(l )-Phenylglycine, but not necessarily other α2δ subunit voltage-gated calcium channel ligands, attenuates neuropathic pain in rats

Pain 125 (2006) 136–142 www.elsevier.com/locate/pain

(L)-Phenylglycine, but not necessarily other a2d subunit voltage-gated calcium channel ligands, attenuates neuropathic pain in rats James J. Lynch III *, Prisca Honore, David J. Anderson, William H. Bunnelle, Kathleen H. Mortell, Chengmin Zhong, Carrie L. Wade, Chang Z. Zhu, Hongyu Xu, Kennan C. Marsh, Chih-Hung Lee, Michael F. Jarvis, Murali Gopalakrishnan Neuroscience Research, Global Pharmaceutical Research and Development, Abbott Laboratories, 100 Abbott Park Road, Abbott Park, IL 60064, USA Received 25 January 2006; received in revised form 17 April 2006; accepted 3 May 2006

Abstract Gabapentin and pregabalin have been demonstrated, both in animal pain models and clinically, to be effective analgesics particularly for the treatment of neuropathic pain. The precise mechanism of action for these two drugs is unknown, but they are generally believed to function via initially binding to the a2d subunit of voltage-gated Ca2+ channels. In this study, we used a pharmacological approach to test the hypothesis whether high affinity interactions with the a2d subunit alone could lead to attenuation of neuropathic pain in rats. The anti-allodynic effects of gabapentin and pregabalin, along with three other compounds – (L)-phenylglycine, m-chlorophenylglycine and 3-exo-aminobicyclo[2.2.1]heptane-2-exo-carboxylic acid (ABHCA) – discovered to be potent a2d ligands, were tested in the rat spinal nerve ligation model of neuropathic pain. Gabapentin (Ki = 120 nM), pregabalin (180 nM) and (L)-phenylglycine (180 nM) were shown to be anti-allodynic, with respective ED50 values of 230, 90 and 80 lmol/kg (p.o.). (L)-Phenylglycine was as potent as pregabalin and equi-efficacious in reversing mechanical allodynia. In contrast, two ligands with comparable or superior a2d binding affinities, m-chlorophenylglycine (Ki = 54 nM) and ABHCA (150 nM), exhibited no anti-allodynic effects at doses of 30–300 lmol/kg (p.o.), although these compounds achieved substantial brain levels. The data demonstrate that, at least in the rat spinal nerve ligation model of neuropathic pain, (L)-phenylglycine has an anti-allodynic effect, but two equally potent a2d subunit ligands do not. These results suggest that additional mechanisms, besides a2d interactions, may contribute to the effects of compounds like gabapentin, pregabalin and (L)-phenylglycine in neuropathic pain.  2006 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. Keywords: Gabapentin; Pregabalin; Phenylglycine; Allodynia; Analgesia; Brain; Spinal nerve ligation

1. Introduction Gabapentin (Neurontin) is indicated, in the United States, for the management of postherpetic neuralgia and as an adjunctive therapy in the treatment of partial seizures in patients with epilepsy. Although originally developed as an antiepileptic, attention has been increasingly focused on its potential for the treatment of neuro*

Corresponding author. Tel.: +1 847 937 7664; fax: +1 847 938 5286. E-mail address: [email protected] (J.J. Lynch III).

pathic pain. More recently, pregabalin (Lyrica) has demonstrated efficacy in preclinical models of neuropathic pain, and it is approved for the management of diabetic peripheral neuropathy and postherpetic neuralgia in humans. The mechanism(s) of action for the analgesic effects of gabapentin and pregabalin is unknown; however, it has been hypothesized that binding to the a2d subunit of voltage-gated Ca2+ channels may play an important role. The evidence supporting this hypothesis includes the following observations: (i) both gabapentin and pregabalin have been shown to selectively bind to the a2d

0304-3959/$32.00  2006 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.pain.2006.05.012

