Evaluation of peripheral nerve regeneration in Murphy Roths Large mouse strain following transection injury

Abstract

Aim

Murphy Roths Large (MRL/MpJ) mice have demonstrated the ability to heal with minimal or no scar formation in several tissue types. In order to identify a novel animal model, this study sought to evaluate whether this attribute applies to peripheral nerve regeneration.

Materials & methods

This was a two-phase study. 6-week-old male mice were divided into two interventional groups: nerve repair and nerve graft. The MRL/MpJ was compared with the C57BL/6J strain for evaluation of both functional and histological outcomes.

Results

MRL/MpJ strain demonstrated superior axon myelination and less scar formation, however functional outcomes did not show significant difference between strains.

Conclusion

Superior histological outcomes did not translate into superior peripheral nerve regeneration in MRL/MpJ strain.

Peripheral nerve injuries remain to be a major healthcare burden with estimated incidence of 1 in 1000 individuals affected per year [1]. Injured peripheral nerves, as opposed to central nerves, have the capacity to regenerate [2,3]. However, despite having a detailed knowledge of the peripheral nerve injury healing processes and an improvement in advanced microsurgical repair techniques, outcomes of peripheral nerve repair have largely remained unchanged over recent decades. Nerve repair can be performed either by direct suturing, via autologous nerve bridging, or via a variety of decellularized nerve allografts or conduits [4–8]. Regardless of these efforts, functional recovery is still seldom optimal and incomplete peripheral nerve recovery results in full or partial loss of motor or sensory function, neurogenic pain, muscle atrophy and overall decreased quality of life. Animal models play a key role in exploring the mechanisms involved in peripheral nerve regeneration and are crucial to advance our understanding of the pathological process as they provide potential to target and modify those pathways to potentially improve patient outcomes.

In general, wound repair process consists of inflammatory response, reepithelization, and scar tissue development. Adult mammalian tissue injury tends to initiate wound repair processes rather than the regenerative mechanisms typical for prenatal period; this primarily provides injured tissue stability by creating mechanical scaffolds. Peripheral nerve regeneration processes demonstrate no exception from scar tissue development. As a result, intraneural scar tissue is created while attempting to find a fine balance between axonal sprouting and scar formation [9]. Intraneural fibrosis occurs within neural intra-substance and can surround critical neural structures on all levels. Although scar formation is an integral part of the tissue repair process, excessive perineural and intraneural scarring deteriorates and disrupts successful axonal regeneration by placing the repair under stress, altering nerve gliding, causing tissue tension and nerve tethering, and fostering increased fibrotic activity [10]. It can alter perineural microvasculature, impede axonal migration, and thus impair overall nerve regeneration [11]. Although nerve regeneration normally occurs at speed of 1 to 3 mm per day, the mechanical barrier created by scarring may result in an additional 20 to 40 days for axons to traverse across the repair [10]. In addition to mechanical impairment, excessive fibrosis may result in decreased nerve conduction velocities [12]. This deterioration is based on limited remyelination by Schwann cells which normally accelerates the nerve conduction velocity 20–100-fold [13]. Ultimately, reinnervation of the correct target is rarely achieved, resulting in abnormal function [14].

Murphy Roths Large (MRL/MpJ) is the parent and control strain for the Murphy Roths Large/lymphoproliferative (MRL/MpJ-Fas^lpr) mouse. These animals were established by selective interbreeding of the B6 (0.3%), C3H (12.1%), AKR (12.6%) and Large (75%) strains [15]. MRL/MpJ-Fas^lpr carry a mutation in the Fas gene with both strains originally being developed as a model for autoimmune disorders [16]. Despite the lack of the Fas mutation, the MRL/MpJ strain exhibits autoimmune disorders, although these are manifested later in life [17]. Interestingly, MRL/MpJ demonstrate superior healing features with minimal or even scar-less wound healing in certain body areas [15]. The first report of enhanced healing in MRL/MpJ animals occurred after the spontaneous closure of identifying ear-punch holes [18]. In contrast to adult wound healing, MRL/MpJ animals displayed a capacity of wound closure considered to be a form of regeneration with embryogenic epidermal healing containing blastema-like structures. The lack of scar tissue and formation of a circular blastema in adult rodents represented a unique situation for the MRL/MpJ strain and its ancestral Large strain [18]. This process posed multiple histological similarities to amphibian limb regeneration and late embryogenesis and was subsequently confirmed by other investigators [19–21]. Reduced inflammatory response, rapid reepithelization and reduced fibrosis was also demonstrated after corneal alkali burn injuries [22]. Furthermore, MRL/MpJ animals demonstrated a failure to develop the expected arthritic changes following trauma and were indistinguishable from the uninjured contralateral side [23]. However, none of these regenerative features have been observed in excisional back wounds [21], thermal skin injury [24], brain cortex stab lesions [25], or adult digit amputation [26].

