A total of 96 healthy male Sprague-Dawley (SD) rats (age, 4–6 weeks; weight, 120–150 g) were purchased from the Model Animal Research Center of Nanjing University (Nanjing, China). They were kept under a temperature of 22–24°C, a humidity of 50–60% and a 12 h light-dark cycle. All rats had ad libitum access to food and water. Rats were randomly divided into two groups: An SNL group (n=60) and a sham group (n=36). The 60 rats in the SNL group received a unilateral, tight ligation and transection of the left L₅ spinal nerve using a modification of the surgical procedure previously described by Kim and Chung (12). In brief, under anesthesia with 2% pentobarbital sodium administered intraperitoneally (i.p., 50 mg/kg; Sigma-Aldrich; Merck Millipore, Darmstadt, Germany) and following exposure of the left L₅ spinal nerve, the left L₅ and L₄ spinal nerves were identified. The left L₅ spinal nerve was separated and tightly ligated with 5–0 silk sutures between the DRG and the conjunction to form the sciatic nerve and transected just distal to the ligature. The 36 sham rats underwent exposure, while they did not undergo ligation or transection of the spinal nerve. The present study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The experimental animal protocol was reviewed and approved by the Institutional Animal Care and Use Committee of Nanjing University (Nanjing, China).

All experimental rats were subjected to behavioral tests at 1 day prior to as well as 1, 3, 7 and 14 days following surgery. All tests were repeated three times.

To assess tactile allodynia, rats in clear plastic cages with a wire mesh bottom were tested using an up-down method as described by Choi et al (13). A filament (Supertips™ flexible Von Frey hairs, IITC Life Science, Woodland Hills, CA, USA) with a suitable, calibrated buckling weight was applied to the plantar surface of the hind paw with pressure causing the filament to bend. Absence of paw lifting after 4 sec led to the use of the next filament with increased weight, whereas paw lifting indicated a positive response and led to the use of the next weaker filament. This paradigm continued until the withdrawal threshold was estimated as the smallest fiber size that evoked at least three withdrawal responses during five consecutive applications with the same fiber.

A von Frey mechanical pain threshold detector (Model 2390 Electric von Frey aesthesiometer; IITC Life Science) was used to assess mechanical hyperalgesia. A paw-flick response was elicited by applying increasing force (in g) using rigid tips (400 g; Model 2391; IITC Life Science) focused on the plantar surface of the hindpaw. The force applied that elicited a reflex removal of the hindpaw was recorded.

Thermal hyperalgesia was determined by comparison of paw withdrawal latency to thermal stimuli between operated and non-operated sides. Rats were exposed to a radiant source generated and controlled by an IITC Model 390 Analgesia Meter (10 V; 30 W; spot dimensions, 4×6 mm; IITC Life Science). The light radiation intensity was set at 50% with intervals of 30 sec (in order to avoid extensive damage).

At 15 days after surgery, rats in the SNL or sham group were anesthetized with 2% pentobarbital sodium (50 mg/kg i.p., Sigma-Aldrich; Merck Millipore) for the dissociation of DRG neurons. The L₄ and L₅ DRGs and attached dorsal roots were removed and the connective tissue capsule was dissected using microdissection scissors in ice-cold, oxygenated, buffered Dulbecco's modified Eagle's medium (DMEM, High Glucose, Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) containing 3.7 g/l NaHCO₃ (pH 7.2). DRG fragments were rinsed with DMEM three times and subsequently transferred into 1.5 ml DMEM containing 2 mg/ml type I collagenase (Sigma-Aldrich; Merck Millipore) and 1.2 mg/ml type I trypsin (Sigma-Aldrich; Merck Millipore) and aerated with 5% CO₂ and 95% O₂ at 36.5°C for 25 min until complete dissociation was confirmed microscopically. DRG neurons were rinsed in an oxygenated, standard extracellular solution containing 145 mmol/l tetraethylammonium chloride, 0.8 mmol/l MgCl₂, 5 mmol/l CaCl₂, 20 mmol/l 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES; Sigma-Aldrich, Merck Millipore) and 10 mmol/l D-glucose (the pH was adjusted to 7.3 with Tris base) to terminate the enzyme reactions. Individual neurons were dissociated by passing DRG fragments through a set of fire-polished glass pipettes. Finally, neurons were plated onto culture dishes consisting of extracellular solution through the coated metallic screen (250 mesh, WS Tyler, Mentor, OH, USA). Cells were allowed to attach to the culture dishes for 2–3 h at room temperature.

