Abstract

Objectives

Patients suffering from neuropathic pain are difficult to treat and many methods are used to resolve this issue. In this study, we used a model of neuropathic pain comprising rats with chronic constriction injury (CCI) on the left sciatic nerve to investigate the chronic effect of gabapentin via intrathecal administration. We also observed the expression of dorsal spinal protein kinase C gamma subunit (PKCγ) and other pain-related molecules in the spinal area which included cyclooxygenase 2 (COX2), c-Fos and cyclic AMP-dependent transcription factor (ATF3) in the neuropathic pain animals.

Methods

Male Sprague-Dawley (SD) rats (250–380 g) were randomly assigned to four groups, i.e., control, gabapentin (Gaba), MK801, and gabapentin plus MK801 (Gaba+M) groups. A PE-5 catheter was inserted into the lumbar spine area via the cervical spine area. CCI was performed the following day after the intrathecal catheter implant surgery. Gabapentin (1.05 μmol/day) was then given the following day after CCI surgery. Intrathecal gabapentin was administrated for 14 consecutive days. Pain-related behavior was assessed every 2 days thereafter by measuring the latency of foot withdrawal elicited by noxious radiant heat or Von Frey microfilament applied to the hind-paw plantar surface. MK801 (30 μg/day), an N-methyl-D-aspartate (NMDA) receptor blocker, was also added for potential effect. The tissue of dorsal horn of the lumbar spine was harvested on the 14^th day for the expression of COX2, c-Fos, ATF3 and PKCγ with Western blotting, and positive finding protein was then checked on 7^th day for further evaluation.

Results

The beneficial effect of intrathecal gabapentin of statistic significance on thermal duration and mechanical microfilament appeared after 7-day and 11-day consecutive treatment, respectively. Furthermore, the NMDA receptor blocker also potentiated the effect on the behavior of thermal and mechanical stimulations. Gabapentin had no effect on the expression of COX2, c-Fos and ATF3. Interestingly, the expression of PKCγ in the spinal cord was initially inhibited by gabapentin on the 7^th day but was potentiated on the 14^th day.

Conclusions

Our results indicate that chronic intrathecal gabapentin has beneficial effects on the behaviors of both thermal and mechanical stimulations in the neuropathic pain animals and the NMDA blocker can potentiate this effect. Furthermore, gabapentin has biphasic effect on the expression of PKCγ in the spinal cord on Day 7 and Day 14 for the model rats with CCI.

Keywords

anticonvulsants; gabapentin; gamma; injections;intrathecal; neuralgia; neuropathic pain; protein kinase C; spinal;

1. Introduction

Gabapentin, an antiepileptic agent, was found to possess antinociceptive action in 1996.¹ In pain clinics, gabapentin has been considered for treating postherpetic neuralgia.² It has also been reported to be used to target the α2δ1 subunit of voltage-dependent calcium channels.³ In animal models, gabapentin prevents hyperalgesia,⁴ allodynia⁵ and also responds to neuropathic pain in several models.1, 6 It also decreases pain-related responses after peripheral inflammation but does not alter immediate pain-related behaviors.⁷ For patients suffering from neuropathic pain, gabapentin was thought to be able to reduce central sensitization,⁸ postherpetic neuralgia,⁹ and postoperative or posttraumatic thoracic pain.¹⁰ Meanwhile, the N-methyl-D-aspartate (NMDA) receptor blocker, MK801, was also reported to potentiate the antinociceptive effect of gabapentin.⁷ However, few studies investigated the chronic intrathecal use of gabapentin in neuropathic pain. Chu et al¹¹ have shown the beneficial effect of continuous infusion of intrathecal gabapentin on neuropathic pain for a 7-day treatment. However, the effect on the chronic use of NMDA block with gabapentin remains unknown.

The chronic influence of gabapentin on pain-related molecules has not been fully investigated. Gabapentin is suggested as an alternative to the cyclooxygenase-2 (COX2) inhibitor clinically.¹² c-Fos and activating transcription factor-3 (ATF3) have been mentioned to be related to neuropathic pain after nerve injury.13, 14 Furthermore, protein kinase C gamma isoform (PKCγ) are thought to be widely spread in spinal neuron¹⁵ and upregulated during pain condition.¹⁶ It has also been proved that activation of spinal PKCγ could activate the ascending pain transmission.¹⁷ However, little data are available for the impact of long-term intrathecal gabapentin upon pain-related proteins. Our study investigated the effect of intrathecal gabapentin on spinal PKCγ expression during 14-day observations. Our data indicated that chronic intrathecal gabapentin administration could have a beneficial effect on neuropathic animals and NMDA receptor blocker potentiated its effect. Gabapentin also inhibited the expression of spinal PKCγ in the first week but potentiated it in the second week. NMDA receptor blocker also inhibited spinal PKCγ expression in the second week.