J.J. Lynch III et al. / Pain 125 (2006) 136–142

137

subunit, which could result in inhibition of calcium currents and neurotransmitter release (Gee et al., 1996; Bryans and Wustrow, 1999; Taylor, 2004); (ii) R217A mice, in which the a2d-1 subunit has been altered to disrupt gabapentin and pregabalin binding, exhibit diminished responsiveness to the analgesic effects of pregabalin in pain assays (Bramwell et al., 2004; Taylor, 2004); and (iii) up-regulation of a2d-1, development of mechanical allodynia, and gabapentin sensitivity all appear to be positively correlated in rat models of mechanical allodynia (Luo et al., 2002; Li et al., 2004). While the a2d subunit has thus been implicated in the analgesic effects of gabapentin and pregabalin (Taylor, 2004; Frampton and Foster, 2005), whether analgesia results exclusively from a2d subunit interactions remains to be determined. Urban et al. (2005) reported that the stereoselectivity of two gabapentin analogs at the a2d subunit did not completely account for the mechanism of action of gabapentin in various animal pain models. Evidence has also been presented to suggest that, in addition to high affinity a2d binding, interaction with the L neutral amino acid transporter system (to facilitate central nervous system access) was necessary for the in vivo effects of several compounds (Belliotti et al., 2005). The recent synthesis of a series of a-amino acid ligands that display a range of binding affinities provides an additional strategy to examine the role of a2d interactions in antinociception (Mortell et al., 2006). In this study, we used a pharmacological approach to test the hypothesis whether high affinity interactions with the a2d subunit alone could lead to attenuation of neuropathic pain in rats. The selective a2d subunit ligands utilized had binding affinities and brain levels comparable, or superior, to those of gabapentin and pregabalin. Our results demonstrate that, in the rat spinal nerve ligation model of neuropathic pain, the ad ligand, (L)-phenylglycine, had robust anti-allodynic effects while two equally potent a2d subunit ligands did not. The data suggest that potent a2d subunit binding interactions alone do not necessarily translate into anti-allodynic effects, and that additional mechanisms likely contribute to the antinociceptive effects of compounds like gabapentin, pregabalin and (L)-phenylglycine.

Studies were carried out in accordance with guidelines outlined by the Committee for Research and Ethical Issues of IASP and reviewed and approved by the Institutional Animal Care and Use Committee of Abbott Laboratories. Male Sprague–Dawley rats were purchased from Charles River Laboratories (Wilmington, MA) and were housed in AAALAC-approved facilities at Abbott Laboratories in a temperature-regulated environment with lights on between 06:00 and 18:00 h. Food and water were available ad libitum, except that the animals were food deprived overnight prior to their day of analgesia testing and dosing for pharmacokinetic studies. Gabapentin (hydrochloric acid salt), pregabalin (free base), (L)-phenylglycine (hydrochloric acid salt) and m-chlorophenylglycine (free base; racemic) were synthesized at Abbott Laboratories. 3-exo-Aminobicyclo[2.2.1]heptane-2exo-carboxylic acid (ABHCA; free base; racemic) was purchased from Acros Organics (Belgium). On the day of dosing, the compounds were dissolved as in an aqueous mixture of 30– 100% PEG400 (Aldrich Chemical Co., Inc., Milwaukee, WI) and administered orally at a volume of 2–5 ml/kg via a gastric intubation needle. Brain and plasma concentrations of compounds were assessed using a satellite group of animals.

2. Materials and methods

2.3. Behavioral assessment

2.1. Radioligand binding

2.3.1. Spinal nerve ligation model Rats (110–140 g body weight) had unilateral (left side) tight ligation surgery of the L5 and L6 spinal nerves performed as previously described by Kim and Chung (1992). Rats were anesthetized with halothane/oxygen anesthesia (5% for induction, 2–3% for maintenance). Using aseptic techniques, the L5 and L6 spinal nerves were tightly ligated with 5–0 silk suture distal to the dorsal root ganglion. Care was taken to avoid ligating the L4 spinal nerve. Following spinal nerve ligation, a minimum of 7 days of recovery was allowed prior to testing, and only rats of less than approximately 30 days post-surgery were utilized.

2.1.1. Membrane preparation Crude mouse cortical membranes were solubilized using Tween 20 via a procedure similar to that described by Gee et al. (1996). Specifically, the cerebral cortices from frozen mouse brains (Pel Freez Biologicals, Rogers, AR) were homogenized at 4 C in 50 ml of Hepes buffer (10 mM Hepes, 1 mM EDTA, and 1 mM EGTA, pH 7.4) with 0.32 M sucrose. The homogenate was centrifuged at 1000g for 10 min, the pellets discarded and the supernatant spun at 30,000g for 20 min. The resulting pellets

