Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Biochemical and electrophysiological studies have demonstrated a role for potassium channels (K-channels) in modulating neuronal excitability (Cook, 1988; Rudy, 1988; Wible and Brown, 1994). Blockade of K-channels prevents the efflux of potassium from neurons resulting in depolarization and increased release of neurotransmitters (Glover, 1981). In this regard, the K-channel blocker 4-aminopyridine (4-AP) (Rudy, 1988) and its structural analogs increase the release of several neurotransmitters, including acetylcholine (ACh) (Damsma et al., 1988), noradrenaline (NE) (Hu and Fredholm, 1991), dopamine (DA) (Scheer and Lavoie, 1991; Dawson and Routledge, 1995), and serotonin (5-HT) (Schechter, 1997).

Despite many biochemical and electrophysiological studies investigating the effects of 4-AP on K-channels, relatively few studies have investigated the behavioral effects of 4-AP. 4-AP does, however, induce wet dog shakes and repetitive paw movements, produces antidepressant like effects in several animal models of depression, and produces convulsions at high doses (Fragoso-Veloz and Tapia, 1992; Gorman et al., 1995; Trella et al., 1995). In addition, 4-AP attenuates the effects of drugs acting at several neurotransmitter systems. In this regard, 4-AP attenuates the locomotor activating effects of d-amphetamine, cocaine, and scopolamine, as well as increases in punished responding produced by chlordiazepoxide and pentobarbital (Rosenzweig-Lipson et al., 1996, 1997a). Furthermore, 4-AP also attenuates the increases in behavioral activity (grooming, rearing, sniffing, and locomotion) produced by morphine, and the presynaptic D₂ cataleptic effects produced by low doses of quinpirole (Pei et al., 1993; Rosenzweig-Lipson and Barrett, 1995). Taken together, these results suggest that 4-AP produces centrally mediated behavioral effects consistent with K-channel modulation and that these effects interact with or are modulated by neurotransmitter systems involving DA, 5-HT, NE, ACh, -aminobutyric acid (GABA), and opioids.

The purpose of the present study was to determine whether 4-AP can serve as a discriminative stimulus (DS) in rats and to evaluate the specificity of the DS effects of 4-AP using substitution and antagonism studies. Rats were trained to discriminate 4-AP from saline using a two-lever drug discrimination procedure. After the DS was established, the effects of structural analogs of 4-AP, other K-channel blockers (quinine, linopiridine), and indirect DA, NE, 5-HT, ACh agonists, and GABA_A agonists and antagonists were evaluated in substitution studies to determine the specificity of the interoceptive cue produced by 4-AP. In addition to the indirect effects of 4-AP in mediating neurotransmitter release, 4-AP binds at high concentrations to several receptors, including noradrenergic (₁, ), serotonergic (5-HT_1A, 5-HT_2A), dopaminergic (D₁, D₂), and muscarinic (M₂) receptors (Drukarch et al., 1989). Therefore, substitution and antagonism studies were conducted using direct DA, NE, 5-HT, ACh agonists and antagonists and GABA_A agonists. The present results show that 4-AP and its structural analogs fully or partially substitute for 4-AP. Agonists of many different neurotransmitter systems did not substitute for 4-AP, and most antagonists did not attenuate the DS effects of 4-AP. However, the benzodiazepine agonists, chlordiazepoxide (CDP) and diazepam, attenuated the DS of 4-AP, suggesting a link between these systems. The present results demonstrate that the K-channel blocker 4-AP can be trained as a DS in rats and the DS effects of 4-AP are likely mediated through blockade of voltage-dependent K-channels. The results also suggest a novel interaction between benzodiazepines and K-channels.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Subjects. Thirty-six male Sprague-Dawley rats weighing 280 to 330 g (Charles River, Wilmington, MA) were housed individually and maintained on a 12-h light/dark schedule with water available ad libitum in their home cages. The animals were maintained at approximately 85% of free feeding body weight by restricting food (Purina Rodent Chow; Ralston-Purina, Scottsdale, AZ) intake following experimental sessions.

