of YFP axons in donor to recipient nerves in both syngeneic and allogeneic groups was similar and increased over time demonstrating axonal progression into the transplanted nerve. Data are represented as mean with standard deviation error bars. VCA, vascularized composite allotransplantation.

Histomorphometric Analysis

Eight weeks after live imaging, the sciatic nerve as well as muscle and skin (when available) were harvested and fixed in cold, buffered 3% (weight/volume) cold glutaraldehyde (Polysciences, Inc., Warrington, Pennsylvania) solution for 24 hours, then postfixed in 1% osmium tetroxide, dehydrated in graded ethanol solutions and embedded in Araldite 502 (Polysciences, Inc.). One-micrometer-thick tissue cross sections were obtained using an LKB III Ultramicrotome (LKB-Produkter AB, Bromma, Sweden) and stained with toluidine blue. Skin and muscle were qualitatively observed by light microscopy to confirm an immune response. Sciatic nerve distal to the nerve coaptation site was evaluated under light microscopy for neural architectural characteristics. To quantify axonal regeneration, an observer blinded to the experimental groups used automated digital image analysis software to calculate the total number of myelinated axons.⁸

Data Analysis and Statistics

The mean ± standard deviation is represented in all data. Analysis of variance was performed to determine differences between groups with post hoc comparisons determined by a Newman-Keuls test using statistical software (Statistica v.6; StatSoft, Inc., Tulsa, Oklahoma). Statistical significance was established at P < .05.

Results

Transplantation Survival Rates and Other Complications

Animal models of VCA have inherent complications due to the extensive surgical procedures. Therefore, surgical procedures were modified dynamically over the course of the study to minimize death while avoiding changes to experimental variables being considered—nerve rejection and regeneration. Overall, 16 of 34 mice survived the procedures and were included in the present study.

Surgeries were performed sequentially from groups 1 to 4. Hypovolemic shock has been considered a major reason for animal death post limb transplant.¹⁶ Thus, our initial procedures (groups 1 and 2) utilized 0.95 to 1.15 mL of the total supplemented fluid over the course of surgery (Table 2). As this volume of fluid resulted in a high death rate (40%-43%), we considered hypervolemic syndrome as an additional complication resulting in death. By decreasing supplemented fluid volume to 0.65 mL over the course of surgery, the death rate was substantially lowered to 16.7% in groups 3 and 4.

Autophagia or self-mutilation occurred in 7 of 34 mice (20.6%). Four mice were killed due to the wounds. An additional 3 mice (3 of 34, 8.8%) were killed within 72 hours post transplantation due to failed vascular anastomosis.

Transplantation Rejection

Considering only mice that survived the transplantation procedures, transplanted limbs generally experienced edema for 1 week before resolution (Figure 1a, in pilot study). Group 1, which had received syngeneic transplants, showed no signs of tissue rejection (Table 2, Figure 1b). Normally, allogeneic transplants without immunosuppressant treatment experience tissue rejection (erythema to eschar formation) as early as postoperative day 5.¹⁴ Group 2, which received allogeneic transplants and incomplete immunosuppression with FK506, showed signs of delayed rejection by 2 weeks post transplantation in all mice (Figure 1c). In groups 3 and 4, which had received semi-allogeneic transplants and FK506 treatment, mice showed the same delayed rejection signs around 2 weeks post transplantation as group 2 (Figures 1d and 1e). Under incomplete immunosuppression, all mice exhibiting delayed rejection lacked typical acute rejection signs consisting of edema and erythema. These mice instead exhibited necrosis from the skin edge in the conjoined transplant location (Figures 1c, 1d, and 1e, red arrows). Over 8 weeks, mice undergoing delayed rejection exhibited a loss of bone, muscle, skin, and tendon. Yet grossly, nerve tissue remained intact. Therefore, the extent of nerve regeneration during delayed rejection was analyzed further.

Figure 1.

Representative images of mouse VCA complications and rejection signs. (a) Edema in the grafted limb started 2 days post surgery and resolved by 7 days. (b-e) Tissue rejection signs displayed by 2 weeks post operation and were characterized by necrosis at conjunction skin edges (red arrows). (b) Syngeneic transplant at 8 weeks. (c) Allogeneic at 2 weeks. (d and e) Semi-allogeneic transplants at 2 weeks.

