Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Substantial research has implicated -adrenergic receptors in the mechanism of action of antidepressant drugs. Stimulation of either 1- or 2-adrenergic receptors is sufficient to produce antidepressant-like effects on behavior in animal models such as learned helplessness, the forced-swim test, and differential-reinforcement-of-low rate (DRL) behavior (Martin et al., 1986; Finnegan et al., 1987; Frances and Simon, 1987; O'Donnell, 1987, 1990, 1993). In addition, repeated treatment with antidepressants from distinct pharmacological classes results in the down-regulation of central -adrenergic receptors (Mobley and Sulser, 1981; Frazer and Conway, 1984; Ordway et al., 1991); the reduction in the density of 1-adrenergic receptors is much greater than that of 2-adrenergic receptors (Ordway et al., 1991). This suggests a greater involvement of the 1 subtype in the mediation of the effects of antidepressant drugs that act on noradrenergic systems.

Norepinephrine, the major endogenous-adrenergic transmitter in the brain, has a greater affinity for 1-adrenergic receptors than 2-adrenergic receptors (Minneman et al., 1979; O'Donnell et al., 1984). Furthermore, although down-regulation of 2-adrenergic receptors via repeated administration of clenbuterol, a 2-adrenergic agonist, reduces the behavioral effects of 2-selective-adrenergic agonists; antidepressants as well as 1-adrenergic agonists remain fully effective (O'Donnell, 1990). The results of drug discrimination studies also are more consistent with a role for the 1 subtype in the mediation of the effects of antidepressant drugs. It was found that antidepressant drugs, as a class, do not substitute for the discriminative stimulus effects of the 2-adrenergic agonist clenbuterol in rats, suggesting that administration of these drugs does not result in the stimulation of 2-adrenergic receptors in vivo (Makhay and O'Donnell, 1999). In contrast, desipramine does substitute in rats trained to discriminate centrally administered (i.e., i.c.v.) isoproterenol. Through the use of subtype-selective-adrenergic agonists and antagonists, it has been shown that the discriminative stimulus effects of i.c.v. isoproterenol are mediated by 1-adrenergic receptors (Crissman et al., 2001), even though this agonist exhibits similar affinity for and efficacy at 1- and 2-adrenergic receptors (O'Donnell et al., 1984). 1-Adrenergic receptor mediation also has been observed for the effects of i.c.v. isoproterenol on DRL behavior (O'Donnell et al., 1994). The ability of desipramine to substitute for the discriminative stimulus effects of i.c.v. isoproterenol suggests that its administration does result in the stimulation of central 1-adrenergic receptors in vivo (Crissman et al., 2001). These data suggest that antidepressants that interact with noradrenergic neurons produce their effects, in part, via the 1-adrenergic receptor. The present study examined whether antidepressants from different pharmacological classes generalize to the discriminative stimulus effects of centrally administered isoproterenol.

Rats were trained to discriminate i.c.v. administration of 10 µg of isoproterenol from artificial cerebral spinal fluid (aCSF). Once the acquisition criterion was achieved, substitution tests were carried out with antidepressant drugs and related compounds. Drugs tested were the tricyclic antidepressants desipramine and protriptyline, the monoamine oxidase inhibitor phenelzine, the norepinephrine uptake inhibitor nisoxetine, the serotonin uptake inhibitor fluoxetine, the atypical antidepressants bupropion, buspirone, mirtazapine, trazodone, and venlafaxine, the novel, putative antidepressants N-nitro-L-arginine (L-NNA), a nitric-oxide synthase inhibitor, and N-acetyl-L-tryptophan 3,5-bis benzyl ester (L-AT), a neurokinin-1 receptor antagonist. Those drugs that generalized to isoproterenol's cue were tested after pretreatment with the 1-adrenergic antagonist betaxolol to verify mediation of the discriminative stimulus effects by 1-adrenergic receptors.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Subjects. Male Sprague-Dawley rats, weighing 250 to 350 g at the beginning of the experiment (n = 4-8/dose determination), were housed individually in polycarbonate cages containing wood shavings in a room that was kept at a constant temperature (22°C) and on a 12-h on/12-h off light cycle (light on at 6:00 AM). Rats had free access to food, except during the experimental sessions. Access to water was limited to 25 ml after daily test sessions. All experiments were carried out according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals (revised 1996).

