Peripheral contributions to the mechanical hyperalgesia following a lumbar 5 spinal nerve lesion in rats

Neuroscience 165 (2010) 221–232

PERIPHERAL CONTRIBUTIONS TO THE MECHANICAL HYPERALGESIA FOLLOWING A LUMBAR 5 SPINAL NERVE LESION IN RATS J. H. JANG,a1,2 B. H. LEE,a,b,c2 T. S. NAM,a,b,c J. W. KIM,a D. W. KIMa,c* AND J. W. LEEMa,b,c*

PID block or by L5 nerve stimulation. These results suggest that neighboring C-afferents remaining intact after partial nerve injury play a critical role in the development of mechanical hyperalgesia through interaction with degenerating afferents, and also via peripheral sensitization by PID. © 2010 IBRO. Published by Elsevier Ltd. All rights reserved.

a Department of Physiology, Yonsei University College of Medicine, Seoul 120-752, South Korea b Brain Research Institute, Yonsei University College of Medicine, Seoul 120-752, South Korea c Brain Korea 21 Project for Medical Science, Yonsei University College of Medicine, Seoul 120-752, South Korea

Key words: nerve injury, antidromic stimulation, mechanical hyperalgesia, neuropathic pain, injury discharge, Wallerian degeneration.

Abstract—Using lumbar 5 (L5) dorsal root rhizotomy-bearing rats, we examined the extent to which L5 spinal nerve lesion (SNL)-induced mechanical hyperalgesia was governed by two peripheral components, that is Wallerian degeneration (WD) and peripherally-propagating injury discharge (PID). The contribution of WD to SNL-induced hyperalgesia was studied by excluding PID with lidocaine treatment that blocked nerve conduction temporarily, but completely at the time of injury, whereas PID was examined separately by using brief tetanic electrical stimulation of the spinal nerve mimicking PID. Following the disappearance of L5 rhizotomyinduced transient hyperalgesia, L5 SNL resulted in long-lasting mechanical hyperalgesia as early as one day post-SNL despite a PID block, highlighting the role of WD. In a comparative experiment, a delayed onset of hyperalgesia (7 days) was measured in L3 rhizotomy-bearing rats following L3 SNL with a PID block, in which injured fiber (L3) was separated from intact fibers (L4 and L5) anatomically until they meet at the peripheral terminals, supporting the importance of interactions between degenerating and adjacent intact fibers for WD-induced hyperalgesia. Tetanic electrical stimulation of decentralized L5 spinal nerve resulted in mechanical hyperalgesia developing within 1 day and persisting for 7 days. This hyperalgesia was prevented by lidocaine blockade of the L5 nerve, and was unaffected by lidocaine blockades of the central inputs from L3 and L4 fibers during L5 nerve stimulation, suggesting the mediation of PID-induced hyperalgesia by sensitization, not activation, of peripheral terminals of adjacent intact afferents. The similar hyperalgesia was also observed following electrical stimulation of decentralized L3 spinal nerve. Prior elimination of L4 C-fibers by local capsaicin prevented hyperalgesia induced either by L5 SNL with a

Damage to the peripheral nerve leads to Wallerian degeneration (WD) of the nerve fibers distal to the injury site. The interactions between degenerating fibers and adjacent intact afferent fibers during WD have been proposed to be one of the multiple underlying mechanisms of injury-induced neuropathic pain (Li et al., 2000; Li et al., 2002; Sheth et al., 2002; Obata et al., 2004). Considerable evidence has indicated that such interactions are mediated by pro-inflammatory cytokines such as tumor necrosis factor-␣ (TNF-␣) and interleukins (Sorkin et al., 1997; Leem and Bove, 2002; Shamash et al., 2002; Zelenka et al., 2005; Campana et al., 2006; Uceyler et al., 2007). Besides WD, nerve damage generates injury-produced discharges that arise from the injury sites and last for several minutes, and ectopic discharges that arise from proximal sites of injured nerves, such as the neuroma and the dorsal root ganglion (DRG), and last for several weeks (Wall et al., 1974; Govrin-Lippmann and Devor, 1978; Chung et al., 1992; Michaelis et al., 2000). It is known that both injury and ectopic discharges, which propagate centrally to the spinal cord or the brain stem, contribute in part to neuropathic pain (Ossipov et al., 2002; Sun et al., 2005; Jang et al., 2007). Thus, when considering the contribution of nerve damage-induced WD to neuropathic pain, the possible involvement of centrally-propagating injury-induced discharges should be excluded. Dorsal rhizotomy, when performed prior to peripheral nerve injury, is one way of excluding the involvement of such central input of injury discharges. Injury discharge also propagates antidromically along afferent fibers to peripheral terminals, which contribute to the development of neuropathic pain through the release of glutamate and neuropeptides (deGroot et al., 2000; Jang et al., 2004a,b). Thus, in the presence of a blockade of injury-induced central inputs by a prior dorsal rhizotomy, both peripherally-propagating injury discharge (PID) and WD are considered to be peripheral contributions to nerve

1 Present address: Department of Anesthesia, 3000 ML, University of Iowa Hospitals and Clinics, Iowa City, IA 52242, USA. 2 These authors contributed equally to this work. *Correspondence to: J. W. Leem or D. W. Kim, Department of Physiology, Yonsei University College of Medicine, C.P.O. Box 8044, Seoul 120-752, Korea. Tel: ⫹822-2228-1709 or ⫹822-2228-1703; fax: ⫹822393-0203. E-mail addresses: [email protected] (J. W. Leem), [email protected] (D. W. Kim). Abbreviations: CAPs, compound action potentials; DRG, dorsal root ganglion; L5, lumbar 5; PBS, phosphate-buffered saline; PID, peripherally-propagating injury discharge; PWTs, paw withdrawal thresholds; RTX, resiniferatoxin; SNL, spinal nerve lesion; TNF-␣, tumor necrosis factor-␣; WD, Wallerian degeneration.

0306-4522/10 $ - see front matter © 2010 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2009.09.082

221

222

J. H. Jang et al. / Neuroscience 165 (2010) 221–232

injury-induced neuropathic pain. In this condition, the role of WD alone can be studied in the presence of a conduction blockade of PID at the time of injury. Meanwhile, the role of PID can be studied separately by mimicking it with brief tetanic electrical stimulation of the nerve (Vatine et al., 1998). Our earlier studies on rats with lumbar 5 (L5) rhizotomy showed that mild hyperalgesia lasted for 7 days and that an L5 spinal nerve lesion (SNL) performed after the disappearance of such mild hyperalgesia produced long-lasting mechanical hyperalgesia (Jang et al., 2004a,b). Using a rat model with prior L5 rhizotomy, we investigated to what extent the two processes, WD and PID, contributed to L5 SNL-induced mechanical hyperalgesia. In addition, it has been determined which of the uninjured A- and Cafferents in the L4 spinal nerve is important for each contribution. Two contributions have been studied separately by measuring mechanical hyperalgesia induced either by the L5 SNL performed in the presence of a nerve block with lidocaine or by brief tetanic electrical stimulation of the L5 spinal nerve. In the comparative experiments using L3 rhizotomy-bearing rats, mechanical hyperalgesia elicited by L3 SNL with lidocaine blockade, or by electrical stimulation of L3 spinal nerve, was measured.

EXPERIMENTAL PROCEDURES Experimental animals Adult male rats (150 –250 g; Sprague–Dawley, Harlan) were used in this study. The animals were housed in groups of three to four, with food and water available ad libitum. All animals were acclimated to a light-dark cycle for approximately 1 week before the surgery and behavioral testing. All experiments were carried out in accordance with the approval of the Institutional Animal Care and Use Committee of Yonsei University, Seoul, Korea.