The concept of superior healing being associated with decreased scar formation, as demonstrated in MRL/MpJ strain's ear and corneal injury, might lead into improved healing after peripheral nerve injury. Buckley et al. demonstrated a significantly higher density of regenerated cutaneous nerves when compared with controls within the first 3 weeks following ear-punch holes described previously [27]. However, there is limited data investigating the peripheral nerve healing capacity of the MRL/MpJ strain and reports documenting the functional regenerative properties of the peripheral nervous system are lacking.

The purpose of this study is to address this deficiency by utilizing accepted methods to evaluate the functional outcomes of peripheral nerve regeneration in the MRL/MpJ strain. We seek to test the hypothesis that the MRL/MpJ mouse strain will demonstrate superior functional and histological outcomes in comparison to the control strain, C57BL/6J, after peripheral nerve transection injury and subsequent repair.

Materials & methods

Animal model

All experiments were approved by the University Committee on Animal Resources at our institution (approval number: 102171/2018-034). All animals were acclimated for at least 7 days before experimentation. Mice were group-housed in vivarium under standard controlled conditions. MRL/MpJ (Jackson Laboratories, USA) mouse strain served as a study group. C57BL/6J (wild-type) (Jackson Laboratories), animal model commonly used in studies investigating different aspects of peripheral nerve regeneration, was chosen as the control strain given the lack of specific mutations that may both enhance or alter regeneration of the peripheral nerve [28–33]. In addition, C57BL/6J has been shown to express high resistance to self-mutilation and, therefore, to be an ideal model to study peripheral nerve regeneration [34].

Study design

This was a two-phase study. During the first phase, 132 8-week-old mice were included in the study. 12 mice, MRL/MpJ (n = 6) and C57BL/6J (n = 6), were sacrificed at the beginning of the study and used for baseline functional and histological analysis. 120 mice, MRL/MpJ (n = 60) and C57BL/6J (n = 60), were equally divided into two surgical intervention groups based on surgical repair technique: direct repair (DR) and nerve graft (NG). Animals included in the first study phase were assessed with walking track analysis (WTA) to calculate Sciatic Function Index (SFI) and various histological analyses over the course of 12 weeks. Second phase of the study included 40 eight-week-old mice that underwent identical surgical interventions as animals involved in the first study phase and were divided into following groups: MRL/MpJ DR (n = 10), MRL/MpJ NG (n = 10), C57BL/6J DR (n = 10) and C57BL/6J NG (n = 10). Those animals were evaluated for outcome measures with CatWalk gait analysis system and electromyography (EMG) weekly over the course of 6 weeks. Figures 1 & 2 summarize the experimental settings during the first and second study phase.

Figure 1. . Study design.

Bilateral sciatic nerves from 12 mice, MRL/MpJ (n = 6) and C57BL/6 (n = 6), were obtained for baseline histological analysis. During the first study phase, 120 mice, MRL/MpJ (n = 60) and C57BL/6J (n = 60), were then equally divided into two surgical intervention groups based on surgical repair technique: DR and NG. Likewise, second study phase included 40 mice, MRL/MpJ (n = 20) and C57BL/6J (n = 20).

DR: Direct repair; NG: Nerve graft.

Figure 2. . Experimental Timeline.

First study phase (A) had total duration of 12 weeks. Functional outcomes (walking track-based SFI) were assessed at baseline and at postoperative weeks (POW) 01, 03, 06, 09, and 12; while histological outcomes at baseline and at POW 03, 06, 09, and 12 (red X). Second study phase lasted for the duration of 6 weeks (B). Functional outcomes (CatWalk-based SFI and EMG outcomes) were assessed at baseline and then weekly at POW 01, 02, 03, 04, 05 and 06.

Sciatic nerve injury & repair

All mice that underwent surgical intervention were anesthetized with 1.5–2% isoflurane according to the standard operating protocol. A 20 mm curvilinear incision was made over the posterolateral aspect of the right proximal hind-limb. The subcutaneous tissue was dissected to expose the right sciatic nerve (ScN). For mice in the DR group, the ScN was transected and immediately repaired under microscope magnification using epineural 10-0 Ethilon^® sutures (Ethicon, USA) at the 12 and 6 o'clock positions. For mice in the NG group, a 5 mm segment of ScN was excised, reversed, and inset into the previously divided ScN using epineural 10-0 Ethilon^® sutures at the 12 and 6 o'clock positions at both ends of the graft (Figure 3). In all animals the surgical field was irrigated, and the incision was closed with 5-0 Vicryl Rapide™ (Ethicon, USA) interrupted cutaneous sutures. Strict aseptic technique was maintained during the surgery and mice were monitored post-operatively. At postoperative weeks (POW) 3, 6, 9 and 12 (Figure 3), six mice from each group involved in the first study phase were euthanized and the right ScN was harvested for histological analysis.

Figure 3. . Sciatic nerve intervention detail.

Right sciatic nerve (ScN) was skeletonized prior to intervention (A, D). For mice in the DR group, the ScN was transected (B) and immediately repaired (C) under microscope magnification using epineural 10-0 sutures at the 12 and 6 o'clock positions. For mice in the NG group, a 5 mm segment of ScN was excised, reversed, and inset into the previously divided ScN using epineural sutures at the 12 and 6 o'clock positions at both ends of the graft (F).