Apparatus. Ten experimental chambers (model E10-10; Coulbourn Instruments, Allentown, PA) were each equipped with a house light, two levers, which required a downward force equivalent to 15 g (0.15 N) to constitute a response, and a centrally located water dipper, which delivered a 0.02-ml water reinforcer when schedule requirements were met. A MED Associates (St. Albans, VT) operating system was used to record responses and control schedule contingencies.

Discrimination Training. Rats were trained to alternate daily between the response levers under a fixed-ratio 1 schedule of reinforcement (FR1) during daily 20-min sessions. The reinforcement contingency was increased incrementally to an FR10 schedule, i.e., every 10th response was reinforced. Once unbiased lever pressing under this schedule was established, each rat was implanted with a cannula into the right lateral ventricle (see below) for i.c.v. administration of the training drug (10 µg of isoproterenol dissolved in 10 µl of aCSF) or the aCSF vehicle. These solutions were administered using a syringe pump; the 28-gauge infusion cannula was left in place for an additional minute after administration of the drug or vehicle. After i.c.v. injection, the rats received 0.3 ml of saline i.p. This was done to minimize the possibility that the i.p. administration of test drugs would alter stimulus cues in any subsequent tests that involved this route of administration.

Cannula Implantation. Rats were anesthetized (120 mg/kg ketamine and 6 mg/kg xylazine) and placed in a stereotaxic frame. A 22-gauge guide cannula (Plastics One, Roanoke, VA) was implanted in each rat's right lateral ventricle (0.8 mm relative to bregma, 1.5 mm relative to the midline structure, 3.6 mm relative to dura) and cemented to the skull. The rats were allowed to recover at least 1 week, after which behavioral testing was resumed. Cannula placement was verified by dye infusion in three rats before surgeries were performed on trained animals.

Test Sessions. Once the training criterion was met, a series of generalization and antagonism tests were carried out. Test sessions were conducted only after 3 days of accurate discrimination of vehicle and drug. During a test session, responses were recorded until completion of an FR10 on one lever or a period of 20 min expired; no reinforcement was delivered. Immediately after completion of the FR10, the rat was removed from the experimental chamber and returned to its home cage. Those antidepressants that substituted for isoproterenol's discriminative stimulus were retested in the presence of 1 mg/kg betaxolol. This dose of betaxolol was shown to completely antagonize the discriminative stimulus effects of isoproterenol (Crissman et al., 2001).

Data Analysis. Drug discrimination results are expressed as the mean percentage of animals' responses on the isoproterenol-appropriate lever. The effects of drugs were considered to generalize to the discriminative stimulus effects of isoproterenol when their administration resulted in greater than 80% isoproterenol-appropriate responding. For the calculation of ED₅₀ values for the generalization tests, dose-response curves were subjected to nonlinear regression analysis (Draper and Smith, 1966; O'Donnell, 1990). t tests were used to assess the statistical significance of the antagonistic effects of betaxolol; statistical significance was assumed when p < 0.05.

Drugs. Mirtazapine (Organon NV, Oss, The Netherlands), venlafaxine (Wyeth-Ayerst, Princeton, NJ), fluoxetine, nisoxetine, protriptyline, buspirone, trazodone, N^G-nitro-L-arginine, phenelzine, bupropion, N-acetyl-L-tryptophan 3,5-bis benzyl ester, isoproterenol, desipramine, and betaxolol (Sigma-Aldrich, St. Louis, MO) were used. All drugs are hydrochloride salts, except isoproterenol, which is a bitartrate salt, phenelzine, which is a sulfate salt, and L-NNA and L-AT, which are free bases. Isoproterenol was administered i.c.v.; all other drugs were administered i.p. Isoproterenol and aCSF were administered 10 min before the start of the session. The other drugs (i.e., those used in substitution tests) were administered 20 min before the start of the test session, except for the 1-adrenergic antagonist betaxolol, which was administered 25 min before testing.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Establishment of Discrimination Baseline. The rats used in this study met the training criterion after an average of 28 ± 3 sessions. Once the training criterion was reached, drug appropriate responding was maintained during daily maintenance training sessions.