Surgical procedures All surgical procedures were carried out under enflurane anesthesia (2–3% enflurane-O2 mixture). Animals used were initially subjected to an L5 or L3 dorsal rhizotomy before receiving subsequent surgical manipulations. For the L5 dorsal rhizotomy, a longitudinal skin incision was made to expose the L3–L6 vertebral segments, and a hemilaminectomy was performed at the left L5 segments. Care was taken to avoid any mechanical trauma to the spinal cord, the dorsal roots and the DRG. The dura matter was opened, and the left L5 dorsal root was exposed and sectioned 2–3 mm proximal to the L5 DRG. A small portion of the distal end of the dissected root was sectioned for removal. For the L3 dorsal rhizotomy, the same procedure as described above was performed, except that a hemilaminectomy was performed at the left L3 vertebral segments to expose the left L3 dorsal root. A thin nerve branch connecting the L3 and L4 spinal nerves was sectioned at the time of the L3 rhizotomy. Wounds were closed by suturing the overlying muscles and skin in layers. For all rats that underwent dorsal rhizotomies, it was confirmed that the rhizotomy was indeed performed on the L5 dorsal root and that the disconnected parts of the dorsal root were not reconnected at the end of the scheduled behavioral tests. Neuropathic surgery was done by ligation of the left L5 spinal nerve as described by Kim and Chung (1992), but the ligated nerve was sectioned distally (spinal nerve lesion, SNL). Briefly, under enflurane anesthesia, a skin incision was made over the lumbar spine, and the left transverse process of L6 vertebra was

exposed and removed to access the L5 spinal nerve. The exposed spinal nerve was tightly ligated with 6-0 silk thread and sectioned about 1 mm distal to the ligation. For the left L3 SNL, the L3 nerve was exposed by removing the left transverse process of the L4 vertebra, ligated, and sectioned. The exposed L3 and L5 spinal nerves were also subjected to the local lidocaine treatment or electrical stimulation. In some experiments in which the left L4 spinal nerve was treated with lidocaine or capsaicin, the left L5 transverse process was removed to expose the nerve. After the surgeries were completed, the overlying muscles and skin were sutured in layers. The animals were provided with postoperative care for recovery.

Experimental designs Experiments were carried out in rats that received a left L5 dorsal rhizotomy in which a mild mechanical hyperalgesia, if present, had worn off within a week of the rhizotomy. Rats that did not show the abolition of hyperalgesia, which was observed in approximately 15% of rats that had received L5 rhizotomy in the present study, were excluded. The L5 rhizotomy-bearing rats were assigned to the following experimental groups, with the exception of experiment 4. Experiment 1 was designed to test if peripheral contributions would produce mechanical hyperalgesia following SNL in the absence of centrally conducting injury discharge. This group consisted of rats that received a left L5 rhizotomy followed a week later by a sham-L5 SNL, a left L5 rhizotomy followed a week later by a left L5 SNL, and a sham-L5 rhizotomy followed a week later by a left L5 SNL (n⫽10 for each). For sham operations, all surgeries were done exactly as the L5 SNL or L5 rhizotomy but without cutting the spinal nerve or dorsal root. Experiment 2 was performed to test the contribution of WD to SNL-induced hyperalgesia. This group included left L5 rhizotomybearing rats in which the ipsilateral L5 SNL was performed on post-rhizotomy day 7 in the presence of local lidocaine [2% in phosphate-buffered saline (PBS), pH 7.4] or PBS (n⫽10 for each). Experiment 3 was performed to test if PID contributed to SNL-induced hyperalgesia through peripheral sensitization. Rats in this group were treated the same as rats in experiment 2, except for treating the left L5 spinal nerve with electrical stimulation instead of SNL, in the presence of local lidocaine or PBS treatment (n⫽9 for each). It was also tested if a weaker strength of electrical L5 nerve stimulation, which activates A-fibers not Cfibers, still led to changes in mechanical sensitivity (n⫽9). Also included were L5 rhizotomy-bearing rats receiving a left L5 nerve stimulation in the presence of local treatment of the left L3 and L4 spinal nerves with lidocaine or PBS (n⫽9 for each). Experiment 4 was designed to support experiments 2 and 3. This group included rats that received a left L3 rhizotomy, which was followed a week later by a left L3 SNL in the presence of local lidocaine or PBS treatment (n⫽9 for each). Also included were L3 rhizotomy-bearing rats that received electrical stimulation of the left L3 spinal nerve in the presence of local lidocaine or PBS treatment (n⫽9 for each). Experiment 5 was designed to measure the role of adjacent intact C-afferents in WD-induced hyperalgesia following SNL. Rats in this group included left L5 rhizotomy-bearing rats that received a prior treatment of left L4 spinal nerve with either capsaicin or vehicle, which were then followed 6 weeks later by a left L5 SNL in the presence of local lidocaine treatment (n⫽9 for each). Our previous studies have shown that waiting 6 weeks after capsaicin treatment was required for complete abolition of established hyperalgesia by the elimination of capsaicin-sensitive L4 C-afferents (Jang et al., 2007). Experiment 6 was designed to determine if adjacent intact C-afferents are essential for PID-induced hyperalgesia following SNL. This group included left L5 rhizotomy-bearing rats that had received a prior treatment on the left L4 spinal nerve with either

J. H. Jang et al. / Neuroscience 165 (2010) 221–232 capsaicin or a vehicle, which were followed 6 weeks later by an electrical stimulation of the left L5 spinal nerve (n⫽9 for each).

Behavioral testing Each rat was placed in a plexiglass cage (8⫻8⫻20 cm3) above a wire mesh bottom that allowed full access to their paws. Behavioral accommodation was allowed for 15 min before starting the behavioral testing. Mechanical hypersensitivity was evaluated by measuring paw withdrawal thresholds (PWTs) upon the application of a von Frey filament using the Dixon’s up-down testing paradigm (Dixon, 1980). An ascending series of von Frey filaments of incremental force (0.3, 0.5, 0.7, 2.5, 3.7, 5.2, 8.0, and 15.0 g) was applied perpendicular to the middle of tori of the hind paw (mainly, L4 dermatome), not on the keratinized foot pads, for 2–3 s until each filament bent slightly, starting with a 2.5 g stimulus. For the PWT measurement, the 50% withdrawal threshold was determined by the calculation from the formula, log (50% threshold)⫽Xf⫹␬␦; where Xf⫽value of the final von Frey filament (log unit), ␬⫽correction factors (from calibration table), and ␦⫽mean differences between stimuli (Chaplan et al., 1994). The 18 g pressure of 50% threshold measurement (the maximum score according to up-down method) was selected as the cutoff value. The baseline levels of PWTs were obtained 1 h before surgery. After the dorsal rhizotomy, the testing was done for all animals on post-rhizotomy days 1, 3, and 7. From post-rhizotomy day 7, for experiments 1, 2, 3, and 4, the testing was done initially once a day for a week and then once a week thereafter up to post-treatment day 42. For experiments 5 and 6, rats received the first treatment followed 6 weeks later by the second one; the testing was initially done on day 1, and then once a week thereafter for up to 6 weeks after the first treatment and then was preformed once a day for a week and once a week thereafter for up to 7 weeks after the second treatment. Behavioral testing was performed by investigators who were blind to the surgeries and treatment that the animals had received.

Brief tetanic electrical stimulation Animals were under enflurane anesthesia throughout this electrical stimulation experiment. The left L5 or L3 spinal nerve of L5 or L3 rhizotomy-bearing rats was exposed as described above. A piece of parafilm was placed underneath the exposed spinal nerve, isolating it from the surrounding tissue. A pair of silver wires that were very thin and flexible (0.005 inches in diameter) were placed to gently loop around the spinal nerve (3 mm apart). The other ends of the wires were connected to a stimulus isolation unit (SI-1850, World Precision Instruments, New Haven, CT, USA). Square-wave pulses (0.5 ms, 4 Hz) were applied with graded levels of current to determine the lowest level of current needed to elicit a muscle twitch, or the threshold current. The current pulses at strengths 200 times a threshold current (2– 4 mA, 200⫻ TH) were delivered for 5 min to activate both A- and C-fibers in the spinal nerve. All procedures were performed carefully under direct observation with a microscope so as not to allow wire electrodes to contact any other nerve or tissue except for the stimulating nerve. Although tetanic electrical stimulation of the spinal nerve resulted in hind limb jerking, no pulling of the nerve and a little movement at the contact between the loop-electrode and the nerve were observed because of the flexibility of the wire electrodes. In some experiments, tetanic electrical stimualtion at a lower level of current strength (50⫻ TH, 0.1 ms pulse, 4 Hz), which also elicited hind limb jerking without activating C-fibers, was used to see if the jerk itself caused an increase in mechanical sensitivity of hind paw. After the stimulation experiment was completed, the electrodes were removed and the wound was sutured for postoperative care.

223

Local lidocaine or capsaicin treatment For local lidocaine treatment, a piece of parafilm was placed underneath the nerve under study and a cotton pad (5 mm in width) soaked in lidocaine (2% in PBS) was placed around the spinal nerve. An attached cotton pad and the affected region of the nerve were wrapped together with parafilm to prevent leakage of any solution. Then, another cotton pad was applied outside wrapped parafilm to prevent possible spread of the solution to neighboring tissue or nerve, and then left for 15 min. After removing the applied cotton pad and Parafilm, SNL or electrical stimulation was performed at the middle of the lidocaine application site. As a control, a PBS-soaked pad was applied. The capsaicin was applied locally to the left L4 spinal nerve in the same manner as in the lidocaine treatment, except for applying a cotton pad soaked in 1.5% capsaicin (Sigma, St. Louis, MO, USA) in an aqueous vehicle (5% Tween 80, 5% EtOH, 90% PBS) for 30 min. As the control, the vehicle solution was applied. All treatment procedures were done under microscopic observation. After the surgery for the capsaicin treatment was completed, the animals were provided with postoperative care for recovery.

