Differential Effects of Antipsychotic Drugs on Serotonin-1A Receptor-Mediated Disruption of Prepulse Inhibition

Although several studies have suggested a role for serotonin receptors in schizophrenia, most studies have focused on serotonin-2A (5-HT_2A) receptors. Recently, there has been increased interest in a possible role of 5-HT_1A receptors in schizophrenia as well. For example, some studies have suggested a link between a 5-HT_1A receptor C(-1019)G polymorphism and schizophrenia (Huang et al., 2004). Post mortem research has shown changes in the density of the 5-HT_1A receptor, predominantly in the cortex from patients with schizophrenia. Thus, when using either homogenate binding or autoradiography, there was a 15 to 80% increase in 5-HT_1A receptor binding density in the frontal cortex, mostly in Brodmann's area (BA) 10 (prefrontal cortex) but also in BA9, BA44, and BA46 (Bantick et al., 2001). There was a tendency for increased density of 5-HT_1A receptor binding in a number of other brain regions, but unlike the frontal cortex, findings have been less consistent (Bantick et al., 2001).

It is well established that atypical antipsychotics with combined D₂ and 5-HT_2A receptor antagonism are clinically effective in schizophrenia; however, the combination of D₂ receptor antagonism and 5-HT_1A receptor agonism has received less attention as an important receptor profile for antipsychotic treatment (Wadenberg and Ahlenius, 1991; Ichikawa and Meltzer, 1999; Bantick et al., 2001). Atypical antipsychotic drugs, such as aripiprazole, clozapine, and ziprasidone, have a moderate affinity for 5-HT_1A receptors (Ichikawa and Meltzer, 1999; Bantick et al., 2001) and display a low incidence of extrapyramidal side effects (Rollema et al., 2000; Cosi and Koek, 2001). Animal studies have shown that treatment with 5-HT_1A receptor agonists selectively increases dopamine release in the prefrontal cortex, while reducing or not affecting dopamine release in the striatum (for review, see Ichikawa and Meltzer, 1999; Bantick et al., 2001). This is important for schizophrenia because increasing cortical dopamine release may result in an improvement of the negative symptoms of schizophrenia, whereas the lack of such an effect in the striatum may result in a low incidence of extrapyramidal side effects (Ichikawa and Meltzer, 1999; Rollema et al., 2000). Furthermore, treatment with clozapine selectively increased dopamine release in the rat prefrontal cortex and ventral hippocampus, an effect at least partially mediated by 5-HT_1A receptors (Chung et al., 2004), potentially inhibiting serotonergic transmission via activation of 5-HT_1A autoreceptors (Bantick et al., 2001).

Several other of the proposed beneficial 5-HT_1A receptor components of antipsychotic drug action have been deduced from indirect and in vitro studies, such as in binding assays or cell cultures (Newman-Tancredi et al., 2005; Bruins Slot et al., 2006). Importantly, however, these atypical antipsychotics tend to have a rich pharmacology, including affinity for dopamine (e.g., D₂), 5-HT (e.g., 5-HT_1A, 5-HT_2A), and several other receptor subtypes, the combination of which is likely to be responsible for their clinical efficacy. Therefore, the specific importance of 5-HT_1A receptor agonist or antagonist properties in the antipsychotic profile of these drugs in vivo remains unclear.

In contrast to possible beneficial actions of 5-HT_1A receptor activation (see above), some of these effects resemble a schizophrenia-like state, rather than an antipsychotic action. For example, it is well described that administration of the prototypical 5-HT_1A receptor agonist, 8-hydroxy-di-propylaminotetralin (8-OH-DPAT), causes a disruption of prepulse inhibition (PPI) (Rigdon and Weatherspoon, 1992; Sipes and Geyer, 1995; Gogos and Van den Buuse, 2004), a measure of sensorimotor gating that is also deficient in schizophrenia (Braff and Geyer, 1989; Kumari and Sharma, 2002). The effect of 8-OH-DPAT on PPI could be blocked by pretreatment with the selective 5-HT_1A receptor antagonist, (+)WAY 100,135 (Sipes and Geyer, 1995; Czyrak et al., 2003), confirming that this disruption is mediated by stimulation of 5-HT_1A receptors rather than 5-HT₇ receptors for which 8-OH-DPAT also has affinity (Shen et al., 1993). However, there is limited information on the effect of antipsychotic drugs on 5-HT_1A receptor-mediated disruptions of PPI. Therefore the aim of the present study was to compare six antipsychotic drugs, haloperidol, risperidone, clozapine, olanzapine, amisulpride, and aripiprazole, with respect to their ability to modulate the action of 8-OH-DPAT on PPI. We also tested the dopamine D₂ receptor antagonist, raclopride, as comparison for haloperidol, and the 5-HT_1A receptor partial agonists, MDL 73,005EF and buspirone. We measured startle amplitude, as well as PPI at a short interstimulus interval (ISI; 30 ms) and a longer ISI (100 ms). The results show differential effects of antipsychotic drugs on 5-HT_1A receptor-mediated disruption of PPI with a major component of this interaction likely to be blockade of dopamine D₂ receptors.

