The Fibular Nerve Injury Method: A Reliable Assay to Identify and Test Factors That Repair Neuromuscular Junctions

The common fibular nerve, also called the common peroneal nerve, arises from the sciatic nerve above the popliteal fossa, where it swings around the head of the fibula to the anterior aspect of the leg (Figure 1A). There it branches into the superficial and deep fibular nerves, together supplying the dorsiflexors of the foot and toes (anterior tibialis, extensor digitorum longus and brevis, and extensor halluces longus muscles), and the everters of the foot (peroneus muscles). This nerve also carries sensory fibers that project to the dorsum of the foot and lateral aspect of the lower half of the leg. It is a relatively thin structure composed of motor and sensory axons. This nerve branch follows a predictable anatomic course. Lateral to the knee, the nerve is mostly superficial as it runs over the tendon of the lateral head of the gastrocnemius muscle (Figure 1 and Figure 2). This location serves as a stable landmark that can be easily reached with a small incision, limiting damage to skin and fascia (Figure 1A). The smaller diameter of this nerve, when compared to the sciatic and tibial nerves, makes it possible to crush all axons using less force, reducing the likelihood of completely severing the more superficial axons.

Mice expressing yellow fluorescence protein (YFP) only in neurons¹⁷ were used to optimize the crush procedure on the common fibular nerve and unambiguously visualize regenerating axons. The nerve was crushed at the edge of the gastrocnemius tendon most proximal to the target muscles because this site is more accessible from the location of the incision. This site also serves as a reliable anatomical landmark, making it possible to compare regeneration of NMJs among animals of the same age and sex. Compressing the nerve for 5 sec using a fine forceps results in the disappearance of YFP from the site of injury (Figure 2B). However, the connective tissue and cells residing in the epineurium remain contiguous, serving as a conduit for rapid and precise regeneration of axons to their original targets (Figure 2B). In 70 days old female mice, this injury is sufficient to cause degeneration of all axonal segments distal from the neuronal soma (Figure 4B).

To determine the reliability and reproducibility of this injury method, reinnervation of the extensor digitorum longus (EDL) muscle was examined. This muscle was chosen for several reasons: 1) It is proximal yet physically separated from the incision and nerve injury sites (Figure 3A). Hence, the muscle is only altered by degeneration of severed innervating axons. Its proximity to the crush site minimizes the time needed for it to be reinnervated and muscle atrophy. 2) It is primarily composed of fast type skeletal muscle fibers, which are more susceptible to aging and diseases. 3) Its NMJs undergo significant structural changes during the progression of diseases and aging that can be attenuated by exercise and caloric restriction. 4) It is readily accessible for live imaging and molecular manipulation (Figure 3D). 5) It can be easily separated into its four phalangeal components that can be whole-mounted making it possible to fully image all of its innervating axons and their connections using light microscopy (Figure 3B).

Reinnervation of previously vacated post-synaptic sites after crushing the fibular nerve on the right leg was assessed in three 70 days old female mice expressing YFP in motor axons. Post-synaptic sites were visualized using fluorescently tagged alpha-bungarotoxin (BTX), which binds selectively and with high affinity to muscles nAChRs. Muscles were deemed to be denervated, partially or fully reinnervated following these criteria: 1) In denervated muscle fibers, motor axons were completely missing from post-synaptic sites and less than 5% colocalization between the axon and AChRs was observed. 2) Partially innervated muscle fibers were categorized by some but incomplete apposition of motor axons with post-synaptic sites and 5-95% colocalization between the axon and AChRs was observed. 3) In fully innervated muscle fibers, there was nearly perfect apposition between motor nerve endings and post-synaptic sites and greater than 95% colocalization between the axon and AChRs was observed. Individual NMJs were excluded from counts if their endplates lay perpendicular to the imaging plane or the whole NMJ was not visualized. Using these criteria, high concordance in the rate and degree of reinnervation was observed among all mice examined. At 4 days post-crush, muscles were found completely denervated in all animals examined (Figure 4B). This finding shows that crushing the common fibular nerve for 5 sec, as described above, is sufficient to sever all axons. By 7 days post-crush, nerve endings were in the process of reoccupying previously vacated post-synaptic sites (Figure 4C, 4E-F). However, most muscle fibers were still found only partially innervated. With additional days post-crush, nerve endings continued to differentiate into presynaptic sites and NMJs were found fully reinnervated by 12 days (Figure 4D, 4E-F). Importantly, there was little variability among mice denervated for the same length of time (Figure 4E-F), demonstrating that crushing the fibular nerve can be used as an assay to compare reinnervation of muscles between animals of the same age and sex.

The ability to faithfully compare regenerating NMJs between animals provides opportunities to understand the cellular features associated with each step required to fully repair this synapse. To examine the architecture of denervated and regenerating NMJs from the 70 days old mice used above for comparing rates of reinnervation, high-resolution confocal microscopy images of NMJs were obtained. As expected, this analysis revealed a number of changes that occur in α-motor axon nerve endings as they reinnervate muscle fibers (Figure 4G-I). Axonal growth cones appear to expand and begin to branch out as they contact postsynaptic sites (Figure 4G). These axonal branches then grow over specific regions of the postsynapse, culminating in the near complete juxtaposition of the axon nerve ending with the postsynaptic site (Figure 4H-I). Additional structural features in regenerating motor nerve endings that highly resemble those found during development were observed, including multiple axons competing for the same target (Figure 4H), culminating in only one axon innervating one muscle fiber by 12 days post-crush. Additionally, axonal branches extending beyond post-synaptic sites, referred herein as sprouts, at all stages post-injury were observed (Figure 4I). These sprouts were highly prevalent, even in fully innervated NMJs suggesting that the final phase of axonal repair involves retraction of extrajunctional axonal branches. Despite these obvious changes at nerve endings, post-synaptic sites remained mostly indistinguishable at all stages post-injury compared to those in uninjured muscles, including at 4 days post-crush when muscle fibers are found completely denervated. There was no obvious sign of fragmentation or significant loss and redistribution to extra-synaptic regions of AChRs. These findings strongly indicate that the fibular nerve crush method can be used as an assay to identify and test molecular, pharmacological and life-style interventions that promote repair of nerve endings at synapses, including the NMJ.