Recordings of compound action potentials The recordings of compound action potentials (CAPs) from the sciatic nerves in naive rats were performed as described previously (Jang et al., 2007) to confirm a nerve block by lidocaine. Briefly, the animal was anesthetized with urethane (i.p., 1.5 g/kg initially, followed by additional doses of 100 mg/kg as required). Tracheotomy tube insertion and a polyethylene catheter insertion into the external jugular vein were carried out for artificial ventilation and administration of muscle relaxant (pancuronium, 2 mg/ kg), respectively. After paralyzing the animal, the left L5 spinal nerve was exposed for applying electrical stimulation, and the left sciatic nerve at the mid-thigh region for recording CAPs. Ten traces of CAPs evoked by the single pulse current stimuli (0.5 ms, 4 mA, 0.3 Hz) were averaged for analysis. Recordings of A- and C-fiber CAPs were obtained before and at various time points after the lidocaine treatment performed on the stimulation site.

Statistical analysis To determine the differences between the different treatment groups on a given testing day, data were analyzed using a Mann– Whitney rank sum test or a Kruskal Wallis analysis of variance (ANOVA) followed by a Dunn’s test for multiple comparisons. The differences from baseline within a given treatment group were analyzed using a Friedman repeated measures ANOVA followed by a Dunn’s test for multiple comparisons. Bonferroni corrections were made. P⬍0.05 was considered to be statistically significant. All data are represented as means⫾SEMs.

RESULTS Mechanical hyperalgesia following L5 SNL in rats receiving a prior L5 dorsal rhizotomy As shown in Fig. 1, an L5 dorsal rhizotomy resulted in a partial decrease in PWTs of the affected hind paws on post-rhizotomy day 1, which was followed by a complete return to the basal level within a week. The recovered PWTs in rhizotomy-bearing rats were maintained throughout the entire 42 day observation period. For another group of rhizotomy-bearing rats, L5 SNL ipsilateral to the rhizotomy, when performed after rhizotomy-induced partial decreases of PWTs had returned to the baseline, resulted in a rapid decrease in PWTs of the affected hind paws on

224

J. H. Jang et al. / Neuroscience 165 (2010) 221–232

Fig. 1. Changes in mechanical sensitivity of the hind paws in rats receiving L5 rhizotomy alone, L5 rhizotomy followed by L5 SNL, and L5 SNL alone. The L5 dorsal rhizotomy (DR) produced a mild, transient mechanical hyperalgesia by showing a partial drop in PWTs of the affected hind paws and a complete return to the baseline within a week (open circles, n⫽10). In rats receiving L5 rhizotomy, L5 spinal nerve lesion (SNL) ipsilateral to the rhizotomy produced a severe, long-lasting mechanical hyperalgesia by showing a rapid decrease of PWTs that remained lowered throughout the observation period (closed circles, n⫽10), which is comparable to hyperalgesia that developed following L5 SNL alone (open triangles, n⫽10). The preoperative test period is denoted as P and postoperative time for dorsal rhizotomy as D and for SNL as S. Data are represented as means⫾SEMs. * P⬍0.05 versus baseline. # P⬍0.05 versus L5 DR⫹sham.

fibers as compared to degenerating fibers in injured L5 spinal nerve. Given that interactions at contact points between degenerating and intact fibers during WD serve as an immediate cause of mechanical hyperalgesia, L3 SNL with lidocaine treatment would result in a delayed onset of hyperalgesia as compared with L5 SNL with lidocaine treatment. This possibility is tested on L3 rhizotomy-bearing rats. As shown in Fig. 2B, L3 SNL performed in the presence of local lidocaine led to a PWT decrease in the affected hind paws; which started after post-SNL day 7, reached the significant level on post-SNL week 2, and remained lowered up to post-SNL week 6 (lidocaine-treated group; P⬍0.05 vs. baseline, Friedman RM ANOVA, Dunn’s test). However, L3 SNL with PBS treatment resulted in a PWT drop, which reached a significant level within 1 day of

post-SNL day 1. The decreased PWTs remained lowered throughout the entire observation period, which were significantly different from the baseline (P⬍0.05, Friedman RM ANOVA, Dunn’s test) as well as from PWTs in rhizotomy-bearing rats with a sham L5 SNL (P⬍0.05, Kruskal Wallis ANOVA, Dunn’s test). These PWT decreases were quite similar in magnitude and time-course to PWT decreases induced by L5 SNL in sham rhizotomy-bearing rats. Lidocaine-resistant SNL-induced mechanical hyperalgesia To study the contribution of WD to L5 SNL-induced hyperalgesia, changes in mechanical sensitivity of the hind paws of L5 rhizotomy-bearing rats following L5 SNL were measured in the condition of lidocaine blockades performed at the time of L5 SNL. As shown in Fig. 2A, L5 SNL of rhizotomy-bearing rats resulted in a significant decrease in PWTs of the affected hind paws on post-SNL day 1. This decrease in PWTs remained lowered throughout the 42 day observation period. When an L5 SNL was performed in the presence of a lidocaine-block (lidocaine-treated group), PWTs decreased similarly in both magnitude and time-course as the PBS-treated group. The decreased PWTs in both the PBS and lidocaine-treated groups were significantly different from their corresponding baselines (P⬍0.05, Friedman RM ANOVA, Dunn’s test). In the periphery, L3 fibers are in contact with L4 fibers at the peripheral terminals whereas the contact point of L5 fibers with L4 fibers begins in the sciatic nerve. Thus, it is expected that degenerating L3 fibers after L3 SNL would require a much longer time to make contact with intact L4

Fig. 2. Mechanical hyperalgesia of rhizotomy-bearing rats following SNL in the presence of a nerve block. (A) After the return of L5 rhizotomy-induced partial drops of PWTs to the baseline, ipsilateral L5 SNL produced the long-lasting mechanical hyperalgesia even if the lesioned site was treated with lidocaine to prevent the generation of the injury discharge by SNL (closed circles, n⫽10). This mechanical hyperalgesia was comparable to that seen in the control group in which L5 SNL was performed in the presence of PBS treatment on the lesioned site (open circles, n⫽10). (B) The L3 SNL of L3 rhizotomybearing rats, in which the L3 SNL site was treated with lidocaine, produced mechanical hyperalgesia that developed with a 7 d delayed onset (closed circles, n⫽9). In the control group in which the L3 SNL site was treated with PBS, L3 SNL produced mechanical hyperalgesia that developed within a day of SNL (open circles, n⫽9). The denotations for post-operative time are the same as in Fig. 1. Data are represented as means⫾SEMs. * P⬍0.05 versus baseline. # P⬍0.05 versus lidocaine-treated group.

J. H. Jang et al. / Neuroscience 165 (2010) 221–232

225

Fig. 3. Peripheral nerve block by lidocaine. (A) Typical traces of A- and C-fiber CAPs recorded from the left sciatic nerves of naive rats, which were elicited by electrically stimulating the left L5 spinal nerve with single pulse current stimuli (0.5 ms, 4 mA, 0.3 Hz). Recordings were obtained before and at various time points after the lidocaine treatment performed on the stimulation site. Recordings of full sized A- and C-fiber CAPs are in the left column, and those expanded in time and amplitude in the right column. The scale bars denote 100 ␮V in amplitude and 50 ms in time for the left traces, and 10 ␮V and 25 ms for the right traces. The A- and C-fiber CAPs are indicated by arrows with A and C, respectively, in the top traces. The stimulus artifact can be seen in the second trace from the top of the right column, in which the evoked A- and C-fiber CAPs are absent. (B) Histograms representing the percent amplitudes of A- and C-fiber CAPs at various post-treatment times versus the pre-treatment value (n⫽8). Both A- and C-fiber CAPs were blocked immediately after the lidocaine treatment, and reappeared to reach the baseline level at 90 and 150 min post-lidocaine for A- and C-fiber CAPs, respectively. Data are represented as means⫾SEMs. * P⬍0.05 versus baseline.