Apparatus. Experimental sessions took place in two-lever operant chambers that were enclosed in ventilated sound-attenuating cubicles equipped with white noise to block extraneous sounds (Med Associates Inc., George, VT). Precision dustless pellets (45 mg; Bioserv, Frenchtown, NJ) could be delivered to food troughs to serve as reinforcers. Each press of a lever produced an audible click and was recorded as a response. The operant chambers were controlled and monitored by Epson Equity 386 computers (Epson, Long Beach, CA) with software from Med Associates Inc.

Procedure. Three groups of 12 rats were trained to discriminate 4-AP from saline using a two-lever food reinforced drug discrimination procedure. Rats were initially trained to lever press on a fixed ratio (FR) schedule of food presentation (only one lever at a time presented) with a gradual increase in the FR (final FR = 30) over the training sessions. Rats were then trained to respond on one lever following a saline injection (s.c.) and on the other lever following a 4-AP injection (s.c.) using a double alternation injection procedure (s, d, d, s). When the injection schedule started, both levers were in the chamber and the FR was dropped to 5 with a gradual increase to FR 30. At the final schedule, completion of 30 consecutive responses on the injection appropriate lever resulted in a food pellet reinforcer. Responding on the incorrect lever reset the FR for the injection appropriate lever. Training and test sessions were 30 min in duration. Criteria for establishing 4-AP as a discriminative stimulus were that the first ratio (30 consecutive responses) was completed on the injection appropriate lever and that there was >90% injection appropriate responding for the total session for five consecutive sessions. For the first group of 12 animals, the initial training dose was 0.3 mg/kg 4-AP with a pretreatment time of 15 min, which was spent in the home cage (20 sessions). The dose of 4-AP was increased to 0.56 mg/kg and then to 1.0 mg/kg with the same pretreatment time, because the rats were not displaying stimulus control. After 40 more sessions, the pretreatment time was increased to 30 min. Nine rats then met criteria within 30 sessions. Following an increase to 1.7 mg/kg, the remaining three rats reached criteria within 10 sessions. Total training time averaged 100 sessions (range: 69-106). For groups 2 and 3, the training dose was 1.0 mg/kg 4-AP with a pretreatment time of 30 min. The rats in group 2 reached criteria in approximately 50 sessions (range: 28-89), and those in group 3 in approximately 32 sessions (range: 17-64). When rats met their criterion, substitution and antagonism experiments began. Because there were no qualitative differences between the rats trained at 1.0 or 1.7 mg/kg in substitution or antagonism experiments, the data were grouped across training doses.

Drug Testing. Substitution or antagonism studies were conducted once or twice a week in individual rats, with a full dose-response curve in 4 to 12 rats for each compound. For substitution studies, a dose of test compound was administered at an appropriate pretreatment time before the test session. For antagonism studies, a dose of the test compound was administered as a pretreatment before an injection of the training dose of 4-AP (1.0 mg/kg, s.c.). Pretreatment times for both substitution and antagonism studies were selected for different drugs based on previous studies. On test days, completion of 30 consecutive responses on either lever resulted in a food pellet reinforcer. Compounds were tested up to rate-decreasing doses (<50% of drug injection baseline rate) or until the observation of adverse behavioral effects (i.e., convulsions). Compounds were administered i.p., s.c., or p.o. in volumes of 1 ml/kg of body weight. Test sessions were conducted only if an animal met criteria (first ratio correct and >90% responding on the injection appropriate lever) on the previous day and on 4 of 5 previous training days. Training sessions did not take place on days after test sessions to allow for a wash-out period for test compounds.

