Interaction of local anaesthetics with recombinant μ, κ, and δ-opioid receptors expressed in Chinese hamster ovary cells

British Journal of Anaesthesia 85 (5): 740±6 (2000)

LABORATORY INVESTIGATIONS Interaction of local anaesthetics with recombinant m, k, and dopioid receptors expressed in Chinese hamster ovary cells K. Hirota1, H. Okawa1, B. L. Appadu1, D. K. Grandy2 and D. G. Lambert1* 1

University Department of Anaesthesia, Leicester Royal In®rmary, Leicester, LE1 5WW, UK. 2 Oregon Health Sciences University, Portland, OR, USA *Corresponding author Local anaesthetics potentiate epidural or intrathecal opioid analgesia via a poorly de®ned mechanism. In this study, we have examined the interaction of local anaesthetics (lidocaine, bupivacaine and its optical isomers, tetracaine, procaine and prilocaine) with recombinant m-, k-, and d-opioid receptors expressed in Chinese hamster ovary cells (CHO-m, k, and d, respectively). Lidocaine produced a concentration-dependent displacement of radiolabelled opioid antagonist [3H]diprenorphine ([3H]DPN) binding with the following rank order of inhibitor constant (Ki): k (210 mM) > m (552 mM) > d (1810 mM). Procaine, prilocaine, tetracaine and bupivacaine also displaced [3H]DPN binding in CHO- m with Ki values of 244, 204, 43 and 161 mM respectively. Lidocaine produced a concentration-dependent and naloxone-insensitive inhibition of cAMP formation in all cell lines including untransfected cells. Concentration producing 50% inhibition of maximum was m, 1.32 mM; k, 2.41 mM; d, 1.27 mM; untransfected, 2.78 mM. When lidocaine (300 mM) was co-incubated with spiradoline (k-selective) and [D-Ala2, MePhe4, Gly(ol)5] enkephalin (DAMGO m-selective) in CHO-k and m cells we did not observe an additive interaction for cAMP formation. In contrast, there was an apparent inhibitory action of the combination at the k receptor. This study suggests that clinical concentrations of local anaesthetics interact with m and k but not d opioid receptors. As there was no synergism between local anaesthetics and opioids we suggest that the interaction of these agents in the clinical setting does not occur at the cellular level. Br J Anaesth 2000; 85: 740±6 Keywords: anaesthetics local; measurement techniques, radioligand binding; metabolism, cAMP Accepted for publication: June 9, 2000

Clinically, local anaesthetics and opioids are commonly used in combination and are often administered into the epidural or intrathecal space in an attempt to manage postoperative pain and the pain of childbirth. These two classes of agents are reported to produce synergistic antinociceptive actions1±7 such that addition of local anaesthetic can reduce the dose of opioid used and hence reduce the number and severity of any adverse reactions. Local anaesthetic agents produce a use dependent block of voltage dependent Na+ channels, and hence, reduce axonal conduction.8 Opioid receptors are classi®ed as m, k and d and all have been cloned and expressed in a variety of cells and display pharmacology consistent with the receptors previously identi®ed. Opioid receptor activation activates an inwardly rectifying K+ channel (Kir-hence

promoting an ef¯ux of K+) and closes voltage sensitive Ca2+ channels resulting in membrane hyperpolarization, reduced Ca2+ in¯ux, and hence, reduced neurotransmission. Adenylyl cyclase is also inhibited leading to reduced cAMP formation.9±14 This reduced cAMP formation is often measured as a biochemical index of opioid receptor activation but may also be involved in reduced neuronal ®ring via an interaction with the inwardly rectifying K+ channel (Ih).15 The mechanism and site(s) of opioid-local anaesthetic interaction are yet to be fully described. Several potential sites are possible. These include: changes in opioid pharmacokinetics (i.e. changes in tissue pH) produced by local anaesthetics;5 potentiation of the inhibitory effect of opioids on neurotransmitter release via modulation of

Ó The Board of Management and Trustees of the British Journal of Anaesthesia 2000

Local anaesthetics and opioid receptors

second messenger systems such as adenylyl cyclase;12 14 16 or actions at voltage sensitive Ca2+ channels.17 18 A direct interaction of local anaesthetics with opioid receptors has also been suggested.7 In this study, we have tested the latter hypothesis, that local anaesthetics interact with opioid receptors and that this may modulate the formation of cAMP. In order to avoid the interpretation problems associated with cell and tissue studies of a heterogeneous population of endogenous receptors, we have utilized Chinese hamster ovary (CHO) cells expressing recombinant m-, k-, and d-opioid receptors (CHO-m, k, and d, respectively).

