Injury of Muscular but not Cutaneous Nerve Drives Acute Neuropathic Pain in Rats

Abstract

Acute pain is a common complication after injury of a peripheral nerve but the underlying mechanism is obscure. We established a model of acute neuropathic pain via pulling a pre-implanted suture loop to transect a peripheral nerve in awake rats. The tibial (both muscular and cutaneous), gastrocnemius–soleus (muscular only), and sural nerves (cutaneous only) were each transected. Transection of the tibial and gastrocnemius–soleus nerves, but not the sural nerve immediately evoked spontaneous pain and mechanical allodynia in the skin territories innervated by the adjacent intact nerves. Evans blue extravasation and cutaneous temperature of the intact skin territory were also significantly increased. In vivo electrophysiological recordings revealed that injury of a muscular nerve induced mechanical hypersensitivity and spontaneous activity in the nociceptive C-neurons in adjacent intact nerves. Our results indicate that injury of a muscular nerve, but not a cutaneous nerve, drives acute neuropathic pain.

Introduction

Acute neuropathic pain is a common complaint in clinical practice, but it is often underappreciated and undertreated in patients who experience trauma, surgery, and illness. In about one-half of patients, acute neuropathic pain develops into chronic neuropathic pain lasting for years, which greatly damages physical and mental health [1].

Growing evidence in animal models has demonstrated that it is injury of the muscular nerve, not the cutaneous nerve that matters in chronic neuropathic pain [2–4]. A significant proportion of patients who recover from poliovirus infection, which mainly targets muscular nerves and leads to progressive muscle weakness, suffer from chronic neuropathic pain, indicating that pathology of muscular nerves contributes to chronic pain [5, 6]. Besides, the sural nerve, a cutaneous branch of the sciatic nerve, is widely harvested in nerve repair surgery and most patients report no chronic pain after grafting [7].

However, it remains unclear whether muscular nerves also play a significant role in the generation of acute neuropathic pain. Some limitations have restricted the research on acute pain in animal models. To ensure that animals are awake and avoid the influence of anesthetics, pain-related behaviors are usually observed at least hours or days after surgery, leaving the acute neuropathic pain that occurs immediately after nerve injury rarely investigated. Previous studies have demonstrated that chronic neuropathic pain is associated with the release of inflammatory cytokines and the overexpression of sodium channels [8–15]. In fact, it usually takes hours or even days for significant neurophysiological changes to develop, implying that different mechanisms drive acute neuropathic pain. In a previous study, we applied a novel method to transect the L5 spinal nerve in awake rats to study acute neuropathic pain. Using this novel method, we observed the rapid onset of spontaneous pain and evoked pain hypersensitivity. The outburst of pain-related behavior was synchronous with the spontaneous activity and hyperexcitability of nociceptive neurons in the adjacent intact L4 dorsal root ganglion (DRG) [16].

In this study, we continued to apply this modified surgical method to transect peripheral nerves in awake rats to investigate the role of muscular nerves in acute neuropathic pain. The tibial (both muscular and cutaneous), gastrocnemius–soleus (GS) (muscular only), and sural (cutaneous only) nerves were each transected in awake rats by quickly pulling a pre-implanted suture loop surrounding the nerve. Then we measured the paw withdrawal threshold (PWT) to mechanical stimuli and the duration of spontaneous foot lifting, investigated the hypersensitivity of nociceptive C-neurons by in vivo electrophysiological recording, and evaluated neurogenic inflammation by Evans blue extravasation. This study was designed to help unravel the mechanisms underlying acute neuropathic pain.

Materials and Methods

Animals

Adult female Sprague-Dawley rats (180 g–200 g) provided by the National Institutes for Food and Drug Control, China, were housed under a 12-h light/12-h dark cycle at room temperature (23 °C ± 2 °C). All rats had free access to rodent chow and water. All experiments were performed in Peking Union Medical College, Beijing, China and were approved by the Institutional Animal Care and Use Committees of the Chinese Academy of Medical Sciences and the Institute of Basic Medical Sciences (Project #211-2014). A total of 137 rats were used: 33 for PWT measurement, 36 for spontaneous foot lifting measurement, 32 for in vivo electrophysiological recording, 21 for Evans blue extravasation, and 15 for cutaneous temperature measurement.