Materials and Methods

Animals. Experiments were done on male Sprague-Dawley rats (body weight, 300–400 g) which were obtained from the breeding colony at the Department of Pathology, University of Melbourne. After arriving at the institute, the rats were allowed at least 1 week of acclimation before testing commenced.

Protocol. Around 2 to 3 days before the start of the experiments, all rats were acclimated to the PPI procedure once without any treatments. After this “pretest,” rats received two treatments per test (pretreatment with antipsychotic drug or saline versus treatment with 8-OH-DPAT or saline) and were tested six times with 3- to 4-day intervals: saline/saline; antipsychotic drug, low dose/saline; antipsychotic drug, high dose/saline; saline/8-OH-DPAT; antipsychotic drug, low dose/8-OH-DPAT; and antipsychotic drug, high dose/8-OH-DPAT. One separate cohort of 7 to 11 rats was used for each antipsychotic drug, except MDL 73,005EF and buspirone, which were tested at only one dose and were combined in one experiment. The sequence of treatment was pseudorandomized so that at the end of the series of experiments, all rats in a cohort had received all treatment combinations. Injection volumes were 1 ml/kg body weight. Antipsychotic drugs or saline were injected i.p. 30 min before injection of 0.5 mg/kg 8-OH-DPAT or saline, which was injected s.c. approximately 5 min before the animals were placed in the PPI enclosures.

Drugs. 8-OH-DPAT was obtained from Tocris Cookson Inc. (Bristol, UK) and dissolved at 0.5 mg/ml in 0.9% saline. Haloperidol was obtained from Sigma-Aldrich (St. Louis, MO), dissolved in saline, and injected at 0.05 and 0.25 mg/kg. Raclopride was obtained from Astra Hässle AB (Mölndal, Sweden), dissolved in saline, and similarly injected at 0.05 and 0.25 mg/kg. Clozapine was obtained from BDG Synthesis (Wellington, New Zealand), dissolved in a minimal amount of 0.1 N HCl, and diluted to the required 1 or 5 mg/kg dose. For pretreatment with olanzapine, we used Zyprexa Zydis wafers (Eli Lilly, West Ryde, NSW, Australia), containing 5 or 15 mg of olanzapine each, which were dissolved in saline to prepare the required 1 and 5 mg/kg doses. For pretreatment with risperidone, we used 1 mg/ml Risperdal solution (Janssen-Cilag, Lane Core, NSW, Australia) undiluted (1 mg/kg) or diluted in saline to obtain the required dose of 0.2 mg/kg. For pretreatment with amisulpride, we used Solian 400 tablets (Sanofi-Synthelabo Australia, North Ryde, NSW, Australia), containing 400 mg of amisulpride each, which were dissolved in saline to obtain the required doses of 10 or 50 mg/kg. For pretreatment with aripiprazole, we used Abilify tablets (Bristol-Myers Squibb Co., Noble Park North, Victoria, Australia), containing 15 mg of aripiprazole each, which were dissolved in saline to obtain the required 1 and 5 mg/kg doses. MDL 73,005EF and buspirone were obtained from Sigma and dissolved in saline at 1 and 5 mg/kg, respectively. All drug doses were obtained from the literature or preliminary experiments in the laboratory.

Startle Amplitude and PPI. PPI of the acoustic startle response was measured using eight automated SR-Lab startle chambers (San Diego Instruments, San Diego, CA). The startle chambers (38-cm width × 41-cm length × 58-cm height) were isolated to minimize extrinsic sound sources, well lit, and ventilated. A speaker positioned centrally in the roof of the chamber presented all test sounds. Rats were placed in an acrylic Plexiglas cylinder of 8.8 cm in diameter, with a length of 19.5 cm, which was closed at either end. The Plexiglas cylinder was attached to a platform with a piezoelectric transducer to detect whole-body movements within the cylinder. Presentation of sounds and the recording of responses were automated using SR-Lab software (San Diego Instruments), which was controlled by a computer in an adjacent room.