were resuspended in Hepes buffer and centrifuged at 30,000g for 20 min. The final pellets were resuspended in 20 ml of Hepes buffer with 25% glycerol and 5 ml of 2% Tween 20 and were stirred for 60 min on ice. After centrifugation at 75,000g for 90 min, the soluble fractions were frozen at 70 C prior to use. 2.1.2. Binding assay (L)-[3H]leucine binding to mouse cortical membranes was conducted as previously described with minor modifications (Brown et al., 1998). Membranes were incubated in a total assay volume of 0.25 ml containing 20 nM (L)-[3H]leucine (59 Ci/mmol, NEN Perkin-Elmer Life Sciences, Boston, MA) in 10 mM Tris–Cl, pH 7.5, for 120 min at room temperature. Specific radioligand binding was defined by subtracting the non-specific binding, as defined by the inclusion of 100 lM L-leucine, from total binding. Binding was terminated by rapid vacuum filtration through glass fiber filter plates (Millipore, Bedford, MA) presoaked with 0.3% polyethylenimine using a 96-well filtration apparatus (Packard Instruments, Meriden, CT). Filters were rapidly rinsed with 2 ml of ice-cold 50 mM Tris–Cl, pH 7.5, and radioactivity bound to the filters was assessed using a Packard TopCount. 2.2. Animals, compounds and dosing

138

J.J. Lynch III et al. / Pain 125 (2006) 136–142

2.3.2. Mechanical allodynia testing Mechanical allodynia was measured using calibrated (force; g) von Frey filaments (Stoelting, Wood Dale, IL) according to an up–down procedure (Chaplan et al., 1994). Briefly, the rats were placed into individual Plexiglass containers (20 · 12.5 · 20 cm) with wire mesh floors, which allowed access to the ventral side of the paw. A von Frey filament was slowly presented perpendicularly to the plantar surface of the hind paw, and held in this position for 6–8 s with enough force to cause a slight buckle in the filament. Positive responses included sharp withdrawal in reaction to the stimulus, or flinching behavior immediately after removal of the stimulus. The 50% withdrawal threshold was determined using the up–down method of Dixon (Dixon, 1980), and the value reported for each hind paw tested was the mean for two readings. Compounds were administered p.o., and their effects on mechanical allodynia were determined over time by measuring hind paw withdrawal thresholds at the time points indicated. 2.4. Blood plasma and brain concentrations For the determination of plasma and brain concentrations of the parent compound, naı¨ve rats were dosed with 100 lmol/kg of the compounds, p.o., and sacrificed at 15, 45, and 120 min post-dosing. For the determination of plasma concentrations, blood was collected into heparinized tubes and then centrifuged, and the separated plasma was frozen at 20 C until analysis. For the determination of brain concentrations, the rats were decapitated at the various time points, and the brains were immediately removed and rapidly freed from blood vessels as much as possible. The resulting brain tissues were immediately frozen at 20 C, weighed and homogenized, and then the homogenate was stored at 20 C. For analysis, compounds were extracted from the samples via liquid–liquid extraction and were quantified by liquid chromatography/mass spectroscopy. 2.5. Data analysis and statistics In the radioligand binding experiments, seven log concentrations of the compounds were tested in duplicate, and IC50 values were determined by non-linear regression using GraphPad Prism (v 3.02; GraphPad Software, Inc., San Diego, CA). Ki values were calculated from the IC50 values using the Cheng–Prusoff equation, in which Ki = IC50/(1 + ([Ligand]/ Kd)). Under the present assay conditions, the Kd of (L)-[3H]leucine was determined to be 111 nM (n = 3). Data were expressed as means ± SEM. In the behavioral experiments, ED50 values were estimated by linear regression using GraphPad Prism, and statistical comparisons were made using the Kruskal–Wallis test followed by Dunn’s multiple comparisons test. Statistical significance was determined at P < 0.05.

3. Results (L)-[3H]leucine binding to solubilized mouse cortex was displaced by both gabapentin and pregabalin with Ki values of 120 and 180 nM, respectively. These values are comparable to those previously reported for displacement of [3H]gabapentin binding to pig brain