20°C for 90 min as described previously.22 Non-speci®c binding was de®ned in the presence of naloxone 10 mM. Following incubation, each sample was ®ltered (and washed) under vacuum through Whatman GF/B ®lters using a Brandel cell harvester. Filter retained radioactivity was extracted for at least 8 h in 4 ml of scintillation ¯uid. In displacement studies, the interaction of lidocaine (10±6 to 10±2 M) with m, k and d-opioid receptors, and of bupivacaine (10±6 to 3310±3 M) and its enantiomers (S(±) and R(+): 10±6 to 10±3 M), procaine (10±6 to 310±2 M), prilocaine (10±6 to 3310±2 M), and tetracaine (10±6 to 10±2 M) with m-opioid receptor were determined by displacement of ~0.5 nM [3H]DPN.

Materials and methods Measurement of cAMP formation

Sources of reagents All tissue culture media and supplements were from Life Technologies (UK). With the exception of bupivacaine isomers all local anaesthetic agents were from Sigma (Poole, UK). S(±)bupivacaine (Batch OA859/15) and R(+)bupivacaine (Batch OA314/14) were from Astra Pain Control (Sweden). [3H]Diprenorphine (DPN) (speci®c activity 41 Ci mmol±1) was from Amersham International (Bucks, UK). [2,8±3H]cAMP (28.4 Ci mmol±1) was from NEN DuPont (Boston, MA). All other drugs were from Sigma (Poole, UK) or Calbiochem (Notts., UK). Other reagents were of the highest purity available.

Membrane preparation and cell culture CHO-m, k, d cells19±21 and untransfected CHO wild-type (CHO-wt) cells (i.e. cells not expressing the plasmid that encodes the receptor of interest and, therefore, acting as negative controls) were grown for experimentation in Hams F12 medium supplemented with penicillin 100 iu ml±1, streptomycin 100 mg ml±1, fungizone 2.5 mg ml±1 and fetal calf serum 10%. (Stock cultures also contained G418 200 mg ml±1.) Cultures were maintained at 37°C in 5% CO2/ humidi®ed air at 37°C, fed every 2±3 days and passaged every 7 days. Experiments were performed at days 5±7 after subculture. All cells were harvested for use by the addition of saline 0.9% containing HEPES (10 mM)/EDTA (0.02%). Cells were homogenized at 4°C using a tissue Tearor (setting 5.5330-s bursts) in 50 mM Tris±HCl buffer (pH 7.4). The homogenate was centrifuged at 18 000 3g for 10 min and the pellet resuspended in Tris±HCl buffer. This procedure was repeated twice more. Membranes were prepared and used fresh daily.

[3H]DPN binding The binding of the non-selective opioid receptor antagonist [3H]DPN was performed in 1 ml volumes of Tris±HCl buffer containing approximately 200 mg of membranes at

Whole cells (CHO-m, k, d and wt) suspended in 0.3 ml Krebs±HEPES buffer, pH 7.4 were incubated in the presence of isobutylmethylxanthine (1 mM) with or without (for the basal) forskolin (10 mM) at 37°C for 15 min. To obtain a concentration response curve for lidocaine inhibition of cAMP formation, cells were incubated additionally with or without lidocaine (10±5 to 3310±2 M). In order to determine any naloxonesensitivity of lidocaine inhibition of cAMP formation, CHO-m, k or d cells were incubated with lidocaine 2 mM in the presence or absence of naloxone 10 mM. In addition some cells were co-incubated with lidocaine (300 mM) and subtype selective opioid agonists: (DAla2, MePhe4, Gly(ol)5) enkephalin (DAMGO: 100 nM) for m and spiradoline (2 nM) for k. Reactions were terminated by the addition of 20 ml HCl (10 M), 20 ml NaOH (10 M) and 180 ml Tris buffer (1 M, pH 7.4). The concentration of cAMP was measured in the supernatant following centrifugation (13 000 r.p.m./2 min) using a speci®c protein binding assay as described previously.22

