Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The 5-HT_1B receptor is a G_i-protein coupled receptor that regulates the release of 5-HT via modulation of K⁺ or Ca²⁺ channels in 5-HT neurons (Engel et al., 1986; Maura et al., 1986). The receptor also exists as a presynaptic heteroreceptor, additionally regulating cholinergic, dopaminergic, glutamatergic, and GABAergic transmission (Tanaka and North, 1993; Hartig et al., 1996). A single amino acid residue difference in the transmitter binding domain of the receptor distinguishes the pharmacology of the receptor between human and rat or mouse receptor (threonine in the human and asparagine in the rodent; Oksenberg et al., 1992), whereas such differences seem to be minimal between guinea pigs and humans. As a result, many compounds display marked differences in affinity between the human and rat forms of the receptor (Oksenberg et al., 1992). It should be noted that with compounds that do bind preferentially to the rat form of the receptor, such as NAS-181 (Stenfors et al., 2000), the 5-HT_1B receptors seem to regulate release of 5-HT as it does in the guinea pig (Stenfors et al., 2001). However, because of these species differences, it was necessary to carry out functional assays in species containing more human-like receptors, such as the guinea pig and primate used in the present studies (Zgombick et al., 1998; Alberts et al., 2000). Although the sequence of the pigeon form of the receptor remains unknown, the in vivo pharmacology described in the present studies would suggest that it is closer to the human and guinea pig forms of the receptor than it may be to the rat form.

Serotonergic agents, such as the selective serotonin reuptake inhibitors (SSRIs) offer safety advantages over older medications for the treatment of depression, such as the tricyclic antidepressants, but are associated with their own spectrum of undesired effects. These include sexual dysfunction, acute heightened anxiety, and sleep disturbances (Goldstein and Goodnick, 1998). An improvement in acceptability and compliance may be possible if serotonergic neuronal function could be enhanced by a mechanism other than inhibition of the serotonin transporters. Antagonism of the 5-HT_1B autoreceptors may offer this possibility. It is believed that these receptors function as inhibitory autoreceptors, to limit serotonin release as synaptic concentrations increase. In addition, the regional distribution of 5-HT_1B receptors differs from that of the 5-HT transporter, and it is possible that as a result, perhaps this class of compounds might present fewer, or at least different undesired effects than compounds interacting with the transporter. For example, although there are high numbers of 5-HT transporters in dorsal raphe nucleus (Bengel et al., 1997), few 5-HT_1B binding sites are present in this region (Bonaventure et al., 1998; Sari et al., 1999), despite high levels of 5-HT_1B message. Thus, 5-HT_1B receptors do not seem to be involved in the somatodendritic autoinhibitory regulation. An additional difference between agents interacting with the transporter and those with the 5-HT_1B receptor is that after acute administration with SSRIs, there is an initial decrease in synaptic 5-HT conferred by feedback inhibition upon the autoreceptor population (Blier and de Montigny, 1983; Hjorth and Sharp, 1990), which is ultimately overcome, perhaps by desensitization of the autoreceptors (Blier and de Montigny, 1983, 1985). It has been speculated that the delayed onset of therapeutic effects of SSRIs is related to the delayed restoration of basal levels of 5-HT. Direct blockade of the 5-HT_1B autoreceptors, however, results in no initial decrease in synaptic 5-HT (Roberts et al., 1997; Stenfors et al., 1999) and an immediate increase in turnover (Stenfors et al., 1999, 2001), suggesting the possibility that this class of drugs may in fact provide a faster onset of action than current antidepressants.

A number of preclinical assays exist to determine the potential for anxiolytic or antidepressant activity. In one such procedure, punished responding, behavior is suppressed by a noxious event, such as by delivery of electrical current, and clinically useful anxiolytics reverse the behavioral suppression (McMillan and Leander, 1975). Several investigators have previously used separation-induced vocalization in guinea pig pups (Hennessy et al., 2001), although this procedure has been less extensively characterized than suppressed responding procedures. The procedures used to determine putative antidepressant activity include assessment of pharmacological activity in operant procedures, such as differential reinforcement of low rates schedules (DRL), in which behavioral efficiency is increased by antidepressant administration (O'Donnell and Seiden, 1985; Sokolowski and Seiden, 1999), or in response duration differentiation (RDD) schedules, in which mean lever hold durations are increased by antidepressants (Hudzik and McMillan, 1994; Kinney et al., 1998). The RDD procedure, although not used nearly as much as other procedures for assessing antidepressant potential, has clearly demonstrated efficacy of several classes of antidepressants (e.g., SSRIs, monoimine oxidase inhibitors, selective dopamine reuptake inhibitors), and furthermore, the rank potencies in the procedure agree with their clinical potencies (Hudzik and McMillan, 1994). A model more closely aligned with the clinical aspects of depression is learned helplessness, in which animals are conditioned to fail to escape from a noxious stimulus (Seligman, 1978). This effect on escape failures is blocked by antidepressants. All of these models have been extensively utilized and validated in rats and, in the case of suppressed responding, in pigeons and squirrel monkeys as well (McMillan and Leander, 1975; Sepinwall et al., 1978; Patel and Migler, 1982). The pharmacological differences between rodent and human 5-HT_1B receptors have made assessment of compounds in such rodent models difficult, and hence, to assess the anxiolytic and antidepressant potential of 5-HT_1B antagonists, it was necessary to adapt the standard rodent models to guinea pigs and pigeons. This also meant that many frequently used procedures for assessing anxiolytic or antidepressant activity, such as the elevated plus maze or forced swim test could not be used, because they are difficult to adapt to guinea pigs and other nonrodent test species.

