Psychopharmacology

, Volume 169, Issue 3, pp 234–239

Reversal of phencyclidine-induced prepulse inhibition deficits by clozapine in monkeys

Authors

    • The Nathan S. Kline Institute for Psychiatric Research
    • Department of PsychiatryNew York University School of Medicine
  • Shobhit S. Negi
    • The Nathan S. Kline Institute for Psychiatric Research
  • Scott V. Gerum
    • The Nathan S. Kline Institute for Psychiatric Research
  • Daniel C. Javitt
    • The Nathan S. Kline Institute for Psychiatric Research
    • Department of PsychiatryNew York University School of Medicine
Original Investigation

DOI: 10.1007/s00213-003-1533-8

Cite this article as:
Linn, G.S., Negi, S.S., Gerum, S.V. et al. Psychopharmacology (2003) 169: 234. doi:10.1007/s00213-003-1533-8

Abstract

Rationale

Prepulse inhibition (PPI) of the acoustic startle reflex is a measure of sensorimotor gating, which occurs across species and is deficient in severe neuropsychiatric disorders such as schizophrenia. In monkeys, as in rodents, phencyclidine (PCP) induces schizophrenia-like deficits in PPI. In rodents, in general, typical antipsychotics (e.g. haloperidol) reverse PPI deficits induced by dopamine (DA) agonists (e.g. apomorphine), but not those induced by N-methyl-d-aspartate (NMDA) receptor antagonists [e.g. phencyclidine (PCP)], whereas atypical antipsychotics (e.g. clozapine) reverse PPI deficits induced by DA agonists and NMDA antagonists. However, some discrepancies exist with some compounds and strains of rodents.

Objectives

This study investigated whether a typical (haloperidol, 0.035 mg/kg) and an atypical (clozapine, 2.5 mg/kg) antipsychotic could be distinguished in their ability to reverse PCP-induced deficits in PPI in eight monkeys (Cebus apella).

Methods

First, haloperidol dose was determined by its ability to attenuate apomorphine-induced deficits in PPI. Then, haloperidol and clozapine were tested in eight monkeys with PCP-induced deficits of PPI. Experimental parameters were similar to standard human PPI procedures, with 115 dB white noise startle pulses, either alone or preceded by 120 ms with a prepulse 16 dB above the 70 dB background noise.

Results

Clozapine reversed PCP-induced PPI deficits. In contrast, haloperidol did not significantly attenuate PCP-induced PPI deficits even at doses that significantly attenuated apomorphine effects.

Conclusions

In this primate model, clozapine was distinguishable from haloperidol by its ability to attenuate PCP-induced deficits in PPI. The results provide further evidence that PPI in nonhuman primates may provide an important animal model for the development of novel anti-schizophrenia medications.

Keywords

ApomorphineClozapineHaloperidolNMDA receptorNon-human primate modelPhencyclidine (PCP)Prepulse inhibition (PPI)SchizophreniaNeuroleptic sensitizationSensorimotor gating

Introduction

Prepulse inhibition (PPI) of the acoustic startle reflex is a measure of sensorimotor gating which occurs across species (Ison et al. 1973; Graham 1975; Braff et al. 1978; Braff and Geyer 1990; Swerdlow et al. 1999; Geyer et al. 2001; Javitt and Lindsley 2001; Linn and Javitt 2001). PPI is deficient in severe neuropsychiatric disorders such as schizophrenia (Braff et al. 1978, 2001; Braff and Geyer 1990) and these deficits correlate closely with cognitive dysfunction (Perry et al. 1999). Over the last 2 decades, rodent models of pharmacologically induced PPI deficits have been developed to examine and identify antipsychotic medications (reviewed in Geyer et al. 2001). In general, typical antipsychotics (e.g. haloperidol), reverse PPI deficits induced by dopamine (DA) agonists (e.g. apomorphine), but not those induced by N-methyl-d-aspartate (NMDA) receptor antagonists [e.g. phencyclidine (PCP)], whereas atypical antipsychotics (e.g. clozapine) reverse PPI deficits induced by DA agonists and NMDA antagonists. However, some discrepancies exist with some compounds and strains of rodents (Swerdlow et al. 1997, 1999; Geyer et al. 2001) and methods for measuring PPI differ between rodents and humans. We have demonstrated PPI of acoustic startle in New and Old World monkeys and have shown that PCP disrupts PPI in monkeys, similar to its effects in rodents (Javitt and Lindsley 2001; Linn and Javitt 2001). The goal of this study was to determine whether typical and atypical antipsychotics could be distinguished in their ability to reverse PCP-induced deficits in PPI in monkeys. The study had two components: first, to identify the dose of haloperidol to use for the second component of the study, which compared haloperidol with clozapine in their ability to attenuate PCP-induced deficits in PPI. The haloperidol dose was determined by its ability to significantly attenuate PPI deficits induced by the dopamine agonist apomorphine in a subset of animals. The clozapine dose was chosen as the lowest dose that attenuated PPI deficits induced by PCP in a different subset of subjects.