For all experiments, we used a PPI session that consisted of 104 trials with a variable intertrial interval of 12 to 28 s (average, 19 s). The first 3 min of the session were a 70-dB background noise presentation, allowing further acclimation to the test environment. The session commenced and finished with eight 115-dB, 40-ms pulse-alone trials. Together with two blocks of eight pulse-alone trials from the main part of the session, these startle blocks were used to calculate average startle amplitude and startle habituation across the session. The main part of the session included eight of each of the following prepulse-pulse trials: PP2P115, PP4P115, PP8P115, and PP16P115 at 30 ms between the start of the 20-ms prepulse and start of the pulse and at 100 ms between the start of the prepulse and the pulse. PP2, PP4, PP8, and PP16 indicate 2, 4, 8, or 16 dB above the 70-dB background, i.e., 72-, 74-, 78-, and 86-dB prepulses. In addition to the pulse-alone and prepulse-pulse trials, the session included eight “NOSTIM” trials, where no startle stimulus was presented, to check for nonspecific movement artifacts. The sequence of trials within the session was pseudorandomized and the same for all eight simultaneously tested rats and for all consecutive drug experiments.

Data Analysis. All data are expressed as mean ± S.E.M. Startle amplitudes were calculated as the average values of each of the four blocks of startle stimuli, allowing analysis of both startle magnitude and startle habituation. However, to simplify the data presentation, startle habituation will not be addressed here. Prepulse inhibition values were obtained and plotted for each of the prepulse levels for each ISI. Figures 1, 2, 3, 4, 5, 6, 7, 8 show these results for each prepulse intensity, whereas Fig. 9 shows a summary of the results with only the average of all prepulse intensities included. First, analysis of variance (ANOVA) with repeated measures was used to assess main treatment effects on PPI. Within-animal repeated measures factors were antipsychotic dose effect (three levels), 8-OH-DPAT effect (two levels), and prepulse level (four levels). Where appropriate, one-way ANOVA and least-significant difference test were then used to assess differences between treatments for individual prepulse intensities, i.e., to assess effects of antipsychotics or 8-OH-DPAT. When P < 0.05, differences were considered statistically significant.

Results

Haloperidol (Fig. 1; Table 1). Combined analysis of startle data for both haloperidol doses revealed a trend for a main effect of dose [F(2,12) = 2.9; P = 0.089]. Compared with saline, haloperidol at 0.25 but not 0.05 mg/kg significantly reduced startle amplitudes [F(1,6) = 15.9; P = 0.007]. 8-OH-DPAT treatment tended to cause an increase in startle responses; however, this effect was not significant (Table 1).

Although haloperidol had no effect on PPI at the 30-ms interval, there were significant main effects of 8-OH-DPAT treatment [F(1,6) = 31.0; P = 0.001] and of prepulse level [F(3,18) = 125.0; P < 0.001]. However, there were no significant treatment effects at any of the individual prepulse intensities (Fig. 1).

Combined analysis of all haloperidol doses and all prepulse levels at the 100-ms interval showed significant main effects of dose [F(2,10) = 17.7; P = 0.001], 8-OH-DPAT [F(1,5) = 35.0; P = 0.002], and prepulse [F(3,15) = 113.0; P < 0.001]. At PP2, the disruption of PPI by 8-OH-DPAT was significant after pretreatment with saline and 0.05 mg/kg haloperidol but was blocked after pretreatment with 0.25 mg/kg haloperidol (Fig. 1). At PP4, the effect of 8-OH-DPAT was significant after pretreatment with 0.05 mg/kg haloperidol, although not after saline pretreatment, and again was blocked after pretreatment with 0.25 mg/kg haloperidol (Fig. 1). Likewise, at PP8 and PP16, the effect of 8-OH-DPAT was significant after pretreatment with saline and 0.05 mg/kg haloperidol and was reduced after pretreatment with 0.25 mg/kg haloperidol (Fig. 1). At none of the prepulse intensities did haloperidol treatment significantly alter PPI on its own.

Raclopride (Fig. 2; Table 1). To confirm and extend the observation that haloperidol pretreatment could block the action of 8-OH-DPAT on PPI, we also tested raclopride, another (putative) antipsychotic drug with a predominantly dopamine D₂-blocking mode of action. Analysis of startle data showed that neither raclopride nor 8-OH-DPAT significantly affected startle (Table 1), although 8-OH-DPAT treatment tended to cause an increase in startle responses. Raclopride had no effect on PPI at the 30-ms interval, although there was a significant main effect of prepulse level [F(3,18) = 110.4; P < 0.001]. As with haloperidol, there were no significant treatment effects at any of the individual prepulse intensities (Fig. 2).