SNL, and persisted at low levels until post-SNL week 6 (PBS-treated group; P⬍0.05 vs. baseline, Friedman RM ANOVA, Dunn’s test). The PWTs of lidocaine-treated rats were significantly different from PBS-treated rats on postSNL days 1 through 7 (P⬍0.05, unmatched Mann–Whitney rank sum test). The blockade of impulse conduction by local lidocaine treatment was confirmed by interruption of peripheral propagation of CAPs recorded from the sciatic nerve of naive rats. As shown in Fig. 3A complete blockade of both A-fiber and C-fiber CAPs occurred immediately after the lidocaine treatment. The marked blockade of C-fiber CAPs persisted up to 90 min post-lidocaine (n⫽8, P⬍0.05 vs. baseline, Friedman RM ANOVA, Dunn’s test). The blocked C-fiber CAP was completely recovered to baseline 150 min postlidocaine. The A-fiber blockade, which persisted up to 30 min post-lidocaine, was returned to the baseline level 90 min post-lidocaine. Spinal nerve stimulation-induced mechanical hyperalgesia The effects of brief tetanic electrical stimulation of the L5 spinal nerve at a high level of current strength (200⫻ TH,

0.5 ms-pulse, 4 Hz, 5 min), which would activate both Aand C-fibers, on the mechanical sensitivity of hind paws in L5 rhizotomy-bearing rats were investigated. As shown in Fig. 4A, electrical stimulation of the L5 spinal nerve without lidocaine treatment (PBS-treated group) resulted in a significant decrease in PWTs of affected hind paws, which lasted for 7 days from post-stimulation days 1 through 7, as compared with the baseline (P⬍0.05, Friedman RM ANOVA, Dunn’s test). Such decreases were also significant in comparison with PWTs of lidocaine-treated rats in which the L5 spinal nerve was electrically stimulated in the presence of a nerve block with lidocaine (P⬍0.05, unmatched Mann– Whitney rank sum test). In additional experiments, tetanic electrical stimulation of L5 spinal nerve at a lower level of current strength (50⫻ TH, 0.1 ms, 4 Hz, 5 min), which activates A-fibers only in association with hind limb jerking, showed no significant reduction of PWTs throughout the post-stimulation period, indicating that the jerk itself was not involved in producing mechanical hyperalgesia. We examined if L5 nerve stimulation still induces hyperalgesia with the blockade of central inputs from L3 and L4 afferent fibers during stimulation in order to test the involvement of peripheral sensitization. It is possible that

226

J. H. Jang et al. / Neuroscience 165 (2010) 221–232

stimulation-induced hyperalgesia is simply a result of the activation of L3 and L4 afferents during stimulation because peripherally released substances from L5 peripheral terminals are likely to cause activaton of neighboring afferents and subsequent central sensitization and hyperalgesia. As shown in Fig. 4B, in the presence of local lidocaine treatment on L3 and L4 spinal nerves, electrical stimulation of the L5 spinal nerve resulted in a significant decrease in PWTs of affected hind paws on post-stimulation days 1 through 7 as compared with the baseline (lidocaine-treated rats; P⬍0.05, Friedman RM ANOVA, Dunn’s test). Similar PWT decreases after electrical stimulation of the L5 spinal nerve were observed in the absence of lidocaine treatment, and they were significantly different from the baseline on post-stimulation days 1 through 7 (PBS-treated rats; P⬍0.05, Friedman RM ANOVA, Dunn’s test). No significant difference was observed between PWTs of lidocaine and PBS-treated rats throughout the entire observation period after stimulation. Brief tetanic electrical stimulation of L3 spinal nerve at a high level of current strength mimicking PID elicited by L3 SNL was used to investigate its effects on the mechanical sensitivity of hind paws in L3 rhizotomy-bearing rats. As shown in Fig. 4C, electrical stimulation of the L3 spinal nerve without lidocaine treatment (PBS-treated group) resulted in a significant decrease in PWTs of affected hind paws, which lasted for 7 days from post-stimulation days 1 through 7, as compared with the baseline (P⬍0.05, Friedman RM ANOVA, Dunn’s test) as well as PWTs of the lidocaine-treated group (P⬍0.05, unmatched Mann–Whitney rank sum test). In contrast, electrical stimulation of the L3 spinal nerve in the presence of lidocaine treatment (lidocaine-treated group) led to no significant changes in PWTs of affected hind paws throughout the entire 42 day observation period. Effects of a prior elimination of intact L4 C-fibers on the development of peripheral components of L5 SNL-induced mechanical hyperalgesia Fig. 4. Mechanical hyperalgesia of rhizotomy-bearing rats following an electrical stimulation of the spinal nerve. (A) Brief tetanic electrical stimulation (ES) of L5 spinal nerve in L5 rhizotomy-bearing rats at a high level of current strength (200⫻ TH, 0.5 ms-pulse, 4 Hz, 5 min), which would activate both A- and C-fibers, induced mechanical hyperalgesia that lasted for 7 d (open circles, n⫽9). When the L5 spinal nerve was electrically stimulated in the presence of a nerve block of the L5 nerve with lidocaine, no mechanical hyperalgesia developed (closed circles, n⫽9). Tetanic electrical stimulation of the L5 spinal nerve at a lower level of current strength (50⫻ TH, 0.1 ms-pulse, 4 Hz, 5 min), which would activate A-fibers only, induced no mechanical hyperalgesia (open triangles, n⫽9). (B) In the presence of lidocaine treatment on the L3 and L4 spinal nerves, the L5 nerve stimulation at a high level of current strength induced mechanical hyperalgesia on post-stimulation days 1 through 7 (closed circles, n⫽9), which was comparable to that developed following the L5 nerve stimulation without lidocaine blockades of L3 and L4 spinal nerves (open circles, n⫽9). (C) Electrical stimulation of L3 spinal nerve in L3 rhizotomybearing rats at a high level of current strength induced mechanical hyperalgesia that lasted for 7 d (open circles, n⫽9). When the L3 spinal nerve was electrically stimulated in the presence of a nerve block of the L3 nerve with lidocaine, no mechanical hyperalgesia developed (closed circles, n⫽9). Postoperative time for the electrical stimulation is denoted as E. Data are represented as means⫾SEMs. * P⬍0.05 versus baseline. # P⬍0.05 versus lidocaine-treated group.

To determine which of the intact A- and C-afferents adjacent to degenerating fibers are important for the initiation of mechanical hyperalgesia induced by WD, lidocaine-resistant L5 SNL-induced hyperalgesia affected by a prior elimination of intact L4 C-fibers was examined. As shown in Fig. 5A, L5 SNL in the presence of a lidocaine-block of the L5 spinal nerve resulted in PWT drops of the affected hind paws throughout the 49 day observation period, which were significantly different from the baseline (PBS-treated group, P⬍0.05, Friedman RM ANOVA, Dunn’s test) in L5 rhizotomy-bearing rats without capsaicin treatment on the L4 spinal nerve. However, L5 SNL in the presence of a lidocaine-block, when performed on L5 rhizotomy-bearing rats receiving capsaicin treatment on the L4 spinal nerve, resulted in no significant changes in PWTs of the affected hind paws as compared with the PWTs of PBS-treated rats throughout the observation period (P⬍0.05, unmatched Mann–Whitney rank sum test). The same experimental protocol as above has been used to investigate whether intact C-afferents are neces-

J. H. Jang et al. / Neuroscience 165 (2010) 221–232

227

DISCUSSION

Fig. 5. Effects of a prior elimination of intact L4 C-fibers on the development of peripheral components of L5 SNL-induced mechanical hyperalgesia. (A) In L5 rhizotomy-bearing rats not receiving the local capsaicin treatment on the L4 spinal nerve, an L5 SNL in the presence of a lidocaine-block of the L5 spinal nerve induced long-lasting mechanical hyperalgesia (open circles, n⫽9). In contrast, L5 SNL in the presence of a lidocaine-block of the L5 nerve, when performed on L5 rhizotomy-bearing rats receiving capsaicin treatment on the L4 spinal nerve, induced no mechanical hyperalgesia (closed circles, n⫽9). (B) Brief tetanic electrical stimulation of the L5 spinal nerve at a high current strength in L5 rhizotomy-bearing rats not receiving capsaicin treatment on the L4 nerve induced mechanical hyperalgesia that lasted for 7 d (open circles, n⫽9). Meanwhile, L5 nerve stimulation of L5 rhizotomy-bearing rats receiving capsaicin treatment on the L4 spinal nerve induced no mechanical hyperalgesia (closed circles, n⫽9). Postoperative time for the capsaicin or vehicle treatment is denoted as C. Data are represented as means⫾SEMs. * P⬍0.05 versus baseline. # P⬍0.05 versus capsaicin-treated group.

sary for initiating mechanical hyperalgesia induced by injury discharge propagating peripherally. As shown in Fig. 5B, electrical stimulation of the L5 spinal nerve in L5 rhizotomy-bearing rats without receiving capsaicin treatment led to PWT drops of the affected hind paws, which were significantly different from the baseline on post-stimulation days 1 through 7 (PBS-treated group, P⬍0.05, Friedman RM ANOVA, Dunn’s test). Meanwhile, L5 nerve stimulation performed on L5 rhizotomy-bearing rats receiving capsaicin treatment on the L4 spinal nerve led to no significant changes in PWTs of the affected hind paws as compared with PWTs of PBS-treated rats throughout the observation period (P⬍0.05, unmatched Mann–Whitney rank sum test).