AR-A000002 is a novel selective 5-HT_1B antagonist that increases 5-HT turnover and release in vivo (Stenfors, 2001; Stenfors et al., 2001). The compound has a K_i value of 0.47 nM against the 5-HT_1B receptor and exhibits 10-fold selectivity against 5-HT_1D receptors, and >300-fold selectivity against 60 other receptors studied (Stenfors, 2001). In the present study, the effects of AR-A000002 are characterized in a number of preclinical assays to confirm the in vivo pharmacological profile of this compound and to determine its potential utility as an antidepressant and anxiolytic agent.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Subjects

All subjects were individually housed (exceptions below) in standard caging within the animal facilities of AstraZeneca R & D (Rochester, NY) (pigeons: suppressed responding, DRL), AstraZeneca R & D (Boston, MA) (guinea pigs: learned helplessness), AstraZeneca R & D (Wilmington, MA) (guinea pigs: suppressed responding), AstraZeneca R & D (Sodertalje, Sweden) (guinea pigs: drug discrimination, DRL, vocalization), and McLean Hospital (Belmont, MA) (squirrel monkeys) and were maintained under a 12-h light/dark cycle. Guinea pigs used for drug discrimination and DRL studies were housed two per cage, and guinea pig pups used for the separation-induced vocalization studies were housed with their mothers. Further details are described in the sections for individual procedures. Temperature and humidity were maintained according to Association for Assessment and Accreditation of Laboratory Animal Care standards in all AstraZeneca facilities.

5-HT_1B/1D Agonist (CP135,807) Drug Discrimination

Sixteen male Duncan-Hartley guinea pigs (Mollegaard, Ejby, Denmark) were used. Subjects weighed 450 g at the start of the experiment and 950 g by its conclusion. Animals were maintained on a restricted diet by providing 20 g of feed per day. Water was available at all times, except during experimental testing.

Animals were initially trained in standard rat two-lever operant chambers to respond under a fixed ratio (FR) 32 schedule of food presentation (45-mg food pellet), delivered to a cup located inside the chamber (each 32nd lever press was reinforced by food pellet delivery). Once reliable responding occurred on both levers in the operant chamber, defined as less than 25% variation in the number of lever presses made for several consecutive days, responding on only one of the levers was made contingent upon the injection that preceded the session. The subjects were administered either an injection of the 5-HT_1B/D agonist CP135,807 (3 mg/kg) or of vehicle 45 min before each session. Responding on one lever in the operant chamber was reinforced if the session was preceded by an s.c. injection of CP135,807 (e.g., left-hand lever), and responding on the opposite lever in the chamber (e.g., right-hand lever) was reinforced if the session was preceded by a vehicle injection. Daily injections were randomized, except that no single injection type could be administered for three consecutive days. Once injection-appropriate responding predominated, the schedule was again changed, such that daily sessions consisted of two components: sessions began with a fixed interval (FI) 120-s schedule, such that the first response occurring after 120 s had elapsed resulted in food delivery. After this initial FI, the schedule was a superimposed FR20, variable interval 20 s (VI20), in which the 20th or greater response occurring after an average of 20 s (range 10-30 s) resulted in food delivery. This ensured a constant rate of responding throughout daily 15-min sessions.

Once animals achieved a criterion of 9 of 10 consecutive days in which the injection-appropriate lever was chosen (20 responses were completed on the injection-appropriate lever before completion of 20 responses on the incorrect lever), drug testing began. Test sessions consisted of only the first 2 min of a regular training session (the FI component), and responding was not reinforced on these days. AR-A000002 (3, 10, and 30 mg/kg) was administered s.c. 45 min before CP135,807, which in turn was administered 45 min before operant testing. The highest dose of AR-A000002 was also administered on its own to determine whether any agonist-like effects were noted. The proportion of responses on the CP135,807-appropriate lever and the total number of responses emitted were recorded. ANOVA was used (within subjects design) to compare performance under training conditions (responding to the CP135,807 cue) with combinations of CP135,807 and AR-A000002.

Anxiolytic Screens

Separation-Induced Vocalization. The subjects were Duncan-Hartley guinea pig pups. Pregnant mothers (approximately 4 weeks pregnant) were purchased from Mollegaard. The mothers were housed singly in large cages, with free access to food and water. Guinea pigs were born about 4 weeks after arrival in the laboratory and tested at least 5 days after birth for the first time. The pups were then tested once a week (between 2:00 and 5:00 PM). All pups were tested repeatedly until the study was completed a month later.