membranes (Belliotti et al., 2005). (L)-Phenylglycine, m-chlorophenylglycine and ABHCA also displaced binding, with Ki values of 180, 54 and 150 nM, respectively. m-Chlorophenylglycine was the most potent among the five compounds tested. The binding affinities (and SEM values) are summarized in Table 1. Oral administration of gabapentin and pregabalin to rats with spinal nerve ligation significantly increased ipsilateral paw withdrawal thresholds (i.e., those for the left hind paw, which was ipsilateral to the spinal nerve ligation) at 60 min post-dosing. As shown in Fig. 1, both gabapentin and pregabalin exhibited dose-dependent efficacy, with pregabalin (ED50 = 90 lmol/kg) approximately 3-fold more potent than gabapentin (ED50 = 230 lmol/kg). Pharmacokinetic studies determined that gabapentin (dosed at 100 lmol/kg (p.o.)) reached average plasma levels of 10.77 lg/ml at 2 h post-dosing, at which time total brain levels of gabapentin were approximately 3-fold lower (3.78 lg/g, or 22 lM) (see Table 2). Pregabalin, on the other hand, reached maximal plasma levels at 15 min post-dosing, while its brain levels reached highest levels (2.22 lg/g, or 14 lM) at 2 h post-dosing. (L)-Phenylglycine was found to be as potent and efficacious as pregabalin in attenuating mechanical allodynia, with an ED50 value of 80 lmol/kg. Maximal plasma and brain levels of (L)-phenylglycine reached 15.62 lg/ml and 6.30 lg/g (or 14 lM), respectively: values that were comparable to those of gabapentin and pregabalin. In contrast to (L)-phenylglycine, pregabalin and gabapentin, doses of 30–300 lmol/kg (p.o.) m-chlorophenylglycine and ABHCA failed to affect paw withdrawal thresholds (Fig. 1). In order to eliminate pretreatment time as a factor for the lack of effects of compounds like m-chlorophenylglycine in attenuating mechanical allodynia, a time course study was performed using m-chlorophenylglycine and (L)-phenylglycine, the latter as a positive control. Again, paw withdrawal thresholds were not significantly altered in animals dosed with 300 lmol/kg m-chlorophenylglycine at the post-dosing times of 15, 45 and 120 min (Fig. 2). In contrast, significant efficacy was observed at all three time points tested after dosing with 300 lmol/kg (L)-phenylglycine. To assess whether adequate brain levels of m-chlorophenylglycine and ABHCA were achieved following oral dosing, satellite groups of rats were dosed, and both brain and plasma concentrations were determined at the same time points as were utilized in the time course study shown in Fig. 2. The data presented in Table 2 demonstrate that m-chlorophenylglycine and ABHCA attained maximal brain levels of 3.98 and 2.68 lg/g (or 21 and 17 lM), respectively, after oral administration, and these concentrations were approximately 389- and 113-fold higher than their respective Ki values. Indeed,

J.J. Lynch III et al. / Pain 125 (2006) 136–142

139

Table 1 (L)-[3H]leucine binding affinity and mechanical allodynia ED50 values Name

Ki (nM)

Structure

pKia (SEM)

ED50b (lmol/kg (p.o.))

NH2 Gabapentin

CO2H

6.92 (0.09)

230

180

6.74 (0.09)

90

180c

6.74 (0.21)

80

54c

7.27 (0.05)

>300

150

6.81 (0.11)

>300

NH2

CH3

Pregabalin

120c

CO2H

H3C

NH2 (L)-Phenylglycine

CO2H NH2 Cl

m-Chlorophenylglycine

CO2H

CO2H

ABHCA

NH2 ABHCA, 3-exo-aminobicyclo[2.2.1]heptane-2-exo-carboxylic acid. a n P 3. b n = 6–12 rats/dose. c Binding affinity values from Mortell et al. (2006).

comparable brain levels and ratios were observed with (L)-phenylglycine (14 lM; 78-fold), gabapentin (22 lM; 183-fold) and pregabalin (14 lM; 78-fold). 4. Discussion The primary mechanism of action for the analgesic effects of gabapentin and pregabalin remains unknown;

Paw Withdrawal Threshold (g)

15

10

* vehicle (L)-phenylglycine pregabalin gabapentin m m-chlorophenylglycine ABHCA A-784534

*

** *

*

5

0

Vehicle 10

100

1000

Dose (μ μmol/kg, p.o.) Fig. 1. Ipsilateral paw withdrawal thresholds to mechanical stimulation in spinal nerve ligation rats. Studies were conducted 60 min after treatment with varying doses (as indicated) of gabapentin, pregabalin, (L)-phenylglycine, m-chlorophenylglycine or ABHCA, p.o. Range (delineated by two dotted horizontal lines) represents mean ipsilateral and contralateral paw values for baseline data. Means ± SEM; n = 6–12 rats/group for each treatment group, and n = 34 for vehicle. *P < 0.05 versus vehicle control.