Drugs. 4-AP, 2-aminopyridine (2-AP), 3-aminopyridine (3-AP), 2,3-diaminopyridine (2,3-DIAP), 3,4-diaminopyridine (3,4-DIAP), 2,6-diaminopyridine (2,6-DIAP), 4-dimethylaminopyridine (4-DMAP), and pyridine (Sigma Chemical Co., St. Louis, MO) were dissolved in 0.9% saline solution with small amounts of 1 N HCl added to bring the pH to 7. Haloperidol (Sigma), clozapine (Research Biochemicals International, Natick, MA), and diazepam (Elkins-Sinn, Cherry Hill, NJ) were dissolved in small amounts of 1 N HCl and diluted with sterile H₂O. Ketanserin tartrate (Jannssen Pharmaceutica, Beerse, Belgium); propranolol, venlafaxine, WAY-100635 (Wyeth-Ayerst Research, Princeton, NJ); CDP, imipramine, quinpirole HCl, SKF-81297, idazoxan, mCPP, cirazoline HCl, mecamylamine, 8-OH-DPAT HBr (Research Biochemicals International); fluoxetine (Lilly Laboratory, Indianapolis IN); and linopiridine (DuPont, Wilmington, DE) were dissolved in sterile H₂O. SCH-23390 (Schering Corp., Bloomfield, NJ) and prazosin (Research Biochemicals International) were dissolved in small amounts of methanol and diluted with sterile H₂O. Pinacidil, yohimbine, cromakalim, pentobarbital sodium, 9-amino-1,2,3,4-tetrahydroacridine HCl (tacrine) (Research Biochemicals International); phenytoin (Warner Lambert, Ann Arbor MI); and BAY K 8644 (Miles Pharmaceuticals, West Haven, CT) were dissolved in 90% sterile H₂O and 10% Tween 80. Scopolamine HBr, clonidine, DOI, dizolcipine, phencyclidine HCl, methamphetamine, and d-amphetamine (Research Biochemicals International); ondansetron (Glaxo, Middlesex, UK); naloxone (Endo Laboratories, Garden City, NY); quinine, pentylenetetrazole (Sigma); fenfluramine (Wyeth-Ayerst); morphine SO₄ (Merck, Rahway, NJ); and nisoxetine and tomoxetine (Lilly Laboratories) were dissolved in 0.9% saline solution. Flumazenil (Hoffman LaRoche, Basal, Switzerland) was dissolved in polyethylene glycol 200. SB-200646 (Wyeth-Ayerst) was suspended in 10% methyl cellulose.

Analysis of Results. Rates of responding were calculated by dividing the total number of responses by the total time of the session. Baseline drug lever rates were calculated by averaging rates of drug injection days preceding a test day. Baseline saline lever rates were calculated by averaging rates of saline injection days preceding a test day. Percentage of drug lever responding was calculated by dividing the number of responses on the drug appropriate lever by the total number of responses in the session multiplied by 100. Compounds were considered to fully substitute in an individual animal if there was >80% drug appropriate lever responding. Compounds were considered to partially substitute for 4-AP in an individual animal if there was between 40 and 80% drug appropriate lever responding. Compounds were considered to fully attenuate the effects of the training dose of 4-AP in an individual animal if drug lever responding was reduced to <20%. Compounds were considered to partially attenuate the effects of the training dose of 4-AP in an individual animal if drug lever responding was reduced to 20 to 60%. If an animal did not receive a reinforcer during a test session, the percentage drug lever responding was not analyzed. However, the rate of responding was included in the analysis. ED₅₀ values for 4-AP were calculated based on nonlinear regression using statistical software (JMP, Cary, NC).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Control Performances. During days preceding test sessions, animals averaged 98 to 100% responding on the 4-AP appropriate lever after administration of 1.0 mg/kg 4-AP, and 0 to 1% 4-AP appropriate lever responding after the administration of saline. Response rates were similar after the administration of saline (2.07 ± 0.1 responses/s; mean ± S.E.) or after the administration of 4-AP (1.81 ± 0.12 responses/s; mean ± S.E.).

Effects of 4-AP. In all three groups, 4-AP (0.01-1.7 mg/kg) engendered dose-related increases in the percentage of responses on the drug appropriate lever, with all animals showing full substitution at one or more doses of 4-AP (Fig. 1). There was no difference between the three groups trained at 1.0 mg/kg 4-AP, with ED₅₀ values of 0.31 mg/kg (CI: 0.21-0.47), 0.39 mg/kg (CI: 0.14-1.09), and 0.35 mg/kg (CI: 0.25-0.43) for groups one, two, and three, respectively. The rats trained at 1.7 mg/kg had an ED₅₀ value of 1.08 mg/kg (CI: 0.69-1.67). The highest dose of 4-AP (1.7 mg/kg) decreased rates of responding, with decreases of 34 to 46% in the three groups.