Materials and methods

Subjects

Subjects were eight adult female capuchin monkeys (Cebus apella) drawn from our existing social colonies. Details of animals, caging and husbandry, which follow NIH guidelines for enhanced environmental enrichment to promote primate psychological wellbeing, are described in Linn et al. (2001). The monkeys originated from the wild and so were of a heterogeneous genetic background. Four of the monkeys had previous long-term exposure to the typical antipsychotic fluphenazine decanoate ending approximately 10 years prior to the onset of this experiment (Linn et al. 2001). Four of the monkeys were antipsychotic naive.

PPI sessions and treatments

Experimental procedures and equipment for PPI sessions were similar to those described in Linn and Javitt (2001). Animals were placed in standard primate restraint chairs and stimuli were delivered and data collected using a San Diego Instruments EMG Startle Reflex System augmented with an external amplifier and a single field speaker suspended 12 cm in front of the subject. Responses were measured from right supraorbital and lateral canthal needle electrodes.

Experimental parameters were similar to standard human PPI procedures. Stimuli were 115 dB white noise startle pulses of 40 ms duration, either alone or preceded by 120 ms with a 20 ms prepulse 16 dB above the 70 dB background noise. Trial types were presented pseudorandomly with intertrial intervals ranging from 8 to 23 s. Test sessions were initiated with three startle pulse alone stimuli.

For antipsychotic experiments, subjects were treated with the atypical antipsychotic clozapine (2.5 mg/kg IM) or the typical antipsychotic haloperidol (0.035 mg/kg IM) 30 min prior to the first of two PPI sessions. Following the first session monkeys were treated with PCP (0.12 mg/kg IM) and, 10 min post-PCP injection, a second PPI session (antipsychotic+PCP) was conducted. This treatment sequence was necessary due to the short duration of PCP (and apomorphine, see below) effects. This method also controlled across conditions for habituation to stimuli and duration in the test apparatus for the relevant comparisons (antipsychotic+PCP, PCP alone, saline), as each was the second PPI session of each experiment. Antipsychotic treatment order was balanced across subjects (half received clozapine first and half received haloperidol first). All procedures were approved by the Institutional Animal Care and Use Committee.

Doses of haloperidol and clozapine were based on published reports from PPI studies with rodents (Bakshi et al. 1994; Johansson et al. 1995), behavioral studies in Cebus apella monkeys (Casey 1993), our prior experience administering antipsychotics to monkeys (e.g. Linn et al. 2001) and our own pilot experiments. Additionally, the haloperidol dose was determined based on its ability to significantly attenuate PPI deficits induced by apomorphine (0.4 mg/kg IM) in this species of monkey (experimental parameters were identical to those used for PCP sessions; see Results below). The clozapine dose was well tolerated and did not produce excess sedation or disruption of PPI. As four of the subjects had prior exposure to the typical antipsychotic fluphenazine decanoate, we noted any signs of dyskinesias or dystonias using ratings similar to those used in our previous experiments (Lifshitz et al. 1991; Linn et al. 2001).

The PCP dose was the same as has been shown, repeatedly, to induce PPI deficits in this species of primate and to cause a mild behavioral response without general impairment or sedation (Linn and Javitt 2001). However, as the original PCP alone experiments (PCP1) were conducted prior to antipsychotic+PCP experiments, we ran a second set of PCP alone experiments (PCP2) after completion of the antipsychotic treatment experiments. This was to verify that the PCP dose would still effectively induce deficits in PPI and to eliminate the possibility that any antipsychotic+PCP effects were due to habituation to this dose of PCP. The order of experimental PPI sessions was as follows: balanced baseline and PCP alone (PCP1) or saline PPI sessions, balanced clozapine or haloperidol alone and antipsychotic+PCP PPI sessions, a second round of baseline and PCP alone (PCP2) PPI sessions. Data from PCP1 and saline PPI sessions have been previously reported in Linn and Javitt (2001). At least 2 weeks separated successive PPI experiments for all subjects.