Suppressed Responding in the Pigeon: Acute and Subchronic Studies. Six male Racing pigeons, ranging in weight from 400 to 480 g, were used. Animals were maintained at 85 to 90% free-feeding weight by restricted, post-testing feeding and were trained and tested in standard pigeon operant chambers (Ralph Gerbrands Co., Arlington, MA; MED Associates, St. Albans, VT). The chambers were equipped with a house light situated at the top of the chamber and three translucent response keys on the front panel. Only responses on the center key had any programmed consequences. At programmed times, responses (pecks) on the center key resulted in operation of a food magazine, which delivered access to grain from a receptacle at the bottom of the front panel. The chambers were also equipped with shock generators (MED Associates) that delivered electrical current to a cable that was fed through the tops of the chambers.

Suppressed Responding: Guinea Pigs. Eleven male Duncan-Hartley guinea pigs were used. Subjects weighed 730 to 940 g at the time of testing and were food restricted to 85% of free-feeding weight by postsession feeding with approximately 25 g of standard guinea pig chow per day. Diet was also supplemented with daily alfalfa (to supplement vitamin C), and all animals had free access to water, except during experimental testing. Subjects were individually housed throughout the course of the experiments under a 12-h light/dark cycle. Animals were trained and tested in standard two-lever rat operant chambers (MED Associates). The chambers were fitted with two retractable response levers, and a stimulus lamp over each of the two levers. A pellet dispenser delivered 45-mg food pellets (P.J. Noyes Company, Inc., Lancaster, NH) formed from standard guinea pig chow, to a cup located inside of the chamber, below and between the two response levers. A lamp at the top and back of the chamber served as a house light. The grid floors of the operant chambers were interfaced to shock generators and scramblers (MED Associates). All events in the chambers were controlled and monitored by a microprocessor.

Suppressed Responding: Squirrel Monkeys. Three male squirrel monkeys (Saimiri sciureus) were used, weighing between 0.7 and 1.0 kg at the time of testing. Animals were individually housed and were given restricted access to food. Monkeys were trained and tested in standard small primate operant chambers to respond under a multiple fixed ratio schedule of food reinforcement. Under the multiple schedule, components during which white or red stimulus lights were illuminated alternated throughout the session. In the presence of the white stimulus lights, completion of 30 lever presses resulted in the delivery of a 190-mg banana-flavored food pellet and initiated a 10-s time-out period during which the chamber was dark and responding had no programmed consequences. In the presence of the red stimulus lights, completion of the FR30 response requirement resulted in the delivery of a food pellet and initiated a 10-s time-out period. Additionally, a second schedule was superimposed under which completion of every 50th response in the presence of red stimulus lights produced a brief electric shock to the monkey's tail (200 ms; 0.5-2.0 mA). Shock intensity was adjusted for individual monkeys to levels that suppressed responding in the red stimulus light components to less than 10% of nonsuppressed response rates, i.e., response rates in the white stimulus light components.

Antidepressant Screens

DRL25/30: Pigeon. Seven male white Carneux pigeons, ranging in weight from 400 to 540 g, were used in the present experiments. Animals were maintained at 80 to 85% of free-feeding weight by restricted, postsession feeding.

DRL72: Guinea Pigs. Sixteen male, pigmented three-color guinea pigs were used (Harlan Sera-Lab, Crawley Down, Sussex, UK), weighing around 950 g at the time of testing. Food intake was restricted to 20 g/animal/day. Animals had free access to water, except during experimental testing. Animals were given supplemental access to fresh cabbage and hay as a source of additional vitamin C.

RDD: Guinea Pig. Six male Duncan-Hartley guinea pigs were used. Animals were maintained at 90% of initial free-feeding weights by postsession feeding.

Learned Helplessness: Guinea Pig. Seventy male Hartley guinea pigs were used. Subjects were 350 to 425 g in weight at the time of testing, were fed ad libitum, and were housed under a 12-h light/dark cycle.

Drugs

For the drug discrimination, separation-induced vocalization, and DRL studies in guinea pigs, 2'-methyl-4'-(5-methyl-[1,2,4]oxadiazol-3-yl)-biphenyl-4-carboxylic acid [4-(5-methoxy-3(4-methyl-piperazin-1-yl)-phenyl]amide (GR127935) (1-10 mg/kg), and 3-(N-methylpyrrolidin-2R-ylmethyl)-5-(3-nitropyrid-2-ylamino)-lH-indole (CP135807) and (R)-N-[5-methyl-8-(4-methylpiperazin-1-yl)-1,2,3,4-tetrahydro-2-naphthyl]-4-morpholinobenzamide (AR-A000002) (3-30 mg/kg) were provided as the free bases and dissolved in the minimum amount of glacial acetic acid. The volume was made up with 0.9% saline solution, which served as the vehicle control in all of the experiments. Zolpidem base (Synthelabo Group distributed by Astra Läkemedel, Sodertalje, Sweden; 1-10 mg/kg), diazepam base [2-ml ampoules; Roche Diagnostics, Indianapolis, IN; 5.0 mg/ml in Abbott's cocktail (1-10 mg/kg), fluoxetine HCl (Eli Lilly & Co., Indianapolis, IN; 1-10 mg/kg), citalopram HCl (H. Lundbeck A/S, Copenhagen, Denmark; 0.3-3 mg/kg), and desipramine HCl (10 mg/kg)] were all dissolved in saline. All drugs were administered subcutaneously (except diazepam, which was given intraperitoneally) 1 h before sessions, and except where otherwise specified above and below. Doses were calculated as the free bases.