Note. VCA, vascularized composite allotransplantation.

Assessment of Nerve Regeneration During Delayed Tissue Rejection

To understand how delayed tissue rejection affects nerve regeneration, only mice exhibiting signs of delayed rejection in transplanted limbs were considered for further analysis of nerve regeneration in allogeneic and semi-allogeneic transplants. Tissue rejection was confirmed using qualitative histology revealing lymphocyte infiltration in all tissues (data not shown). To assess the extent of axonal progression, the sciatic nerve was imaged biweekly for up to 8 weeks post transplant in Thy1-YFP mice, and fluorescence intensity measured in donor and recipient nerve (Figure 2a). In group 1, the syngeneic graft with no tissue rejection, axonal progression was visible over the time course of imaging and demonstrated robust axonal regeneration into the donor nerve (Figure 2a). In group 2, which displayed delayed tissue rejection, axonal progression into the donor nerve demonstrated similar trends to group 1 with robust axonal rejection despite ongoing tissue rejection (skin and muscle) in the limb (Figure 2b). Quantification of the axonal progression by measuring fluorescence intensity revealed similar levels of relative donor nerve intensity increasing with time (Figure 2c). In fact, group 2 showed greater levels of relative donor axon intensity compared with group 1 by 8 weeks, which clearly demonstrated that axonal progression into donor nerve was not affected by ongoing tissue rejection in the limb.

To assess the extent of SC migration, the sciatic nerve was imaged biweekly for up to 8 weeks as before using S100-GFP mice as recipients and donor transplants. In groups 3 and 4, no detectable SC migration occurred (Figure 3). Despite ongoing tissue rejection of muscle, bone, skin, and tendon, group 4 (which received S100-GFP limbs) contained SCs with robust fluorescence demonstrating that even SCs remained viable in the nerve.

Figure 3.

Semi-allogeneic VCA models demonstrate SC viability over 8 weeks, but a lack of migration.

Note. Serial images captured of S100-GFP mice (representative images from group 4 [S100-GFP as donor are shown]). Despite signs of tissue rejection beginning at 2 weeks post transplantation (right panels), the imaging signal in the donor nerve containing fluorescent SCs was bright suggesting viable SCs but no migration (change of intensity in the recipient). The right panels show the limb tissue rejection corresponding with the imaged nerve regeneration. “P” indicates proximal and “D” distal in the nerve complex. Red arrows indicate suture line in the complex. VCA, vascularized composite allotransplantation; SC, Schwann cell.

After the completion of imaging, the nerves were harvested for quantification of myelinated axon numbers. In the transplanted limb (distal to the coaptation of the sciatic nerve), all groups showed robust axonal regeneration based on myelinated axon numbers (Figure 4). Overall, the data indicate that despite extensive tissue rejection in the transplanted limb, nerve tissue was minimally affected and yielded robust axonal regeneration.

Figure 4.

Histomorphometric analysis of distal nerve in VCA models.

Note. Allogeneic and semi-allogeneic models demonstrated robust axonal regeneration despite ongoing tissue rejection. No significant differences were found in the total number of regenerated myelinated axon numbers. Data are represented as means with standard deviation error bars. VCA, vascularized composite allotransplantation.

Discussion

Current reconstructive strategies for major tissue loss face challenges. These include poor functional outcomes, high patient morbidity, prolonged rehabilitation, and high cost of multiple operations. In complex injuries not amenable to conventional reconstruction with native tissues, VCA seeks to achieve tissue restoration and functional outcomes that lead to improved quality of life. Despite increasing clinical interest in VCA, the process of neurological recovery following VCA is not well understood, and a need for comprehensive study to understand nerve regeneration is critical to advance the field. Therefore, our study assessed axonal regeneration and SC migration and viability using transgenic mouse VCA models to gain further insight.