Surgery

The day of acute nerve transection was defined as day 0. On day − 2 (2 days before transection), after the behavioral test, rats were anesthetized with pentobarbital sodium (50 mg/kg intraperitoneally; Sigma-Aldrich Corp., St. Louis, MO). Under sterile conditions, a longitudinal incision was made in the lateral skin of the right thigh, then the biceps femoris muscle was bluntly separated to expose and isolate the sciatic nerve with its three branches: tibial, common peroneal, and sural. The GS is a branch of the tibial nerve that innervates the gastrocnemius and soleus muscles. A suture loop was placed around the target branch with its two ends running through a 2 mm-long sterilized transparent tube. A knot in the distal end of the tube was made to prevent the tube from slipping out. The tube was embedded under the skin and then the incision was closed in layers. To assess the influence of the surgery itself, the behavioral test was repeated 1 day before transection (day − 1). On day 0, the target nerve was transected in awake rats by quickly pulling the pre-implanted suture loop while holding the transparent tube steady (Fig. 1A). When the target nerve was completely transected by the suture loop, the intact loop was pulled out. After all experiments, we checked the target nerve to confirm it was completely transected.

Fig. 1.

Nerve branches and plantar territories. A Diagram of the sciatic nerve and its branches: tibial (both muscular and cutaneous), gastrocnemius-soleus (GS; muscular only) and sural (cutaneous only). In nerve transection, the transparent tube was held steady and the suture loop was pulled rapidly. B The plantar hind paw is divided into three territories: lateral, middle, and medial, innervated by the sural, tibial, and saphenous nerves, respectively. Paw withdrawal thresholds were measured at sites 1, 2, and 3 (dashed circles) in the corresponding territories.

Behavioral Test

Paw Withdrawal Threshold

Before all tests, rats were placed in an acrylic glass box to habituate to the environment and observed for 1 h/day for 3 days. Before each test, habituation was allowed for 10 min.

The plantar surface of the rat hind paw can be divided into three territories: medial (saphenous nerve receptive field), middle (tibial nerve receptive field), and lateral (sural nerve receptive field), with overlaps between adjacent territories [17]. A calibrated electronic von Frey filament (Electronic von Frey 2390-5 Anesthesiometer; IITC Life Science, Woodland Hills, CA) was applied perpendicular to each of the sites 1, 2, and 3 on the plantar surface, and held for ~ 3 s (Fig. 1B). The PWTs of sites 1, 2, and 3 represented those of the lateral, middle, and medial territories, respectively. Sharp paw withdrawal with licking or shaking was regarded as a positive response. The average of three repeated measurements was taken as the final PWT. The average PWTs for days − 4, − 3, and − 2 were defined as the baseline.

Spontaneous Pain Duration

The habituation process in spontaneous pain assessment was the same as for the PWT test. Behaviors such as spontaneous foot lifting, biting, and licking were regarded as indicators of spontaneous pain [18]. To observe the behaviors clearly, a mirror was placed below the glass box at an angle of 45° to the experimental platform. A high-definition camera was placed in front of the box to capture direct images and indirect inverted images in the mirror. The cumulative duration (in seconds) of spontaneous foot lifting, biting, and licking the right hind paw per 30 min was recorded as the quantitative assessment of spontaneous pain. The average of spontaneous pain duration on days − 3 and − 2 (before implanting the loop) was defined as the baseline.

Electrophysiological Recording

L4 Dorsal Root Ganglion Exposure

L4 DRG exposure and in vivo electrophysiological recording were as previously reported [19]. Briefly, rats were anesthetized with pentobarbital sodium (initial dose of 50 mg/kg intraperitoneally followed by supplementary doses of 20 mg/kg as needed), the L5 transverse process was removed, and a laminectomy was made from L1 to L6. Oxygenated artificial cerebrospinal fluid at 35 °C was dripped onto the surface of the L4 DRG during surgery and recording. Under a dissecting microscope, the perineurium and epineurium of the L4 DRG were carefully removed, then the rat was transferred to the recording platform. A pool was made by attaching the skin to a metal ring.

