Abstract
The present study investigated whether spinal administration ofS-(+)-3-isobutylgaba (S-(+)-3-IBG) or its stereoisomer, R-(−)-3-isobutylgaba (R-(−)-3-IBG), are effective in reducing the hyperalgesia and swelling observed after injection of kaolin and carrageenan into the knee joint of the rat. The effects of pretreatment and post-treatment of S-(+)-3-IBG,R-(−)-3-IBG and artificial cerebrospinal fluid (aCSF) on the swelling, pain-related behavior scores and the heat hyperalgesia induced by knee joint inflammation were compared. Infusion of eitherS-(+)-3-IBG or R-(−)-3-IBG through a microdialysis fiber, implanted in the dorsal horn of the spinal cord, for 1.5 h before injection of kaolin and carrageenan resulted in a 20 to 30% reduction in joint swelling compared with aCSF-treated controls, and prevented the development of heat hyperalgesia and spontaneous pain. In contrast, infusion of either stereoisomer after the development of inflammation reduced the hyperalgesia but did not reduce the amount of joint swelling compared with aCSF-treated animals. In summary, S-(+)-3-IBG and R-(−)-3-IBG are effective antihyperalgesic agents when administered both before and after joint inflammation. In addition, if administered before injection of kaolin and carrageenan into the knee joint this drug can attenuate joint inflammation. Both the antihyperalgesic and anti-inflammatory properties of this drug probably are mediated through a central neurogenic mechanism.
Acute experimental arthritis can be induced by injection of kaolin and carrageenan into the knee joint (Sluka and Westlund, 1993a). The inflammatory agent, carrageenan, causes plasma extravasation and edema after the release of neuropeptides (Lam and Ferrell, 1993) and other inflammatory mediators (Herbert and Schmidt, 1992; Birrell and McQueen, 1993; Birrell et al., 1993) into the joint cavity. Concomitant with the injury to the joint tissue, both peripheral and central sensitization occurs (Coggeshall et al., 1983;Schaible and Schmidt, 1985, 1988; Schaible et al., 1991;Dougherty et al., 1992). This peripheral and central sensitization is manifested in the awake rat as hyperalgesia (Sluka and Westlund, 1993a), which can be quantified easily by measuring a reduction in paw withdrawal latencies to a noxious radiant heat source.
Hyperalgesia induced by injection of kaolin and carrageenan in the rat can be blocked by the spinal administration of antagonists of the GABAA receptor, NMDA receptor, non-NMDA excitatory amino acid receptors and neurokinin 1 receptors (Sluka and Westlund, 1993b; Sluka et al., 1993, 1994, 1997; Reeset al., 1995). Both secondary hyperalgesia and allodynia are mediated by changes within the central nervous system as a result of increased afferent barrage from the site of injury. Thus, it is thought that in this model of secondary hyperalgesia GABA, NMDA, non-NMDA and neurokinin 1 receptors are involved in the maintenance of central sensitization. Spinal administration of either a GABAA or non-NMDA excitatory amino acid receptor antagonist before the injection of kaolin and carrageenan (pretreatment) results in a reduction in the amount of swelling as well as in the amount of hyperalgesia observed 4 h after injection (Sluka and Westlund, 1993b; Sluka et al., 1993). It has been proposed that this reduction in swelling is caused by the blockade of DRR, because GABAA and non-NMDA antagonists can prevent the generation of DRR induced by knee joint inflammation (Reeset al., 1995). DRR are action potentials fired antidromically back down the primary afferent fibers to the periphery which are thought to contribute to the release of neuropeptides in the joint resulting in potentiation of the inflammation and further afferent activation. Thus, we previously have proposed that DRR are one of the mechanisms involved in the vicious cycle of pain and inflammation (Sluka et al., 1995). One hypothesis of this study was that S-(+)-3-IBG may block DRR. If this is the case, pretreatment with this drug should reduce the amount of swelling after injection of kaolin and carrageenan.
S-(+)-3-IBG is a more potent analog of gabapentin, an anticonvulsant currently in clinical use as an add-on therapy in patients with partial seizures resistant to conventional therapies (seeGoa and Sorkin, 1993; Taylor, 1995 for reviews). Although gabapentin originally was designed as a GABA analog, it interacts with neither GABAA nor GABAB receptors (Bartoszyk and Reimann, 1985), although more recent studies show that its effects probably are mediated through the α2δ subunit of voltage-dependent calcium channels (Gee et al., 1996). Because recent electrophysiological studies have shown that voltage-dependent calcium channels are involved in the development of hyperalgesia after inflammation (Neugebauer et al., 1996; Nebe et al., 1997), the second hypothesis of the present study was thatS-(+)-3-IBG could block hyperalgesia after inflammation of the knee joint.