View larger version (15K): [in this window] [in a new window]   Fig. 1.   The effects of 4-AP in rats trained to discriminate 4-AP from saline. Abscissa: dose, log scale; ordinates: (top panel) percentage of responses on the 4-AP-associated lever, (bottom panel) response rate (responses/s). Points and error bars represent means ± S.E. , group 1 (training dose, 1.0 mg/kg); , group 1 (training dose, 1.7 mg/kg); , group 2; , group 3.

4-AP Time Course. The training dose of 4-AP was tested at pretreatment times of 5, 10, 15, 30, 60, 120, and 240 min. Responding gradually shifted from the saline-associated lever to the 4-AP-associated lever over the first 30 min (Fig. 2). The DS effects of 4-AP peaked at 30 and 60 min and gradually over the next several hours, such that by 240 min responding had shifted back to the saline-associated lever. Full substitution was observed in one rat for all pretreatment times tested.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Time course of the effects of 1.0 mg/kg 4-AP in rats trained to discriminate 4-AP from saline. Abscissa: time (minutes). Other details as in Fig. 1.

Effects of Other Aminopyridines. Like 4-AP, 3-AP (1.0-10.0 mg/kg), and 2,3-DIAP (1.0-17.8 mg/kg) engendered dose-related increases in the percentage of responses on the 4-AP-associated lever, with >90% 4-AP lever responding for the group of rats (Fig. 3). Full substitution was observed in 5 of 5 rats following the administration of 3-AP (5.6 mg/kg) and in 9 of 10 rats following the administration of 2,3-DIAP (10.0 mg/kg). 2-AP (3.0-17.8 mg/kg) also engendered dose-related increases in the percentage of responses on the 4-AP-associated lever, with approximately 80% (17.8 mg/kg) responding on the 4-AP-associated lever. Full substitution was observed in 7 of 9 rats following administration of 2-AP (17.8 mg/kg). Doses of 3-AP (5.6 mg/kg), 2,3-DIAP (10.0 mg/kg), and 2-AP (17.8 mg/kg), which produced maximal substitution decreased response rates for the group of rats tested. For this group of aminopyridines, the rank order of potency was 4-AP > 3-AP > 2,3-DIAP > 2-AP.

View larger version (29K): [in this window] [in a new window]   Fig. 3.   The effects of structurally related aminopyridines in rats trained to discriminate 4-AP from saline. , 2-AP; , 3-AP; , pyridine; , 2,3-DIAP; , 2,6-DIAP; , 3,4-DIAP; , 4-DMAP. Other details as in Fig. 1.

Effects of Ion Channel Modulators. The K-channel blockers, quinine (30.0-100.0 mg/kg) and linopiridine (0.1-5.6 mg/kg), and the Ca²⁺ channel activator (±)-BAY K 8644 (0.3-1.0 mg/kg) were tested in substitution studies (Table 1). Neither quinine nor (±)-BAY K 8644 substituted for 4-AP up to doses that markedly decreased response rates. Linopiridine produced full substitution in only 1 of 4 rats after a dose of 5.6 mg/kg linopiridine. Although higher doses were not tested due to lack of supply, lower doses of linopiridine have been shown to attenuate the effects of scopolamine (Buxton et al., 1994; Fontana et al., 1994), and higher doses of linopiridine (10 mg/kg) induce tremor and mortality in rats (DeNoble et al., 1990).

View larger version (15K): [in this window] [in a new window]   Fig. 4.   The effects of 1.0 mg/kg 4-AP alone and after pretreatment with cromakalim () and pinacidil () in rats trained to discriminate 4-AP from saline. Points above SAL and 4-AP represent the effects of saline or 4-AP alone. Other details as in Fig. 1.