however, it is hypothesized that these effects are mediated via binding to the a2d subunit of the voltagegated Ca2+ channel. This interaction could, in principle, modify voltage-gated Ca2+ channel activity, leading to a decrease in Ca2+ influx, which, in turn, could decrease synaptic release of excitatory neurotransmitters. This hypothesis would suggest that other potent a2d subunit ligands would also be analgesic, and this has been demonstrated to be true with select analogs structurally related to gabapentin and pregabalin (Field et al., 1997; Sterns et al., 2004; Belliotti et al., 2005). Data from the current study and from other published reports, however, suggest that the a2d hypothesis may not account for the primary mechanism of action for the analgesic effects of compounds like gabapentin and pregabalin. In the current study, and in agreement with previously published results, both gabapentin and pregabalin attenuated mechanical allodynia in spinal nerve ligation rats (Field et al., 1999, 2000; Urban et al., 2005). In addition, our studies demonstrate for the first time that (L)-phenylglycine, another potent a2d subunit ligand with structural similarity to gabapentin and pregabalin, was anti-allodynic (Table 1; Figs. 1 and 2). (L)-Phenylglycine exhibited in vivo efficacy (ED50 value of 80 lmol/kg (p.o.)), as well as in vitro affinity (Ki value of 180 nM), comparable to that of pregabalin. As reported elsewhere, (L)-phenylglycine was also analgesic in the rat complete Freund’s adjuvant model of chronic inflammatory pain with an efficacy profile similar to that of gabapentin (Mortell et al., 2006). Because some

140

J.J. Lynch III et al. / Pain 125 (2006) 136–142

Table 2 Comparison of brain and plasma levels for gabapentin, pregabalin and other compounds at 0.25, 0.75 and 2 h following oral dosing (100 lmol/kg) in rats Name

Time post-dosing (h)

Plasma (lg/ml)

Brain (lg/g)

Brain/plasma ratio

Gabapentin

0.25 0.75 2

4.23 ± 0.51 10.83 ± 0.50 10.77 ± 0.58

0.62 ± 0.09 3.41 ± 0.28 3.78 ± 0.16

0.15 0.31 0.35

Pregabalin

0.25 0.75 2

18.42 ± 1.85 NA 9.39 ± 0.24

0.64 ± 0.06 NA 2.22 ± 0.06

0.03 NA 0.24

(L)-Phenylglycine

0.25 0.75 2

15.62 ± 1.77 10.03 ± 0.61 4.12 ± 0.69

5.88 ± 0.39 6.30 ± 0.55 3.35 ± 0.33

0.38 0.63 0.81

m-Chlorophenylglycine

0.25 0.75 2

7.59 ± 1.81 5.18 ± 1.16 1.65 ± 0.29

3.98 ± 0.40 1.76 ± 0.06 0.42 ± 0.08

0.52 0.34 0.25

ABHCA

0.25 0.75 2

14.91 ± 0.32 15.13 ± 0.36 10.20 ± 0.47

0.74 ± 0.21 1.56 ± 0.40 2.68 ± 0.12

0.05 0.10 0.26

NA, not available. ABHCA, 3-exo-aminobicyclo[2.2.1]heptane-2-exo-carboxylic acid. Plasma and brain concentrations are expressed as means ± SEM (n = 3).

phenylglycine derivatives are known antagonists of metabotropic glutamate receptors (mGluR) (Miller et al., 2003), we investigated the functional activity of (L)-phenylglycine, as well as m-chlorophenylglycine, with human mGluR1 and mGluR5 expressed in stable cell lines and determined that these compounds were inactive at concentrations from 1 pM to 100 lM (unpublished results). In contrast to the anti-allodynic effects observed with gabapentin, pregabalin and (L)-phenylglycine in the current study, in vivo efficacy was not observed with two

Paw Withdrawal Threshold (g)

15

*

*

*

10 vehicle 300 μ umol/kg (L)-phenylglycine 300 μ umol/kg m-chlorophenylglycine m-chlorophenylglycine

5

0

0

0.5

1

1.5

2

Time Post-Dosing (hr) Fig. 2. Time course study of ipsilateral paw withdrawal threshold after treatment with 300 lmol/kg (L)-phenylglycine, 300 lmol/kg m-chlorophenylglycine, or vehicle, p.o. Range (delineated by two dotted horizontal lines) represents mean ipsilateral and contralateral paw values for baseline data. Means ± SEM; n = 5–6 for each treatment group and vehicle. *P < 0.05 versus appropriate vehicle control.