Therefore, the present study investigated whether spinal administration of S-(+)-3-IBG or its stereoisomer are effective in reducing the hyperalgesia and swelling observed after injection of kaolin and carrageenan into the knee joint of the rat. The S-(+)-3-IBG or its stereoisomer was administered either after the development of inflammation (post-treatment) or before the inflammation was induced (pretreatment). Both isomers of the drug were used to test whether the actions of this drug were stereospecific.
Methods
Placement of microdialysis fibers.
All experiments were approved by the Animal Care and Use Committee at our institution. Sixty male Sprague-Dawley rats (235–380 g) were anesthetized with sodium pentobarbital (Nembutal; 50 mg·kg−1i.p.). A microdialysis fiber [200 μm outside diameter (o.d.), 45,000 MW Cut-off, Hospal AN69] was coated with epoxy resin, except for a 2-mm section. A small midline incision was made in the back at the level of the last rib. The muscle was then removed from around the T13 vertebra and a hole drilled in both lateral aspects. The microdialysis fiber then was passed transversely through the dorsal horn of the spinal cord between lumbar segments L3 and L6 so that the permeable 2-mm section of the fiber lay in the dorsal horn. The microdialysis fiber was connected to PE20 tubing (Becton Dickinson, San Jose, CA) which then was tunneled under the skin to the nape of the neck. The fiber was stabilized with dental cement. aCSF was pumped through the tubing at a rate of 5 μl·min−1 for 1 h before the PE20 tubing was sealed and the animals allowed to recover. Once the rats were awake they were examined for motor deficits; any rat which had motor deficits was automatically excluded from the study. No other problems were encountered in animals implanted with microdialysis fibers by this method. Previous studies in this laboratory have shown no differences in behavioral scores measured before and after fiber implantation. Many of the fiber placements were checked histologically, and no differences in scores attributable to fiber placement were detected.
Behavioral testing and assessment of arthritis.
As a measure of heat hyperalgesia animals were tested for paw withdrawal to radiant heat according to the protocol of Hargreaves et al. (1988). On the day after fiber placement animals were housed in small lucite cubicles on an elevated glass plate. Radiant heat was applied to the plantar surface of the heel of the hindpaw until the rat lifted the paw. The time at which this occurred was considered the paw withdrawal latency. Both paws were tested independently at 5-min intervals, for a total of five trials. A mean of these five readings was used as the PWL. In the post-treatment group, the animals were tested before the induction of arthritis in the knee joint (baseline), 4 h postinduction and 1.5 h after drug infusion, i.e.,5.5 h after induction of arthritis. In pretreatmentrats, PWL was measured before administration of any drugs (baseline) and after the drug had been infused for 1.5 h (postdrug) at which time kaolin and carrageenan were injected into the knee joint. PWL was measured for a final time 4 h after induction of arthritis. The experimenter was naive to the expected actions of the drug.
A decrease in the PWL to noxious radiant heat in an animal with knee joint inflammation is indicative of secondary hyperalgesia. The term hyperalgesia is used to describe “an increased response to a stimulus which previously is normally painful” (Merskey and Bogduk, 1994) and secondary is used to illustrate that the hyperalgesia was measured distant to the site of injury (Lewis, 1942).
To estimate the temperature at which the rats withdrew their paws from the radiant heat source the temperature of the surface of the glass plate after heating was measured with a calibrated thermistor.
The circumference of the knee joint also was measured with a flexible tape measure before injection of kaolin and carrageenan (baseline). After arthritis was induced two independent observers also scored the rats to note the extent of guarding of the hindpaw of the arthritic limb. To quantify these changes, the animals were graded by a subjective pain-rating scale (0–5) modified from the scale described by Attal et al. (1990) where: 0 is normal, 1 is curling of the toes, 2 is eversion of the paw; 3 is partial weight bearing, 4 is non-weight bearing and guarding and 5 is avoidance of any contact with the hindlimb.
Induction of arthritis.
Rats were anesthetized briefly with sodium methohexital (Brevital; 60 mg·kg−1i.p.) after the baseline behavioral test (post-treatment group) or after infusion of the drug (pretreatment group). The knee joint then was injected with 3% kaolin and 3% carrageenan suspended in sterile saline (0.1 ml; pH 7.4) and flexed manually until the rat awoke (approximately 5–10 min).