Identification of C-neurons and Their Receptive Fields

The receptive field of a neuron was identified by applying peripheral stimuli. Axonal conduction velocity was determined by dividing the distance between the receptive field and the cell body by the latency of the action potential. A C-neuron was identified by a velocity < 2 m/s, then mechanical (50 mN von Frey filament for 3 s), warm (51 °C fine tube for 5 s), and cold stimuli (0 °C ice-water for 10 s) were applied to identify the subtype of the C-neuron.

Mechanical stimuli were applied to the C-neuron receptive field using a Q-tip, light brush, and von Frey filaments in ascending order (5 mN, 10 mN, 30 mN, and 50 mN). In vivo recording started 20 min before nerve transection and lasted for 2 h. If action potentials were recorded in C-neurons when mechanical stimuli were delivered, the stimuli were labeled as positive.

Criteria for Defining a Spontaneously Active C-neuron

For an identified C-neuron, a continuous recording started 5 min before transection and lasted for 4 h without any external stimulation. If an identified C-neuron discharged continuously during this period, it was classified as spontaneously active. Any discharge associated with electrode insertion and lasting < 30 s was classified as “injury discharge” and terminated as a failure.

Compensatory A-fiber Input

After identifying a spontaneously active C-neuron, stimulation mimicking A-fiber strength (10 Hz, 0.5 mA for 3 min) was delivered to the proximal segment of the transected nerve through a pair of silver hook electrodes.

Evans Blue Extravasation

Evans blue (1%, 2 mL in saline, intravenously; Sigma-Aldrich, St. Louis, MO) was injected into the caudal vein 2 min before nerve transection. Transcardial perfusion with 0.1 mol/L phosphate-buffered saline was carried out 30 min after transection. After perfusion, photographs of the dorsal paws were taken, and then the dorsal skin of the ipsilateral and contralateral paws was removed. Extravasated dye was measured using a previously reported method [20]. Briefly, the removed skin was incubated overnight with N, N-dimethylformamide at 55 °C to completely dissolve the Evans blue into the solvent. Then, the Evans blue absorbance value was measured by a microplate spectrophotometer at 630 nm. The concentration of Evans blue was normalized to a reference sample (0 μg/mL–9.6 μg/mL).

Cutaneous Temperature Measurement

Rats were placed prone with an electrical temperature sensor in the rectum to measure the rectal temperature. Another two electrical temperature sensors were closely attached to the skin surface on different plantar territories of the hind paw with adhesive tape. The nerves were exposed as described above and were transected using fine scissors. Temperature recording started 5 min before transection and ended 30 min later.

Statistical Analysis

All data are presented as the mean ± SEM. Differences of PWT at different time points were analyzed using one-way analysis of variance (ANOVA) followed by the Bonferroni post hoc test. Student’s t test was applied to compare the mechanical threshold of C-neuron discharge and Evans blue extravasation between different groups. The χ² test was applied to compare the incidence of spontaneous discharge between different groups. Two-way (time and group) ANOVA with the Bonferroni post hoc test was applied to compare repeated measurements of cutaneous temperature at different time points among different groups. A statistically significant difference was defined as a two-sided P value < 0.05. IBM SPSS Statistics for Windows (version 21.0, Armonk, NY) was used for analysis.

Results

Animal Behavior

When rats were awake on the day after implanting the tube (day − 1), they behaved almost normally, except for occasionally licking at the wound. On day 0, after acute transection of the tibial nerve and GS nerve, the rats immediately became restless, walked around clumsily and limped on the injured right hind paw with licking and biting. Hours later, there was a tendency to recover with a decreased licking and biting frequency. In contrast, rats behaved normally after sural nerve transection and sham transection.