other potent ligands, m-chlorophenylglycine and ABHCA, with affinities (Ki values of 54 and 150 nM, respectively) comparable, or superior, to that of gabapentin or pregabalin (Fig. 1). The pharmacokinetics of m-chlorophenylglycine and ABHCA (which would include their transportation and distribution into the central nervous system, and as demonstrated by adequate brain levels) as well as their pretreatment time were unlikely to have contributed to their lack of analgesic effects (Table 2; Fig. 2). Therefore, although m-chlorophenylglycine and ABHCA were determined to be potent a2d ligands with plasma and brain concentrations and ratios comparable to those of the other three compounds tested, m-chlorophenylglycine and ABHCA were not efficacious in this model of neuropathic pain, thus suggesting that binding to a2d alone does not necessarily result in analgesia. It should be noted, however, that target occupancy cannot necessarily be assumed based solely upon relatively high brain concentrations: e.g., high brain levels would also be observed with compounds that readily cross the blood–brain barrier but either were unbound or were bound to non-a2d-related sites. Nevertheless, the relatively potent in vitro binding affinities observed with the compounds tested in this study suggest it unlikely that a2d subunit binding did not occur once the compounds penetrated into the central nervous system. Over the years, based on functional studies, a number of non-a2d-mediated mechanisms of action have been proposed for compounds such as gabapentin. For example, a body of evidence, including a recent publication using a rat spinal nerve ligation model of neuropathic

J.J. Lynch III et al. / Pain 125 (2006) 136–142

pain similar to that utilized in the current study (Urban et al., 2005), exists to indicate that gabapentin is a c-aminobutyric acid subtype B (GABAB) selective receptor agonist and that this mechanism of action may underlie gabapentin’s analgesic effects (Bertrand et al., 2001, 2003; Ng et al., 2001). In addition, it has been demonstrated that GABA levels are increased in the brains of rodents and humans after administration of gabapentin, and that the nitric oxide pathway may play a role in the analgesic effects of gabapentin in some animal pain models (Loscher et al., 1991; Petroff et al., 1996; Dixit and Bhargava, 2002). Gabapentin has also been shown to be analgesic (possibly via its inhibition of glutamate release) during chronic constriction injury in the rat, a neuropathic pain model that exhibits little to no up-regulation of a2d subunit in both dorsal root ganglion and spinal cord tissue (Luo et al., 2002; Coderre et al., 2005). With the discovery of two gabapentin analogs, (1S,3R)3-methylgabapentin and (1R,3R)3-methylgabapentin, that differ in their a2d subunit binding affinity by greater than 200-fold, it was originally demonstrated that a2d subunit binding affinity had a positive correlation with analgesic efficacy in two animal models of neuropathic pain (Bryans et al., 1998; Field et al., 2000). However, it has now been shown, in a number of other animal pain models, that the stereoselectivity of 3-methylgabapentin for the a2d subunit is unlikely related to the analgesic effects of these two diastereoisomers and that the role of the a2d subunit in gabapentin-induced analgesia still remains ambiguous (Urban et al., 2005). It has additionally been proposed that the mechanism of action for gabapentin is via its interaction with the L neutral amino acid (LNAA) transporter system, which is utilized for the transport of endogenous agents such as (L)-leucine (Thurlow et al., 1993, 1996a,b; Su et al., 1995). The hypothesis follows that the binding of gabapentin to this transporter inhibits the uptake of glycine, which may in turn result in anticonvulsive/analgesic effects (Thurlow et al., 1996a). However, the LNAA transporter hypothesis, the strength of which lies mainly on pharmacological results with the transporter-selective ligand 2-( )-endoamino-bicycloheptane-2-carboxylic acid (BCH), was largely abandoned after molecular biology results were reported on the purification and sequencing of the a2d subunit, which was then hypothesized to be the site of action for gabapentin (Gee et al., 1996; Thurlow et al., 1996a; Brown et al., 1998). Interestingly, recent data suggest that interactions at both the LNAA transporter and the a2d subunit may be necessary, such that gabapentin and its analogs require both transport across the blood–brain barrier as well as binding to the a2d subunit of the voltage-gated calcium channel, respectively, to be effective (Belliotti et al., 2005). It should be noted that although we utilized (L)-[3H]leucine for displacement analysis in the present study, the Ki values of the compounds tested directly