For the DRL and punished responding studies in pigeons, and RDD studies in guinea pigs, AR-A000002 (0.1-1 mg/kg) was dissolved in distilled water with 1 to 2 drops of 85% lactic acid. Amitriptyline HCl (1 and 3 mg/kg) and fluoxetine HCl (1 and 10 mg/kg) were dissolved in 0.9% saline. Diazepam (0.1-3 mg/kg) was dissolved in Abbott's cocktail (40% propylene glycol, 10% ethanol, and 50% water) or was provided already dissolved in this vehicle. Drugs were administered i.m. into the breast muscle in a volume of 1 ml/kg 45 min before sessions. Corresponding vehicle was appropriate for the compound being tested.

For the suppressed responding studies in the squirrel monkey, AR-A000002 (0.1-1 mg/kg) was dissolved in distilled water with 1 to 2 drops of 85% lactic acid. Fluoxetine (1-78 mg/kg) was dissolved in distilled water. Diazepam (0.1-3 mg/kg) was dissolved in Abbott's cocktail. Drugs were administered i.m. into the thigh muscle in a volume of 1 ml/kg b.wt. 45 min before sessions. Vehicle control for AR-A000002 and fluoxetine HCl consisted of distilled water with 1 to 2 drops of lactic acid (for AR-A000002) or without acid (for fluoxetine). Vehicle control for diazepam consisted of the Abbott's cocktail.

For the suppressed responding and learned helplessness studies in guinea pigs, AR-A000002 was dissolved in distilled water with 1 to 2 drops of 85% lactic acid (which served as vehicle control) and administered s.c. in a volume of 1 mg/kg b.wt. 45 min before (suppressed responding) and immediately after (learned helplessness) testing.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Drug Discrimination. Guinea pigs trained to discriminate 3 mg/kg CP135,807 from saline emitted 95% of their responses on the lever associated with drug when administered CP135,807 before sessions. There was a main effect of drug treatment upon response choice [F(3,12) = 10.9, P < 0.001]. Thirteen percent of responses occurred on the lever associated with drug when vehicle was administered before sessions (Fig. 1). AR-A000002 dose dependently blocked the discriminative cue produced by CP135,807 (Fig. 1). AR-A000002 did not produce CP135,807-like discriminative effects when administered alone after a dose of 30 mg/kg (data not shown).

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Mean proportion of responses emitted on the CP135,807-appropriate lever in 16 guinea pigs trained to discriminate 3 mg/kg CP135,807 from vehicle. Effects of vehicle administration (left-most column) of the training dose of CP135, 807 (second column) and of the training dose combined with 3 mg/kg (third column) and 10 mg/kg (fourth column) AR-A000002. *, P < 0.05, compared with vehicle, Dunnett's post hoc after significant omnibus main effect of drug treatment.

Separation-Induced Vocalizations. Guinea pig pups separated from their mothers emitted from 150 to 175 calls/min under control (nondrug) conditions (Fig. 2). The benzodiazepine ligands diazepam [5 and 10 mg/kg; F(4,19) = 74, P < 0.001] and zolpidem [3 and 10 mg/kg; F(4,16) = 14, P < 0.001] dose dependently blocked the number of calls emitted by the pups. The 5-HT_1B antagonists AR-A000002 [F(4,12) = 4.4, P < 0.05] and GR127,935 [F(4,28) = 4.3, P < 0.01] both also reduced the number of calls at 30 and 10 mg/kg, respectively (Fig. 2), although not to the same degree as the benzodiazepine ligands. Neither SSRI fluoxetine [F(4,28) = 1.2, P = 0.31] nor citalopram [F(4,16) = 0.4, P = 0.9] produced any significant reduction in calls after the doses tested.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Mean calls during a 2-min observation period in groups of four to eight guinea pig pups in the separation-induced vocalization paradigm. Animals were tested first without any treatment (pre), after vehicle (sal), and after doses (x-axes, mg/kg) of diazepam, zolpidem, AR-A000002, GR127935, citalopram, and fluoxetine. *, significant difference from vehicle control (repeated measures ANOVA/Dunnett's post hoc test; within-subjects design).