DR: Direct repair; NG: Nerve graft.

Sciatic function index

During the first study phase, functional outcome analysis was evaluated by calculating the sciatic function index (SFI) according to established and validated methodology by walking track analysis (WTA) [35,36]. Analysis was performed preoperatively at baseline, and at POWs 1, 3, 6, 9, and 12. In brief, a standard 2-inch-wide by 3-foot-long walking track apparatus was utilized. Animals had their left hind-paw painted with blue ink and their right hind-paw with black ink. Subsequently animals were placed at one end of the corridor to allow them to spontaneously traverse to the other, leaving inked pawprints on blotting paper. The pawprints were digitally scanned, blinded, and scored by two independent researchers. Several measurements were taken from the pawprints as described by Meek et al. [37]: distance from heel to toe, the print length; the distance from the first to the fifth toes, the toe spread; and the distance from the second to the fourth toes, the intermediary toe spread. Based on these measurements, corresponding factors (print length factor [PLF], toe spread factor [TSF], and intermediary toe spread factor [ITF]) and resultant SFI was calculated according to Bain et al. [38] as follows:

$$\text{SFI}=(-38.3 \times \text{PLF}) + (109.5 \times \text{TSF}) + (13.3 \times \text{ITF} -8.8)$$

Electromyography

Electromyography (EMG) is well validated and objective method of nerve regeneration monitoring [39–43]. EMG evaluated injured nerve target muscle function during various time points with the maximum amplitude, latency and velocity of compound motor action potential to estimate number of regenerated axons across the site of ScN injury. Five disposable stainless needle electrodes (14 × 0.38 mm, Natus^® Medical Incorporated, CA, USA) were used for each measurement. Anode and cathode electrodes were placed on both sides of sciatic notch. Recording electrode was placed aligning the gastrocnemius muscle and reference electrode subcutaneously next to the Achilles tendon. Finally, ground electrode was inserted over animal's back area. Evoked surface EMG was recorded bilaterally from posterolateral hind limb. Three EMG measurement were performed for each hind limb at each time point. Non-injury side was used as a control.

CatWalk gait system analysis

CatWalk™ XT automated gait analysis system was used to assess locomotor recovery changes [44]. This system allows to monitor gait changes when animal crosses walkway with illuminated transparent glass floor. High-speed and high-resolution camera captured paw prints and gait pattern in the setting of animal movement. After acquisition, data were automatically classified by incorporated software to obtain SFI.

Specimen processing

Right ScN specimens were collected at each euthanasia end point. After anesthesia, the animals were trans-cardially perfused with 50 ml 4% paraformaldehyde (PFA) diluted in phosphate-buffered saline (PBS) and 2.5% glutaraldehyde (GA). Right ScN samples were harvested 1 mm proximal and distal to the repair site for the DR groups, and 1 mm proximal and distal to the interposed nerve autograft in the NG groups. Specimens were placed in individual 1 ml vials containing 4% PFA and 2.5% GA diluted in PBS for 24 h at 4°C. Following rinsing with 0.1 M Millonig's buffer, samples were stained in osmium tetroxide solution (1-part 4% Osmium, 1 part 6% potassium ferrocyanide, and 2 parts 0.1 M Millonig's buffer) for two-hours. We then proceeded with identical processing method as we previously described in Leckenby et al. [45]. Serial dehydration at the temperature of 4°C was performed while specimens were agitated on a BioRocker™ (labForce Thomas Scientific, USA) through a graded series of 20, 50, 70, 90 and 100% ethanol for 5 min at each concentration. Specimens were placed at 4°C in pure acetone for 15 min following dehydration and then washed with pure propylene oxide twice for 10-min durations [45].

The EMBed-812 embedding kit (Electron Microscopy Science, USA) was adjusted to create a soft epoxy-resin mixture, according to Glauert's formula [46]. Room temperature embedding began with a 1:2 mix (epoxy:propylene oxide) for 1 h, then exchanged for a 1:1 mix for an additional hour before being placed in pure epoxy for 12 h, constantly agitated on a rocker. Specimens were then transferred to a fresh mixture of epoxy for 1 h prior to transfer to silicon molds and baked at 60°C for 24 h [45].

Sectioning & imaging

All embedded blocks were initially semi-thin sectioned to a thickness of 2 μm for light microscopy. The semi-thin sections were placed on glass microscope slides and stained with toluidine blue solution (0.25 g sodium borate, 0.25 g toluidine blue 0.25 ml distilled water). Adequate fixation and osmication of the specimens were thus confirmed prior to ultra-thin sectioning for electron microscopy.

Specimens were prepared for ultra-thin sectioning by trimming the block faces to minimize surface area contacting a trim 20^® diamond knife (DiATOME, USA). UC7 Ultra-microtome (Leica, Germany) in combination with an ultra 45° diamond knife and water reservoir (DiATOME, USA) was used to complete ultra-thin sectioning at a thickness of 60 nm prior to view under electron microscope [45]. Thin sections were cut onto grids, stained with uranyl acetate and lead citrate, examined with a Hitachi H-7650 transmission electron microscope, and digitally photographed.