Combined analysis of all raclopride doses and all prepulse levels at the 100-ms interval showed significant main effects of 8-OH-DPAT [F(1,6) = 30.3; P = 0.002] and prepulse [F(3,18) = 52.2; P < 0.001]. In this cohort of rats, there were no treatment effects at PP2 (Fig. 2). Similar to the haloperidol experiment, at PP4, the disruption of PPI by 8-OH-DPAT was significant after pretreatment with 0.05 mg/kg raclopride, although not after saline pretreatment, and was blocked after pretreatment with 0.25 mg/kg raclopride (Fig. 2). At PP8, the effect of 8-OH-DPAT was significant after pretreatment with saline and 0.05 mg/kg raclopride and was blocked after pretreatment with 0.25 mg/kg raclopride (Fig. 2). At PP16, the effect of 8-OH-DPAT was significant only after saline pretreatment. At none of the prepulse intensities did raclopride pretreatment significantly alter PPI on its own.

Clozapine (Fig. 3; Table 1). To assess whether the blocking action of haloperidol and raclopride extended to atypical antipsychotic drugs, we also tested clozapine, olanzapine, risperidone, amisulpride, and aripiprazole. Clozapine significantly reduced startle responses [F(2,16) = 8.4; P = 0.003], and this effect was significant for both the 1 [F(1,8) = 15.8; P = 0.004] and 5 [F(1,8) = 10.2; P = 0.013] mg/kg doses. 8-OH-DPAT treatment caused a slight but significant increase of startle amplitude [F(1,8) = 6.0; P = 0.040] (Table 1).

At the 30-ms interval, ANOVA showed the expected main effects of 8-OH-DPAT [F(1,8) = 7.7; P = 0.024] and prepulse level [F(3,18) = 110.4; P < 0.001]. Moreover, ANOVA revealed that clozapine treatment resulted in complex modulation of PPI dependent on the prepulse intensity [dose × prepulse interaction, F(6,48) = 3.3; P = 0.009]. Further analysis at different prepulse intensities revealed that there was a clear tendency for 5 mg/kg clozapine to reduce PPI at this ISI, allowing an apparent enhancement of PPI by 8-OH-DPAT (Fig. 3). Thus, average PPI after saline- or 8-OH-DPAT treatment was 22 ± 9 and 32 ± 3%, respectively, after saline pretreatment, compared with -1 ± 11 and 30 ± 4%, respectively, after clozapine pretreatment. At PP2, there was no effect of 8-OH-DPAT after saline or 1 mg/kg clozapine pretreatment; however, after 5 mg/kg clozapine, PPI tended to be reduced (P = 0.072), revealing a significant PPI enhancing effect of 8-OH-DPAT (Fig. 3). At PP4, PPI was significantly reduced by 5 mg/kg clozapine, again allowing a significant effect of 8-OH-DPAT (Fig. 3). At PP8, the effect of 5 mg/kg clozapine was not significant, although again an apparent effect of 8-OH-DPAT appeared after this dose of the antipsychotic (Fig. 3). At PP16, there were no significant treatment effects, although the difference between 8-OH-DPAT after saline pretreatment and after 5 mg/kg clozapine pretreatment reached trend level (P = 0.070) (Fig. 3).

The effect of clozapine on PPI at the 100-ms ISI did not show the same reducing influence as seen at the 30-ms ISI; however, there was a dose × 8-OH-DPAT × prepulse interaction [F(6,48) = 2.8; P = 0.020], again suggesting complex interacting effects of clozapine and 8-OH-DPAT on PPI dependent on the prepulse intensity (Fig. 3). ANOVA also showed main effects of 8-OH-DPAT [F(1,8) = 11.5; P = 0.009] and of prepulse intensity [F(3,24) = 36.3; P < 0.001]. At PP2, the significant effect of 8-OH-DPAT was blocked by both the 1 and 5 mg/kg doses. On the other hand, no effect of clozapine at either dose was observed at PP4, whereas at PP8 and PP16, both doses appeared to enhance the effect of 8-OH-DPAT (Fig. 3).

Olanzapine (Fig. 4; Table 1). Olanzapine significantly reduced startle responses [F(2,14) = 5.1; P = 0.021] at the 5 mg/kg dose [F(1,7) = 43.6; P < 0.001] but not the 1 mg/kg dose. 8-OH-DPAT treatment tended to cause an increase in startle amplitude, an effect that became significant when data for saline and 5 mg/kg olanzapine were analyzed [F(1,7) = 8.1; P = 0.025] (Table 1).