Effects of Noradrenergic Drugs. The alpha-1 agonist cirazoline (0.03-1.0 mg/kg), the alpha-1 antagonist prazosin (1.0-5.6 mg/kg), the alpha-2 agonist clonidine (0.01-0.1 mg/kg), and the alpha-2 antagonists yohimbine (3-10 mg/kg) and idazoxan (1-30 mg/kg); the norepinephrine (NE) reuptake inhibitors tomoxetine (1.0-30.0 mg/kg) and nisoxetine (10.0-30.0 mg/kg); and the nonspecific NE/5-HT reuptake inhibitor imipramine (10.0-30.0 mg/kg) were tested in substitution studies (Table 2). Cirazoline, prazosin, and clonidine did not substitute for 4-AP up to doses that decreased response rates. Yohimbine, but not idazoxan, increased the percentage of responses on the 4-AP-associated lever, with approximately 46% 4-AP lever responding for the group of rats. Full substitution was observed in 2 of 7 rats at both 5.6 and 10 mg/kg. A third rat partially substituted (63-79%) at both 5.6 and 10 mg/kg. Idazoxan did not substitute for 4-AP up to a dose of 30 mg/kg. Full substitution was observed in 1 of 4 rats at 10 and 17.8 mg/kg.

View larger version (13K): [in this window] [in a new window]   Fig. 5.   The effects of norepinephrine reuptake inhibitors in rats trained to discriminate 4-AP from saline. , nisoxetine; , tomoxetine; , imipramine. Other details as in Fig. 1.

View larger version (12K): [in this window] [in a new window]   Fig. 6.   The effects of 1.0 mg/kg 4-AP alone and after pretreatment with prazosin () in rats trained to discriminate 4-AP from saline. Points above SAL and 4-AP represent the effects of saline or 4-AP alone. Other details as in Fig. 1.

Effects of Serotonergic Drugs. The 5-HT_1A agonist 8-OH-DPAT, the 5-HT_1A antagonist WAY-100635, the 5-HT_2A/2C agonist DOI, the 5-HT_2A antagonist ketanserin, the 5-HT_2C agonist mCPP, the 5-HT_2C/2B antagonist SB-200646, the 5-HT₃ antagonist ondansetron, the 5-HT releaser fenfluramine (0.3-10.0 mg/kg), the 5-HT reuptake inhibitor fluoxetine (10.0-100.0 mg/kg), and the 5-HT/NE reuptake inhibitor venlafaxine (3.0-56.0 mg/kg) were tested in substitution studies (Table 3). 8-OH-DPAT, WAY-100635, ketanserin, SB-200646, and ondansetron did not substitute up to rate-decreasing or pharmacologically active doses. mCPP increased the percentage of responses on the 4-AP-associated lever, with approximately 66% 4-AP lever responding for the group of rats. Full substitution was observed in 3 of 6 rats at 1 and 1.7 mg/kg. DOI increased the percentage of responses on the 4-AP-associated lever, with approximately 40% 4-AP lever responding for the group of rats. At the dose of DOI that showed partial substitution (1.0 mg/kg), 1 of 2 rats showed full substitution; this dose of DOI markedly reduced response rates in the other two animals. Fenfluramine, fluoxetine, and venlafaxine did not substitute for 4-AP up to doses that markedly decreased response rates. However, full substitution was observed in 1 of 4 rats at a dose of 1.0 mg/kg fenfluramine and in 1 of 4 rats at a dose of 56 mg/kg venlafaxine.

Effects of Dopaminergic Drugs. The D₁ agonist SKF-81297 (0.1-3.0 mg/kg), the D₂ agonist quinpirole (0.0003-1.0 mg/kg), the DA releasers d-amphetamine (1.0-3.0 mg/kg) and methamphetamine (0.1-3.0 mg/kg), and the atypical antipsychotic clozapine were tested in substitution studies. SKF-81297, quinpirole, d-amphetamine, methamphetamine, and clozapine did not fully substitute for 4-AP up to doses that markedly decreased response rates (Table 4). Full substitution was observed in 2 of 4 rats following varying doses of quinpirole (0.01, 0.3 mg/kg). Full substitution was observed in 3 of 6 rats at either 1 or 3 mg/kg clozapine.