2. Methods and statistical analysis

2.1. Animal preparation

Male SD rats (250–380 g), purchased from National Cheng Kung University laboratory animal center, were randomly assigned to four groups (i.e., control, gabapentin (Gaba), MK801, gabapentin plus MK801 (Gaba+M), n = 6 in each group). All procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of our university. Isoflurane (2%) was used for anesthesia during the intrathecal catheter implantation procedure. A PE-5 catheter was then inserted into the lumbar enlargement of the spinal cord via the foramen magnum in the cervical area. All CCI procedures were performed on the left side on the following day after the intrathecal surgery. Only normal rats without prominent neurological symptoms were included for the CCI procedure. The CCI was produced by exposing and isolating the sciatic nerve at mid-thigh level, and tying three loosely-tied knots in it with 4-0 chromic thread. Muscle and skin were closed in layers. Gabapentin was then given on the following day after CCI.

2.2. Experimental protocol

Intrathecal administration of gabapentin (1.05 μmol/day, Sigma Corp) was carried out by slow infusion (about 1 min) for 14 consecutive days via the implanted catheter. The drug was given in 10 μL water solution, followed by 10 μL of normal saline to flush the catheter. MK801 (30 μg/10 μL, Sigma Corp, dissolved in 10% DMSO) was also given 30 min before gabapentin injection. Saline was used as a control.

2.3. Behavioral studies

All behaviors were measured 2 hours after drug administration. The observer was blinded to the management of the animals. Pain-related behavior was assessed by measuring the latency of foot withdrawal elicited by noxious radiant heat applied to the hind-paw plantar surface (cutoff value 30 seconds). Three measurements at intervals within 1 hour were performed on the lesion side and their mean was considered as the value for each animal. The withdrawal threshold of mechanical stimulation was determined similarly to the method described previously.¹⁸ Briefly, the rats were placed in an individual Plexiglass housing (18 cm × 8 cm × 8 cm) with wire mesh flooring, and allowed to explore and groom until they settled down. A set of von Frey filaments (Stoeling Company, Kiel, WI, USA) with bending forces ranging from 0.6 g to 26 g were applied perpendicularly in an ascending order to the plantar surface of the left hind-paw. Hind-paw withdrawal was considered as positive response. If there was no response, the next filament with upgraded bending force was applied. The 50% withdrawal produced by a bending force was recorded and determined thrice using Dixon’s up–down method and the average value was considered as the mechanical withdrawal threshold of the paw.¹⁹

2.4. Protein expression with Western blotting

At the end of experiment, the animals were first deeply anesthetized by 5% isoflurane and then blood was flashed away with icy normal saline via an intracardiac catheter. The spinal cord tissue of the dorsal-half horn at the lumbar region was then harvested. The tissue was homogenized in a lyses buffer. The lysate was centrifuged and the protein concentration was determined by protein assay. The protein sample was then subjected to SDS-polyacrylamide gel electrophoresis and transferred to PVDF membrane for COX2 (Santa Cruz, CA, USA, 1:200), c-Fos (Santa Cruz, CA, USA, 1:200), ATF3 (Santa Cruz, CA, USA, 1:500), and PKCγ (BD Transduction Lab, USA, 1:1000) expression and β-actin (Millipore, USA, 1:5000) was used for internal control. Specific antibody–antigen complex was detected with an enhanced chemiluminescence system. The density of each band was scanned and measured with imaging analysis software (Image J, NIH) after the background was subtracted and the ratio of band was calculated for statistic analysis.

2.5. Statistical analysis

All values are expressed as mean ± standard error mean (SEM). All behavior data of paw withdrawal latencies for thermal or mechanical threshold (Von Frey) were subjected to passing the normality test first, then were set for suitable parametric statistics and analyzed with the Student t test (two groups) or one-way analysis of variance (ANOVA) with repeated measures followed by Newman-Keuls test for post hoc analysis. A value of p < 0.05 was considered to be statistically significant.