Statistical analysis All data are expressed as mean (SEM). The log concentration of displacers producing 50% displacement of speci®c binding (IC50) and the concentration of local anaesthetic producing 50% maximum inhibition of forskolin stimulated cAMP formation (IC50) were obtained by computer assisted curve ®tting (GRAPHPAD-PRISM). IC50 values from binding experiments were additionally corrected for the competing mass of [3H]DPN according to Cheng and Prusoff23 to yield the inhibitor constant (Ki). Further terminology follows IUPHAR recommendations where Bmax is de®ned as the maximum speci®c binding of a ligand determined in a radioligand binding assay (an estimate of the number of receptors), and Kd is the equilibrium dissociation constant (calculated as ligand concentration at 0.5 Bmax). Additivity data are analysed

741

Hirota et al.

using an unpaired t-test with P<0.05 considered to represent a signi®cant difference.

Results [3H]DPN binding Saturation analyses performed previously with these cells yielded Bmax and Kd values of (mean (SEM)) 262 (3), 70 (2), 279 (4) fmol mg protein±1 and 85 (5), 91 (6), 372 (15) pM in CHO-m, k, and d respectively.24 In displacement studies, lidocaine produced a concentration-dependent displacement of [3H]DPN binding to all opioid receptor subtypes (Fig. 1A) with the following rank order potency (mean Ki in mM): k (210) > m (552) > d (1810). A range of other local anaesthetics also displaced [3H]DPN binding to m receptors (Fig. 1B) with Ki values shown in Table 1. Whilst it was not possible to produce full displacement curves for R(+) bupivacaine and S(±) bupivacaine with the quantity available to us there was a suggestion of some stereoselectivity (R(+) more potent than S(±)) in CHO-m and k but not d (Fig. 2).

Inhibition of forskolin-stimulated cAMP formation Lidocaine produced a concentration-dependent (Fig. 3A) but naloxone-insensitive (Fig. 3B) inhibition of forskolin stimulated cAMP formation in all cell lines including untransfected CHO-wt cells. The IC50 values were not signi®cantly different among the cell lines examined (Table 2). The interaction of lidocaine with the k and m receptor occurred at clinically achievable concentrations and was probed further. As our initial premise was to determine if there was an interaction between opioids and local anaesthetics, we examined the effects of lidocaine in the presence of the selective k and m agonists spiradoline (2 nM) and (D-Ala2, MePhe4, Gly(ol)5) enkephalin (DAMGO, 100 nM); these concentrations were chosen from a previous study24 and were close to their IC50 for inhibition of cAMP formation. Lidocaine at 300 mM was chosen as this produced a signi®cant displacement (50±70%, Fig. 1) of the binding of [3H]DPN but only a small inhibition (<20%, Fig. 3) of cAMP formation. Both spiradoline and DAMGO inhibited forskolin stimulated cAMP formation in CHO-k and m cells respectively. This was naloxone (10 mM) sensitive (data not shown). The measured combination of lidocaine and spiradoline was signi®cantly lower than the value predicted. A similar trend was observed for the lidocaine±DAMGO combination but this failed to reach statistical signi®cance (Table 3). Due to the direct inhibitory action of lidocaine on cAMP formation it was not possible to probe this effect further.

Fig 1 (A) Lidocaine concentration-dependently displaced [3H]DPN from CHO-k, m and d with the following rank order: k > m > d. (B) Interaction of a range of local anaesthetics with m-opioid receptor. Concentration± response curves are corrected for the competing mass of [3H]DPN according to Cheng and Prusoff.23 All data are mean (SEM) (n=5).