Mechanical Allodynia of the Lateral Plantar Paw (Sural Nerve Receptive Field)

Compared with baseline, the PWTs of the lateral territory showed no significant changes on day − 1 in the four groups (P > 0.05). After tibial nerve transection, the PWT of the lateral territory decreased dramatically, from baseline (13.6 g ± 0.4 g) to the range of 1.2 g ± 0.3 g to 2.4 g ± 0.9 g (P < 0.05, n = 9). After GS nerve transection, a moderate decline in the PWT of the lateral territory occurred, from baseline (14.1 g ± 0.2 g) to the range of 4.5 g ± 0.8 g to 6.8 g ± 0.7 g (P < 0.01, n = 10). After sural nerve transection, only the PWTs at 5 h, 3 days, and 7 days slightly but significantly decreased (P < 0.05, n = 8). In the sham group, the PWTs in the lateral territory were between 12.9 g ± 0.5 g and 14.3 g ± 0.8 g with significant differences at 5 h and 7 days (P < 0.05, n = 6) (Fig. 2A).

Fig. 2.

Paw withdrawal thresholds (PWTs) in each territory and spontaneous pain duration. A PWTs of the lateral territory. In the tibial and GS transection groups, PWTs significantly decreased immediately after transection and lasted for 14 days. In the sural and sham groups, PWTs showed no significant change at most time points. B PWTs of the medial territory. PWTs decreased significantly after GS nerve transection. In the tibial, sural, and sham groups, the PWTs at most time points showed no significant change. C PWTs of the middle territory. PWTs increased significantly after tibial nerve transection and dropped after GS nerve transection with a potential recovery tendency after 1 day. D Spontaneous pain duration. After tibial nerve transection and GS nerve transection, the spontaneous pain duration extended significantly, and a recovery tendency occurred over time. In the sural and sham groups, spontaneous pain duration did not significantly change at most time points [time points vs baseline, *P < 0.05, **P < 0.01 (tibial group); ^#P < 0.05, ^##P < 0.01 (GS group); ^&P < 0.05, ^&&P < 0.01 (sural group); ^{^}P < 0.05, ^{^^}P < 0.01 (sham group); arrowheads, implantation of suture loop and tube without nerve transection on day − 2; arrows, nerve transection].

Mechanical Allodynia of the Medial Plantar Paw (Saphenous Nerve Receptive Field)

Compared with baseline, the PWTs of the medial territory on day − 1 showed no significant differences in the four groups. After tibial nerve transection, the PWT of the medial territory was significantly lower than baseline at 5 h post-transection with a threshold of 8.6 g ± 0.8 g (P < 0.05, n = 9). After GS nerve transection, the PWTs of the medial territory dropped from baseline (12.8 g ± 0.3 g) to the range of 5.8 g ± 0.6 g to 8.9 g ± 0.5 g (P < 0.01, n = 10). After sural nerve transection, the PWTs at most time points were not significantly changed. Only the PWT on day 3 was significantly decreased to 10.2 g ± 0.5 g (P < 0.01, n = 8). In the sham group, there were no significant changes in the PWTs of the medial territory (P > 0.05, n = 6) (Fig. 2B).

Mechanical Allodynia of the Middle Plantar Paw (Tibial Nerve Receptive Field)

Compared with baseline, the PWTs of the middle territory were not significantly changed on day − 1 in all groups. After tibial nerve transection, rats lost sensation to mechanical stimuli applied to the middle territory. Because of paralysis and numbness, the rats did not respond to mechanical stimuli even if the force of the von Frey filament was up to 20 g. Paws were elevated passively by the von Frey filament with PWTs between 18.6 g ± 1.4 g and 23.1 g ± 0.4 g (P < 0.05, n = 9). After GS transection, the PWTs of the middle territory declined significantly from baseline 14.8 g ± 0.3 g to the range of 6.9 g ± 0.9 g to 12.8 g ± 1.5 g (P < 0.05, n = 10). In the sural group, the PWTs of the middle territory at different time points were not significantly changed (P > 0.05, n = 8). In the sham group, the PWTs of the middle territory remained between 13.4 g ± 0.8 g and 15.7 g ± 0.6 g with only that on post-transection day 7 higher than baseline (P < 0.05, n = 6) (Fig. 2C).