3. Results

Fig. 1 shows the thermal behavior response after 14-day gabapentin treatment for the left CCI hind-paw. The radiant heat behavior showed that in the gabapentin group (Gaba) and MK801 plus gabapentin group (Gaba+M) the heat latency could be prolonged after Day 7 [10.03 ± 0.41 (Con), 13.27 ± 0.73 (Gaba), 13.57 ± 0.25 (Gaba+M), p < 0.01 vs. Con] and could be sustained to Day 14 [12.67 ± 0.48 (Con), 16.50 ± 0.58 (Gaba), 19.97 ± 1.65 (Gaba+M), p < 0.01 vs. Con] after drug administration. Meanwhile, MK801 plus gabapentin also showed a greater effect than gabapentin alone after Day 9 to Day 11 [Day 9: 14.50 ± 0.23 (Gaba), 16.67 ± 0.51 (Gaba+M), p = 0.003; Day 11: 14.65 ± 0.55 (Gaba), 18.37 ± 1.21 (Gaba+M), p = 0.019] but not after Day 14 [16.50 ± 0.58 (Gaba), 19.97 ± 1.65 (Gaba+M), p = 0.076]. In Fig. 2, the mechanical response in comparison with control, also showed statistic difference since Day 11 (14.67 ± 0.16 vs. 15.32 ± 0.07, p < 0.01) or Day 9 (15.00 ± 0.09 vs. 15.77 ± 0.20, p = 0.038) for Gaba group or Gaba+M group, respectively. Also, coadministration of MK801 and gabapentin also produced much potential effect than gabapentin alone did on Day 7 (16.19 ± 0.20 vs. 15.21 ± 0.03, p < 0.01), Day 9 (15.77 ± 0.20 vs. 15.27 ± 0.11, p = 0.04) and Day 14 (16.31 ± 0.32 vs. 15.43 ± 0.07, p = 0.02) on mechanical allodynia-like behavior but not on Day 11 (16.06 ± 0.38 vs. 15.32 ± 0.08, p = 0.09). Fig. 3 shows the effect of intrathecal gabapentin on spinal expression of various pain related protein. It clearly shows that gabapentin has had no effect on the expression of COX2 [Fig. 3A, 1.00 ± 0.02 (N: Naive), 1.00 ± 0.08 (C: control), 1.16 ± 0.13 (G: gabapentin)], c-Fos [Fig. 3B, 1.00 ± 0.01 (N), 0.94 ± 0.01 (C), 0.96 ± 0.02 (G)] and ATF3 [Fig. 3C, 1.00 ± 0.07 (N), 0.98 ± 0.01 (C), 0.97 ± 0.01 (G)] on Day 14 and COX2 also had not changed on Day 7 [figure not shown, 1.00 ± 0.03 (N), 0.98 ± 0.01 (C), 1.08 ± 0.01 (G)]. However, the different effect of intrathecal gabapentin administration on the expression of PKCγ was clearly demonstrated in Fig. 4. Gabapentin significantly suppressed the expression of spinal PKCγ in the spinal cord on Day 7 [Fig. 4A, 1.00 ± 0.11 (N), 2.11 ± 0.02 (C), 0.49 ± 0.01 (G)] but increased its expression on Day 14 in the CCI model [Fig. 4B, 1.00 ± 0.07 (N), 1.15 ± 0.08 (C), 1.81 ± 0.34 (G)]. Meanwhile, the expression of PKCγ was also significantly suppressed after administering NMDA receptor blocker alone on Day 14 [Fig. 4B, 0.55 ± 0.16 (M: MK801) vs. 1.00 ± 0.07 (N) p = 0.03; 0.88 ± 0.10 (G+M: gabapentin plus MK801) vs. 2.10 ± 0.24 (G) p = 0.04].

Download full-size image

Fig. 1. Thermal behavior response for animals in the CCI model after intrathecal gabapentin administration for 14 days. Gabapentin (Gaba) can prolong the thermal duration from Day 7 and coadministration of MK801 (Gaba+M) further prolonged thermal duration to Day 9 and Day 11. # and * denote the statistical difference (p < 0.05) compared to the control group or gabapentin group (Gaba), respectively. n = 6 in each group. Data were presented with mean + SEM.

Download full-size image

Fig. 2. Mechanical behavior responses for von Frey microfilament in CCI model after intrathecal gabapentin administration for 14 days. Gabapentin (Gaba) improved the mechanical sensation on Days 11 and 14 and, similar to Fig. 1, coadministration of MK801 (Gaba+M) can potentiate the effect from Day 9 (vs. control group) or Day 7 (vs. gabapentin group). # and * demonstrate the statistical difference (p < 0.05) compared with the control group or gabapentin group (Gaba), respectively. n = 6 in each group. Data are presented as mean ± SEM.