Administration of drugs.
All drugs were dissolved in aCSF (pH 7.4, adjusted by bubbling with 95% CO2/5% O2) and infused through the microdialysis fiber at a rate of 5 μl·min−1. Based on previous in vitro estimates a maximum of about 1 to 10% of the drug passes from the dialysis fiber into the spinal cord (Sluka and Westlund, 1993c; Sluka et al., 1997).
The animals received either S-(+)-3-IBG,R-(−)-3-IBG or aCSF. In the post-treatmentgroup, drugs were infused at concentrations of 0.1, 0.9 and 10 mg·ml−1 (n = 6 for each treatment group). The most effective dose in the post-treatment group, 10 mg·ml−1, was used for thepretreatment group. Thus the pretreatment group received a single dose of 10 mg·ml−1 ofS-(+)-3-IBG, R-(−)-3-IBG or aCSF (n = 6 for each treatment group). The drugs were a gift from Parke-Davis and were synthesized at Parke-Davis Research Laboratories, a Division of Warner-Lambert (Ann Arbor, MI).
Statistical analysis.
A one-way analysis of variance was used to assess whether there was a dose-dependent effect after drug administration. A post hoc t test was carried out when appropriate. A P value of less than .05 was used to indicate significance for all comparisons unless otherwise specified.
The withdrawal response and circumference data were distributed normally (Kolmogorov-Smirnov test). Therefore, to assess the effects of the drug treatment on PWL and circumference after kaolin/carrageenan inflammation, pairwise comparisons with each control were made by paired t tests. Unpaired t tests were used to make comparisons between treatment groups at the same time point. Baseline values were normalized to 100% expressed as means ± S.E.M. for illustrations.
Nonparametric tests were used to analyze the behavioral score which is discontinuous. Pairwise comparisons within the same treatment group were made with Wilcoxon’s signed-rank test; comparisons between treatment groups were made with the Mann-Whitney U test.
Results
At the outset of the experiment the PWL and knee joint circumference were established for each rat. The mean PWL and knee joint circumference were 9.99 ± 0.22 sec (n = 60) and 5.4 ± 0.02 cm (n = 60), respectively. Calibration of the radiant heat source revealed that 9.99 sec after the radiant heat was started the temperature of the glass plate was approximately 50°C.
Effect of intra-articular injection of kaolin and carrageenan.
Four hours after injection of kaolin and carrageenan the PWL to noxious radiant heat decreased to 80% of the baseline value (table 1; n = 42), which indicates the presence of secondary hyperalgesia. This decrease was significant (paired t test, P < .01). Calibration of the radiant heat source revealed that after induction of arthritis the rats withdrew the paw of the arthritic limb when the glass plate reached a temperature of 45°C.
Four hours after inflammation of the knee joint there was a 15% increase in knee joint circumference compared with the baseline measurement. This increase was significant (P < .05, pairedt test, table 1). After inflammation there was also a change in the rats’ posture (decreased weight bearing on the swollen limb, and curling of the toes) reflected by the significantly increased spontaneous pain rating score (Wilcoxon, P < .05) (table 1). The mean spontaneous pain rating score 4 h after induction of arthritis was 2.2 ± 0.13.
Effect of S-(+)-3-IBG or R-(−)-3-IBG infused into the spinal cord after the development of acute arthritis.
Microdialysis infusion of S-(+)-3-IBG orR-(−)-3-IBG 4 h after injection of kaolin and carrageenan reduced the hyperalgesia to radiant heat normally observed after inflammation of the knee joint, in a dose-dependent manner (analysis of variance, P < .03 for each drug; table 2). The PWL values after infusion of the lowest dose of either drug were not different from those recorded 4 h after induction of arthritis (table 2; B − A,i.e., PWL after drug minus PWL at 4 h).
In contrast, infusion of the highest dose, 10 mg·ml−1, of either isomer of IBG through the microdialysis fiber 4 h after the knee joint was inflamed resulted in a return of the PWL to the baseline value (table 2, B − A; fig. 1, upper panel), whereas the PWL obtained with infusion of aCSF remained significantly reduced from baseline. The effect observed after microdialysis infusion of the 10 mg·ml−1 dose of S-(+)-3-IBG was significantly different from that observed after infusion of the 0.1 mg·ml−1 dose (post hoc t test, P < .02). Furthermore, although there was a tendency for S-(+)-enantiomer to be more effective at reducing the hyperalgesia to radiant heat than theR-(−)-enantiomer, this difference was not significant (unpaired t test, P > .1).