Suppressed Responding: Pigeons. Under control conditions, rates of responding in pigeons under the unpunished component of the multiple schedule averaged 0.84 ± 0.1 S.E.M. responses/s. Rates of responding under the punished component averaged 0.41 ± 0.07 S.E.M. responses/s. Across vehicle control days, an average of 84 was delivered (Table 1), which ranged from 64 to 94. The effects of AR-A000002, fluoxetine, and diazepam on rates of responding in both the punished and unpunished components are shown in Fig. 3. Diazepam produced a significant increase in punished responding after the 0.3- and 1-mg/kg doses, with a maximal effect after the higher dose of 450% of control values (P < 0.01 versus control). This increase in punished responding was paralleled by a much smaller increase in unpunished rates of responding, demonstrating some specificity of the anxiolytic effect. The number of shock deliveries increased in a dose-related manner after diazepam administration, to a maximum of 207 after the 1-mg/kg dose (Table 1). At the highest diazepam dose tested (3 mg/kg), rates of responding were considerably suppressed.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Effects of AR-A000002 (top), fluoxetine (middle), and diazepam (bottom) on mean (±S.E.M.) percentage of control response rates in a suppressed responding in pigeons (n = 6). Filled circles denote percentage of control rates during suppressed components, and open circles denote percentage of control rates during unsuppressed components. Horizontal lines indicate 99% confidence limits about control rates in the suppressed component, and horizontal dashed lines the 99% confidence limits about control rates in the suppressed components. Horizontal axis: dose, milligrams per kilogram s.c. *, significant difference from control (t test).

Subchronic Treatment with AR-A000002. Fig. 4 shows the effects of vehicle, acute AR-A000002 (0.1 mg/kg) and of daily administration of 0.1 mg/kg AR-A000002 on unsuppressed (open circles) and suppressed (filled circles) responding. On the 1st day of AR-A000002 administration, there was a modest, nonsignificant increase in rates of punished responding, as was previously shown in the initial generation of the dose-effect curve (Fig. 3). However, by the 2nd day of AR-A000002 administration, there was a 1.5-fold increase in efficacy from the previous day, now resulting in a statistically significant increase in suppressed responding. This increase in suppressed responding was maintained over the following 5 days of daily dosing and had fully returned to prebaseline levels by the following vehicle control day.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Effects of 0.1 mg/kg AR-A000002 on mean (±S.E.M.) percentage of control response rate in the suppressed responding in pigeons after daily administration, 45 min before sessions (n = 6). x-axis: ctl, effects after vehicle administration. A, effects of 0.1 mg/kg after acute administration, 1 to 7 (consecutive day of subchronic administration). Filled circles denote percentage of control rates during suppressed components, and open circles denote percentage of control rates during unsuppressed components. The horizontal lines indicate 99% confidence limits about control rates in the suppressed component. The filled and unfilled symbols to the right of the figure show the effects of vehicle control on the following week expressed as percentage of the previous week's control, demonstrating that behavior returned to baseline after chronic treatment.

Suppressed Responding: Guinea Pigs. Average rates of responding on control days ranged from 1.0 to 1.4 responses/s in the unsuppressed component and from 0.1 to 0.38 responses/s in the suppressed component. AR-A000002 (0.3 mg/kg) engendered a significant increase in the percentage of control rates of suppressed responding, whereas 0.1 and 1 mg/kg diazepam was active in increasing rates of suppressed responding (Table 2). Neither drug significantly altered rates of responding in the unsuppressed component across the dose range studied.

Suppressed Responding: Squirrel Monkeys. On control days, responding in the presence of the red stimulus light (suppressed component) was suppressed to rates that were less than 5% of the mean response rates in the presence of white stimulus lights. Control, nonsuppressed response rates ranged from 2.49 to 3.32 responses/s with a mean of 3.04 ± 0.45 responses/s. Control suppressed rates of responding ranged from 0.00 to 0.16 responses/s with a mean of 0.07 ± 0.08 responses/s.

DRL: Pigeons. Under control (vehicle) conditions, subjects earned, on average, 1.7 to 3.5 reinforcers, and averaged slightly over 100 responses. Figure 5, top, shows the effects of amitriptyline and fluoxetine on reinforcers earned in the session. Both amitriptyline (3 mg/kg) and fluoxetine (10 mg/kg) evoked a 4-fold increase in the number of reinforcers earned. These drugs evoked the response without altering the total number of responses emitted (Table 4), indicating that the effect was not secondary to motor slowing. Of the AR-A000002 doses tested, the 1-mg/kg dose was similarly effective in increasing the number of reinforcers earned. This, too, occurred in the absence of alterations in the numbers of responses emitted (Table 4).

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Mean (±S.E.M.) effects of doses (x-axes) of amitriptyline and fluoxetine (top) and of AR-A000002 (bottom) on the number of reinforcers earned (y-axes) in pigeons responding under a DRL25 or DRL30 schedule of food presentation (n = 7). *, significant difference from mean after vehicle administration (t, P < 0.05).

DRL: Guinea Pigs. Guinea pigs, on average, earned 8 to 12 reinforcers (75-100 responses emitted) in a 60-min session under control conditions. Fluoxetine (10 mg/kg) and desipramine (10 mg/kg) produced small, but significant increases in the numbers of reinforcers earned in the session (Fig. 6). AR-A000002 (30 mg/kg) produced a comparable increase in reinforcers. Of the drugs, only fluoxetine significantly reduced response rates to about 75% of the vehicle control values (data not shown).

View larger version (10K): [in this window] [in a new window]   Fig. 6.   Mean ± S.E.M. response rates (y-axis) in guinea pigs responding under a DRL72 schedule of food presentation (n = 16) after saline administration, and after doses of fluoxetine (10 mg/kg), desipramine (10 mg/kg), and AR-A000002 (30 mg/kg).