Discriminative Stimulus Effects of Antidepressants with Noradrenergic Activity. Figure 1 shows the results of substitution tests obtained during sessions when seven different compounds with noradrenergic activity were tested for their ability to substitute for the 10-µg i.c.v. training dose of isoproterenol. Protriptyline, desipramine, nisoxetine, phenelzine, mirtazapine, venlafaxine, and bupropion engendered dose-related increases in the percentage of isoproterenol-appropriate responding. At the highest dose tested for each drug, greater than 90% isoproterenol-appropriate responding was observed. Based on ED₅₀ values (Table 1) the rank-order potency of these drugs for producing isoproterenol-appropriate responding was bupropion > protriptyline > phenelzine > venlafaxine > desipramine > mirtazapine > nisoxetine. At the dose ranges tested, none of these drugs produced large increases in the latency to complete the FR10 schedule.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Effects of protriptyline (n = 7), desipramine (n = 5), venlafaxine (n = 5-7), mirtazapine (n = 4-5), bupropion (n = 6), phenelzine (n = 5), and nisoxetine (n = 6) in rats trained to discriminate 10 µg of isoproterenol from aCSF. Abscissae: dose, log scale; ordinates: mean (± S.E.M.) percentage of responses on the isoproterenol-appropriate lever (top) and mean (± S.E.M.) latency (in seconds) to complete the fixed ratio 10 requirement (bottom). When rats failed to complete the FR10, a 1200-s latency was assumed. Animals were tested 20 min after drug administration.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Effect of the 1-selective-adrenergic antagonist betaxolol (n = 4) on the percentage of isoproterenol-appropriate responding in rats elicited by antidepressants with noradrenergic activity. Betaxolol (1 mg/kg i.p.) was administered 5 min before i.p. injection of the challenge drugs, which were administered 20 min before behavioral testing. Doses of drugs tested were protriptyline, 3 mg/kg; desipramine, 1 mg/kg; venlafaxine, 1 mg/kg; mirtazapine, 5.6 mg/kg; bupropion, 1 mg/kg; phenelzine, 1 mg/kg; and nisoxetine, 5.6 mg/kg. Lever selection is shown as the percentage of isoproterenol-appropriate responding (means ± S.E.M.). The mean latency to complete the fixed ratio requirement in tests with betaxolol pretreatment was 26.2 ± 2.2; in the absence of betaxolol pretreatment, the mean latency was 90.6 ± 20.7.

Discriminative Stimulus Effects of Antidepressants with Serotonergic Activity. Figure 3 shows generalization results and latencies when ranges of doses for three different antidepressants with serotonergic activity were tested for generalization to isoproterenol's discriminative stimulus effects. The selective serotonin uptake inhibitor fluoxetine failed to substitute for isoproterenol at doses ranging from 0.01 to 3 mg/kg. When administered 10 mg/kg fluoxetine, only one of six rats completed the fixed ratio requirement; this rat responded on the isoproterenol lever. The weak norepinephrine-serotonin uptake inhibitor trazodone produced 47% isoproterenol-appropriate responding at a dose of 3 mg/kg. When administered 10 mg/kg trazodone, all four rats failed to complete the fixed ratio requirement. Buspirone, a partial agonist at 5-hydroxytryptamine_1A receptors, resulted in only 62 and 35% substitution at the 1- and 3-mg/kg doses, respectively.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Effects of fluoxetine (n = 8), buspirone (n = 5), and trazodone (n = 4-6) in rats trained to discriminate 10 µg of isoproterenol from aCSF. Abscissae: dose, log scale; ordinates: mean (± S.E.M.) percentage of responses on the isoproterenol-appropriate lever (top) and mean (± S.E.M.) latency (in seconds) to complete the fixed ratio 10 requirement (bottom). When rats failed to complete the FR10, a 1200-s latency was assumed. Animals were tested 20 min after drug administration.