Neurons were studied 8–10 h after dissociation. Cells were visualized with an inverted microscope and their soma diameter was determined as small (<30 µm), medium (30–40 µm) or large (>40 µm) with the aid of a calibrated microscope eyepiece reticle (Olympus Corporation, Tokyo, Japan). The subsequent analysis focused on the medium-diameter neurons, which were considered as the nociceptive afferent (Aδ and C fibres, slow-conducting velocities) and low-threshold sensory afferent (14).

Whole-cell configuration of the patch-clamp technique was used for current clamp and voltage-clamp electrophysiological recordings using the EPC-9 patch-clamp amplifier (HEKA Elektronik, Lambrecht, Germany) with an AgCl reference electrode. Electrical stimuli pulses and current recording were controlled and performed using PULSE+PULSEFIT ver. 8.79 (HEKA Elektronik). Channel currents were sampled using the ITC-16 data acquisition system (InstruTech; HEKA Elektronik). Recordings were acquired at 20 kHz, filtered at 2 kHz and sampled using a computer with Digidata 1200 A/D acquisition board (HEKA Elektronik) for offline analyses. Sigmaplot ver. 10.0 (Systat software, Inc., San Jose, CA, USA) was used for data analysis. The current-voltage curves were generated using Igor pro 4.0 (WaveMetrics, Inc., Lake Oswego, OR, USA) software.

Micro-electrodes were pulled from glass capillaries (diameter, 1.6 mm; Beijing Xianqu Weifeng Technology Development Co., Beijing, China) using a pipette puller (PP-830; Narishige Scientific Instrument Lab., Tokyo, Japan). The diameter of the micro-electrode tip was 1–2 mm. The resistance of the internal solution [CsCl 135 mmol/l, MgCl₂ 2 mmol/l, EGTA 11 mmol/l, CaCl₂ 1 mmol/l, HEPES 20 mmol/l, magnesium-adenosine triphosphate (Mg-ATP) 5 mmol/l, guanosine 5′-triphosphate lithium salt (Li-GTP) 0.4 mmol/l, pH 7.3], filled into pipettes was 3–6 MΩ. Micropipettes were positioned on the membrane and gentle suction was applied to obtain a resistant seal in the GΩ range (>1 GΩ). Following rupture of the membrane with suction at negative pressure, entering the whole-cell recording configuration, neurons were allowed to reach adequate equilibration between the internal pipette solution and the cell interior for 10 min. The slow capacitance, system resistance, series resistance, capacitance current and leakage current were electronically compensated with the compensation circuitry of the amplifier during current recording. Effective data was defined as a steady current and series resistance <20 MΩ.

Drugs were added to extracellular solution, bathing the dissociated DRG neurons. Fluid changes and continuous perfusions were achieved through a gravity-dependent flow system. Stocks of GBP, ω-conotoxin MVIIC, ω-conotoxin MVIIA, CdCl₂ and Li-GTP (Sigma-Aldrich; Merck Millipore) were prepared in distilled water and stored at −20°C.

The current density of certain cells was obtained from the inward current standardized by the membrane capacitance and compared with those of other cells. The voltage-dependent current curve was fit according to the Boltzmann equation, G/G_max=1/{1+exp[(Va_1/2-Vm)/Ka]}, where G is the conductance calculated as G=I/(V_m-V_r), where I is the current, V_m is the command voltage, V_r is the reversal potential, Va_1/2 is the voltage at the half maximal activation current and Ka is the slope factor describing the voltage dependence of the conductance. The curve of the steady-state inactivation current was fit according to the Boltzmann equation, I/I_max = 1 / {1 + exp [(V - Vi_1/2) / Ki]}, where I is the current, V is the pre-stimulus voltage, Vi_1/2 is the voltage at the half maximal inactivation current and Ki is the slope factor of inactivation.

SPSS statistical software ver. 16.0 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. Values are expressed as the mean ± standard deviation. Two-way analysis of variance was used for comparison among groups and the q test was used for pairwise comparison. Statistical analysis between any two groups was performed using the group t-test and a paired t-test was used for comparison of pre- and post-drug application within groups. P<0.05 was considered to indicate a statistically significant difference.