RDD: Guinea Pigs. Under baseline, guinea pigs emitted 290 ± 47 responses and earned 136 ± 19 reinforcers. Hence, 46 ± 6% of responses were reinforced. Mean response durations approximated the minimum required response duration, averaging 1.13 ± 0.14 s. The effects on mean response duration of fluoxetine and amitriptyline (top) and of AR-A000002 are shown in Fig. 7. Both of drugs produced increases in mean response durations. As with the standards, AR-A000002 (0.3 mg/kg) also increased mean response durations. Amitriptyline produced flattening of the relative frequency distribution at doses that increased mean response duration (Fig. 8), indicating some disruption of overall responding, but there was relatively less flattening produced by fluoxetine and AR-A000002.

View larger version (16K): [in this window] [in a new window]   Fig. 7.   Effects of doses (x-axes) of fluoxetine and amitriptyline (top) and of AR-A000002 (bottom) on mean response duration (y-axes) in guinea pigs responding under the RDD1 to 1.3 schedule. n = 6 at each data point. *, significant difference from vehicle control (CTL) (t, P < 0.05).

View larger version (22K): [in this window] [in a new window]   Fig. 8.   Effects of the doses that increased mean response duration (from Fig. 7) on the distributions of response durations. x-axes: mean ± S.E.M. percentage of total response durations. y-axes, response duration (seconds), in successive 0.1-s intervals. Hatched columns represent responses falling within the durations that were reinforced (1-1.3 s).

Learned Helplessness: Guinea Pig. Table 5 shows the mean and S.E.M. escape failures on the 1st day of avoidance training. None of the drug-treated groups differed significantly from the vehicle-treated group (B). However, by the 2nd day of avoidance training, the effect of drug treatment became apparent. Figure 9 shows the effects of the various treatments on escape failures in the second avoidance training session. There was a significant main effect of treatment level (F = 4.29, P < 0.002). In the no-induction group (A), the number of trials in which animals failed to escape shock averaged only 1.4 ± 1.1. This contrasts with the number of escape failures after induction and saline administration (16.8 ± 7), and these two values were significantly different (Dunn's post hoc, P < 0.05). Imipramine (17.8 mg/kg) treatment (group C) also produced fewer escape failures than the saline-treated group, as did 1 mg/kg AR-A000002 (group E). AR-A000002 at a dose of 0.3 mg/kg (group D) produced a trend toward an effect that was not statistically significant (P = 0.06), but the highest dose used (5 mg/kg, group F) was without effect.

View larger version (12K): [in this window] [in a new window]   Fig. 9.   Mean escape failures (y-axis) on the 2nd day of avoidance training in groups of guinea pigs (A-F, x-axis) in the learned helplessness procedure. *, significant difference from vehicle control group A (ANOVA, Dunn's post hoc). n = 11 to 12 for each data point. A, vehicle control. B, group not submitted to inescapable shock. C, group treated with imipramine, 17.8 mg/kg. D to F, groups treated with 0.3, 1.0, and 5.0 mg/kg AR-A000002, respectively.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

AR-A000002 is a selective antagonist at 5-HT_1B receptors. Its in vivo activity at the receptor was evidenced in the present study by blockade of the discriminative stimulus effects of the 5-HT_1B/1D agonist CP135,807, although lacking any agonist effects on its own. Although this study provides some evidence that the compound blocks the receptor in vivo, better pharmacological characterization of the discriminative stimulus effects of CP135,807 (i.e., demonstration of nonblockade of the stimulus effects of the agonist with 5-HT_1A antagonists) would have strengthened this study. However, it can be noted that in a separate group of animals trained to discriminate the 5-HT_1A agonist 8-OH-DPAT from vehicle (J. E. Evenden, unpublished observations) administration of AR-A000002 did not result in generalization from the 8-OH-DPAT stimulus, nor blockade of the 8-OH-DPAT stimulus, suggesting some pharmacological specificity of the effects of AR-A000002. Finally, AR-A000002 has been shown to block the hypothermia induced by an agonist (Stenfors et al., 2001), further supporting that argument that the compound blocks 5-HT_1B receptors in vivo. However, this raises the question that the discriminative stimulus effects of CP135,807 might have been based on hypothermia, which is an interesting possibility that could be explored in future studies.

The results of the present study support the concept that antagonists at the 5-HT_1B receptor may possess therapeutic potential as novel antidepressants and anxiolytics. Given that blockade of the presynaptic 5-HT_1B receptor evokes a net enhancement of 5-HT function (Stenfors et al., 1999, 2001), in a manner similar to SSRIs, a similar profile of activity might be expected in behavioral assays used to evidence therapeutic activity. Efficacy of AR-A000002 in both the guinea pig and pigeon versions of the DRL schedule, and in the RDD and in the learned helplessness procedures in the guinea pig, would suggest antidepressant potential for the compound. Efficacy in the guinea pig pup vocalization paradigm and in the pigeon and guinea pig suppressed responding procedures, in which SSRIs had no significant effects, would also suggest greater anxiolytic potential than the SSRIs. The results of the guinea pig pup vocalization paradigm should be interpreted cautiously, because the vehicle control points were not always appropriate for the test condition (e.g., saline was uniformly used, which did not control for diazepam). Additionally, the procedure required very high doses of the benzodiazepine, diazepam to evoke a reduction in vocalizations, and hence may not have been optimized for this species.