Using L5 rhizotomy-bearing rats with mild mechanical hyperalgesia that lasted for 7 days, we demonstrated that an L5 SNL performed after the abolition of such mild hyperalgesia resulted in a robust, long-lasting, mechanical hyperalgesia that developed as early as one day after SNL despite lidocaine-induced nerve blockades at the time of the injury. This lidocaine-resistant component of hyperalgesia was probably caused by WD. Hyperalgesia also developed in L3 rhizotomy-bearing rats by L3 SNL with a lidocaine block, although its onset was delayed for at least 7 days. We consider this onset discrepancy as evidence that supports the idea that the contribution of WD to nerve injury-induced hyperalgesia results from interactions between degenerating (L5 or L3) and adjacent intact (L4) fibers. Brief tetanic electrical stimulation of the decentralized L5 spinal nerve, mimicking PID, resulted in mechanical hyperalgesia developing within a day and persisting for a week. This antidromic stimulation-induced hyperalgesia was unaffected by lidocaine blockades of the central inputs from L3 and L4 fibers during L5 nerve stimulation, suggesting the mediation of PID-induced hyperalgesia by sensitization, not activation, of peripheral terminals of adjacent intact afferents. The similar hyperalgesia was also observed following electrical stimulation of decentralized L3 spinal nerve. Prior elimination of intact L4 C-fibers prevented the development of both lidocaine-resistant L5 SNL-induced and L5 nerve stimulation-induced mechanical hyperalgesia, suggesting a critical role of intact Cnociceptive afferents in the initiation of peripheral components of nerve injury-induced hyperalgesia. A transient mild mechanical hyperalgesia following a dorsal rhizotomy The present data support the previous results provided by our laboratory and others, which indicate that an L5 rhizotomy produces mild, transient mechanical hyperalgesia of the affected hind paw that is absent within a week of the rhizotomy (Jang et al., 2004a,b; Sukhotinsky et al., 2004; Jang et al., 2007). However, these results are in contradiction to other previous studies showing that dorsal rhizotomy itself produces robust neuropathic pain behavior lasting longer than a week (Colburn et al., 1999; Eschenfelder et al., 2000; Li et al., 2000). Although the reasons for this discrepancy are unclear, our data selection may resolve this conflict at least in part. In our studies, 15% of the rats that had received rhizotomy showed rigorous pain behavior which lasted much longer than our presented results (data not shown). However, these rats were excluded for the purpose of our studies. This data selection makes it difficult to directly compare our results with other’s contradictory results. Considering the excluded data, our results are not far from the report by Li et al. (2000) that showed a decrease of paw withdrawal threshold became much smaller a few days after rhizotomy. It is also possible that mechanical trauma to the spinal cord, ventral root or the DRG in the process of performing the dorsal root rhizotomy causes prolonged hyperalgesia. In fact, in the beginning of

228

J. H. Jang et al. / Neuroscience 165 (2010) 221–232

the preliminary experiments we observed that about 50% of rats that had L5 rhizotomy showed severe mechanical hyperalgesia that lasted longer than a week implying the importance of delicate surgery. The mechanism by which the rhizotomy induces a transient mechanical hyperalgesia is not clear so far. The sensitization of central (Svendsen et al., 1997) and peripheral (see discussion below) neurons by the barrage of impulses elicited briefly at the time of the root transection is probably involved. In our study, a mild transient mechanical hyperalgesia induced by the rhizotomy vanished within a week and did not reappear thereafter throughout the 12 weeks observation period. We speculate that this is a result of the complete return of sensitized neurons to the normal state. Thus, after the behavioral effect of the rhizotomy wears off, central or peripheral consequences due to manipulations on spinal nerves, such as lesions and stimulation, would not be confounded by a prior rhizotomy. The involvement of WD in mechanical hyperalgesia following SNL We demonstrated using rats with a prior L5 rhizotomy that an L5 SNL results in mechanical hyperalgesia similar in severity and time course to that induced by L5 SNL alone in spite of a nerve block with lidocaine at the injury site to interrupt the peripheral propagation of injury-elicited discharge (Fig. 2A). The lidocaine treatment used in the present study shows a temporary but complete block of Aand C-fiber conduction (Fig. 3). The lidocaine-resistant L5 SNL-induced hyperalgesia develops within one day of the injury and lasts for over 42 days. Given that peripheral component of SNL-induced mechanical hyperalgesia is governed mainly by WD and PID, the role of WD can be deduced from the inability of lidocaine to prevent L5 SNLinduced mechanical hyperalgesia. However, since the WD contribution is not tested directly, the possible involvement of some other factors than WD in SNL-induced hyperalgesia cannot be excluded. It also holds true that the conclusion drawn from this experimental approach is derivative in nature. Nonetheless, we hypothesize that the lidocaineresistant injury-induced hyperalgesia is caused by WD. The reason for this is that the onset of hyperalgesia after nerve injury with lidocaine treatment is likely dependent on the contact between degenerating and adjacent intact axons during WD (see below). Pro-inflammatory cytokines are likely to be involved in the lidocaine-resistant L5 SNL-induced hyperalgesia that has an early onset. In support of this, levels of TNF-␣ and interleukin-1␤ (IL-1␤) are found to rapidly increase within hours of nerve injury at the injury site (Shamash et al., 2002; Campana et al., 2006; Uceyler et al., 2007). TNF-␣ and IL-1␤ are observed to elicit spontaneous activities within minutes when applied to DRG, axons or peripheral terminals (Sorkin et al., 1997; Binshtok et al., 2008), which may contribute to hyperalgesia by releasing algesic substances such as substance P and ATP. Abundant evidence suggests that glutamate, substance P and ATP act as mediators implicated in enhanced pain sensitivity caused by the sensitization of dorsal horn neurons

(Dougherty et al., 1993; Rusin et al., 1993; Gu and MacDermott, 1997; Stanfa et al., 2000) as well as nociceptors (Carlton et al., 1998; Hamilton et al., 1999; Tsuda et al., 2000; Jarvis et al., 2001; Hilliges et al., 2002; Zhang et al., 2006). In addition, these cytokines directly sensitize nociceptors when applied on DRG neurons or peripheral terminals (Sorkin et al., 1997; Binshtok et al., 2008), and produce mechanical hyperalgesia within an hour when injected into hind paw (Cunha et al., 1992; Woolf et al., 1997; Binshtok et al., 2008). Therefore, it is conceivable that pro-inflammatory cytokines produced within hours of the nerve injury is implicated in mechanical hyperalgesia by eliciting spontaneous activity that causes immediate release of algesic substances responsible for the sensitization of spinal neurons and nociceptors, and also by directly sensitizing nociceptors. Indeed, intact nociceptors intermingled with degenerating fibers became sensitized to mechanical stimuli (Shim et al., 2005) with spontaneous activities (Wu et al., 2001, 2002; Shim et al., 2005) after a partial nerve injury. However, it still remains to be determined if this can happen in the presence of blockade of PID. In the comparative experiments, when the WD contribution was isolated by excluding PID with lidocaine treatment, the onset of mechanical hyperalgesia was delayed after L3 SNL (Fig. 2B). This delayed onset may result from a delayed contact between degenerating L3 and intact L4 axons. For peripheral nerve fibers in the rat hind limb, the majority of fibers in the L3 spinal nerve run within the femoral nerve and its distal branch saphenous nerve to innervate the medial aspect of the sole of the foot (L3 dermatome). Most fibers in the L4 and L5 spinal nerves join together to run intermingled within the sciatic nerve and its distal branch tibial and sural nerves to innervate the middle (L4 dermatome) and lateral (L5 dermatome) parts of the sole, respectively (Swett and Woolf, 1985; Swett et al., 1991). The nerve fibers in two nerve trunks, the sciatic and femoral nerves with their respective nerve branches, do not connect peripherally until they get closer for interacting in their terminals at the boundary between the L3 and L4 dermatomes on the foot skin. From this anatomical point of view, when the L3 spinal nerve is damaged, it is expected that degenerating L3 axons would require a much longer time to make contact with nearby intact L4 axons as compared to degenerating axons in injured L5 spinal nerve, and that the contact might be limited at the peripheral terminals. This delayed contact between injured and intact axons may account for a delayed onset of mechanical hyperalgesia following L3 SNL with lidocaine treatment, supporting the importance of interactions between degenerating and adjacent intact fibers during WD for nerve injury-induced hyperalgesia. Mechanical hyperalgesia following brief antidromic electrical stimulation of the spinal nerve The stimulation parameters used here to mimic injury discharge produced by SNL are based on the observation that injury discharge generated by the axotomy of C-fibers persists for 4 min at a discharge frequency in the range of