All SNL-treated rats that underwent electrophysiological analysis exhibited clear behavioral indications of allodynia, mechanical hyperalgesia and thermal hyperalgesia. Prior to surgery, no significant differences in the mean foot withdrawal threshold were observed between the contralateral and ipsilateral feet of the mice in each group as well as between the SNL and sham groups. Mean withdrawal thresholds obtained from the feet ipsilateral to the SNL decreased significantly on the first post-operative day and remained significantly lower than the pre-operative value for up to 14 post-operative days (P<0.05; Fig. 1A and B). By contrast, there were no significant changes in the withdrawal threshold in paws contralateral to the SNL between the groups or for either foot in the sham-operated group. The latency for foot withdrawal in SNL rats significantly decreased in response to thermal stimuli delivered to the ipsilateral foot on the first post-operative day and remained lower during the entire observation period, compared with the non-operated side in the SNL rats (P<0.05; Fig. 1C). However, latency was not affected by the sham operation.

After entering the whole-cell recording configuration, inward currents were activated under the command voltage from −70 to +40 mV at 80-msec intervals in 10 mV increments and with a holding potential (V_H) of −90 mV. There were no low-voltage-activated (LVA) components and no fast component of inactivation in the inward currents. Activation of inward currents occurred at −40 mV and peak inward currents occurred at 0 mV according to the current density-potential (I–V) curve. Inward currents were completely suppressed following the addition of 200 µmol/l CdCl₂ with the same stimuli, which was in accordance with the properties of HVA-Ca²⁺ currents (Fig. 2).

Voltage steps of 80 msec were applied to various test potentials between −70 and +40 mV in 10-mV steps with 5-sec intervals from a V_H of −90 mV. This was performed to examine current-voltage associations in axotomized, adjacent neurons and control neurons from the L₄ and L₅ DRGs and attached dorsal roots of sham-operated rats. Peak inward calcium current levels were measured, corrected for cell size and subsequently expressed as the peak current density (pA/pF). Peak current densities of control, adjacent and axotomized neurons were 144±24, 104±17 and 75±17 pA/pF, respectively. Compared with that of control neurons, the peak current density of adjacent and axotomized neurons decreased significantly (P<0.05) and the peak current density of axotomized neurons was significantly lower than that of adjacent neurons (P<0.05). The activation voltage of the peak HVA-Ca²⁺ current density in control and adjacent neurons was 0 mV, while that of axotomized neurons was −10 mV (Fig. 3).

Steady maximal activation of HVA-Ca²⁺ currents was achieved with a voltage of V= 10 or 0 mV at 80-msec intervals and the dose-dependent inhibition effect of GBP was subsequently determined using 0.1, 1, 10, 100 and 300 µmol/l GBP. Changes in peak inward Ca²⁺ flux (ICa) were recorded and the inhibition of currents with GBP at various doses was calculated. The peak ICa was significantly enhanced in cells exposed to higher concentrations of GBP (10, 100 and 300 µmol/l; P<0.05), in comparison with cells exposed to lower concentrations of GBP (0.1 and 1 µmol/l). GBP significantly inhibited the peak inward ICa in a dose-dependent manner. Furthermore, the inhibitory effects of GBP on the peak current of axotomized neurons at concentrations of 10, 100 and 300 µmol/l were significantly higher than those on the current of control and adjacent neurons (P<0.05). There were no significant differences in the inhibitory effects on the peak current between control and adjacent neurons at various concentrations of GBP (P>0.05; Table I).

The membrane V_H was set at −90 mV and calcium currents were elicited by a series of command voltages (range, −70 to +40 mV) with successive increments of 10 mV at 80-msec intervals. As shown in Fig. 4, the midpoint potentials for the activation curves (Va_1/2) of control, adjacent and axotomized neurons were −16.23±1.92, −18.03±0.37 and −20.73±0.33 mV, respectively, with the Va_1/2 of axotomized neurons being significantly lower than that of the control and adjacent neurons (P<0.05; Fig. 4C); however, no difference was observed between those of the control and adjacent neurons (P>0.05; Fig. 4C). To study the kinetics of channel inactivation, calcium currents were generated by applying a membrane V_H at −90 mV for 10 msec, followed by series of command voltages from −70 to +20 mV with successive increments of 10 mV at 500-msec intervals. Subsequently, command voltages of −10 mV at 80-msec intervals were applied. The midpoint potentials for the inactivation curves (Vi_1/2) of control, adjacent and axotomized neurons were −31.93±1.03, −31.45±1.87 and −32.73±1.37 mV, respectively. There were no significant differences in the midpoint potentials among the three types of neurons (P>0.05; Fig. 4D).