The spontaneous pain score also was significantly reduced by infusion of either stereoisomer (Wilcoxon test, P < .05). After infusion of the 10 mg·ml−1 concentration the paw posture was almost baseline, 0 (fig. 1, middle panel).
However, infusion of either S-(+)-3-IBG orR-(−)-3-IBG for 1.5 h, 4 h after the development of acute inflammation, did not reduce the amount of swelling; the circumference of the knee joint after drug infusion was not significantly different from the knee joint circumference in animals in which aCSF was infused (fig. 1, lower panel).
Effect of S-(+)-3-IBG or R-(−)-3-IBG infused through the spinal cord before the development of acute arthritis.
Microdialysis infusion of 10 mg·ml−1 of S-(+)-3-IBG or 10 mg·ml−1 of R-(−)-3-IBG or aCSF through the dorsal horn of the spinal cord alone did not change the PWL in the heat hyperalgesia test when compared with baseline values (fig. 2, upper panel).
The PWL of control rats, pretreated with aCSF, was reduced significantly 4 h after injection of kaolin and carrageenan (P < .01, paired t test). However, no secondary heat hyperalgesia developed in the rats in which 10 mg·ml−1 of S-(+)-3-IBG orR-(−)-3-IBG was infused through the spinal cord for 1.5 h before the injection of kaolin and carrageenan (fig. 2, top panel). Furthermore, pretreatment with S-(+)-3-IBG orR-(−)-3-IBG prevented the development of abnormal paw posture indicative of spontaneous pain (fig. 2, middle panel).
Infusion of S-(+)-3-IBG or R-(−)-3-IBG through the spinal cord for 1.5 h before the induction of arthritis significantly reduced (P < .05; unpaired t test) the amount of swelling typical after injection of kaolin and carrageenan by approximately 30%, when compared with rats in which aCSF was infused through the microdialysis fiber (fig. 2, bottom panel). Although theS-(+)-enantiomer tended to be more effective at reducing the amount of swelling than the R-(−)-enantiomer, the difference between the two was not significant (unpaired ttest, P > .07).
Discussion
The results from this study show that injection of kaolin and carrageenan into the knee joint of the rat results in an acute arthritis which is characterized by secondary heat hyperalgesia, swelling of the knee joint and spontaneous pain. Microdialysis infusion of S-(+)-3-IBG or R-(−)-3-IBG through the dorsal horn of the spinal cord after inflammation of the knee joint reduced the amount of heat hyperalgesia and the spontaneous pain observed in a dose-dependent manner, but did not alter the amount of swelling of the knee joint when compared with rats treated with aCSF. Microdialysis infusion of either enantiomer for 1.5 h before the injection of kaolin and carrageenan (pretreatment) did not change the baseline responses to the noxious radiant heat stimulus. However, pretreatment with either enantiomer reduced the amount of swelling observed and blocked the development of secondary hyperalgesia and spontaneous pain normally observed after inflammation.
The antihyperalgesic effects of S-(+)-3-IBG are consistent with other studies. Singh and colleagues (1996) demonstrated that systemic administration of the analog of S-(+)-3-IBG, gabapentin, reduced thermal hyperalgesia after carrageenan inflammation of the paw. Other studies have shown that gabapentin also can block hyperalgesia and allodynia in rat models of neuropathic pain (Xiao and Bennett, 1995). More recently Field et al. (1997)demonstrated that S-(+)-3-IBG is antihyperalgesic in the formalin test and in a carrageenan-induced inflammatory pain model.
The antihyperalgesic properties of gabapentin and its analog,S-(+)-3-IBG, are thought to be centrally mediated. Stanfa and colleagues (1997) recently demonstrated that carrageenan-induced sensitization of dorsal horn neurons can be blocked by gabapentin. A more recent behavioral study showed that a dose which blocked hyperalgesia when administered intrathecally was ineffective when administered peripherally (Field et al., 1997). Our study confirms that the antihyperalgesic property of S-(+)-3-IBG is, at least in part, centrally mediated because this method of drug administration, microdialysis infusion into the dorsal horn, has been shown to limit drug delivery to the spinal cord (Sluka et al., 1994).