It should, however, be noted that the studies in the present series using guinea pigs as subjects represent some of the first behavioral pharmacology studies to be conducted in this species. Although most of the procedures used in the present study have been used extensively with rats as subjects, their adaptation to the guinea pig was novel. Furthermore, although the studies all contained internal controls with either known antidepressants or anxiolytics, their predictive validity has yet to be established. On the other hand, it was remarkable how similar the behavioral patterns generated in guinea pigs were to those reported previously for the rat. For example, in rats responding under RDD1 to 1.3 schedule, levels of accuracy generally range from 30 to 50%, which was noted in the present study for the guinea pigs. The normal distribution of response durations about a mean that is near to the minimum reinforced response duration noted in rats (Hudzik and McMillan, 1994, 1995; Kinney et al., 1998) was evident in guinea pigs in the present study. Additionally, fluoxetine (as well as AR-A000002) did not produce a disruption of the shape of the response-duration distribution (similar to the previous study in rats; Hudzik and McMillan, 1994) consistent with the known mild or absent sedative effects with fluoxetine, and therefore possibly speaking to lack of sedation with AR-A000002. However, one difference between the rat and the guinea pig drug effects in the RDD was that the magnitude of the effects upon the mean response duration, which is the primary dependent variable, was not as great as in the rat, suggesting that the schedule conditions may not have been optimal for this species. Variables that influenced the magnitude of the antidepressant-like response might include the size, shape, and force required to operate the operant manipulanda with respect to animal size, as well as schedule considerations, such as the minimum lever-press duration required for reinforcement (Hudzik et al., 1999).

Interestingly, the results of the present study may help to clarify effects previously noted in a study using the RDD procedure in rats (Kinney et al., 1998). In that study, the noradrenergic -receptor antagonist propranolol was shown to have measurable effects upon mean response duration that were consistent with antidepressant action. Although the compound is not antidepressant in humans, propranolol does function as a 5-HT_1B antagonist against the rat (but not human) form of the receptor (Oksenberg et al., 1992), which may have contributed to the effect noted in that study.

A survey of the literature revealed that remarkably few studies of DRL responding in pigeons have been conducted. A possible reason for this became apparent during training of pigeons to the task, because they would frequently cease to peck (extinguish) when the minimum time required to elapse from the previous response (IRT) was increased. It became clear that 30 s would be a maximum IRT tolerated by the pigeon, resulting in a difference in the schedule requirements for rats [e.g., 72 s; O'Donnell and Seiden, 1985) and pigeons (25-30 s) in the present study]. However, the baseline rates of responding and the rates of reinforcement were similar among species, perhaps allowing for determination of drug efficacy (increases in reinforcement rate). This suggests that it is the baseline engendered and not the absolute temporal requirements of this schedule that are important in the demonstration of drug efficacy. Similarly, in guinea pigs responding under the DRL72 schedule, rates of responding and reinforcement were fairly similar to those of the pigeon and rat, speaking further to the generality of the behavioral baseline engendered by the schedule, and of the drug effects noted therein.

As with the drug discrimination, RDD, and DRL72, this report contains the first studies using guinea pigs in a learned helplessness procedure. Similar to the other procedures, the behavioral baselines engendered did not differ appreciably from those historically observed in the rat. The procedure did differ, however, from the standard rat procedure, in that 2 days of both conditioning and of avoidance training was required to elicit escape failures, as well as subsequent reversal thereof. Allowing for some procedural adjustments, together the studies presented in the present article demonstrate the feasibility of application of standard techniques in behavioral pharmacology to the guinea pig and their use in profiling the antidepressant and anxiolytic potential of a compound for which testing in rats and mice would not be appropriate.

Although AR-A000002 had consistent qualitative antidepressant and anxiolytic-like effects in most of the procedures reported here, there were marked potency differences of AR-A000002 among the behavioral procedures used in the present study. For example, efficacy in the guinea pig RDD, pigeon suppressed responding, and learned helplessness was shown after doses at or below 1 mg/kg. In contrast, in the DRL72 and in the drug discrimination, the potency was far lower (e.g., 10-30 mg/kg). Although species, pharmacokinetic, strain, and procedural differences may have played a role in influencing the potency of AR-A000002, it is also clear that full, analogous dose-effect functions were not generated in all studies once efficacy had been obtained after at least one dose. For example, in the guinea pig DRL schedule, only a single high dose was tested, and it remains unclear whether lower doses might have been efficacious. This also leaves open the possibility of a bimodal or of a very wide dose-effect function for AR-A000002, which should be addressed in future studies.