Despite recent advances, very little progress has been made in identifying muscle-derived factors required and sufficient to maintain and repair the NMJ. We therefore asked if the common fibular nerve injury method can be used to identify molecules altered at specific stages of the reinnervation process and with potential roles in repairing the NMJ. As a proof of principal, expression analysis of two NMJ-associated genes, the AChR gamma subunit and the muscle-specific kinase, MuSK¹⁸ was performed. As demonstrated with quantitative PCR, these genes are increased following denervation and decreased as NMJs are reinnervated (Figure 5a-b). These genes are upregulated at 4 days post-crush, as previously reported using completely denervated muscles¹⁹. As nerves reinnervate muscle fibers, levels of these genes decrease and rapidly return to baseline. Among animals examined, the pattern of expression for both genes is nearly identical, demonstrating a close correlation between molecular and cellular changes at regenerating NMJs. This method therefore provides unique opportunities to identify molecular changes associated with different stages of NMJ regeneration.

Figure 1: Common Fibular Nerve Anatomy. (A) Schematic detailing the course of the sciatic nerve and its terminal branches in relation to superficial and bony landmarks. ScN = Sciatic Nerve, TN = Tibial Nerve, SrN = Sural Nerve, CPN = Common Fibular Nerve, SPN = Superficial Fibular Nerve, DPN = Deep Fibular Nerve. The desired incision site is indicated. (B) Diagram of common fibular nerve location in relation to surrounding muscles with the skin removed. Relative incision site is shown overlying the investing fascia. TA = Tibialis Anterior Muscle EDL = Extensor Digitorum Longus Muscle. (C) Exposed sciatic and common fibular nerve. CPN crosses over the gastrocnemius muscle (Gn) tendon at the level of the knee. Crush site relative to tendon shown. Please click here to view a larger version of this figure.

Figure 2: Common Fibular Nerve Crush Surgery. (A) The initial incision need only be a few cm in length. After penetrating through the skin and superficial fascia, the incision must be continued through the investing/deep fascia lying between the Biceps Femoris and TA muscles. (A1) The common fibular nerve is easily visualized without microscopy through the incision (Gn = Gastrocnemius Muscle). (A2) The intersection of the common fibular nerve and gastrocnemius tendon can be easily visualized if the incision is widened (necessary only for photographic purposes). The crush should be placed at the location marked by the red line perpendicular to the nerve and parallel to the Gn tendon. (B) The CPN is even more easily visualized with YFP transgenic mice under a fluorescent dissecting scope. (B1) The nerve loses fluorescence at the site of crush, allowing for confirmation of a full crush injury. (B2) A full CPN crush can still be visualized with the naked eye and room lighting. The nerve will become translucent. This image highlights the fact that the epineurium and superstructure of the CPN remain intact while simultaneously limiting damage to nearby tissue and blood vessels. Please click here to view a larger version of this figure.

Figure 3: Anatomy of the Extensor Digitorum Longus (EDL) Muscle. (A) 3 dimensional image of the EDL in relation to the bones of the mouse hind limb. Generated using JAtlasView (B) Partially dissected EDL muscle separated into its four phalangeal divisions and marking the digits controlled by each division. (C) YFP labeled branch of the deep fibular nerve as it innervates the end-plate band of the EDL division controlling the second digit. (D) Axonal branches and their contact sites can be readily identified, allowing for consistent examination of selected NMJs. Please click here to view a larger version of this figure.

Figure 4: Similar Rates of Reinnervation Between Animals of the Same Age and Sex. Analysis of NMJs in the EDL muscle after crushing the fibular nerve. (A) Post-synaptic sites are completely occupied by axons in uninjured mice. (B) At 4 days post-crush, muscles are found completely denervated. (C) Axons begin to reinnervate muscles 7 days post-crush and (D) completely reoccupy postsynaptic sites by 12 days. (E–F) The rate of NMJ re-occupancy is nearly indistinguishable between animals denervated for the same length of time. (G–I) Representative images of re-differentiating axonal nerve endings into matured presynaptic sites. Reoccupation of postsynaptic sites follows a predictable pattern, starting with the growth cone developing branches that eventually fully occupy postsynaptic sites without extending beyond the NMJ region.Similar to development, postsynaptic sites are initially innervated by multiple re-growing axons but only one axon remains once the NMJ has completely regenerated. (H) Examples of exuberant axonal sprouting, partial innervation without full overlap between pre and postsynaptic apparatuses, and innervation by multiple axons indicated. 70 day-old female mice were examined; Scale Bar = 50 µm (A-D), 20 µm (G-I). Error bar = SEM. N = 3 mice. Please click here to view a larger version of this figure.

Figure 5: The Fibular Nerve Crush Method as an Assay to Identify Candidate Genes Involved in Repairing the NMJ. (A–B) The mRNA level of two NMJ-associated genes, the AChR gamma subunit and MuSK, are similarly altered in the TA muscle of mice denervated for the same length of time. With increasing time post nerve injury, levels of both transcripts decrease, returning to baseline, corroborating histological analysis showing progressive reinnervation of NMJs. Each line represents an individual mouse. Error bar = standard error of the mean of technical replicates. Please click here to view a larger version of this figure.