Discussion In this study we have shown a clear interaction of a range of local anaesthetic agents encompassing both ester and amide forms with recombinant opioid receptors. Few studies examine both receptor and post-receptor events. For lidocaine there appeared to be some subtype selectivity; k (210 mM) > m (552 mM) > d (1810 mM). In general, the calculated local anaesthetic Ki for [3H]DPN displacement from m-opioid receptors (and lidocaine in CHO-k cells) fell within the clinical concentration range encountered during spinal anaesthesia,17 25 26 although we could not ®nd similar

742

Local anaesthetics and opioid receptors Table 1 Ki values (this study mean (95% CI)) for local anaesthetic binding to the m receptor in this study and for other G-protein coupled receptors (b2adrenoceptor, b2AR; Histamine, H1; muscarinic, mACh) compared with relative (to procaine=1) C-®bre conduction block. Local anaesthetic agents are ranked according to their hydrophobicity. a[3H]DHP binding.29 b[3H]mepyramine binding.30 c[3H]QNB binding.31 dFrom Strichartz and Berde.8 ND, not determined Local anaesthetic

Ki mM (95%CI) this study

Ki b2ARa

pKi H1b

mAChc pKi

Rel. cond. blockd

Tetracaine Bupivacaine Lidocaine Prilocaine Procaine

42.8 (35.8±51.7) 161.3 (147.2±177.0) 552.1 (427.6±712.9) 203.7 (184.1±225.4) 243.8 (178.6±332.7)

2.5 79.4 251.2 158.5 63.1

52.5 ND ND ND 107.2

109.6 9772 3715 1585 870.9

8.0 8.0 2.0 3.0 1.0

Fig 3 (A) Lidocaine produced a concentration-dependent inhibition of forskolin-stimulated cAMP formation in CHO cells expressing m, k, d opioid receptors (CHO-m, k, d) and in untransfected wt cells. (B) Lidocaine-induced inhibition of forskolin-stimulated cAMP formation was naloxone insensitive. All data are mean (SEM) (n=4±5).

Fig 2 Interaction of bupivacaine isomers with m- (A), k- (B) and d (C)opioid receptors. All data are mean (SEM) (n=5). For CHO-m and -k, there is a suggestion of stereoselectivity (i.e. the curve for S(±)bupivacaine tends to the right).

concentrations for procaine and prilocaine. However, we were unable to conclusively de®ne the functional consequences of this interaction.

743

Hirota et al. Table 2 IC50 (mM) for inhibition of cAMP formation by lidocaine. CHO-wt, untransfected CHO wild type cells. Mean (95%CI) from n=4

IC50

CHO-m

CHO-k

CHO-d

CHO-wt

1.32 (0.73±2.38)

2.41 (1.50±3.90)

1.27 (0.71±2.28)

2.78 (1.63±4.73)

Table 3 Interaction of lidocaine (300 mM) with spiradoline (2 nM) and (D-Ala2, MePhe4, Gly(ol)5) enkephalin (DAMGO, 100 nM) in CHO-k and m cells respectively. % inhibition of forskolin stimulated cAMP formation are presented as mean (sEM) (n=4±11). aP<0.05 signi®cantly inhibited compared to paired forskolin (paired Student's t-test); bP<0.05 signi®cantly greater than lidocaine alone (unpaired Student' t-test); cP<0.05 signi®cantly reduced compared with predicted lidocaine+spiradoline (unpaired Student's t-test) Combination CHO-k Lidocaine Spiradoline Lidocaine + Lidocaine + CHO-m Lidocaine DAMGO Lidocaine + Lidocaine +

% Inhibition of cAMP formation

spiradoline (predicted if additive) spiradoline (measured)

9.6 (1.6)a 18.0 (2.3)ab 27.6 (4.1)ab 14.9 (2.8)ac

DAMGO (predicted if additive) DAMGO (measured)