Effects of Cholinergic Drugs. The acetylcholinesterase inhibitor, tetrahydroaminoacridine (tacrine) (1.0-5.6 mg/kg) and the nicotinic antagonist mecamylamine (1.0-10.0 mg/kg) were also tested in substitution studies and did not substitute for 4-AP at the highest dose tested (Table 4). Higher doses of tacrine were not tested due to reports of tremor and mortality at higher doses (Hunter et al., 1989; DeNoble et al., 1990; Yoshida and Suzuki, 1993). Full substitution was observed in 1 of 4 rats after 3 or 10 mg/kg mecamylamine.

Effects of Benzodiazepines and Barbiturates. The benzodiazepine agonist CDP (3.0-30.0 mg/kg) and the benzodiazepine antagonist flumazenil (10.0-30.0 mg/kg) were tested in substitution studies (Table 5). Neither CDP nor flumazenil substituted for 4-AP up to doses that markedly decreased response rates or were well above pharmacologically active doses (DeVry and Slangen, 1986).

View larger version (15K): [in this window] [in a new window]   Fig. 7.   The effects of 1.0 mg/kg 4-AP alone and after pretreatment with CDP () and diazepam () in rats trained to discriminate 4-AP from saline. Points above SAL and 4-AP represent the effects of saline or 4-AP alone. Other details as in Fig. 1.

Effects of Proconvulsants and Anticonvulsants. The proconvulsant pentylenetetrazole (10.0-56.0 mg/kg) was tested in substitution studies and did not fully substitute for 4-AP up to doses that markedly decreased response rates (Table 5). Full substitution for 4-AP was observed in 1 of 4 rats at 30.0 mg/kg pentylenetetrazole. The anticonvulsant phenytoin (1.0-17.8 mg/kg) was tested in antagonism studies and partially attenuated the DS effects of 4-AP, with 4-AP lever responding reduced to 60% at 3 mg/kg. Full attenuation of the DS effects of 4-AP was observed in 2 of 5 rats after a pretreatment with 3.0 mg/kg phenytoin. However, neither higher nor lower doses of phenytoin attenuated the effects of 4-AP in any of the rats tested.

Effects of Opiates and N-Methyl-D-aspartate (NMDA) Antagonists. The µ agonist morphine (0.3-10.0 mg/kg) and the NMDA antagonists dizolcipine (0.03-0.17 mg/kg) and phencyclidine (3.0-5.6 mg/kg) were tested in substitution studies and did not fully substitute for 4-AP up to doses that markedly decreased response rates (Table 5). Full substitution was observed in 1 of 5 rats after 0.17 mg/kg dizolcipine and in 1 of 4 rats after 5.6 mg/kg phencyclidine. Dizolcipine and the µ antagonist naloxone were tested in antagonism studies. Neither compound attenuated the DS effects of 4-AP.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Biochemical and electrophysiological studies have demonstrated that 4-AP blocks KV channels across a range of subtypes, including members of the KV1, KV2, and KV3 families (Grissmer et al., 1994; Mathie et al., 1998). Behavioral studies have shown that 4-AP induces wet dog shakes and repetitive paw movements, produces antidepressant-like effects, and induces convulsions at high doses (Fragoso-Veloz and Tapia, 1992; Gorman et al., 1995; Trella et al., 1995). In addition, 4-AP attenuates the behavioral effects induced by psychomotor stimulants, dopaminergics, anticholinergics, benzodiazepines, barbiturates, and opioids (Kitzman et al., 1982; Rosenzweig-Lipson and Barrett, 1995; Rosenzweig-Lipson et al., 1997a). The present study extends the behavioral effects associated with 4-AP to include DS effects.