To study the effects of GBP on the kinetics of HVA-Ca²⁺ channel activation, I–V associations for the transient calcium channel currents were plotted using 10-mV incremental step pulses from −70 to +40 mV at 80-msec intervals with a membrane V_H at −90 mV in the presence of 100 µmol/l GBP. The Va_1/2 decreased from −16.23±1.92 to −19.69±1.07 mV (P<0.05), −18.03±0.37 to −21.73±0.64 mV (P<0.05) and −20.73±0.33 to −23.94±0.15 mV (P<0.05) in the control, adjacent and axotomized neurons, respectively (Fig. 5). To investigate the effects of GBP on the kinetics of HVA-Ca²⁺ channel inactivation, steady-state inactivation was measured in a conventional way using a 500-msec pre-pulse from −70 to +20 mV in 10-mV increments with a −90-mV membrane V_H for 10 msec followed by a second test pulse at −10 mV for 80 msec in the presence of 100 µmol/l GBP. By addition of 100 µmol/l GBP, the current required to inactivate Va1/2 was decreased from −38.78±2.05 to −31.93±1.03 mV in control neurons, from −36.22±2.80 to −31.45±1.87) mV in adjacent neurons and from −44.23±2.30 to −32.73±1.37 mV in axotomized neurons (Fig. 5).

The voltage dependence of activation shifted in a depolarized direction, whereas the Vi_1/2 remained unchanged in axotomized as well as adjacent uninjured DRG neurons, compared with those in control neurons. The shift in activation reduced the ‘window current’ between inactivation and activation, thereby stopping the sustained inward ICa. Following the addition of GBP, the activation and steady-state Vi_1/2 of three groups all shifted in a hyperpolarized direction. The shift in activation and inactivation reduced the overlap (‘window currents’) between inactivation and activation in control and axotomized neurons, while the ‘window currents’ remained unchanged in adjacent neurons (Fig. 5).

Maximal activation of the HVA-Ca²⁺ currents occurred with V=10 or 0 mV at 80-msec intervals, with a membrane V_H of −90 mV. To individually assess different sub-types of HVA-Ca²⁺ channels present in the same cell, 20 µmol/l nifedipine (L-type calcium channel blocker), 2 µmol/l ω-conotoxin MVIIC (P/Q-type calcium channel blocker) and 2 µmol/l ω-conotoxin MVIIA (N-type calcium channel blocker) were applied to the bath in sequence when stable currents were reached. Similar activation voltages were applied to the same cells to evoke Ca²⁺ currents. Currents subtracted prior to and following application of respective blocker yielded L-, N-, P/Q-type currents. Subsequently, 100 µmol/l GBP was applied to determine its inhibitory efficacy on various sub-types of HVA calcium channel current.

As shown in Fig. 6, the proportion of N-type Ca²⁺ currents in axotomized neurons was significantly higher (50.0±2.7%) than that in control neurons (35.9±1.4%) and adjacent neurons (37.1±2.0%; P<0.05; Fig. 6B). The contribution of L-type channels to the HVA-Ca²⁺ influx in axotomized neurons was markedly reduced, as 15.4±2.3% of the HVA-Ca²⁺ current was blocked by 20 µmol/l nifedipine compared to 27.9±2.5% in control neurons and 26.1±1.8% in adjacent neurons (P<0.05). There was no significant difference in the proportion of each sub-type of HVA-Ca²⁺ channel current between adjacent and control neurons (P>0.05; Fig. 6B). However, the P/Q-type current was similar between the three groups.

The proportion of N-type Ca²⁺ currents in axotomized neurons was significantly higher (81.0±2.8%) than that in control and adjacent neurons (68.8±4.5 and 69.7±3.3%, respectively; P<0.05), suggesting that the N-type Ca²⁺ channel current is the GBP-sensitive sub-type of HVA-Ca²⁺ channel currents (Table II). However, the N- and P/Q-type currents were not affected by GBP.