17.4 24.7 42.1 25.6

Local anaesthetic agents have been shown to interact with a number of differing receptor systems including ligand gated ion channels and G-protein coupled receptors. For example, the GABAA receptor27 and nicotinic receptor28 are ligand gated ion channel coupled receptors whose activity is inhibited by local anaesthetics. Interestingly, in the latter study, glutamate, NMDA and AMPA receptors were unaffected.28 Some examples of the interaction of local anaesthetic agents with G-protein coupled receptors are shown in Table 1.29±31 For this receptor system there does not appear to be any signi®cant correlation with relative conduction blocking potency but the rank order potency at the m opioid receptor follows the hydrophobicity ranking. A possible target for the interaction of local anaesthetics and opioids is the opioid receptor and the elements of the signal transduction cascade associated with this receptor, i.e. adenylyl cyclase, voltage sensitive Ca2+ channels and inwardly rectifying K+ channels. Tejwani and colleagues7 reported that bupivacaine inhibited the speci®c binding of a range of radioligands to m-receptors but surprisingly increased binding to k- and d-receptors. However, in a radioligand binding study, Fraser and Fowler32 reported that local anaesthetics did not interact with k-opioid receptors. The discrepancy between these and our study most likely results from the concentration of local anaesthetic used. In the studies of Tejwani and colleagues7 and Fraser and Fowler32 bupivacaine concentration of less than 10 mM were used. The present data indicates that signi®cant inhibition of radioligand binding to opioid receptors requires bupivacaine concentrations in excess of 10 mM (Ki, 161 mM). In a review article, Franks and Lieb33 suggested that stereoselectivity may prove to be one of the most powerful guides to in vitro targets relevant to anaesthesia. In the

(6.8)a (2.8)a (10.6)a (3.9)a

present study (with the amount of bupivacaine isomers available to us) our data give an indication of some stereoselectivity for bupivacaine at m- and k- but probably not d-opioid receptors (Fig. 3) with the R(+) isomer being more potent than the S(±) isomer. However, this stereoselectivity is the reverse of that observed for neuronal Na+ channels (S(±) more potent than (R(+))34 and we have no explanation for this discrepancy. Despite the conclusions of Fraser and Fowler32 to the contrary, our data suggest that the m- and k-opioid receptors may represent a target for local anaesthetic agents and a site at which the clinically observed local-opioid interaction may occur. In this study, we probed the functional consequences of this interaction further by studying the effects of opioids and lidocaine on the formation of cAMP. Opioid receptor agonists inhibit adenylyl cyclase thus inhibiting cAMP formation through opioid receptors and this may give clues as to an interaction with Gi coupled voltage sensitive Ca2+ channels and Kir. In addition, Wang and colleagues35 reported that cAMP mediated m and d but not k opioid analgesia, based on the reversal of m and dmediated analgesia by intrathecal administration of dibutylcAMP (i.e. arti®cially elevating cAMP). In the present study lidocaine inhibited forskolin-stimulated cAMP formation in a concentration-dependent and naloxone insensitive manner indicating that this inhibitory action on adenylyl cyclase was not mediated via opioid receptors. This was further con®rmed in untransfected CHO cells where lidocaine also inhibited cAMP formation with an essentially identical pIC50 when compared with CHO-k, m and d cells. These data are also consistent with the observed direct interaction of local anaesthetics with the catalytic subunit of adenylyl cyclase leading to reduced cAMP formation.16 Thus, whilst the formation of cAMP is, as would be expected for opioids,

744

Local anaesthetics and opioid receptors

inhibited by lidocaine, this response does not represent agonist action at the opioid receptor. If local anaesthetic agents and opioid receptor activation inhibit cAMP formation by different pathways then it might be reasonable to assume that the combination of local and opioid would produce an additive inhibition. This was clearly not the case (Table 3). Indeed we have found a suggestion of an inhibitory action of lidocaine on k (and possibly m) receptor signaling. We cannot explain the lack of a positive interaction between these two agents at the level of cAMP formation. Our data suggest that the interaction between local anaesthetic agents and opioids in the clinical setting does not result from an interaction with opioid receptor signaling (i.e. at the cellular level). It is well known that opioids inhibit cAMP formation, close voltage sensitive Ca2+ channels and activate inwardly rectifying K+ channels. These combined actions are likely to reduce neurotransmission. It is also well known that local anaesthetic agents inhibit voltage sensitive Na+ channels to reduce axonal conduction. Na+ channel blockade with local anaesthetics produces analgesia. It is worthy of mention that a range of anticonvulsants with analgesic activity also appear to inhibit voltage sensitive Na+ channels. For example, carbamazepine36 and gabapentin37 block Na+ channels and are also useful in the pain clinic. The actions of lamotrigine are equivocal where Na+ channel block36 and facilitation of C®bre responses38 have also been reported. We therefore conclude that the clinical interaction between local anaesthetic agents and opioids may occur as a combination of reduced axonal conduction and neurotransmission in the spinal cord and an interaction at the cellular level is unlikely.