J. H. Jang et al. / Neuroscience 165 (2010) 221–232

0.3–7.2 Hz with a mean of 2 Hz (Blenk et al., 1996). The stimulus intensity used, which is strong enough to activate both A- and C-fibers, is known to result in neurotransmitter release from peripheral terminals of primary sensory afferents (deGroot et al., 2000). We could observe severe red reaction, which is one symptom of neurogenic inflammation induced by peripherally released neuropeptides (Holzer, 1988), in affected hind paws during and for a while after electrical stimulation. It is possible, therefore, that acute injury discharge propagating peripherally triggers the release of algesic substances, leading to inflammation and nociceptor sensitization responsible for mechanical hyperalgesia (Carlton et al., 1998; Hamilton et al., 1999; Tsuda et al., 2000; Jarvis et al., 2001; Hilliges et al., 2002; Zhang et al., 2006). It has been controversial if the spread of nociceptor sensitization occurs or not by antidromic activities such as antidromic nerve stimulation, dorsal root or axon reflex, which is similar to the concept that antidromic injury discharge produces sensitization in neighboring nociceptors. Our hypothesis can be supported by some reports (Chahl and Ladd, 1976; Fitzgerald, 1979; Panopoulos et al., 1983; Li et al., 2008), but not by others (Reeh et al., 1986; Meyer et al., 1988; Schmelz et al., 1996). A direct comparison or conclusion of these arguable data cannot be drawn, as different chemicals at various doses or diverse electrical currents were used for nociceptor activiation in different species, and some reports do not involve mechanical tests. Whatever the reasons for this discrepancy, without the compelling conclusion that spreading of nociceptor sensitization does not exist, it would be reasonable to assume that, in the present study, peripheral sensitization occurs in neighboring nociceptors by antidromic activities to explain mechanical hyperalgesia lasting for over a week following stimulation of decentralized spinal nerve. The stimulation-induced hyperalgesia that has an early onset is maintained for a week and weakens thereafter until it disappears (Fig. 4A, C). Based on these findings, we suggest that PID contributes to SNL-induced mechanical hyperalgesia in early time points after injury through peripheral sensitization, and that the disappearance of hyperalgesia is the result of recovery of sensitized nociceptors to a normal state. One might argue that such hyperalgesia is simply due to the activation, not sensitization, of intact L3 and L4 afferents during stimulation, as peripherally released substances from L5 peripheral terminals are likely to cause activaton of neighboring afferents and subsequent central sensitization and hyperalgesia. However, this is not likely the case because the electrical stimulation of the L5 spinal nerve results in hyperalgesia that lasts for a week, even when immediately preceded by a nerve block with lidocaine of the L3 and L4 spinal nerves to obstruct afferent inputs to the spinal cord (Fig. 4B). The peripherally conducting impulses elicited from the spinal nerve produce mechanical hyperalgesia that is intense and long-lasting as demonstrated by the series of experiments, that is, hyperalgesia induced by SNL or by electrical stimulation of L3 or L5 spinal nerve. However,

229

dorsal rhizotomy which may elicit similar injury discharge induced less severe and short-lasting hyperalgesia. The reasons for this discrepancy are not clear at this point, but might be reconciled in part as follows. First, in the present study, peripherally conducting injury discharge was generated once by dorsal rhizotomy, but twice by SNL. For dorsal rhizotomy, dorsal root was cut and proximal part was cut again for removal. In this case, second injury discharge cannot be propagated to the periphery. However, for SNL, tight ligation was followed by cut of distal part of spinal nerve, which may elicit repeated and more severe injury discharge propagating to the periphery. Second, electrical currents used to stimulate spinal nerve are likely to excite most C-fibers, whereas 18% of C-fibers generate action potentials immediately after axtomy (Blenk et al., 1996). We cannot exclude, therefore, the possibility that mechanical hyperalgesia induced by electrical stimulation was exaggerated to a certain extent. However, a week delay of the onset of L3 SNL-induced hyperalgesia with lidocaine treatment, in which actual injury discharge was blocked, strongly suggests the substantial contribution of PID to the SNL-induced mechanical hyperalgesia. Third, in contrast to dorsal rhizotomy, SNL or electrical stimulation can recruit entire afferent and efferent fibers for the impulse generation, and some of these fibers may exert facilitatory effects on nociception. In support, it has been known that noradrenaline which can be released from postganglionic sympathetic neurons by antidromic activity provokes hyperalgesia and discharge of primary nociceptive afferents in inflamed tissue (Hu and Zhu, 1989; Sato et al., 1993; Baik et al., 2003). Intact C-afferent fibers are critical to initiate peripheral components of L5 SNL-induced mechanical hyperalgesia In a previous study, we observed that mechanical hyperalgesia already established after L5 SNL is completely abolished by local capsaicin treatment on L4 spinal nerve (Jang et al., 2007). Although maximum C-fiber elimination only reached 34% as demonstrated by CAPs, the majority of capsaicin-sensitivie C-fibers, which are known as polymodal nociceptors, were removed. Our findings presented here show that prior elimination of L4 C-afferents with the same capsaicin treatment as before prevents the initiation of peripheral components of L5 SNL-induced mechanical hyperalgesia. These results taken together suggest that intact C-nociceptive afferents are necessary for the induction as well as the maintenance of partial nerve injuryinduced neuropathic pain. However, some previous studies using systemic injection of an ultrapotent capsaicin analog, resiniferatoxin (RTX), suggested that, in conflict with our observations, RTX treatment leads to tactile allodynia lasting for 2 to 3 weeks by causing degeneration of unmyelinated afferents and abnormal sprouting of myelinated afferents into the spinal lamina II (Pan et al., 2003), and that capsaicin sensitive fibers are not critical for tactile allodynia induced by nerve-ligation injury (Ossipov et al., 1999). This discrepancy might be attributed to the difference in the means of

230

J. H. Jang et al. / Neuroscience 165 (2010) 221–232

chemical delivery (systemic vs. local treatment). Indeed, in support of our observations, local perineural RTX treatment itself does not elicit mechanical hypersensitivity until a week after treatment (Neubert et al., 2008), and completely prevents loose ligation-induced increase in withdrawal frequency to von Frey filaments (Kissin et al., 2007). Furthermore, local capsaicin treatment onto sciatic nerve reverses established mechanical hyperalgesia following spinal nerve transection rather than enhance over 4 weeks (Kim et al., 2008). It is probable that local capsaicin treatment induces no or only a small amount of axonal sprouting of A-afferents in the dorsal horn, resulting in little or no mechanical hyperalgesia. In support of this, it has been demonstrated that only very limited sprouting of myelinated afferents into the lamina-II is induced following peripheral axotomy in the rat (Bao et al., 2002). On the other hand, capsaicin-induced excitation or degeneration of peripheral axons, in contrast to the case induced by SNL or electrical stimulation, does not seem to be enough to produce long-lasting hyperalgesia. For this issue, it would be noteworthy that capsaicin and RTX are selective stimulants that only can affect capsaicin sensitive afferents, while the SNL or electrical stimulation of spinal nerve can produce a recruitment of the entire spectrum of axons for activation. The possible explanations are as follows. First, although it has been suggested that degeneration in myelinated efferent fibers is sufficient to induce spontaneous activity in nearby C-fiber afferents and mechanical hyperalgesia (Wu et al., 2002), it is not clear if degeneration of capsaicin sensitive C-fiber only is enough to produce nociceptor sensitization and hyperalgesia. Second, some of the additionally recruited fibers may cause more sensitizing effects by releasing algesic mediators. For examples, noradrenaline released from sympathetic efferent fibers may enhance excitability of nociceptors and hyperalgesia (Hu and Zhu, 1989; Sato et al., 1993; Baik et al., 2003). In addition, sympathectomy that would be caused by SNL is known to induce adrenergic sensitivity in cutaneous C-fiber nociceptors (Bossut et al., 1996). The capsaicin insenstivie C- and myelinated A-fibers possessing neuropeptides (Yu et al., 2008) should also be taken into consideration as possible contributors to SNL- or electrical stimulation-induced mechanical hyperalgesia. In contrast to an emphasis on uninjured axons in the present study and previous reports by others (Li et al., 2000; Lee et al., 2003), the importance of injured fibers has also been suggested in the neuropathic pain model (Sheen and Chung, 1993; Yoon et al., 1996; Sukhotinsky et al., 2004). However, in the present study, contribution of injured spinal nerve cannot be emphasized under our experimental conditions because centrally conducting impulses including acute injury discharge and ectopic discharges from injured fibers cannot directly access the CNS due to prior rhizotomy. Indeed, we previously demonstrated that both injured and uninjured fibers contribute to mechanical hyperalgesia following SNL, and prior dorsal rhizotomy is effective to eliminate contribution of injured fibers (Jang et al., 2007).