Axonal counting, G ratio & axon to extracellular matrix ratio

All images were reviewed and histomorphometric calculations were completed by two independently blinded researchers. Images were processed with ImageJ software (National Institute of Health, Laboratory for Optical and Computational Instruments, USA) for manual axonal counting. Ratio between the axonal inner and outer myelin sheath diameter (G ratio) was calculated utilizing the ImageJ plugin ‘G Ratio’ (http://gratio.efil.de/). G ratio provides information about axonal myelination and is widely utilized as a functional and structural description of the level of myelination [47]. As a scar formation surrogate, axon to extracellular matrix ratio was calculated from electron microscopy (EM) images. Axon to extracellular matrix (ECM) surface area ratio was calculated to assess proportion of ECM containing collagen fibrils.

Statistical analysis

Results are expressed as means with standard deviation. Statistical analysis was performed using unpaired Student's t-test for the direct comparison of pre-operative means. Furthermore, linear regression was utilized to assessment of measure improvement within each group. Two-way ANOVA followed by Tukey's HSD Test for multiple comparisons was performed to compare the effect of different strains and type of surgical repair. Statistical analysis of data sets was performed using Prism 9 software (GraphPad Software, CA, USA). Statistical significance was defined as p < 0.05.

Results

First study phase

Walking track analysis

No statistical difference was demonstrated between the MRL/MpJ and C57BL/6 groups at baseline. Starting at POW 01, t-test assessing difference between SFI means at each time point demonstrated that MRL/MpJ mice had significantly higher SFI compared with C57BL/6 in both DR and NG groups (p < 0.001 and p = 0.0301, respectively), and their outcomes remained marginally superior throughout the whole study duration until they reached scheduled end point at POW 12. For both strains, DR groups provided best functional outcomes at POW 12 with SFI of -52.16 (MRL/MpJ), and -67.32 (C57BL/6), while NG groups achieved best outcomes at POW 09 with SFI of -54.09 (MRL/MpJ) and of -65.77 (C57BL/6). Regression analysis revealed overall improvement in SFI within C57BL/6J groups (p < 0.0001), but no improvement in MRL/MpJ DR and NG groups (p = 0.0753 and p = 0.0541, respectively). Further analysis with ANOVA demonstrated that there were overall superior SFI results in MRL/MpJ strain compared with C57BL/6J stain for both DR and NG over the time of the study duration (p = 0.0004 and p = 0.0035, respectively).

Histological analysis

There was no significant difference in axon counts between the MRL/MpJ and C57BL/6 strains at baseline. Surprisingly, following surgical intervention the number of axons regenerating through the area of injury and repair was significantly lower in the MRL/MpJ strain compared with the C57BL/6 starting at POW 03 (1611 vs 1931; p = 0.0493) in DR groups. However, no difference was found in NG groups at POW 03 (1087 vs 1111; p = 0.87). At POW 06, number of regenerated axons in MRL/MpJ group was lower in both DR group (1684 vs 2830; p < 0.001) and NG group (1509 vs 2427; p = 0.0004). This trend remained significant in both DR and NG groups throughout the rest of the study time-points at POW 09 (DR: 2094 vs 2901; p = 0.0052; NG: 2048 vs 2492; p = 0.0398) and at POW 12 (DR: 2760 vs 2303; p = 0.0164; NG: 2264 vs 2822; p = 0.0446). By the end of the experimental period (POW 12), the axon number observed in all groups recovered from initial decline over time and reached the values similar to baseline. Regression analysis revealed overall increase of axon count in all groups (p < 0.0001). Similarly to SFI results, significant difference of axon count between mouse strains was noted for both DR and NG throughout the study duration.

G ratio results showed no difference between the two strains at baseline. Starting at POW 03, G ratio was lower in DR MRL/MpJ groups (0.63 vs 0.74; p = 0.0086), as well as in NG MRL/MpJ groups (0.65 vs 0.75; p = 0.0037) compared with the C57BL/6 strain, suggesting better myelination of regenerating axons in investigated groups. The trend of decreased G ratio in MRL/MpJ compared with control strain remained consistent for the whole duration of study. This was demonstrated by results at POW 06 (DR: 0.57 vs 0.66; p = 0.0079; NG: 0.60 vs 0.66; p = 0.0438), POW 09 (DR: 0.64 vs 0.71; p = 0.0059; NG: 0.60 vs 0.68; p = 0.0356), and POW 12 (DR: 0.63 vs 0.72; p = 0.0136; NG: 0.65 vs 0.72; p = 0.0243). Although there was no significant difference in G ratio noted within any of the groups throughout the duration of the first study phase, significantly superior G ratio was demonstrated in MRL/MpJ compared with C57BL/6J strain for both DR (p < 0.0001) and NG (p < 0.0001) groups.