Discriminative Stimulus Effects of Two Novel, Putative Antidepressants. Administration of 1 to 10 mg/kg L-AT, a neurokinin-1 receptor antagonist, and 0.3 to 10 mg/kg L-NNA, a nitric-oxide synthase inhibitor, did not result in isoproterenol-like discriminative stimulus effects (Fig. 4). At 10 mg/kg, the neurokinin-1 receptor antagonist resulted in 25% isoproterenol-appropriate responding. L-NNA produced, at most, 40% drug-appropriate responding; at the highest dose tested, 10 mg/kg, it caused a large increase in the latency to complete the fixed ratio schedule (Fig. 4).

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Effects of L-AT (n = 4) and L-NNA (n = 5) in rats trained to discriminate 10 µg of isoproterenol from aCSF. Abscissae: dose, log scale; ordinates: mean (± S.E.M.) percentage of responses on the isoproterenol-appropriate lever (top) and mean (± S.E.M.) latency (in seconds) to complete the fixed ratio 10 requirement (bottom). When rats failed to complete the FR10, a 1200-s latency was assumed. Animals were tested 20 min after drug administration.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Antidepressant drugs that enhance noradrenergic activity substituted in rats trained to discriminate centrally administered isoproterenol from aCSF. This included desipramine, protriptyline, phenelzine, mirtazapine, venlafaxine, and bupropion; the norepinephrine uptake inhibitor nisoxetine also substituted. The ability of these drugs to produce isoproterenol-like discriminative stimulus effects was antagonized by betaxolol, indicating that the effects were mediated by 1-adrenergic receptors. Previous pharmacological characterization of the discriminative stimulus effects of isoproterenol revealed mediation by this subtype of the receptor (Crissman et al., 2001), even though isoproterenol exhibits comparable affinity for 1- and 2-adrenergic receptors (Minneman et al., 1979). Similarly, the antidepressant-like effect of centrally administered isoproterenol on DRL behavior is mediated by 1-adrenergic receptors (O'Donnell et al., 1994).

The present data provide functional evidence that administration of antidepressants with noradrenergic activity, at behaviorally relevant doses, results in activation of central 1-adrenergic receptors in vivo. In contrast, previous results indicate that these drugs do not stimulate 2-adrenergic receptors in vivo. This is evidenced by the inability of desipramine, as well as other antidepressants and norepinephrine uptake inhibitors, to substitute in rats trained to discriminate the 2-adrenergic agonist clenbuterol from saline (Makhay and O'Donnell, 1999). Related neurochemical data also suggest that inhibition of norepinephrine uptake results in the stimulation of central 1-, but not 2,-adrenergic receptors. Repeated or continuous administration of such drugs reduces the density of 1-adrenergic receptors, but generally spares 2-adrenergic receptors (Ordway et al., 1988). This does not reflect an intrinsic inability of the 2-subtype to undergo desensitization because repeated treatment with the 2-adrenergic agonist clenbuterol causes rapid and extensive down-regulation of 2-adrenergic receptors (Frances et al., 1983; O'Donnell, 1990).

There seems to be a relationship between the potency with which the norepinephrine uptake inhibitors substitute for isoproterenol and the potency with which they inhibit norepinephrine uptake in vivo. Fuller (1981) assessed the actions of such drugs in vivo by determining their potency for antagonizing the ability of 6-hydroxydopamine to produce norepinephrine depletions in the mouse heart. The potency orders for the three norepinephrine uptake inhibitors that were tested in that study and the present study are the same: protriptyline > desipramine > nisoxetine (Table 1); however, exact comparisons cannot be made due to the species differences. Protriptyline is 13-fold more potent in the drug discrimination assay. This is not surprising because in comparison of the behaviorally relevant doses of these antidepressants in other tests (O'Donnell and Seiden, 1982, 1984; Finnegan et al., 1987), the i.c.v. isoproterenol drug discrimination model seems to be particularly sensitive.