The central sensitization and A-beta fiber mediated tactile allodynia is a widely accepted hypothesis to explain mechanical hypersensitivity in neuropathic pain (Campbell et al., 1988). Signals along low-threshold A-beta fibers become amplified by central sensitization so that they produce pain. Our observation may not be in direct contradiction with this explanation because we do not exclude the involvement of central sensitization which is likely to be induced by spontaneous activities from sensitized nociceptors. However, this central sensitization would not be initiated or maintained without peripheral sensitization under our experimental conditions. Taken all together, our data suggest that capsaicin sensitive nociceptive afferents, presumably polymodal Cnociceptors, remaining intact after partial nerve injury play a critical role in the initiation of mechanical hyperalgesia, perhaps through the interaction with injured axons undergoing WD and through the sensitization of peripheral terminals by injury discharge. The clinical implications of these in managing neuropathic pain would be to achieve membrane stabilization of axons and to suppress hyperresponsiveness of peripheral terminals in nociceptive afferent neurons. Acknowledgments—We would like to thank Mr. H.S. Park and Mr. S. Cho for their technical assistance. This work was supported by a grant from the Stem Cell Research Center of the 21st Century Frontier Research Program (SC4140) funded by the Ministry of Science and Technology, the Republic of Korea.

REFERENCES Baik E, Chung JM, Chung K (2003) Peripheral norepinephrine exacerbates neuritis-induced hyperalgesia. J Pain 4:212–221. Bao L, Wang HF, Cai HJ, Tong YG, Jin SX, Lu YJ, Grant G, Hokfelt T, Zhang X (2002) Peripheral axotomy induces only very limited sprouting of coarse myelinated afferents into inner lamina II of rat spinal cord. Eur J Neurosci 16:175–185. Binshtok AM, Wang H, Zimmermann K, Amaya F, Vardeh D, Shi L, Brenner GJ, Ji RR, Bean BP, Woolf CJ, Samad TA (2008) Nociceptors are interleukin-1beta sensors. J Neurosci 28:14062–14073. Blenk KH, Janig W, Michaelis M, Vogel C (1996) Prolonged injury discharge in unmyelinated nerve fibres following transection of the sural nerve in rats. Neurosci Lett 215:185–188. Bossut DF, Shea VK, Perl ER (1996) Sympathectomy induces adrenergic excitability of cutaneous C-fiber nociceptors. J Neurophysiol 75:514 –517. Campana WM, Li X, Shubayev VI, Angert M, Cai K, Myers RR (2006) Erythropoietin reduces Schwann cell TNF-alpha, Wallerian degeneration and pain-related behaviors after peripheral nerve injury. Eur J Neurosci 23:617– 626. Campbell JN, Raja SN, Meyer RA, Mackinnon SE (1988) Myelinated afferents signal the hyperalgesia associated with nerve injury. Pain 32:89 –94. Carlton SM, Zhou S, Coggeshall RE (1998) Evidence for the interaction of glutamate and NK1 receptors in the periphery. Brain Res 790:160 –169. Chahl LA, Ladd RJ (1976) Local oedema and general excitation of cutaneous sensory receptors produced by electrical stimulation of the saphenous nerve in the rat. Pain 2:25–34. Chaplan SR, Bach FW, Pogrel JW, Chung JM, Yaksh TL (1994) Quantitative assessment of tactile allodynia in the rat paw. J Neurosci Methods 53:55– 63.

J. H. Jang et al. / Neuroscience 165 (2010) 221–232 Chung JM, Leem JW, Kim SH (1992) Somatic afferent fibers which continuously discharge after being isolated from their receptors. Brain Res 599:29 –33. Colburn RW, Rickman AJ, DeLeo JA (1999) The effect of site and type of nerve injury on spinal glial activation and neuropathic pain behavior. Exp Neurol 157:289 –304. Cunha FQ, Poole S, Lorenzetti BB, Ferreira SH (1992) The pivotal role of tumour necrosis factor alpha in the development of inflammatory hyperalgesia. Br J Pharmacol 107:660 – 664. deGroot J, Zhou S, Carlton SM (2000) Peripheral glutamate release in the hindpaw following low and high intensity sciatic stimulation. Neuroreport 11:497–502. Dixon WJ (1980) Efficient analysis of experimental observations. Annu Rev Pharmacol Toxicol 20:441– 462. Dougherty PM, Palecek J, Zorn S, Willis WD (1993) Combined application of excitatory amino acids and substance P produces longlasting changes in responses of primate spinothalamic tract neurons. Brain Res Brain Res Rev 18:227–246. Eschenfelder S, Habler HJ, Janig W (2000) Dorsal root section elicits signs of neuropathic pain rather than reversing them in rats with L5 spinal nerve injury. Pain 87:213–219. Fitzgerald M (1979) The spread of sensitization of polymodal nociceptors in the rabbit from nearby injury and by antidromic nerve stimulation. J Physiol 297:207–216. Govrin-Lippmann R, Devor M (1978) Ongoing activity in severed nerves: source and variation with time. Brain Res 159:406 – 410. Gu JG, MacDermott AB (1997) Activation of ATP P2X receptors elicits glutamate release from sensory neuron synapses. Nature 389: 749 –753. Hamilton SG, Wade A, McMahon SB (1999) The effects of inflammation and inflammatory mediators on nociceptive behaviour induced by ATP analogues in the rat. Br J Pharmacol 126:326 –332. Hilliges M, Weidner C, Schmelz M, Schmidt R, Orstavik K, Torebjork E, Handwerker H (2002) ATP responses in human C nociceptors. Pain 98:59 – 68. Holzer P (1988) Local effector functions of capsaicin-sensitive sensory nerve endings: involvement of tachykinins, calcitonin gene-related peptide and other neuropeptides. Neuroscience 24:739 –768. Hu SJ, Zhu J (1989) Sympathetic facilitation of sustained discharges of polymodal nociceptors. Pain 38:85–90. Jang JH, Kim DW, Sang Nam T, Se Paik K, Leem JW (2004a) Peripheral glutamate receptors contribute to mechanical hyperalgesia in a neuropathic pain model of the rat. Neuroscience 128: 169 –176. Jang JH, Kim KH, Nam TS, Lee WT, Park KA, Kim DW, Leem JW (2007) The role of uninjured C-afferents and injured afferents in the generation of mechanical hypersensitivity after partial peripheral nerve injury in the rat. Exp Neurol 204:288 –298. Jang JH, Nam TS, Paik KS, Leem JW (2004b) Involvement of peripherally released substance P and calcitonin gene-related peptide in mediating mechanical hyperalgesia in a traumatic neuropathy model of the rat. Neurosci Lett 360:129 –132. Jarvis MF, Wismer CT, Schweitzer E, Yu H, van Biesen T, Lynch KJ, Burgard EC, Kowaluk EA (2001) Modulation of BzATP and Formalin induced nociception: attenuation by the P2X receptor antagonist, TNP-ATP and enhancement by the P2X(3) allosteric modulator, cibacron blue. Br J Pharmacol 132:259 –269. Kim SH, Chung JM (1992) An experimental model for peripheral neuropathy produced by segmental spinal nerve ligation in the rat. Pain 50:355–363. Kim SM, Kim J, Kim E, Hwang SJ, Shin HK, Lee SE (2008) Local application of capsaicin alleviates mechanical hyperalgesia after spinal nerve transection. Neurosci Lett 433:199 –204. Kissin I, Freitas CF, Bradley EL Jr (2007) Perineural resiniferatoxin prevents the development of hyperalgesia produced by loose ligation of the sciatic nerve in rats. Anesth Analg 104:1210 –1216.