No significant difference was demonstrated between the two strains when comparing baseline values for the axon to extracellular matrix ratio (Axon/ECM) (DR: 2.67 vs 2.61; p = 0.48; NG 2.63 vs 2.58; p = 0.42). After ScN transection and reconstruction, initial steep drop in Axon/ECM was followed by gradual increase through all time-points in all groups irrespective of type of reconstruction. Although not significantly different at POW 03 (DR: 0.20 vs 0.17; p = 0.35; NG: 0.09 vs 0.09; p = 0.95), this ratio was found to be increased in MRL/MpJ strain compared with C57BL/6J at POW 06 (DR: 0.59 vs 0.28; p = 0.0064; NG: 0.40 vs 0.26; p = 0.0461), POW 09 (DR:0.76 vs 0.46; p = 0.0003, NG: 0.54 vs 0.31; p = 0.0034), and finally at POW 12 (DR: vs 0.46; p = 0.0003, NG: 0.54 vs 0.31; p = 0.0034). Axon/ECM significantly improved over the study period in each of the study groups (p < 0.0001). While comparing Axon/ECM over the course of 12 post-operative weeks, MRL/MpJ demonstrated significantly better improvement compared with control strain for both DR and NG groups (p < 0.0001). All first study phase data are summarized in Figure 4 and Table 1.

Figure 4. . Phase 1 study results – walking track-based sciatic function index, axon count, G ratio, axon/extracellular matrix ratio.

Table 1. . Phase 1 study results summary. Comparison of MRL/MpJ and C57BL/6 for all the outcomes measured.

Outcome	Group (n = 10)	Strain	Baseline		POW 01		POW 03		POW 06		POW 09		POW 12
			Mean (SD)	p-value	Mean (SD)	p-value	Mean (SD)	p-value	Mean (SD)	p-value	Mean (SD)	p-value	Mean (SD)	p-value
Sciatic function index	Direct repair	C57BL/6	-8.37 (12.0)	0.15	-71.84 (13.9)	<0.0001	-69.43 (16.6)	0.0015	-65.28 (14.5)	0.0256	-67.48 (13.4)	0.0388	-67.32 (11.0)	0.0168
		MRL/MpJ	-5.45 (12.2)		-61.08 (9.6)		-59.8 (14.6)		-56.78 (13.9)		-54.85 (17.3)		-52.16 (16.6)
	Nerve graft	C57BL/6	-7.16 (16.6)	0.17	-72.26 (13.8)	0.0301	-69.33 (12.9)	0.0164	-73.41 (12.0)	<0.0001	-65.77 (12.6)	0.0499	-67.66 (9.4)	0.0412
		MRL/MpJ	-10.43 (11.7)		-66.63 (13.7)		-62.72 (14.6		-56.25 (13.9)		-54.09 (23.8		-56.78 (16.5)
Axon count	Direct repair	C57BL/6	2381 (486)	0.42	–	–	1931 (375)	0.0493	2830 (613)	<0.0001	2901 (760)	0.0052	2760 (524)	0.0164
		MRL/MpJ	2233 (361)		–		1611 (551)		1684 (354)		2094 (857)		2303 (561)
	Nerve graft	C57BL/6	2360 (484)	0.23	–	–	1111 (459)	0.87	2427 (652)	0.0004	2492 (5910	0.0398	2822 (873)	0.0446
		MRL/MpJ	2401 (423)		–		1087 (429)		1509 (790)		2048 (488)		2264 (726)
G ratio	Direct repair	C57BL/6	0.65 (0.03)	0.48	–	–	0.74 (0.06)	0.0086	0.66 (0.06)	0.0079	0.71 (0.05)	0.0059	0.72 (0.04)	0.0136
		MRL/MpJ	0.64 (0.05)		–		0.63 (0.10)		0.57 (0.08)		0.64 (0.06)		0.63 (0.09)
	Nerve graft	C57BL/6	0.63 (0.04)	0.23	–	–	0.75 (0.07)	0.0037	0.66 (0.06)	0.0438	0.68 (0.06)	0.0356	0.72 (0.08)	0.0243
		MRL/MpJ	0.66 (0.05)		–		0.65 (0.07)		0.60 (0.07)		0.60 (0.09)		0.65 (0.05)
Axon:ECM	Direct repair	C57BL/6	2.61 (0.46)	0.84	–	–	0.17 (0.06)	0.35	0.28 (0.06)	0.0064	0.46 (0.06)	0.0003	0.79 (0.12)	<0.0001
		MRL/MpJ	2.67 (0.42)		–		0.20 (0.02)		0.59 (0.08)		0.76 (0.09)		1.17 (0.02)
	Nerve graft	C57BL/6	2.58 (0.37)	0.42	–	–	0.09 (0.02)	0.95	0.26 (0.10)	0.0461	0.31 (0.07)	0.0034	0.38 (0.03)	<0.0001
		MRL/MpJ	2.63 (0.41)		–		0.09 (0.14)		0.40 (0.08)		0.54 (0.16)		1.08 (0.07)

Bold values denote statistical significance.