141

using [3H]gabapentin as the radioligand were comparable to those obtained with (L)-[3H]leucine (gabapentin Ki = 54 nM, pregabalin 130 nM, (L)-phenylglycine 85 nM, and m-chlorophenylglycine 24 nM; unpublished observations). Thus, our data suggest that the compounds evaluated in the current study not only had the opportunity to bind to the LNAA transporter, but also to the a2d subunit. In the present study, we did not address the functionality of these ligand interactions to determine whether differential functional effects (as opposed to simply binding affinity) may have contributed to the differences found in the in vivo efficacy for the ligands tested. In future studies, it would be of interest to further investigate the basis of functional distinctions of compounds such as (L)-phenylglycine, gabapentin and pregabalin versus m-chlorophenylglycine and ABHCA, whose potent a2d binding affinities did not translate to in vivo efficacies. In summary, results from our studies suggest that the small-molecule, amino acid-derivative (L)-phenylglycine has a pharmacologic and analgesic efficacy profile similar to those of pregabalin and gabapentin. In addition, the data suggest that the mechanism of action for the anti-allodynic effects of (L)-phenylglycine, pregabalin and gabapentin, at least in the rat spinal nerve ligation model of neuropathic pain, may be unrelated to their potent a2d subunit binding affinity.

Acknowledgements The authors gratefully acknowledge Loan Miller for providing mGluR selectivity data. Funding was supported by Abbott Laboratories.

References Belliotti TR, Capiris T, Ekhato V, Kinsora JJ, Field MJ, Heffner TG, et al. Structure–activity relationships of pregabalin and analogues that target the a2-d protein. J Med Chem 2005;48:2294–307. Bertrand S, Ng GYK, Purisai MG, Wolfe SE, Severidt MW, Nouel D, et al. The anticonvulsant, antihyperalgesic agent gabapentin is an agonist at brain c-aminobutyric acid type B receptors negatively coupled to voltage-dependent calcium channels. J Pharmacol Exp Ther 2001;298:15–24. Bertrand S, Nouel D, Morin F, Nagy F, Lacaille J-C. Gabapentin actions on Kir3 currents and N-type Ca2+ channels via GABAB receptors in hippocampal pyramidal cells. Synapse 2003;50:95–109. Bramwell SR, Cox PJ, Melrose H, Stott E, Wain L, Corradini L, et al. The analgesic actions of pregabalin are mediated through its binding to the a2d-1 subunit of voltage gated calcium channels. Abstract Viewer/Itinerary Planner. Washington, DC: Society for Neuroscience 2004; Program No. 523.19. Brown JP, Dissanayake VUK, Briggs AR, Milic MR, Gee NS. Isolation of the [3H]gabapentin-binding protein/a2d Ca2+ channel subunit from porcine brain: development of a radioligand binding assay for a2d subunits using [3H]leucine. Anal Biochem 1998;255:236–43.