Extended Spontaneous Pain Duration After Muscular Nerve Transection

In all four groups, no spontaneous pain appeared on the day after implanting the tube (day − 1), and there was no significant difference in spontaneous pain duration between baseline and day − 1. Rats exhibited pain behaviors and high facilitation of spontaneous pain immediately after tibial nerve transection (P < 0.05, except for P = 0.05 on day 5, n = 10). The most frequent spontaneous pain occurred immediately after tibial nerve transection, with an extremely long duration of 319.9 s ± 51.1 s (P < 0.01). Then, rats tended to recover and spontaneous pain duration was attenuated to 39.6 s ± 14.4 s on post-transection day 7. After GS nerve transection, spontaneous pain duration was moderately extended (P < 0.05, except for P > 0.05 on day 7, n = 9), and the longest duration (89.1 s ± 19.6 s) occurred immediately after transection with a tendency to decline afterwards. Compared with baseline (1.1 s ± 1.1 s), spontaneous pain duration slightly increased to 10.0 s ± 5.0 s immediately after sural nerve transection (P < 0.05 at 0 h, n = 9), but no significant spontaneous pain behaviors were observed thereafter (P > 0.05). No spontaneous pain behaviors were observed in the sham group (P > 0.05, n = 8) (Fig. 2D).

Sensitization of Nociceptive C-neurons After Muscular Nerve Transection

We made in vivo extracellular recordings from C-neuron cell bodies in L4 DRGs after nerve transection, noting the conduction velocity and responses to mechanical, warm, and cold stimuli of 22 nociceptive C-neurons (Fig. 3A–G). The receptive fields of nociceptive C-neurons after tibial nerve transection were mainly in the adjacent lateral (sural nerve receptive field) and medial (saphenous nerve receptive field) territories (Fig. 3F). The mechanical threshold to evoke C-neuron discharge significantly decreased within 5 min after tibial nerve transection (post-transection 5.6 mN ± 2.3 mN vs pre-transection 31.4 mN ± 5.9 mN, P < 0.01, n = 6) and GS nerve transection (post-transection 11.6 mN ± 2.9 mN vs pre-transection 32.7 mN ± 3.7 mN, P < 0.01, n = 5). The mechanical threshold did not significantly change after sural nerve transection (post-transection 28.5 mN ± 4.9 mN vs pre-transection 30.8 mN ± 5.1 mN, P > 0.05, n = 5) (Fig. 3H).

Fig. 3.

In vivo responses of nociceptive C-neurons to mechanical stimuli in the tibial, GS, and sural nerve transection groups. A Bright-field image of the L4 DRG surface showing a neuron cell body (arrow) and an extracellular recording electrode (dashed lines) (scale bar, 50 μm). B Conduction velocity (CV) was measured by electrically stimulating the peripheral receptive field on the dorsal surface of the paw (black dot) (bar, 100 ms). C Typical response of the C-neuron to 50 mN mechanical stimulation. Action potentials (APs) in the original recording trace (Ie) are presented as corresponding tic marks below. D Representative response of the C-neuron to nociceptive thermal stimulation (51 °C for 5 s). E Representative response of the C-neuron to cold stimulation (0 °C for 10 s). F Receptive fields of nociceptive C-neurons after tibial (△), GS (×), sural (□), and sham (○) transection. G Action potentials (vertical lines) evoked by different mechanical stimuli—Q-tip (Q), light brush (Br), and von-Frey filaments with 5 mN, 10 mN, 30 mN, and 50 mN bending forces—immediately after nerve transection. H Mechanical thresholds to evoke the discharge of nociceptive C-neurons (*P < 0.05, **P < 0.01, post-transection vs pre-transection).

Increased Spontaneous Activity of Nociceptive C-neurons After Muscular Nerve Transection

The spontaneous activity of L4 DRG C-neurons became evident ~ 3 min after tibial or GS nerve transection (Fig. 4A–C). The percentage of C-neurons with spontaneous activity was significantly higher after tibial (37%, 21/57, P < 0.01) and GS transection (18%, 7/38, P < 0.01) than in the sham group (3%, 1/38) (Fig. 4D). Spontaneous activity was rarely recorded in C-neurons after sural nerve transection (1%, 1/80, P > 0.05). Spontaneous activity was suppressed when continuous A-fiber-strength electrical stimulation (10 Hz, 0.5 mA for 3 min) was applied to the proximal tibial or GS nerve (Fig. 4B, C).