An in vitro binding study showed that S-(+)-3-IBG and its stereoisomer bind to the α2δ subunit of voltage-dependent calcium channels, the R-(−)-3-IBG isomer binding with a 10-fold lower potency than its stereoisomer (Geeet al., 1996). However, in the present study we did not observe any differences in potency between the two isomers, although the S-(+)-enantiomer tended to be more effective. This result is a little surprising but may be because of the small difference in potencies between the two drugs and the relatively small sample size (n = 6). The α2δ subunit appears to be common to all voltage-dependent calcium channels (Isom et al., 1994; Hofmann et al., 1994) where it is thought to increase the expression of calcium channel complexes (Williams et al., 1992; Brust et al., 1993; Isomet al., 1994; Gurnett et al., 1996). Thus, the actions of S-(+)-3-IBG and its stereoisomer may involve more than one type of voltage-dependent calcium channel. For example, electrophysiological studies with calcium channel blockers have shown that both N- and L-type voltage-dependent calcium channels are involved the development of hyperalgesia after carrageenan-induced inflammation (Neugebauer et al., 1996). However, blockade of N- and L-type channels appears to be involved in the generation of pain evoked by noxious mechanical stimulation in normal tissue as well as in the mechanical hyperalgesia associated with inflammation (Neugebaueret al., 1996). More recent work from the same laboratory shows that blockade of P-type calcium channels has inconsistent effects upon pain evoked by noxious mechanical stimulation in normal tissue, but significantly reduces the mechanical hyperalgesia associated with inflammation (Nebe et al., 1997). Because voltage-dependent calcium channels are involved in the release of neurotransmitters,S-(+)-3-IBG and its stereoisomer may prevent the release of transmitters involved in the process of central sensitization, for example, glutamate and substance P.
Although a reduction in the amount of swelling was not observed if the drug was administered after the development of inflammation, administration of the drug before injection of the kaolin and carrageenan significantly reduced the amount of swelling observed at 4 h when compared with aCSF-treated animals. This finding is in contrast to previous studies with this drug which have not shown a reduction in swelling after carrageenan-induced inflammation of the paw (Singh et al., 1996). The route of administration (spinalversus systemic) and the duration of administration (continuous infusion for 1.5 h versus single dose) as well as the different model of inflammation may account for this difference in findings. Again, the α2δ subunit of voltage-dependent calcium channels probably is involved in mediating the anti-inflammatory action of this drug. Blocking the α2δ subunit of voltage-dependent calcium channels may prevent the inward depolarizing current of central primary afferent terminals from reaching threshold. This would prevent the initiation of DRR which have been shown to be present after inflammation, and are thought to be evoked as a result of excessive depolarization of the central terminals of primary afferent fibers (Sluka et al., 1995). Blockade of these DRRs has been proposed as a mechanism by which spinally mediated events can control the amount of swelling induced by injection of kaolin and carrageenan into the knee joint.
Post-treatment with S-(+)-3-IBG or its stereoisomer antagonized the development as well as the maintenance of heat hyperalgesia, whereas a reduction in swelling was observed only if the drug was administered before the onset of inflammation. This implies that the α2δ subunit of voltage-dependent calcium channels are involved in the maintenance of hyperalgesia. In contrast, the α2δ subunits of voltage-dependent calcium channels probably are involved only in the initiation and not the maintenance of peripheral inflammation.
In summary, S-(+)-3-IBG and R-(−)-3-IBG are effective antihyperalgesic agents, attenuating both secondary hyperalgesia and spontaneous pain behavior, when administered both before and after joint inflammation. In addition, when administered before injection of kaolin and carrageenan into the knee joint, this drug can attenuate joint inflammation through a central neurogenic mechanism.
Footnotes
-
Send reprint requests to: Karin N. Westlund High, Department of Anatomy and Neurosciences, University of Texas Medical Branch, 301 University Boulevard, Galveston, TX 77555-1069.
-
↵1 This work was supported by National Institutes of Health grant NS 32778.
- Abbreviations:
- PWL
- paw withdrawal latency
- S-(+)-3-IBG
- S-(+)-3-isobutylgaba
- R-(−)-3-IBG
- R-(−)-3-isobutylgaba
- S.E.M.
- standard error of the mean
- aCSF
- artificial cerebrospinal fluid
- DRR
- dorsal root reflex
- NMDA
- N-methyl-d-aspartate
- GABA
- γ-aminobutyric acid
- Received June 4, 1997.
- Accepted January 20, 1998.
- The American Society for Pharmacology and Experimental Therapeutics