Acknowledgements The authors would like to thank Dr L. A. Devi (Department of Pharmacology, New York University Medical Center, New York, USA) for providing CHO-d cells.

8 9 10 11

12 13 14 15 16 17 18

19 20 21 22

23

References 1 Shapiro LA, Hoffman S, Jedeikin R, Kaplan R. Single-injection epidural anesthesia with bupivacaine and morphine for prostatectomy. Anesth Analg 1981; 60: 818±20 2 Lanz E, Kehrberger E, Theiss D. Epidural morphine: A clinical double-blind study of dosage. Anesth Analg 1985; 64: 786±91 3 Douglas JM, McMorland GH, Janzen JA. In¯uence of bupivacaine as an adjuvant to epidural morphine for analgesia after cesarean section. Anesth Analg 1988; 67: 1138±41 4 Akerman B, ArwestroÈm E, Post C. Local anesthetics potentiate spinal morphine antinociception. Anesth Analg 1988; 67: 943±48 5 Maves TJ, Gebhart GF. Antinociceptive synergy between intrathecal morphine and lidocaine during visceral and somatic nociception in the rat. Anesthesiology 1992; 76: 91±9 6 Penning JP, Yaksh TL. Interaction of intrathecal morphine with bupivacaine and lidocaine in the rat. Anesthesiology 1992; 77: 1186±1200 7 Tejwani GA, Rattan AK, McDonald JS. Role of spinal opioid