Statistical analysis

Paired t statistical comparison of means for startle alone amplitudes and PPI indicated no difference between PCP1 and PCP2 experimental sessions [startle alone, t(7)=1.3, NS; PPI, t(7)=0.79, NS; SPSS software]. Subsequently, for simplicity of statistical comparison and presentation, data from each subject for PCP1 and PCP2 were combined, with the average value, PCP, used for further analysis. Similarly, statistical comparison of means for startle alone and PPI for the three baseline PPI experimental sessions (baselines sessions for PCP1, PCP2 and saline treatment PPI sessions) did not differ and were combined, with the average value, Baseline, used for further analysis [startle alone, F(2,6)=0.665, NS; PPI, F(2,6)=2.24, NS; GLM for repeated measures, SPSS software].

For comparison of treatment conditions PPI was calculated by determining the percentage change between the peak amplitude mean of startle pulse alone and startle pulse with prepulse (percentage change=[(pulse alone)−(prepulse+pulse)]/(pulse alone)×100) (Cadenhead et al. 1999). Use of percentage change also helps to control for variation in amplitude of response among animals and across test days. PPIs for antipsychotic+PCP treatment sessions were compared to saline and PCP alone sessions for these same subjects using GLM repeated measures with within-subjects contrasts (Treatment factor with four levels: PCP alone, saline, clozapine+PCP, haloperidol+PCP). Haloperidol and clozapine alone sessions were similarly compared to the baseline values. Prior or no prior exposure to typical antipsychotics was included as a between-subjects factor (SPSS software). Values in text represent mean±SEM.

Results

Determination of haloperidol dose

The mean startle response amplitudes (n=5 monkeys) for the pulse alone and the pulse+prepulse conditions for the baseline, apomorphine (0.4 mg/kg IM) and haloperidol (0.035 mg/kg IM)+apomorphine are given in Table 1. Apomorphine and haloperidol+apomorphine effects on PPI are shown in Fig. 1. Apomorphine at a dose of 0.4 mg/kg IM induced consistent deficits in PPI [F(1,4)=8.16, P=0.046]. Lower doses of apomorphine did not induce consistent PPI deficits across subjects. Haloperidol at a dose of 0.035 mg/kg IM significantly attenuated apomorphine-induced deficits in PPI [F(1,4)=9.63, P=0.036]. Doses less than 0.035 mg/kg were ineffective in consistently attenuating apomorphine-induced deficits in PPI for all subjects. Doses greater than 0.035 mg/kg tended to produce heavy sedation that abolished startle response in tested subjects. Therefore, the 0.035 mg/kg dose of haloperidol was used for studies with PCP.
Table 1.

Response amplitudes and prepulse inhibition (PPI) for dose determination of haloperidol

Treatment

Startle pulse alone response amplitude (mean±SEM)

Startle with prepulse response amplitude (mean±SEM)

Prepulse inhibition (PPI) (mean±SEM) (percentage of startle)

Baseline (none)

355±95

151±49

59.1±6.6

Apomorphine (0.4 mg/kg)

288±97

237±54

4.4±13*

Haloperidol (0.035 mg/kg)+apomorphine

145±31*

85±53

35.6±6.9**

*Differs from baseline, P=0.041; **differs from apomorphine, P=0.036

Fig. 1.

The dose of haloperidol (0.035 mg/kg IM) was determined by its ability to attenuate apomorphine-induced deficits in PPI. A Differs from baseline (P=0.046). B Differs from apomorphine (P=0.036) but not from baseline (n=5 monkeys)