PPI at the 30-ms ISI showed the expected effect of prepulse intensity [F(3,21) = 102.6; P < 0.001] but was not significantly affected by either olanzapine or 8-OH-DPAT (Fig. 4). PPI at the 100-ms ISI again showed the expected effect of prepulse intensity [F(3,21) = 101.5; P < 0.001] and was significantly reduced by 8-OH-DPAT treatment [F(1,7) = 43.2; P < 0.001]. In addition, there was a main effect of olanzapine of borderline significance [F(2,14) = 3.8; P = 0.048]. At none of the prepulse intensities did olanzapine block the reducing effect of 8-OH-DPAT (Fig. 4). Thus, the effect of 8-OH-DPAT was significant at all prepulse intensities and all pretreatments, except after 1 mg/kg olanzapine at PP2 and 5 mg/kg olanzapine at PP8, which failed to reach significance (Fig. 4). At none of the prepulse intensities did olanzapine pretreatment significantly alter PPI on its own.

Risperidone (Fig. 5; Table 1). Combined analysis of startle amplitudes revealed main effects of risperidone dose [F(2,12) = 10.2; P = 0.003] and 8-OH-DPAT [F(1,6) = 13.6; P = 0.010]. Pretreatment with risperidone slightly reduced startle amplitudes both at the 0.2 [F(1,6) = 18.4; P = 0.005] and 1 [F(1,6) = 7.3; P = 0.036] mg/kg doses, whereas 8-OH-DPAT enhanced startle amplitude after both pretreatments [F(1,6) = 10.8; P = 0.017 and F(1,6) = 13.1; P = 0.011, respectively] (Table 1).

Analysis of PPI data at the 30-ms ISI revealed the expected main effects of 8-OH-DPAT [F(1,6) = 9.9; P = 0.020] and prepulse level [F(3,18) = 109.6; P < 0.001]. Moreover, ANOVA revealed that risperidone pretreatment, as with clozapine, resulted in complex modulation of PPI dependent on the prepulse intensity [dose × prepulse interaction, F(6,36) = 2.6, P = 0.035; dose × 8-OH-DPAT × prepulse interaction, F(6,48) = 2.8, P = 0.023]. This was particularly clear at PP2, where 1 mg/kg risperidone significantly reduced PPI on its own, unmasking significant enhancement of PPI by 8-OH-DPAT (Fig. 5). At other prepulse intensities, neither risperidone nor 8-OH-DPAT affected PPI at the 30-ms ISI (Fig. 5).

Analysis of PPI at the 100-ms ISI revealed a main effect of risperidone dose [F(2,12) = 11.7; P = 0.002], reflecting a general tendency for PPI to be enhanced after risperidone treatment (Fig. 5). In addition, there was the expected disruption of PPI by 8-OH-DPAT [F(1,6) = 36.7; P = 0.001] and main effect of prepulse intensity [F(3,18) = 93.2; P < 0.001]. At PP2, PP8, and PP16, 1 mg/kg risperidone significantly increased PPI on its own while generally not blocking the disruption induced by 8-OH-DPAT (Fig. 5). Thus, the effect of 8-OH-DPAT was significant at all prepulse intensities and all pretreatment doses, except after 0.2 mg/kg risperidone at PP2, after 1 mg/kg risperidone at PP4 (P = 0.069), and after saline pretreatment at PP8 and PP16 (Fig. 5).

Amisulpride (Fig. 6; Table 1). Amisulpride had no effect on startle amplitude (Table 1), whereas 8-OH-DPAT significantly increased startle [F(1,9) = 13.6; P = 0.005]. Analysis of PPI at the 30-ms ISI showed the expected main effect of prepulse intensity [F(3,27) = 48.8; P < 0.001]. 8-OH-DPAT treatment tended to increase PPI but only at lower prepulse intensities [8-OH-DPAT × prepulse interaction, F(3,27) = 6.5; P = 0.002]. However, analysis of individual prepulse intensities did not reveal any significant effects of 8-OH-DPAT or amisulpride (Fig. 6).

Analysis of PPI at the 100-ms ISI revealed marked disruption by 8-OH-DPAT [F(1,9) = 60.6; P < 0.001] and a main effect of prepulse intensity [F(3,27) = 93.1; P < 0.001]; however, there was no effect of amisulpride (Fig. 6). As with clozapine, olanzapine, and risperidone, amisulpride pretreatment did not block the action of 8-OH-DPAT on PPI (Fig. 6). Thus, the effect of 8-OH-DPAT was significant at all prepulse intensities and all pretreatment doses, except after 10 mg/kg amisulpride at PP4 (P = 0.079), PP8, and PP16 (Fig. 6). At none of the prepulse intensities did amisulpride pretreatment significantly alter PPI on its own.