Second study phase

CatWalk analysis

CatWalk-based SFI results showed no significant difference in any time point over the 6-week long second study phase in DR groups, except for POW2 with significantly higher SFI in MRL/MpJ DR compared with C57BL/6J (p = 0.0113). No significant improvement within DR groups over the course of the study was observed (C57BL/6J: p = 0.75; MRL/MpJ: p = 0.097). Within NG groups, only MRL/MpJ strain shown significantly improved values within the course of 6 weeks (p = 0.0034). However, overall no difference in improvement between strains was demonstrated in both DR (p = 0.103) and NG (p = 0.21) groups. Results are summarized in Figure 5.

Figure 5. . CatWalk-based sciatic function results.

Electromyography

In DR groups, linear regression analysis revealed improvement within the groups in all three modalities measured on the operated limb over the course of 6 weeks. C57BL/6 DR group demonstrated increase in Amplitude (p = 0.0165), decrease in Latency (p = 0.0007) and increase in Velocity (p = 0.0007). Similar results were seen in MRL/MpJ DR group for Amplitude (p < 0.0001) and Velocity (p = 0.0318). However, Latency decrease was found not significant (p = 0.0638). Most importantly, two-way ANOVA results revealed no difference in above mentioned measures between the strains (p = 0.65, p = 0.442 and p = 0.41, respectively). In NG groups, linear regression showed improvement in Velocity in both B57BL/6J and MRL/MpJ strains (p = 0.001 and <0.0001, respectively) and Latency in C57BL/6J (p < 0.0001). All other results demonstrated to be non-significant in regards to improvement within DR groups. There was no difference in improvement of any of EMG domains measured between strains of any. Results are summarized in Figure 6.

Figure 6. . Electromyography results.

Discussion

The outcomes of peripheral nerve regeneration remain suboptimal with frequent loss of functional recovery. Animal models of transection injuries have been the subject of extensive research, but peripheral nerve injury outcomes largely remain unchanged over recent decades. The purpose of this study was to investigate whether MRL/MpJ mouse strain represents a potential novel animal model to study peripheral nerve regeneration with new opportunities to identify novel underlying mechanisms that affect functional and histological regeneration following peripheral nerve injury and repair since the development of scarring within the injured and regenerating nerve has been proposed as one potential mechanism responsible for poor outcomes [12].

The MRL/MpJ strain is known to have superior wound healing ability that has been attributed to decreased scar formation in several tissue types [15]. In the initial study reporting scar-less healing following ear-punch injury in the MRL/MpJ strain, abnormal tissue architecture was restored [18]. Buckley et al. reported results specifically focusing on peripheral nerve regeneration and demonstrated a significantly higher density of regenerated nerves following ear punch biopsy injury in comparison to C57BL/6J mouse strain [27]. However, currently there are no reports of functional outcomes or morphological analyses of peripheral nerve regeneration in the MRL/MpJ strain. Therefore, the importance of Buckley et al.‘s study remains unknown. Overall, the purpose of our study was to address this deficit and report the regenerative outcomes of the MRL/MpJ strain following peripheral nerve injury.

Although the results of our study have shown that the MRL/MpJ strain demonstrated slightly higher walking track-based SFI values and superior histomorphology outcomes related to G Ratio and Axon to Extracellular Matrix Ratio in comparison to the C57BL/6J strain, clinical relevance remains unclear. More specifically, functional recovery, represented by SFI in our study, revealed significantly superior results in MRL/MpJ individuals at all time-points following injury and repair with both direct neurorrhaphy and nerve grafting based on t-test comparisons at each time points, including POW 1. Based on our current knowledge of peripheral nerve regeneration, it is not expected that superior regenerative capability will be observed as early as one week following peripheral nerve transection and repair [48,49]. Moreover, SFI data analysis using linear regression did not reveal any significant evolution in SFI improvement within investigated groups during the time of the study and, therefore, conclusion of the MRL/MpJ mouse strain accelerated functional recovery based on marginal difference in SFI between those two strains cannot be made. Obvious reason, why walking track-based SFI values were shown to be significantly higher in investigated strain, remains unclear. However, given very marginal difference and high number of animals included in the first study phase, the results are thought to be based on high power of the study which provided ability to recognize even non-clinically significant differences.

Based on those results, we decided to proceed with second study phase to assess peripheral nerve regeneration capacity with further outcome measure method – needle electromyography and SFI calculated by automated gait analysis system (CatWalk™ XT). During the second study phase, experimental animals were followed and assessed for outcomes over the period of 6 weeks. This time frame was based on the first study phase data that showed significant results by POW 06 in all measured categories (Table 1). Both EMG and CatWalk-based SFI results demonstrated no difference between investigated and control strains. This supports first phase study conclusion that there is overall no clinically significant difference in peripheral nerve regeneration capability between MRL/MpJ and C57BL/6J.