745

24

25 26 27 28 29

receptors in the antinociceptive interactions between intrathecal morphine and bupivacaine. Anesth Analg 1992; 74: 726±34 Strichartz GR, Berde CB. Local anesthetics. In: Miller RD, ed. Anesthesia, 4th Edn. New York: Churchill-Livingstone, 1994; 489±521 Reisine T. Opiate receptors. Neuropharmacol 1995; 34: 463±72 Knapp RJ, Malatynska E, Collins N, Fang L, Wang JY, Hruby VJ, Roeske WR, Yamamura HI. Molecular biology and pharmacology of cloned opioid receptors. FASEB J 1995; 9: 516±25 Dhawan BN, Cesslin F, Raghubir T, Reiseine T, Bradley PB, Portoghese PS, Hamon M. International union of pharmacology. XII. Classi®cation of opioid receptors. Pharmacol Rev 1996; 48: 567±92 Standifer KM, Pasternak GW. G proteins and opioid receptor mediated signaling. Cell Signal 1997; 9: 237±48 Law PY, Loh HH. Regulation of opioid receptor activities. J Pharmacol Exp Therap 1999; 289: 607±24 Connor M, Christie MJ. Opioid recpeptor signalling mechanisms. Clin Exp Pharmac Physiol 1999; 26: 493±99 Ingram SL, Williams JT. Opioid inhibition of Ih via adenylyl cyclase. Neuron 1994; 13: 179±86 Butterworth JF, Brownlow RC, Leith JP, Prielipp RC. Bupivacaine inhibits cyclic-3¢,5¢-adenosine monophosphate production. Anesthesiology 1993; 79: 88±95 Sugiyama K, Muteki T. Local anesthetics depress the calcium current of rat sensory neurons in culture. Anesthesiology 1994; 80: 1369±78 Hirota K, Browne T, Appadu BL, Lambert DG. Do local anaesthetics interact with dihydropyridine binding sites on the neuronal L-type Ca2+ channels? Br J Anaesth 1997; 78: 185±88 Bunzow JR, Zhang G, Bouvier C, Saez C, Ronnekleiv OK, Kelly MJ, Grandy DK. Characterization and distribution of a cloned rat m-opioid receptor. J Neurochem 1995; 64: 14±24 Hjorth SA, Thirstrup K, Grandy DK, Schwartz TW. Analysis of selective binding epitopes for the k-opioid receptor antagonist nor-binaltorphimine. Mol Pharmacol 1995; 47: 1089±94 Cvejic S, Trapaidze N, Cyr C, Devi LA. Thr353, located within the C-terminal tail of the m opiate receptor, is involved in receptor down-regulation. J Biol Chem 1996; 271: 4073±76 Smart D, Hirst RA, Hirota K, Grandy DK, Lambert DG. The effects of recombinant rat m-opioid receptor activation in CHO cells on phospholipase C, [Ca2+]i and adenylyl cyclase. Br J Pharmacol 1997; 120: 1165±71 Cheng YC, Prusoff WM. Relationship between the inhibition constant (Ki) and the concentration of inhibitor which caused 50% inhibition (IC50) of an enzymatic reaction. Biochem Pharmacol 1973; 22: 3099±108 Hirota K, Okawa H, Appadu BL, Grandy DJ, Devi LA, Lambert DG. Allosteric and stereoselective interaction of ketamine with recombinant m, k and d-opioid receptors expressed in Chinese hamster ovary cells. Anesthesiology 1999; 90: 174±83 MoÈrch ET, Rosenberg MK, Truant AT. Lidocaine for spinal anesthesia. Acta Anaesthesiol Scand 1957; 1: 105±15 Covino BG, Scott DB, Lambert DH. Pharmacological considerations. In: Handbook of Spinal Anaesthesia and Analgesia. Philadelphia: WB Saunders, 1994; 71±104 Nordmark J, Rydqvist B. Local anaesthetics potentiate GABAmediated-cl- currents by inhibiting GABA uptake. Neuroreport 1997; 8: 465±68 Lu WY, Bieger D. Inhibition of nicotinic cholinoceptor mediated current in vagal motor neurons by local anesthetics. Can J Physiol Pharmacol 1996; 74: 1265±69 Butterworth J, James RL, Grimes J. Structure-af®nity

Hirota et al.

30 31 32 33 34

relationships and stereospeci®city of several homologous series of local anesthetics for the b2-adrenergic receptor Anesth Analg 1997; 85: 336±42 Horio S, Nagare T, Ishida Y, Moritoki H. Interaction of local anaesthetics with histamine H1 receptors in guinea-pig ileum. J Pharm Pharmacol 1997; 49: 715±20 Taylor WJ, Wolf A, Young JM. The interaction of amine local anesthetics with muscarinic receptors. Br J Pharmacol 1980; 71: 327±35 Fraser GL, Fowler CJ. k-Opioid receptor recognition sites are not modulated by local anaesthetics. Biochem Pharmacol 1995; 49: 883±91 Franks NP, Lieb WR. Molecular and cellular mechanisms of general anaesthesia. Nature 1994; 367: 607±14 Vanhoutte F, Vereecke J, Carmeliet E. Stereoselective effects of the enantiomers of bupivacaine on the electrophysiological

746

35 36 37

38

properties of the guinea-pig papillary muscle. Br J Pharmacol 1991; 103: 1275±81 Wang JF, Ren MF, Xue JC, Han JS. Cyclic AMP mediates mu and delta, but not kappa opioid analgesia in the spinal cord of the rat. Life Sci 1993; 52: 1955±60 Lees G. The effects of anticonvulsants on 4-aminopyridineinduced bursting: in vitro studies on rat peripheral nerve and dorsal roots. Br J Pharmacol 1996; 117: 573±79 Taylor CP, Gee NS, Su TZ, Kocsis JD, Welty DF, Brown JP, Dooley DJ, Boden P, Singh L. A summary of mechanistic hypotheses of gabapentin pharmacology. Epilepsy Res 1998; 29: 233±49 Chapman V, Wildman MA, Dickenson AH. Distinct electrophysiological effects of two spinally administered membrane stabilising drugs, bupivacaine and lamotrigine. Pain 1997; 71: 285±95