The ability of bupropion to substitute for isoproterenol, together with the antagonism of this effect by betaxolol, indicates that this drug activates central 1-adrenergic receptors; however, the manner by which it does so is unclear. Bupropion has been reported to inhibit norepinephrine uptake and down-regulate -adrenergic receptors (Gandolfi et al., 1983; Alhaider and Mustafa, 1988; Ascher et al., 1995); however, there are a number of contradictory reports (Bryant et al., 1983; Ferris and Beaman, 1983; Suranyi-Cadotte et al., 1995). In rats trained to discriminate the norepinephrine uptake inhibitor reboxetine from vehicle, doses of 2.5 and 10 mg/kg bupropion did not substitute (Dekeyne et al., 2001). In contrast, other norepinephrine uptake inhibitors did substitute for reboxetine. This study did not examine the involvement of -adrenergic receptors in the mediation of the discriminative stimulus effects of reboxetine. Overall, these data suggest that bupropion may indirectly stimulate central 1-adrenergic receptors by a mechanism that may not involve inhibition of norepinephrine uptake.

The drugs with primarily serotonergic actions that were tested, fluoxetine, buspirone, and trazodone, did not, in general, substitute for isoproterenol. At 10 mg/kg, fluoxetine administration resulted in isoproterenol-appropriate responding in one of six rats tested; the other rats failed to complete the fixed ratio requirement. Although fluoxetine is known to lose its serotonergic selectivity at higher doses (Perry and Fuller, 1997), and thus might be expected to substitute, the response rate-decreasing effects of these doses likely precluded any such observation. The failure of the serotonergic antidepressant drugs to indirectly stimulate 1-adrenergic receptors in vivo is consistent with findings suggesting that activation of noradrenergic and serotonergic activity represents parallel mechanisms (Miller et al., 1992).

Although the processes of uptake inhibition and receptor down-regulation are common mechanisms for antidepressants acting via the noradrenergic system to produce their effects, research has shown that antidepressant-like effects can result from modification of certain intracellular events, independent of postsynaptic receptor stimulation; the putative antidepressants L-NNA and L-AT are thought to act via such a process. Repeated treatment with L-NNA has been shown to down-regulate cortical -adrenergic receptors to an extent comparable with that seen after repeated treatment with imipramine (Karolewicz et al., 1999). In addition, L-NNA produces an antidepressant-like effect in mice, reducing the time of immobility in the forced-swim test (Harkin et al., 1999). The failure of L-NNA, at behaviorally relevant doses, to substitute for a 1-mediated cue suggests that the antidepressant-like effects of this drug most likely do not involve enhancement of noradrenergic activity and stimulation of central 1-adrenergic receptors.

Neurokinin biosynthesis in particular brain regions is down-regulated after repeated administration of clinically effective antidepressants, raising the speculation that alterations in neurokinin synthesis may contribute to the efficacy of antidepressant drugs (Barden et al., 1983; Brodin et al., 1994; Shirayama et al., 1996). Furthermore, neurokinin-1 receptor antagonists are active in animal models sensitive to antidepressants (Kramer et al., 1998; Papp et al., 2000). In the present study, behaviorally relevant doses of the neurokinin-1 receptor antagonist L-AT failed to generalize to the isoproterenol cue, suggesting that, although this drug produces antidepressant-like effects, it does not do so by directly or indirectly increasing the stimulation of central 1-adrenergic receptors.

The noradrenergic and serotonergic systems seem to act in a parallel manner in mediating antidepressant efficacy (Miller et al., 1992). Although sufficient evidence suggests separate targets of antidepressant activity that ultimately may lead to a common effect, e.g., the modulation of 1-adrenergic receptors or associated signaling molecules (Mason et al., 1993; Papp et al., 1994; Matsumoto et al., 1995; Ye et al., 1997, 2000; Nelson, 1999; Takahashi et al., 1999), their direct mechanisms of action differ. The present results are consistent with such an interpretation. Although drugs that interact with noradrenergic neurons substituted for the discriminative stimulus effects of isoproterenol, drugs that interact with serotonergic neurons did not. Furthermore, although antagonism of nitric-oxide synthase and neurokinin-1 receptors is sufficient to produce antidepressant-like effects, stimulation of the 1-adrenergic receptor does not result from administration of drugs that act via these mechanisms. Thus, although stimulation of central 1-adrenergic receptors is sufficient to produce antidepressant-like effects on behavior (O'Donnell et al., 1994), it does not seem that stimulation of these receptors is an effect shared by antidepressants from all pharmacological classes.