Download full-size image

Fig. 3. Protein expression in the dorsal lumbar area of the spinal cord after intrathecal gabapentin administration for 14 days. (A) COX2, (B) c-FOS and (C) ATF3. N, C, and G represent naive, control (with CCI model but no drug), and gabapentin group (with CCI model), respectively. They all showed no statistical difference after drug administration. n = 4 in each group and data are presented as mean ± SEM.

Download full-size image

Fig. 4. The expression of PKCγ in dorsal lumbar spinal cord for (A) Day 7 and (B) Day 14, respectively. They presented the suppression and overexpression of PKCγ after gabapentin administration. MK801 itself and gabapentin coadministered with MK801 also suppressed the PKCγ expression. N, C, G, M, and G+M represented naïve, control (with CCI model but no drug), gabapentin group (with CCI model), MK801 group and gabapentin plus MK801, respectively. # and * denote the statistical difference (p < 0.05) compared with the naive group (N) or gabapentin group (G), respectively. n = 4–5 in each group and data are presented as mean ± SEM.

4. Discussion

Our results clearly demonstrated that chronic intrathecal gabapentin administration has a beneficial effect on pain behavior and the coadministration of an MNDA antagonist and gabapentin can potentiate the effect on mechanical and thermal behaviors. We also demonstrated that chronic intrathecal gabapentin administration has biphasic effects on the expression of PKCγ in the spinal dorsal horn. Our results also confirmed the previous study about the efficiency of this drug for chronic intrathecal administration on pain behavior but with a reduced dose (480 μg/day vs. 180 μg/day).¹¹ In contrast with that study, we showed that our lower-dose administration produced no effect on behavior over the first 5 days and improved after 7-day consecutive treatment and gradually increased up to Day 14. Furthermore, the NMDA receptor significantly potentiated those effects. There are many studies which demonstrate the beneficial effect of intrathecal gabapentin in neuropathic pain animals. However, there was no study on its intrathecal administration for over 7 days. This study clearly demonstrated the benefit of the use of intrathecal gabapentin for over 7 days.

To investigate the possible mechanism of gabapentin on neuropathic pain, many studies have focused on the voltage-dependent calcium channels as target.³ However, we focused on other possible molecules mentioned previously and demonstrated the relationship in the pain process. Our results show a biphasic effect of gabapentin on spinal PKCγ expression but no effect on COX2, c-Fos or ATF3 was demonstrated. The overexpression of PKCγ in the spinal cord was thought to be related to nerve injury-induced pain behavior.¹⁵ However, in our study, the overexpression was suppressed by gabapentin at the end of first week (Day 7) and then became normal on Day 14. PKCγ was reported to be involved in nociceptive neuroplasticity in the spinal cord.²⁰ However, the real mechanism for this biphasic phenomenon and its relationship with PKCγ on pain behavior process needs further investigation.

The beneficial effect of coadministration of NMDA blocker with gabapentin in chronic treatment was also clearly demonstrated in our results. The synergistic effect had been demonstrated for gabapentin plus NMDA antagonist in formalin rats.⁷ It is reasonable that NMDA antagonist can potentiate the effect of gabapentin in neuropathic animals. However, gabapentin has been shown to enhance the NMDA current in the dorsal horn²¹ which may induce pain. However, our chronic coadministration in vivo demonstrated the benefit. The possible mechanism for this coadministrative effect may be due to the additive suppression on NMDA-gated calcium influx and voltage-dependent calcium current (α2δ1 subunit) of the neurons. NMDA receptors are critical in a dynamic balance of excitatory and inhibitory inputs which will activate the L-type calcium current.²² The α2δ1 subunit of voltage-dependent calcium is also critical for calcium influx.²³ It is quite reasonable for those two drugs to produce additive independent suppression over the calcium influx. More evidence is needed to confirm this hypothesis. Interestingly, MK801 alone can suppress the expression of PKCγ in spinal cord on Day 14 and combines with gabapentin to make its expression normal. However, because the pain behavior was not well correlated for this expression, the role of PKCγ on late phase of neuropathic pain needs further study to resolve this issue. Calcium influx or NMDA receptor/PKCγ signaling pathway may participate in this effect.²⁴

In our study, we demonstrated that chronic intrathecal gabapentin administration can improve both the thermal and mechanical pain behaviors and the effect was potentiated when an NMDA blocker was used simultaneously for neuropathic pain animals. Intrathecal gabapentin had been demonstrated to reduce allodynia with different neuropathic pain²⁵ but it did not show a long-term effect as seen in our study. Our result is the first one that reveals the effect of gabapentin potentiated by MK801 in neuropathic pain in long-term use. The mechanism of this drug on neuropathic pain may be involved in many different pathways. Further studies are needed to resolve the issue and its real mechanism.