Fig. 4.

Rapid-onset of spontaneous C-neuron discharges in the L4 DRG following muscular nerve transection. A Schematic of the recording setup. B, C An initially quiescent C-nociceptive neuron showed spontaneous activity within 3 min after tibial and GS nerve transections, with inhibitory modulation by A-fiber strength stimuli (horizontal bar, electrical stimulation at 10 Hz, 0.5 mA for 3 min; arrowheads, mechanical stimulation in the receptive field). D Percentage of C-neurons exhibiting spontaneous discharges after nerve transection. The number of neurons exhibiting spontaneous discharge/total neurons recorded is indicated above each column (*P < 0.05, **P < 0.01, tibial, GS, and sural groups vs sham).

Increased Evans Blue Extravasation After Muscular Nerve Transection

Evans blue extravasation was evident in the ipsilateral affected skin ~ 10 min after tibial (Fig. 5A) and GS (Fig. 5B) transection. Compared with 6.25 μg/g ± 1.02 μg/g in the contralateral intact skin, the Evans blue concentration was significantly increased to 21.24 μg/g ± 3.26 μg/g in the ipsilateral affected skin after tibial nerve transection (P < 0.01, n = 7). After GS nerve transection, the Evans blue concentration in the ipsilateral affected skin was significantly higher than that the contralateral intact skin (ipsilateral 14.34 μg/g ± 2.08 μg/g vs contralateral 6.12 μg/g ± 1.35 μg/g, P < 0.01, n = 7) (Fig. 5D). After sural nerve transection, there was no significant difference in Evans blue concentration between the ipsilateral affected and contralateral intact skin (P > 0.05, n = 7) (Fig. 5C, D).

Fig. 5.

Evans blue extravasation and cutaneous temperature after nerve transection. A Evans blue clearly extravasated in the affected skin 30 min after tibial nerve transection (left), and the temperature of the lateral territory of the plantar hind paw increased significantly after the transection (right) (*P < 0.05, **P < 0.01, lateral vs middle territory). B Evans blue extravasated significantly after GS nerve transection (left), but with no significant change in the temperature of the ipsilateral paw (right) (*P < 0.05, **P < 0.01, ipsilateral middle vs contralateral middle territory). C No significant Evans blue extravasation (left) and cutaneous temperature change (right) were found following sural nerve transection (*P < 0.05, **P < 0.01, lateral vs middle territory). D Concentration of Evans blue dye in the dorsal skin after tibial, GS, and sural nerve transections (*P < 0.05, **P < 0.01, ipsilateral vs contralateral skin).

Increased Cutaneous Temperature Following Muscular Nerve Transection

In the tibial and sural nerve transection groups, the cutaneous temperature of the lateral and middle territories of the ipsilateral plantar paw was simultaneously recorded. Ten minutes after tibial transection, the temperature of the lateral territory was significantly increased (P < 0.05, n = 5), with no significant change in the middle territory (Fig. 5A). The temperature of the lateral and middle territories was not significantly changed after sural nerve transection (P > 0.05, n = 5) (Fig. 5C).

In the GS transection group, the cutaneous temperature of the middle territory on the ipsilateral and contralateral paws was simultaneously measured. Transection of the GS nerve did not induce a significant temperature change (P > 0.05, n = 5) (Fig. 5B).

Discussion

In clinical practice, acute neuropathic pain frequently occurs after trauma and surgery, and may progress to chronic pain. A pathological muscular nerve is critical in the development of chronic neuropathic pain in rodent models and human subjects [2–7]. Due to the impact of anesthetics, it is difficult to study acute pain in rodent models and it remains unclear whether muscular nerves also play a significant role in acute neuropathic pain. In this study, we successfully applied a modified surgical method to investigate acute neuropathic pain by transecting the nerve with a pre-implanted suture loop, and found that the transection of a muscular nerve, but not a cutaneous nerve, led to the rapid onset of mechanical allodynia of the adjacent skin territory supplied by intact nerves. Spontaneous pain and mechanical allodynia were accompanied by spontaneous discharges and hypersensitivity of nociceptive C-neurons. These findings have important clinical implications. To relieve acute neuropathic pain and prevent chronic pain, muscular nerves should be carefully protected during surgery. Muscular nerves may be a potential target in the management of acute neuropathic pain.