231

Lee DH, Iyengar S, Lodge D (2003) The role of uninjured nerve in spinal nerve ligated rats points to an improved animal model of neuropathic pain. Eur J Pain 7:473– 479. Leem JG, Bove GM (2002) Mid-axonal tumor necrosis factor-alpha induces ectopic activity in a subset of slowly conducting cutaneous and deep afferent neurons. J Pain 3:45– 49. Li D, Ren Y, Xu X, Zou X, Fang L, Lin Q (2008) Sensitization of primary afferent nociceptors induced by intradermal capsaicin involves the peripheral release of calcitonin gene-related peptide driven by dorsal root reflexes. J Pain 9:1155–1168. Li L, Xian CJ, Zhong JH, Zhou XF (2002) Effect of lumbar 5 ventral root transection on pain behaviors: a novel rat model for neuropathic pain without axotomy of primary sensory neurons. Exp Neurol 175:23–34. Li Y, Dorsi MJ, Meyer RA, Belzberg AJ (2000) Mechanical hyperalgesia after an L5 spinal nerve lesion in the rat is not dependent on input from injured nerve fibers. Pain 85:493–502. Meyer RA, Campbell JN, Raja SN (1988) Antidromic nerve stimulation in monkey does not sensitize unmyelinated nociceptors to heat. Brain Res 441:168 –172. Michaelis M, Liu X, Janig W (2000) Axotomized and intact muscle afferents but no skin afferents develop ongoing discharges of dorsal root ganglion origin after peripheral nerve lesion. J Neurosci 20:2742–2748. Neubert JK, Mannes AJ, Karai LJ, Jenkins AC, Zawatski L, Abu-Asab M, Iadarola MJ (2008) Perineural resiniferatoxin selectively inhibits inflammatory hyperalgesia. Mol Pain 4:3. Obata K, Yamanaka H, Dai Y, Mizushima T, Fukuoka T, Tokunaga A, Yoshikawa H, Noguchi K (2004) Contribution of degeneration of motor and sensory fibers to pain behavior and the changes in neurotrophic factors in rat dorsal root ganglion. Exp Neurol 188:149 –160. Ossipov MH, Bian D, Malan TP Jr, Lai J, Porreca F (1999) Lack of involvement of capsaicin-sensitive primary afferents in nerve-ligation injury induced tactile allodynia in rats. Pain 79:127–133. Ossipov MH, Zhang ET, Carvajal C, Gardell L, Quirion R, Dumont Y, Lai J, Porreca F (2002) Selective mediation of nerve injury-induced tactile hypersensitivity by neuropeptide Y. J Neurosci 22:9858 – 9867. Pan HL, Khan GM, Alloway KD, Chen SR (2003) Resiniferatoxin induces paradoxical changes in thermal and mechanical sensitivities in rats: mechanism of action. J Neurosci 23:2911–2919. Panopoulos P, Gazelius B, Olgart L (1983) Alterations in intradental nerve responsiveness induced by electrical stimulation of the inferior alveolar nerve in the cat. Acta Physiol Scand Suppl 515:37– 43. Reeh PW, Kocher L, Jung S (1986) Does neurogenic inflammation alter the sensitivity of unmyelinated nociceptors in the rat? Brain Res 384:42–50. Rusin KI, Bleakman D, Chard PS, Randic M, Miller RJ (1993) Tachykinins potentiate N-methyl-D-aspartate responses in acutely isolated neurons from the dorsal horn. J Neurochem 60:952–960. Sato J, Suzuki S, Iseki T, Kumazawa T (1993) Adrenergic excitation of cutaneous nociceptors in chronically inflamed rats. Neurosci Lett 164:225–228. Schmelz M, Schmidt R, Ringkamp M, Forster C, Handwerker HO, Torebjork HE (1996) Limitation of sensitization to injured parts of receptive fields in human skin C-nociceptors. Exp Brain Res 109:141–147. Shamash S, Reichert F, Rotshenker S (2002) The cytokine network of Wallerian degeneration: tumor necrosis factor-alpha, interleukin1alpha, and interleukin-1beta. J Neurosci 22:3052–3060. Sheen K, Chung JM (1993) Signs of neuropathic pain depend on signals from injured nerve fibers in a rat model. Brain Res 610:62– 68. Sheth RN, Dorsi MJ, Li Y, Murinson BB, Belzberg AJ, Griffin JW, Meyer RA (2002) Mechanical hyperalgesia after an L5 ventral rhizotomy or an L5 ganglionectomy in the rat. Pain 96:63–72.

232

J. H. Jang et al. / Neuroscience 165 (2010) 221–232

Shim B, Kim DW, Kim BH, Nam TS, Leem JW, Chung JM (2005) Mechanical and heat sensitization of cutaneous nociceptors in rats with experimental peripheral neuropathy. Neuroscience 132:193–201. Sorkin LS, Xiao WH, Wagner R, Myers RR (1997) Tumour necrosis factor-alpha induces ectopic activity in nociceptive primary afferent fibres. Neuroscience 81:255–262. Stanfa LC, Kontinen VK, Dickenson AH (2000) Effects of spinally administered P2X receptor agonists and antagonists on the responses of dorsal horn neurones recorded in normal, carrageenaninflamed and neuropathic rats. Br J Pharmacol 129:351–359. Sukhotinsky I, Ben-Dor E, Raber P, Devor M (2004) Key role of the dorsal root ganglion in neuropathic tactile hypersensibility. Eur J Pain 8:135–143. Sun Q, Tu H, Xing GG, Han JS, Wan Y (2005) Ectopic discharges from injured nerve fibers are highly correlated with tactile allodynia only in early, but not late, stage in rats with spinal nerve ligation. Exp Neurol 191:128 –136. Svendsen F, Tjolsen A, Hole K (1997) LTP of spinal A beta and C-fibre evoked responses after electrical sciatic nerve stimulation. Neuroreport 8:3427–3430. Swett JE, Torigoe Y, Elie VR, Bourassa CM, Miller PG (1991) Sensory neurons of the rat sciatic nerve. Exp Neurol 114:82–103. Swett JE, Woolf CJ (1985) The somatotopic organization of primary afferent terminals in the superficial laminae of the dorsal horn of the rat spinal cord. J Comp Neurol 231:66 –77. Tsuda M, Koizumi S, Kita A, Shigemoto Y, Ueno S, Inoue K (2000) Mechanical allodynia caused by intraplantar injection of P2X receptor agonist in rats: involvement of heteromeric P2X2/3 receptor signaling in capsaicin-insensitive primary afferent neurons. J Neurosci 20:RC90. Uceyler N, Tscharke A, Sommer C (2007) Early cytokine expression in mouse sciatic nerve after chronic constriction nerve injury depends on calpain. Brain Behav Immun 21:553–560.

Vatine JJ, Argov R, Seltzer Z (1998) Brief electrical stimulation of c-fibers in rats produces thermal hyperalgesia lasting weeks. Neurosci Lett 246:125–128. Wall PD, Waxman S, Basbaum AI (1974) Ongoing activity in peripheral nerve: injury discharge. Exp Neurol 45:576 –589. Woolf CJ, Allchorne A, Safieh-Garabedian B, Poole S (1997) Cytokines, nerve growth factor and inflammatory hyperalgesia: the contribution of tumour necrosis factor alpha. Br J Pharmacol 121: 417– 424. Wu G, Ringkamp M, Hartke TV, Murinson BB, Campbell JN, Griffin JW, Meyer RA (2001) Early onset of spontaneous activity in uninjured C-fiber nociceptors after injury to neighboring nerve fibers. J Neurosci 21:RC140. Wu G, Ringkamp M, Murinson BB, Pogatzki EM, Hartke TV, Weerahandi HM, Campbell JN, Griffin JW, Meyer RA (2002) Degeneration of myelinated efferent fibers induces spontaneous activity in uninjured C-fiber afferents. J Neurosci 22:7746 –7753. Yoon YW, Na HS, Chung JM (1996) Contributions of injured and intact afferents to neuropathic pain in an experimental rat model. Pain 64:27–36. Yu L, Yang F, Luo H, Liu FY, Han JS, Xing GG, Wan Y (2008) The role of TRPV1 in different subtypes of dorsal root ganglion neurons in rat chronic inflammatory nociception induced by complete Freund’s adjuvant. Mol Pain 4:61. Zelenka M, Schafers M, Sommer C (2005) Intraneural injection of interleukin-1beta and tumor necrosis factor-alpha into rat sciatic nerve at physiological doses induces signs of neuropathic pain. Pain 116:257–263. Zhang Q, Zhao Y, Guo Y, Cao DY, Tang XD, Tian YL, Yao FR, Wang HS (2006) Activation and sensitization of C and Adelta afferent fibers mediated by P2X receptors in rat dorsal skin. Brain Res 1102:78 – 85.

(Accepted 30 September 2009) (Available online 3 October 2009)