Like 4-AP, other aminopyridines block KV channels in biochemical and electrophysiological studies (Yeh et al., 1976; Kirsch and Narahashi, 1978). In addition, 4-AP and aminopyridine analogs release several neurotransmitters, including DA, NE, 5-HT, and ACh (Damsma et al., 1988; Scheer and Lavoie, 1991). In the present study, 4-AP, 3-AP, 2-AP, and 2,3-DIAP produced dose-dependent increases in the percentage of responses on the 4-AP-associated lever and fully substituted for 4-AP, whereas 2,6-DIAP and 3,4-DIAP only partially substituted for 4-AP. 4-DMAP and pyridine did not substitute for 4-AP. In general, doses of the aminopyridines that partially or fully substituted for 4-AP produced modest decreases in response rate. The rank order of potency for producing 4-AP-like DS effects was: 4-AP > 3-AP = 2,3-DIAP > 2-AP > 2,6-DIAP = 3,4-DIAP > 4-DMAP pyridine. The rank order of potency for the aminopyridines differs across biochemical, electrophysiological, and behavioral studies. Differences between in vitro and in vivo studies may reflect the differential ability of the aminopyridines to cross the blood-brain barrier. However, there are also different rank orders of potency across behavioral studies. For example, the rank order of potency of the aminopyridines for producing full or partial substitution for 4-AP in the present study differs from the rank order of potency for decreasing locomotor activity, which differs from the rank order of potency for producing convulsions (Brandsgaard et al., 1995). In addition, all of the aminopyridines decrease locomotor activity and produce clonic convulsions, whereas in the present study not all aminopyridines fully substituted for 4-AP. Although differences in brain penetrance may contribute to the potency differences across aminopyridines, it is unlikely that it contributes to differences in the rank order of the aminopyridines across behavioral effects (i.e., decreasing locomotor activity, producing convulsions, or substituting for 4-AP). Because the aminopyridines block many different subtypes of KV channels, it is possible that these differences may reflect differing affinities and/or efficacies of the aminopyridines at the KV subtypes. Along this line, one behavior may be mediated by one of the KV subtypes, whereas another behavior may be mediated by a different subtype. As compounds are identified that demonstrate subtype selectivity for KV channels, more specific characterization of the KV channels mediating the DS effects of 4-AP will be elucidated.

The DS effects of 4-AP are not mimicked by quinine and linopiridine, which block ATP-dependent, Ca²⁺-activated, or M-type currents (Yeh et al., 1976; Cook, 1988; Fatherazi and Cook, 1991). Blockade of KV channels but not ATP or Ca²⁺-activated K-channels enhances neurotransmitter release (Schechter, 1997). The ATP-activated K-channel openers cromakalim and pinacidil were tested in antagonism studies and generally failed to block the DS properties of 4-AP. Cromakalim appeared to partially attenuate the DS effects of 4-AP at 0.1 mg/kg (40%); however, this effect occurred only at one dose and was not dose-dependent. This may be due to their action at ATP-dependent and not at KV channels (Kajioka et al., 1990; Häusser et al., 1991). Taken together, these results suggest that the DS effects of 4-AP are likely to be mediated via voltage-gated, but not Ca²⁺-dependent or ATP-dependent, K-channels.

The present study suggests that the DS effects of 4-AP can be dissociated from its convulsant properties for several reasons. First, 4-AP serves as a DS at doses lower than those required for producing convulsions. Second, the proconvulsant drug pentylenetetrazole did not substitute for 4-AP. Third, there is no clear relationship between the rank order of aminopyridines for producing 4-AP-like DS effects and for inducing convulsions. Fourth, previous studies have demonstrated that anticonvulsants such as benzodiazepines, phenytoin, and the NMDA antagonist dizolcipine can block 4-AP-induced convulsions (Fragoso-Veloz and Tapia, 1992; Cramer et al., 1994; Stork and Hoffman, 1994; Juhng et al., 1999). Although in the present study, benzodiazepine agonists such as CDP and diazepam dose dependently blocked the DS effects of 4-AP, the anticonvulsant phenytoin partially attenuated the DS effects of 4-AP at only one dose and dizolcipine did not attenuate the DS effects of 4-AP. Taken together, these results suggest that the DS effects of 4-AP are not directly linked to its convulsant properties.