The inflammatory cytokine TNF-α is known to play a key role in the initiation of nerve injury-induced neuropathic pain, but this occurs at least several hours after the injury [13–15]. It also takes hours to days for the aberrant expression of voltage-gated sodium channels to dominate ectopic activity [8–10]. Evidently, previous evidence on inflammatory cytokines, ion channels, and Wallerian degeneration may not provide an appropriate explanation for this rapid onset of pain following muscular nerve transection. Chronic neuropathic pain is common in patients with Guillain-Barré syndrome and diabetic neuropathy, diseases leading to peripheral demyelination in myelinated A-fibers [21, 22]. In a rat model of acute demyelination of the sciatic nerve by cobra venom, C-fibers discharge spontaneously several minutes after selectively blocking the A-fiber afferents [23]. These findings suggest that the loss of tonic A-fiber inputs may lead to the activation of nociceptive C-fibers and the generation of neuropathic pain. Besides, electrical stimulation of deep tissue Aβ afferent fibers, but not cutaneous Aβ afferents, relieves the hyperalgesia evoked by knee joint inflammation in rats [24]. Deep acupuncture also has a better short-term anti-hyperalgesic effect than superficial acupuncture in patients with lateral epicondylalgia [25], suggesting that external stimulation of A-fiber muscular afferents may compensate for the loss of tonic A-fiber inputs and relieve neuropathic pain.

We hypothesize that the sensitization of adjacent intact sensory neurons occurs through a spinal mechanism involving gate-control and the dorsal root reflex, so that the absence of deep tissue A-fiber inputs following muscular nerve transection leads to acute neuropathic pain. According to gate-control theory, there is an interactive balance between myelinated large-diameter afferents and unmyelinated small-diameter afferents, and the large-diameter fibers inhibit small-diameter fibers in the inhibitory substantia gelatinosa [26]. Low-threshold mechanoreceptive primary afferent A-fibers indirectly inhibit neighboring nociceptive C-fibers via inhibitory gamma-aminobutyric acid (GABA)-ergic and glycinergic interneurons [27–29]. The inhibitory neurotransmitters GABA and glycine are released onto the central terminals of C-fibers, resulting in their hyperpolarization and the inhibition of C-fiber discharges. When the tonic discharges from deep tissue (muscular) A-fiber inputs are blocked, such as by transection or demyelination, the disinhibited C-fiber central terminals may depolarize and discharge antidromically to the peripheral terminals. The antidromic discharge of polymodal C-fibers in the intact nerves may therefore induce neurogenic inflammation by releasing substance P and calcitonin gene-related peptide from the nerve terminals [30]. Neurogenic inflammation, in reverse, may further sensitize peripheral terminals of the adjacent C-fibers and evoke orthodromic discharges, evoking spontaneous pain and cutaneous sensitivity. In this study, when the muscular nerve was transected, C-fibers in the nerve were also blocked. Only the C-fibers in the adjacent intact nerves could be activated and induce neurogenic inflammation in the receptive fields via the mechanisms proposed above, leading to spontaneous pain, mechanical allodynia, and a temperature increase in the hindpaw skin territories. This hypothesis can be further explored using spinal electrophysiological recordings. It is also possible that the neighboring nerve fibers might be sensitized by local inflammation from the transected nerve. However, since the spontaneous pain and mechanical allodynia were evoked almost immediately (within minutes) after nerve transection, it is unlikely that local inflammation plays a major role in this process.

In conclusion, injury of muscular nerves, but not cutaneous nerves, drives acute neuropathic pain. Neurons innervating the muscle may be a potential target in the management of acute neuropathic pain.

Conflict of interest

The authors claim that there are no conflicts of interest.