There are many potential advantages to using mouse VCA models including the similarity between the mouse histocompatibility-2 (H-2) and human leukocyte antigen (HLA) systems. There are also well-established protocols for manipulating the immune system in the mouse, methods for exploring the mechanisms behind many tolerance induction models, and access to a variety of transgenic models.¹ Previously, a heterotopic mouse limb transplantation model decreased morbidity, mortality, and required surgical skill.¹⁴ However, this model limits functional assessments including metrics of behavioral and evoked motor responses and potential optical imaging of regeneration. In this study, we utilized an orthotopic transplantation model to take advantage of commercially available Thy1-YFP and S100-GFP transgenic mice, which allow nerve regeneration to be serially monitored using optical imaging modalities.

As it was previously determined that robust axonal regeneration proceeds with strong immunosuppression to prevent overall tissue rejection following VCA,¹⁴ the current study used a sub-immunosuppressive dose of Tacrolimus (FK506) to induce selective and delayed rejection of the VCA. This sub-immunosuppressive drug regimen allowed us to consider the antigenicity of nerve and its regenerative capacity under these conditions. This monotherapy resulted in delayed tissue rejection in most allogeneic and semi-allogeneic groups. These mice showed delayed rejection signs at the junction of the native and transplanted tissue around 2 weeks post transplantation. With increasing time, the rejection progressed through the whole limb by 8 weeks. Unexpectedly, even though the grafted limbs showed strong rejection signs on skin, muscle, tendon, and bone, optical imaging revealed undetectable impact on grafted nerve in the limbs. In addition, histological evaluation of nerve revealed robust axonal regeneration (myelinated axon numbers) and no immunoreactivity (lymphocyte infiltration). This suggests that the monotherapy of sub-immunosuppressive FK506 effectively prevented rejection in the grafted nerve and SCs but did not prevent rejection in other tissues. The phenomenon of rejection in skin, muscle, tendon, and bone yet minimal impact to the nerve demonstrates the weak antigenicity of the nerve. This study is the first to directly show the weak antigenicity of nerve and SCs in a VCA model. However, our study is limited in scope with regard to the impact of this finding, as we chose a single immunosuppressive (FK506) with known neuroregenerative capabilities.^3-5 Future studies will need to consider whether different immunosuppression regimens still produce limited rejection of nerve.

In the nerve allografts, SCs represent the main immunogenic target for acute and chronic rejection by eliciting the host immune response through expression of both major histocompatibility complexes (MHCs) I and II.^9-11 Donor SCs provide regenerative cues that stimulate and sustain regeneration. However, an aspect of nerve regeneration that may be crucial to long-term graft function is the replacement of donor with host SCs to avoid SC loss long after transplantation that could result in nerve block. It was previously shown that the loss of donor SCs in nerve grafts is compensated by recipient SC migration from both the proximal and distal stumps.^6,7,13 Sudden rejection of donor SCs temporarily halts axonal regeneration and stimulates the proliferation and migration of recipient SCs into the allograft. In contrast, induced immune tolerance resulted in virtually no graft rejection over a 28-day time course, thus abolishing recipient SC migration.⁷ However, in VCA, recipient SC migration is unidirectional, occurring only from the proximal nerve stump. Therefore, we considered how SC migration might be affected by delayed rejection. In this study, S100-GFP transgenic mice were used to monitor SC migration. Serial living imaging from 2 to 8 weeks showed no sign of SC migration from either donor-to-host or host-to-donor nerves. This result suggests a lack of SC migration during delayed tissue rejection perhaps due to the presence of healthy or viable SCs in both the donor and host nerves. In addition, the strong imaging signal over 8 weeks corresponds with this idea suggesting that SCs in the donor were not rejected and remained viable despite the marked signs of tissue rejection in the remaining transplanted limb. However, our study was not designed to detect small changes to SC migration and populations, as we used macroscopic optical imaging. Therefore, we were limited to determine minor changes to SC populations during delayed rejection.

Conclusions

In summary, our data demonstrate robust axonal regeneration and SC viability despite gross evidence of tissue rejection in allogeneic transplanted limbs treated with incomplete immunosuppression using Tacrolimus. The data suggest that nerve, including SCs, is less immunogenic than grafted skin, tendon, muscle, and bone. To our knowledge, this study is the first to demonstrate this rejection hierarchy for consideration in VCA models. This work indicates that the use of VCA mouse models to study nerve regeneration must carefully consider nerve regeneration observed, as axonal progression and SC survival post transplantation can be quite resilient despite clear signs of tissue rejection.