Aripiprazole (Fig. 7; Table 1). 8-OH-DPAT significantly enhanced startle amplitude [F(1,7) = 9.0; P = 0.020]; however, there were no effects of aripiprazole pretreatment (Table 1). PPI at the 30-ms ISI showed the main effect of prepulse intensity [F(3,21) = 78.8; P < 0.001] but was not significantly affected by either aripiprazole or 8-OH-DPAT (Fig. 7).

Analysis of PPI data at the 100-ms ISI revealed a significant disruption by 8-OH-DPAT treatment [F(1,7) = 25.4; P < 0.001] and a main effect of prepulse intensity [F(3,21) = 103.1; P < 0.001]. At PP2, 8-OH-DPAT treatment tended to decrease PPI, except in animals which were pretreated with 5 mg/kg aripiprazole; however, these differences did not reach statistical significance (Fig. 7). In contrast, at PP4, 8-OH-DPAT significantly disrupted PPI, and this effect was blocked by 5 mg/kg aripiprazole (Fig. 7). In addition, at PP8 and PP16, 8-OH-DPAT significantly disrupted PPI; however, this was not influenced by aripiprazole pretreatment (Fig. 7).

MDL 73,005EF and Buspirone (Fig. 8; Table 1). There were no significant main effects of MDL 73,005EF on startle amplitude (Table 1). PPI at the 30-ms ISI was slightly but significantly increased by MDL 73,005EF treatment [F(1,16) = 5.7; P = 0.030] in addition to the main effect of prepulse intensity [F(3,48) = 114.0; P < 0.001] (Fig. 8). At PP2, 8-OH-DPAT significantly increased PPI, an effect that was not observed after pretreatment with MDL 73,005EF because of a significant increase in PPI induced by this pretreatment itself (Fig. 8). At PP4, a similar increase of PPI by MDL 73,005EF pretreatment was seen that was close to significance (P = 0.065). At PP16, but not at PP8, there was again a slight but significant increase of PPI after MDL 73,005EF pretreatment. 8-OH-DPAT had no significant effects at PP4, PP8, or PP16 (Fig. 8).

At the 100-ms ISI, again there was a main effect of prepulse intensity [F(3,48) = 96.4; P < 0.001] and the expected marked disruption of PPI by 8-OH-DPAT treatment [F(1,16) = 14.8; P = 0.001]. There was also an overall increase in PPI induced by MDL 73,005EF pretreatment [F(1,16) = 10.3; P = 0.006] but no statistical interaction between the effects of MDL 73,005EF pretreatment and 8-OH-DPAT treatment (Fig. 8), similar to the result in risperidone-treated animals. At PP2, the effect of MDL 73,005EF on baseline PPI was close to significance (P = 0.057). However, although PPI tended to be increased by MDL 73,005EF at other prepulse intensities as well (Fig. 8), these effects did not reach significance. At PP4, 8-OH-DPAT significantly disrupted PPI after both saline pretreatment and MDL 73,005EF pretreatment. At PP8, a similar trend was observed (Fig. 8), although the effects only reached trend level (P = 0.091 and P = 0.077, respectively). At PP16, PPI was slightly but significantly lower after 8-OH-DPAT treatment in saline controls only (Fig. 8).

Pretreatment with 5 mg/kg buspirone did not significantly alter startle amplitude (Fig. 8). Neither buspirone pretreatment nor 8-OH-DPAT significantly affected PPI at the 30-ms ISI (Fig. 8), and only a main effect of prepulse intensity was observed [F(3,48) = 155.7; P < 0.001]. On the other hand, analysis of PPI at the 100-ms ISI revealed main effects of prepulse intensity [F(3,48) = 91.6; P < 0.001], buspirone pretreatment [F(1,16) = 6.7; P = 0.020], and 8-OH-DPAT treatment [F(1,16) = 4.2; P = 0.057]. There was also a significant interaction of buspirone pretreatment with the effect of 8-OH-DPAT [F(1,16) = 7.1; P = 0.017], reflecting blockade of the effect of 8-OH-DPAT treatment in buspirone-pretreated rats (Fig. 8). Thus, although buspirone pretreatment did not affect PPI on its own, it blocked the disruption of PPI by subsequent 8-OH-DPAT treatment at all prepulse intensities (Fig. 8).