A number of agents that have traditionally been used as antidepressants, most notably SSRIs, are now considered to be the clear drugs of choice in the treatment of panic disorder (Uhlenhuth et al., 1998) and are increasingly being used to treat post-traumatic stress disorder. However, onset of therapeutic effect is often delayed by several weeks, and, indeed, the initial response is often characterized by an increase in anxiety, necessitating combination therapy with a benzodiazepine (Uhlenhuth et al., 1998). In agreement with the observed clinical activity, antidepressants as a class demonstrate little, if any, acute anxiolytic efficacy after acute administration in preclinical paradigms. For example, in suppressed responding procedures such as used in the present study, drugs such as fluoxetine, imipramine, desipramine, and amitriptyline are either inactive or in some cases, possess anxiogenic effects (Fontana and Commissaris, 1988, 1989; Bodnoff et al., 1989; Commissaris et al., 1990). After subchronic administration of many of these agents, however, an anxiolytic effect becomes apparent. However, not all antidepressants are active in preclinical screens for anxiety disorders. For example, imipramine, desipramine, amitriptyline, and phenylzine have shown some activity in suppressed responding procedures when administered subchronically (Fontana and Commissaris, 1989; Commissaris et al., 1990), whereas bupropion, trazodone, and mianserin fail to show anxiolytic properties under the same procedure (Commissaris et al., 1990). Additionally, not all studies with imipramine or fluoxetine have demonstrated activity (Beaufour et al., 1999). The fact that AR-A000002 demonstrated anxiolytic efficacy in the present study after acute administration in both the separation-induced vocalization paradigm and suppressed responding in pigeons and guinea pigs suggests some differences from other antidepressants and indicates the potential for a faster onset time than SSRIs when used for anxiety disorders. In agreement with the acute anxiolytic activity found in the present study, Mansbach et al. (1996) tested the effects of the nonselective 5-HT_1B/D antagonist GR127935 in a similar suppressed responding procedure in pigeons, also reporting acute anxiolytic activity.

The very weak activity of AR-A000002 in the squirrel monkey reported in the present study, however, should temper these suggestions of acute anxiolytic potential. There are several possible explanations for this discrepancy. First, the degree of suppression is known to influence the magnitude of anxiolytic activity (Benvenga and Leander, 1996) in mammals, in this case, rat. Responding was suppressed in the squirrel monkey procedure by 98% of the level occurring in the unsuppressed component, whereas in the pigeon procedure, responding was suppressed by only 50 to 75%. Given the relatively modest anxiolytic effects noted in the pigeon suppressed responding procedure and in the separation-induced vocalization, these may have been insufficient to produce a robust anxiolytic effect in the squirrel monkey procedure. A lower level of suppression in the squirrel monkey, or subchronic treatment might have resulted in the appearance of an effect for AR-A000002. Second, it may be the case that there are true species differences in the response to AR-A000002. For example, in pigeons, robust anticonflict effects are often noted with 5-HT_1A agonists (Barrett et al., 1994), whereas such effects are difficult to show in mammals (Gleeson and Barrett, 1990). These compounds, however, are sometimes clinically effective anxiolytics (e.g., buspirone) even if not as effective as benzodiazepines. It also remains possible that in humans, subchronic treatment may be necessary to evoke a strong anxiolytic effect.

These are among the first studies using guinea pigs in a learned helplessness paradigm. Although the procedure was not extensively validated, the effect of AR-A000002 was fairly clear and could not be explained by a direct enhancement of avoidance responding, because there was no relationship between drug effect, avoidance responding, and escape failures (Table 6). For example, the group treated with 5 mg/kg AR-A000002 avoided shocks on the 2nd day of training at about the same level as the animals that were not given uncontrollable shock (group F versus A), and yet group F had the highest level of escape failures and group A the lowest. On day 2 of avoidance training, imipramine-treated (group C) and 5 mg/kg AR-A000002-treated guinea pigs (groups F) had the lowest levels of avoidance responding (Table 6), yet had opposite effects on escape failures (Fig. 9). Additionally, we have since demonstrated activity of the SSRI citalopram in the procedure in guinea pigs (unpublished observations), suggesting that the paradigm may have some validity in this species.

It can be speculated that 5-HT_1B antagonists, because of their ability to facilitate immediate and sustained increases in the turnover of 5-HT, may possess a faster onset of antidepressant action as well. This assertion, however, remains speculative, because there are no preclinical procedures that can accurately predict onset of therapeutic effect after chronic administration. The several days of drug administration during the learned helplessness procedure likely does not parallel the 2- to 4-week requisite administration time for clinical activity of antidepressants, and no other procedure sensitive to chronic treatment with antidepressant exists that is suitable for the species used here. Furthermore, there are currently no antidepressants with fast therapeutic onset times that can be used to validate any such procedure. Hence, it will be necessary to await the clinical data to test this hypothesis.

In summary, AR-A000002 represents the first of a new class of potential anxiolytic and antidepressant compounds to undergo extensive testing in preclinical screens for these activities. Due to the pharmacology of the 5-HT _1B receptor, it was necessary to develop a battery of tests using guinea pigs, pigeons, and monkeys to characterize this compound. These procedures have been validated using known clinically active antidepressants and anxiolytics. AR-A000002 shows behavioral effects in these models consistent with anxiolytic and antidepressant action.