In order to correlate with functional outcomes, the study was designed to obtain histological specimens at multiple time points during the first study phase. At POW 01, the histological results revealed a severely distorted neural architecture in all specimens obtained making it impossible to distinguish regenerating axons from those undergoing Wallerian degeneration, and for this reason, these have been excluded from the overall analysis. In our opinion, morphological changes occurring at POW 01 preclude tissue from accurate assessment and therefore, histology data was analyzed at and following POW 03.

Despite having similar baseline sciatic nerve axon counts, the MRL/MpJ strain demonstrated a significantly lower regenerative axon count across all post-operative weeks in both the DR and NG groups. Axon counting is one of the most common measures evaluating nerve regeneration success [5]. However, this measure should be performed in correlation with functional outcomes, such as SFI or electromyography, as several axon sprouts are emitted from each parent axon in the proximal nerve stump and may not reflect number of effector axons during the first three months that poses the count itself a limited measure [50]. One explanation that accounts for the differences in axon counts we have demonstrated may be related to above mentioned axonal sprouting following injury and is most likely a result of different neurotrophic factors stimulating myelin clearance as well as axonal sprouting [51]. Axon regeneration occurs as a consequence of coordinated action under the influence of multiple biological pathways where Schwann cells and macrophages play a key role [49]. In contrast to central nervous system, Schwann cell myelin containing inhibitory molecules can be cleared more readily which results in superior peripheral nerve regeneration ability as a result from creating a stimulating environment [52]. An alternative mechanism that may be responsible for the diminished axon sprouting may relate to delayed Wallerian degeneration. Several animal models, such as WLD^s and C57BL/OIa, have been found to have spontaneous mutations that prevent distal axons from breaking down resulting in diminished axonal regeneration [53,54]. Despite the clear difference in axon sprouting resulting in different number of axons in regenerated nerve, it does not appear to affect functional outcomes based on results obtained from SFI and electromyography.

Re-myelination performed by myelinating subtype of Schwann cells plays an integral role in nerve recovery process [55]. Myelinating Schwann cells envelop regenerating axons to create a multi-layer myelin sheath necessary for functioning motor nerve and, therefore, myelin assessment is pivotal part of nerve regeneration [56–58]. The MRL/MpJ strain demonstrated a superior re-myelinating ability, as demonstrated by a significantly lower G Ratio, in comparison to the C57BL/6J strain. One explanation for our findings may relate to the expression of c-Jun protein. c-Jun has been found to play a pivotal role in activating repair mechanisms facilitated by Schwann cells [59]. Following injury, the upregulation of c-Jun expression promotes Schwann cell trans-differentiation to a reparative phenotype. This in turn leads to formation of regenerative tracks, support of neuronal survivor and axonal regrowth [60]. Furthermore c-Jun regulates myelin clearance by stimulating macrophages to break down myelin [61] and thus improves Wallerian degeneration regulation resulting in superior regenerative potential of the peripheral nerve. Arthur-Farraj et al. demonstrated that in mice with c-Jun inactivation, there was a diminished ability of Schwann cells to support peripheral nerve degeneration with an associated persistence of myelin debris, an inability to maintain Büngner bands, resulting in increased neuronal death with functional recovery impairment [60]. Although the importance of c-Jun in relation to nerve regeneration is clear, the exact mechanisms remain unknown. Another mechanism thought to be linked to myelination capacity is related to macrophages activity. Macrophages play a key role in the modulation of the inflammatory response and is largely regulated by pro-inflammatory type 1 macrophages (M1) and anti-inflammatory type 2 macrophages (M2) [62]. Activation of both phenotypes are required for successful peripheral nerve regeneration, however, a prolonged period of M1 activation has been shown to restrict nerve growth [63]. In comparison, M2 polarization has been demonstrated to improve growth of axons [64]. Overall, the MRL/MpJ strain has approximately 30% fewer circulating macrophages in comparison to the C57BL/6J strain [65]. Interestingly, this may not be critical for successful wound healing. Davis et al. demonstrated that the depletion of macrophages from C57BL/6J or MRL/MpJ mice had no effect on wound healing [66]. Furthermore, MRL/MpJ mice have less IL-1α (pro-inflammatory) and more IL-4 and IL-10 (anti-inflammatory) at baseline yet following injury, C57BL/6 mice expressed decreased IL-4 and IL-10 whereas MRL/MpJ mice expressed increased levels [23].