Discussion

The aim of the present experiments was to investigate the in vivo interaction of several antipsychotic drugs with central 5-HT_1A receptor mechanisms in a behavioral animal model with relevance to schizophrenia. Thus, we assessed the ability of antipsychotic drugs to modulate the effect of 8-OH-DPAT on PPI, a measure of sensorimotor gating, which is deficient in schizophrenia (Braff and Geyer, 1989; Kumari and Sharma, 2002). Figure 9 summarizes and compares the effect of the different antipsychotics on the action of 8-OH-DPAT. The main finding of our experiments was that there does not seem to be an interaction of most atypical antipsychotic drugs with the effect of 8-OH-DPAT on PPI (Fig. 9). Only treatment with aripiprazole significantly inhibited this effect, although this effect was only seen at some prepulse intensities. It is possible that the partial agonist activity and high affinity of aripiprazole at 5-HT_1A receptors (Newman-Tancredi et al., 2005; Bruins Slot et al., 2006) are responsible for this interaction. Thus, the efficacy of aripiprazole at 5-HT_1A receptors may not be high enough to elicit a disruption of PPI; however, its receptor occupancy is sufficient to block the effect of subsequently administered 8-OH-DPAT. This explanation is supported by the experiment with another partial agonist at 5-HT_1A receptors, buspirone. This compound did not disrupt PPI by itself but blocked the action of subsequently administered 8-OH-DPAT. The result with the 5-HT_1A receptor partial agonist, MDL 73,005EF, was more complex because it tended to increase resting PPI by itself. Previous studies in other paradigms have also shown that pretreatment with partial 5-HT_1A receptor agonists may cause inhibition of the action of 8-OH-DPAT (Boddeke et al., 1992; Buisson-Defferier and Van den Buuse, 1992; Pauwels et al., 1993). Clozapine, olanzapine, risperidone, and amisulpride all have lower affinity at 5-HT_1A receptors than aripiprazole (Newman-Tancredi et al., 2005), which could explain their lack of effect on the disruption of PPI caused by 8-OH-DPAT treatment.

Surprisingly, pretreatment with haloperidol almost completely blocked the effect of the 5-HT_1A receptor agonist. This blockade was also observed with another dopamine D₂ receptor antagonist, raclopride. This effect of haloperidol is unlikely to be due to a direct action at 5-HT_1A receptors because the affinity of this drug for these receptors is low (Newman-Tancredi et al., 2005). Rather, it is likely that 5-HT_1A receptor activation elicits a chain of events in the brain, ultimately leading to “downstream” dopamine D₂ receptor activation, which, similar to treatment with dopaminergic drugs, leads to disruption of PPI. Previously, some behavioral effects of 8-OH-DPAT, such as lower lip retraction, could also be blocked by pretreatment with spiperone or haloperidol, confirming a possible dopaminergic “link” in the behavioral effects of 5-HT_1A receptor activation (Berendsen et al., 1990). This complicates the explanation of the action of antipsychotic drugs on the effect of 8-OH-DPAT. Aripiprazole is reported to have an affinity for dopamine D₂ receptors only slightly lower than haloperidol (pK_i = 8.59 versus 9.01, respectively) (Newman-Tancredi et al., 2005). Aripiprazole is a partial agonist at these receptors (Burris et al., 2002; Shapiro et al., 2003), whereas in rats, it is metabolized in vivo to a full D₂ receptor antagonist (Wood et al., 2006). Thus, the effect of aripiprazole could be explained by its blocking action on dopamine D₂ receptors as well as or rather than an action on 5-HT_1A receptors. Even the effect of buspirone in blocking the 8-OH-DPAT-induced disruption of PPI could have been mediated by its binding to dopamine D₂ receptors. Buspirone displays high affinity for these receptors and has been shown to act as a dopamine D₂ receptor antagonist in several behavioral models (Ryan et al., 1993; Protais et al., 1998). On the other hand, risperidone was reported to have an affinity at dopamine D₂ receptors of 8.70 (Newman-Tancredi et al., 2005), yet in our experiments, there was no statistical interaction of risperidone pretreatment with the disruption of PPI caused by 8-OH-DPAT treatment. This lack of interaction was mainly caused by an increase in resting PPI after risperidone pretreatment, and it should be noted that PPI in animals treated with both risperidone and 8-OH-DPAT was similar to that in controls, which would further support the conclusion that dopamine D₂ receptor blockade is able to inhibit the effect of 5-HT_1A receptor activation on PPI.