Previous studies have shown that administration of 4-AP results in the release of several neurotransmitters, including NE, 5-HT, DA, and ACh (Damsma et al., 1988; Hu and Fredholm, 1991; Xiao et al., 1993; Dawson and Routledge, 1995; Versteeg et al., 1995; Schechter, 1997). To determine whether enhanced levels of these neurotransmitters contribute to the DS effects of 4-AP, indirect agonists (releasers and/or reuptake inhibitors) were evaluated in substitution studies. Indirect 5-HT, DA, and ACh agonists generally did not substitute for 4-AP. However, the NE reuptake inhibitor tomoxetine (50%), but not nisoxetine (20%), partially substituted for 4-AP. In addition, both imipramine (39%) and venlafaxine (19%), which block both NE and 5-HT reuptake, also failed to substitute for 4-AP. Although the effects of imipramine appear to be similar to those of tomoxetine, it should be noted that imipramine very weakly substitutes only at a dose that markedly decreases response rate. Taken together, these results are not consistent with a role for NE reuptake inhibition in mediating the effects of 4-AP, but suggest that tomoxetine blocks K-channels.

In addition to the indirect effects of 4-AP in mediating neurotransmitter release, 4-AP binds at high concentrations to several receptors, including noradrenergic (₁, ), serotonergic (5-HT_1A, 5-HT_2A), dopaminergic (D₁, D₂), and muscarinic (M₂) receptors (Drukarch et al., 1989). To assess the role of these receptors in mediating the DS effects of 4-AP, agonists and/or antagonists for each of these receptors were evaluated in substitution and/or antagonism studies. For the most part, agonists and antagonists from these systems failed to consistently substitute or antagonize the DS effects of 4-AP. However, in several instances a compound either substituted for or antagonized the DS effects of 4-AP. For example, the ₁ antagonist prazosin attenuated the DS effects of 4-AP (50%) at one or more doses, suggesting that ₁ stimulation contributes to the DS effects of 4-AP. However, the ₁ agonist cirazoline did not substitute for 4-AP, suggesting a limited role for ₁ receptors in mediating the DS effects of 4-AP. Interestingly, both DOI and mCPP partially substituted for 4-AP, suggesting a possible role for 5-HT_2A and/or 5-HT_2C receptors in the DS effects of 4-AP. Consistent with a possible role for 5-HT2A receptors, 4-AP induces headshakes that can be antagonized by the 5-HT_2A/2C antagonist ketanserin (Gorman et al., 1995). However, neither ketanserin nor the 5-HT_2B/2C antagonist SB-200646 antagonized the DS effects of 4-AP, suggesting that 5-HT_2A or 5-HT_2C receptors play only a limited role in the DS effects of 4-AP.

4-AP-induced release of NE can be blocked by ₂-adrenoceptor agonists and can be potentiated by the ₂-adrenoceptor antagonist yohimbine (Hu and Fredholm, 1991), suggesting that ₂ receptors are involved in regulating 4-AP's effects on NE systems. Interestingly, yohimbine partially substituted for 4-AP (46%). However, another ₂ antagonist idazoxan was less effective in substituting for 4-AP (28%). In view of this differential profile, it is possible that the potentiation of 4-AP's effects on NE release by yohimbine as well as the partial substitution of yohimbine for 4-AP are due to K-channel blocking properties of yohimbine.

In the present study, the benzodiazepine agonists CDP and diazepam attenuated the DS effects of 4-AP (77% and 58%). The benzodiazepine antagonist flumazenil did not significantly attenuate the DS effects of 4-AP. Neither CDP nor flumazenil substituted for 4-AP when administered alone. Previous studies in our laboratory have shown that the increases in punished responding produced by CDP can be attenuated by pretreatment with 4-AP (Rosenzweig-Lipson et al., 1996). These effects may be explained by the ion flow changes caused by benzodiazepines and K-channel blockers. When 4-AP blocks K-channels, it prevents the efflux of potassium resulting in depolarization (Glover, 1981). Benzodiazepine agonists bind to GABA_A/Cl ion channel complex causing an influx of Cl resulting in hyperpolarization. Thus, the interactions between 4-AP and benzodiazepine agonists may result from the opposite charges associated with blocking K-channels and stimulating benzodiazepine receptors.

The present study demonstrates that blockade of voltage-dependent K-channels engenders unique DS properties. In conjunction with other studies demonstrating that calcium channel openers and blockers can also be trained as discriminative stimuli (Gladstein et al., 1987; De Jonge et al., 1993; de Beun et al., 1996), the present studies extend the utility of drug discrimination procedures to voltage-gated ion channels.