The rationale for selecting the MRL/MpJ strain as an animal model was principally based on their diminished scar formation that has been reported in several tissue types [18–23]. Previous peripheral nerve studies have demonstrated that scar formation typically takes place in a region that is 2.5 mm proximal and distal to the site of injury, with the degree of scar formation correlating closely with severity of injury and degree of functional return [67,68]. In concordance with previously published studies, our results demonstrated that the MRL/MpJ strain had decreased scar formation within the regenerating nerve, inferred by a lower amount of ECM deposition. In our study we elected to choose two forms of nerve repair: direct repair and interpositional nerve grafting. Our rationale is that these injuries are most translatable to those seen in the clinical environment [69]. Many previous studies focus on crush injuries, however a neuropraxia injury will typically recover to a pre-morbid state [70]. Therefore, it may be more prudent to study nerve regeneration in a setting when irreversible functional loss is predicted as is the case in neurotmesis, and particularly when interpositional autologous nerve grafting is required [71,72]. In a previous study by Atkins et al., the impact of scarring on sciatic nerve regeneration was compared in two transgenic strains [12]. After quantifying collagen presence at the site of injury and direct repair, compound action potentials were assessed and myelinated nerve fibers on each side of repair were counted. An increased collagen level negatively correlated with number of axons regenerating across the injury site, the amplitude of compound action potentials, and slowed conduction velocities. In our study, we used axon to extracellular matrix ratio as a surrogate for scar presence. In accordance with findings presented by Atkins et al., we can report that the MRL/MpJ strain, which also demonstrated superior SFI results, exhibited a higher axon to extracellular matrix ratio. Although the amount of collagen in the extracellular matrix was not directly quantified, collagen remains to be the single major component of the ECM and therefore serves as an appropriate surrogate [73]. Despite above mentioned results, the role of superior histological outcomes seen in MRL/MpJ strain remains unclear in the context of no difference in functional nerve regeneration as demonstrated by SFI and EMG results. Therefore, the superior capacity of MRL/MpJ strain to myelinate their regenerating axons along with decreased scar formation does not appear to be clinically relevant in the process of peripheral nerve regeneration.

In summary, the findings of superior myelination and less scar formation within regenerated nerve in MRL/MpJ strain in both the DR and NG setting are supported by multiple traits that have been attributed to this transgenic strain. However, functional outcomes represented by SFI and electromyography did not demonstrate difference between investigated and control strain in peripheral nerve function following sciatic nerve transection injury and repair with either DR or NG. Therefore, a statement proposing that superior histomorphometry attributes observed in our study lead to superior peripheral nerve regeneration capacity cannot be made.

Limitations

The use of SFI as the function evaluation measure following nerve transection is the major limitation of the present study. Despite being considered one of the few validated functional outcome measures to study peripheral nerve regeneration in murine models and being utilized as the ‘gold standard’ to evaluate peripheral nerve function following injury by some authors, SFI is most commonly used to evaluate crush-type of nerve injury and its validity in transection-type of nerve injury has been questioned in the past [74–77]. Possibility of falsely long footprint caused by animal stopping in the corridor or putting excessive weight on hind paws while standing up is one of the limitations [78]. Moreover, in several mice individuals, we were not able to obtain clear print marks due to development of right hind limb contractures. Several other investigators noted similar findings [37,79]. Those concerns were emphasized by our first study phase results which revealed significantly superior outcomes in MRL/MpJ strain. As discussed earlier, there is no reason to expect such a difference as early as at POW1 time point and it is unlikely that those results reflect superior nerve regeneration capacity in investigated strain. Therefore, we proceeded with second phase of the study including electromyography which confirmed that there is no significant difference between investigated and control strain.

Conclusion

Our study did not demonstrate superior axonal regeneration in the MRL/MpJ strain compared with control strain. Our results revealed that there is no significant difference in functional outcomes between MRL/MpJ and C57BL/6J mouse strain based on assessment with SFI and electromyography. SFI was confirmed to be an outcome measure of major limitation following transection nerve injury. This was revealed by statistically significant difference between investigated and control strains on post-operative week 1, without possibility to reach target reinnervation and therefore affect function measured by SFI. Despite lower number of axons regenerating in MRL/MpJ strain, histological analysis demonstrated better myelination and less scar formation in investigated strain as revealed by superior G ratio and Axon/ECM ratio. However, the superior ability to myelinate regenerating axons and produce less scar tissue during the process of peripheral nerve regeneration did not appear to be translated into superior functional outcomes in our study and therefore superior histological outcomes does not appear to be clinically relevant in the process of peripheral nerve regeneration.

Mechanisms responsible for superior myelination and scar formation results in MRL/MpJ mouse strain remain unknown and further research to uncover underlying molecular pathways is warranted.

Summary points.

• This study evaluated the potential of MRL/MpJ mice strain as a novel animal model to investigate peripheral nerve regeneration mechanisms.

• Investigation of peripheral nerve regeneration in total of 172 mice was conducted.

• There were no differences in functional outcomes represented by Sciatic Function Index and electromyography between MRL/MpJ and C57BL/6J strain.

• Lower number of axons within regenerated peripheral nerve was noted in MRL/MpJ strain compared with controls.

• Improved axon myelination with superior G ratio was demonstrated in MRL/MpJ strain.

• MRL/MpJ strain revealed decreased ability for nerve scar formation following peripheral nerve injury and repair.

• Given no difference in functional outcomes, superior histological outcomes seen in MRL/MpJ strain and represented by G ratio and Axon/EMC ratio seem to have no clinical relevance on peripheral nerve regeneration in MRL/MpJ mouse strain.

• Based on histology results, MRL/MpJ strain provides an opportunity for further mechanistic investigation of decreased scar formation and superior re-myelination, however its clinical relevance on peripheral nerve regeneration in the investigated strain remains unclear.