In a recent study, several antipsychotics were tested against the disruption of PPI induced by treatment with the dopamine receptor agonist, apomorphine (Auclair et al., 2006). Of the antipsychotics we also included, pretreatment with haloperidol, risperidone, and olanzapine blocked the effect of apomorphine, whereas clozapine and aripiprazole were less effective (Auclair et al., 2006). Interestingly, when pretreatment with mixed D₂/5-HT_1A ligands was combined with a 5-HT_1A receptor antagonist, the ability to block the action of apomorphine was enhanced. These results support our finding of a functional interaction of activation of D₂ and 5-HT_1A receptors in PPI regulation but also emphasize the complexity of this interaction. It would be reasonable to assume that involvement of dopamine D₂ receptors in PPI is modulated both positively and negatively by 5-HT_1A receptor activation. Further experimentation, for example with local injections into the brain, is needed to elucidate such interactions.

Our experiments also showed effects of antipsychotic drugs by themselves. For example, olanzapine and risperidone pretreatment increased PPI at the 100-ms ISI, whereas clozapine and risperidone induced complex, prepulse-dependent effects at the 30-ms ISI. These results show that for the full interpretation of drug effects on PPI, an extended protocol, including multiple prepulse intensities and ISIs, is preferable. Particularly at the shorter ISIs, modulation of the startle responses is a mix of true PPI and of negative PPI or prepulse facilitation (PPF) (Plappert et al., 2004; Swerdlow et al., 2004a,b). The regulation and functional significance of PPF is poorly understood. Men have been shown to display higher PPI than women, but women displayed higher PPF than men (Aasen et al., 2005). PPF was found to be reduced in patients with schizophrenia and their unaffected siblings (Wynn et al., 2004). Thus, the modulatory effects of antipsychotic drugs on PPF seen in the current study could have relevance for our understanding of the mechanism of action of these drugs in patients with schizophrenia. PPI at longer ISIs is susceptible to attentional mechanisms, whereas PPI at shorter ISIs is a more “automatic” mechanism (Filion et al., 1993; Bohmelt et al., 1999), and these components could be differentially affected in schizophrenia and by antipsychotic drugs. Thus, psychopharmacological effects on PPI need to be interpreted with caution because the results may represent multiple and separate startle modulation mechanisms. In our experiments, 8-OH-DPAT only disrupted PPI at the 100-ms ISI, making it likely that only PPI, not PPF, mechanisms are involved.

PPI reflects a gating mechanism for sensory information and, as such, could be involved in some of the cognitive deficits seen in patients with schizophrenia (Braff and Geyer, 1989; Kumari and Sharma, 2002). Antipsychotic drugs have been shown by some studies to reverse the disruption of PPI seen in patients with schizophrenia. For example, treatment with either olanzapine or amisulpride reversed PPI deficits in patients with schizophrenia (Quednow et al., 2006). Treatment with clozapine (Oranje et al., 2002) and risperidone (Kumari et al., 2002) similarly restored PPI deficits. In contrast, other studies have not found a reversal with antipsychotic treatment, for example treatment with risperidone (Mackeprang et al., 2002; Oranje et al., 2002) or haloperidol or olanzapine (Duncan et al., 2003). Because the cause of PPI deficits in schizophrenia is unknown, the mechanism by which antipsychotic drugs potentially modulate this deficit remains unclear. Therefore, in the present study, we used a clearly defined agonist treatment, 8-OH-DPAT, to induce a disruption of PPI in rats. Our experiments confirm that affinity and efficacy data obtained in vitro in cell lines or membrane assays are difficult to extrapolate into an in vivo situation. Firstly, in addition to being essentially a mix between PPI and PPF, modulation of startle is controlled by a multitude of brain areas and neurotransmitter systems (Koch, 1999; Geyer et al., 2001). Secondly, even with selective pharmacological stimulation by 8-OH-DPAT, multiple receptor systems appear to be involved in the behavioral response; in this case, at least 5-HT_1A receptors and dopamine D₂ receptors. Clearly, for the interpretation of possible clinical effects of new antipsychotic drugs, preclinical in vivo testing is still crucial.

In conclusion, we have compared several antipsychotic drugs with respect to their ability to modulate the effect of 8-OH-DPAT on PPI. Although 8-OH-DPAT consistently disrupted PPI at the 100-ms ISI, only haloperidol and aripiprazole were able to inhibit this effect. The action of these antipsychotics was mimicked by raclopride and buspirone. Our results provide new insight into the interaction of anti-psychotic drugs with central mechanisms involved in PPI, a behavioral model with relevance to aspects of schizophrenia.