142

J.J. Lynch III et al. / Pain 125 (2006) 136–142

Bryans JS, Davies N, Gee NS, Dissanayake VUK, Ratcliffe GS, Horwell DC, et al. Identification of novel ligands for the gabapentin binding site on the a2d subunit of a calcium channel and their evaluation as anticonvulsant agents. J Med Chem 1998;41:1838–45. Bryans JS, Wustrow DJ. 3-Substituted GABA analogs with central nervous system activity: a review. Med Res Rev 1999;19:149–77. Chaplan SR, Bach FW, Pogrel JW, Chung JM, Yaksh TL. Quantitative assessment of tactile allodynia in the rat paw. J Neurosci Methods 1994;53:55–63. Coderre TJ, Kumar N, Lefebvre CD, Yu JSC. Evidence that gabapentin reduces neuropathic pain by inhibiting the spinal release of glutamate. J Neurochem 2005;94:1131–9. Dixit RK, Bhargava VK. Neurotransmitter mechanisms in gabapentin antinociception. Pharmacol 2002;65:198–203. Dixon WJ. Efficient analysis of experimental observations. Ann Rev Pharmacol Toxicol 1980;20:441–62. Field MJ, Bramwell S, Hughes J, Singh L. Detection of static and dynamic components of mechanical allodynia in rat models of neuropathic pain: are they signalled by distinct primary sensory neurones? Pain 1999;83:303–11. Field MJ, Hughes J, Singh L. Further evidence for the role of the a2d subunit of voltage dependent calcium channels in models of neuropathic pain. Br J Pharmacol 2000;131:282–6. Field MJ, Oles RJ, Lewis AS, McCleary S, Hughes J, Singh L. Gabapentin (neurontin) and S-(+)-3-isobutylgaba represent a novel class of selective antihyperalgesic agents. Br J Pharmacol 1997;121:1513–22. Frampton JE, Foster RH. Pregabalin in the treatment of postherpetic neuralgia. Drugs 2005;65:111–8. Gee NS, Brown JP, Dissanayake VUK, Offord J, Thurlow R, Woodruff GN. The novel anticonvulsant drug, gabapentin (Neurontin), binds to the a2d subunit of a calcium channel. J Biol Chem 1996;271:5768–76. Li C-H, Song YH, Higuera ES, Luo ZD. Spinal dorsal horn calcium channel a2d-1 subunit upregulation contributes to peripheral nerve injury-induced tactile allodynia. J Neurosci 2004;24:8494–9. Loscher W, Honack D, Taylor CP. Gabapentin increases aminooxyacetic acid-induced GABA accumulation in several regions of rat brain. Neurosci Lett 1991;128:150–4. Luo ZD, Calcutt NA, Higuera ES, Valder CR, Song Y-H, Svensson CI, et al. Injury type-specific calcium channel a2d-1 subunit up-regulation in rat neuropathic pain models correlates with antiallodynic effects of gabapentin. J Pharmacol Exp Ther 2002;303:1199–205.

Miller JC, Howson PA, Conway SJ, Williams RV, Clark BP, Jane DE. Phenylglycine derivatives as antagonists of group III metabotropic glutamate receptors expressed on neonatal rat primary afferent terminals. Br J Pharmacol 2003;139:1523–31. Mortell KH, Anderson DJ, Lynch III JJ, Nelson SL, Sarris K, McDonald H, et al. Structure–activity relationships of a-amino acid ligands for the a2d subunit of voltage-gated calcium channels. Bioorg Med Chem Lett 2006;16:1138–41. Ng GYK, Bertrand S, Sullivan R, Ethier N, Wang J, Yergey J, et al. c-Aminobutyric acid type B receptors with specific heterodimer composition and postsynaptic actions in hippocampal neurons are targets of anticonvulsant gabapentin action. Mol Pharmacol 2001;59:144–52. Petroff OAC, Rothman DL, Behar KL, Lamoureux D, Mattson RH. The effect of gabapentin on brain gamma-aminobutyric acid in patients with epilepsy. Ann Neurol 1996;39:95–9. Sterns BA, Anker N, Arruda JM, Campbell BT, Chen C, Cramer M, et al. Synthesis and biological evaluation of 6-aryl-6H-pyrrolo[3, 4-d]pyridazine derivatives: high-affinity ligands to the a2d subunit of voltage gated calcium channels. Bioorg Med Chem Lett 2004;14:1295–8. Su T-Z, Lunney E, Campbell G, Oxender DL. Transport of gabapentin, a c-amino acid drug, by system L a-amino acid transporters: a comparative study in astrocytes, synaptosomes, and CHO cells. J Neurochem 1995;64:2125–31. Taylor CP. The biology and pharmacology of calcium channel a2-d proteins. Pfizer satellite symposium to the 2003 Society for Neuroscience meeting. Sheraton New Orleans Hotel New Orleans, LA November 10, 2003. CNS Drug Rev 2004;10:183–8. Thurlow RJ, Brown JP, Gee NS, Hill DR, Woodruff GN. [3H]Gabapentin may label a system-L-like neutral amino acid carrier in brain. Eur J Pharmacol Mol Pharmacol 1993;247:341–5. Thurlow RJ, Hill DR, Woodruff GN. Comparison of the uptake of [3H]gabapentin with the uptake of L-[3H]leucine into rat brain synaptosomes. Br J Pharmacol 1996a;118:449–56. Thurlow RJ, Hill DR, Woodruff GN. Comparison of the autoradiographic binding distribution of [3H]gabapentin with excitatory amino acid receptor and amino acid uptake site distributions in rat brain. Br J Pharmacol 1996b;118:457–65. Urban MO, Ren K, Park KT, Campbell B, Anker N, Stearns B, et al. Comparison of the antinociceptive profiles of gabapentin and 3methylgabapentin in rat models of acute and persistent pain: implications for mechanism of action. J Pharmacol Exp Ther 2005;313:1209–16.