Effects of clozapine and haloperidol on PCP-induced deficits in PPI

The mean startle response amplitude (n=8 monkeys) for the pulse alone and the pulse+prepulse conditions for clozapine (2.5 mg/kg IM) and haloperidol (0.035 mg/kg IM) alone and with PCP (0.12 mg/kg IM) are given in Table 2. Values for baseline, saline and PCP alone are also included for comparison. There was no difference in startle response or PPI between baseline and saline conditions, indicating that response to stimuli did not significantly change between the first and second PPI sessions in an experiment [startle: t(7)=0.12, NS; PPI: t(7)=0.78, NS, Paired t). Figure 2 illustrates the percentage change in PPI versus saline for PCP alone, clozapine+PCP and haloperidol+PCP. Statistical comparison revealed a significant within-subjects effect for treatment [F(3,18)=3.91, P=0.026]. PCP alone (0.12 mg/kg IM) consistently induced deficits in PPI, even after repeated exposure [significantly differs from saline, F(1,6)=33.4, P=0.001]. Clozapine significantly attenuated PCP-induced deficits in PPI [significantly differs from PCP alone, F(1,6)=14.4, P=0.009, but not saline, F(1,6)=0.00, NS]. Haloperidol did not significantly attenuate PCP-induced deficits in PPI [F(1,6)=2.0, NS]. There was no difference in response to startle pulse alone for these four conditions [F(3,18)=2.73, NS]. However, comparison of clozapine and haloperidol alone with baseline indicated that while PPI did not differ [F(2,12)=0.25, NS], there was a difference among conditions in response to startle pulse alone [F(2,12)=5.3, P=0.02]. Contrasts indicated that the response to startle pulse alone was diminished for the clozapine alone condition [F(1,6)=12.75, P=0.01].
Table 2.

Response amplitudes and prepulse inhibition (PPI) for treatment conditions

Treatment

Startle pulse alone response amplitude (mean±SEM)

Startle with prepulse response amplitude (mean±SEM)

Prepulse inhibition (PPI) (mean±SEM) (percentage of startle)

Baseline (none)

324±72

139±45

58.4±4.6

Clozapine

172±44*

73±20

55.4±7.1

Haloperidol

219±40

68±13

61.8±8.1

Saline

319±89

126±45

61.5±4.2

PCP

295±62

172±43

41.9±4.7**

Clozapine+PCP

209±51

75±19

61.4±6.2***

Haloperidol+PCP

220±41

91±23

54.5±9.7

*Differs from baseline, P=0.01; **differs from saline, P=0.001; ***differs from PCP, P=0.009

Fig. 2.

Percentage change in PPI versus saline (dashed line) for PCP alone and for antipsychotic+PCP. Error bars indicate SEM. +Differs from saline, P=0.001; *differs from PCP, P=0.009. n=8 monkeys

Effects of prior exposure to typical antipsychotic medication

PPI and startle response values were not significantly different between monkeys with and without prior exposure to typical antipsychotic medication [treatment×prior antipsychotic exposure: PPI, F(3,18)=0.77, NS; Startle, F(3,18)=0.92, NS).

Dyskinesias

Dyskinesia and dystonia ratings were similar to those used in our previous experiments (e.g. Linn et al. 2001). Three of the four subjects with prior exposure to typical antipsychotic medication displayed mild to moderate limb dystonias and/or oral-buccal dyskinesias following their first acute treatment with haloperidol (0.035 mg/kg IM). The fourth monkey had some mild tremors but these could not be discerned from tremors that are sometimes observed when untreated animals are placed in the experimental chamber.

In contrast, none of the four typical antipsychotic naive subjects displayed any dystonias or dyskinesias after two or more acute treatments with haloperidol. A fifth monkey, used in the apomorphine/haloperidol pilot study, did not show any symptoms after the first treatment but displayed moderately strong dystonias and dyskinesias following the second acute treatment with haloperidol.

Discussion

Antipsychotics serve as the mainstay treatment for schizophrenia. Clozapine, a paradigmatic "atypical" antipsychotic, shows greater efficacy than typical agents while causing less in the way of extrapyramidal side effects. Neurochemical mechanisms responsible for the differential effects of clozapine are poorly understood. However, clozapine appears to be differentiated from typical antipsychotics by its ability to stimulate NMDA receptor-mediated neurotransmission in vivo (Arvanov and Wang 1999). Nevertheless, there are relatively few neurophysiological measures that distinguish atypical from typical antipsychotics, especially in primates. The present study demonstrates that, in monkeys, clozapine reverses PPI-disruptive effects of the NMDA antagonist PCP whereas haloperidol is ineffective. This distinction is consistent if clozapine and haloperidol data are compared with either of the PCP alone conditions. Effects were not due to inadequate haloperidol dosing since the same dose of haloperidol that failed to significantly reverse effects of PCP did reverse PPI-disruptive effects of the dopamine agonist apomorphine. Higher doses of haloperidol (used in our dose-determining pilot with apomorphine) tended to cause heavy sedation that abolished startle response.

Given the relatively small sample size it is possible that the difference in PPI between haloperidol+PCP and PCP alone conditions may reach significance with an increasing number of subjects. However, with a larger n, the difference in PPI between clozapine+PCP and haloperidol+PCP may also achieve significance. The present study, with n=8, was able to distinguish between clozapine and haloperidol by their ability to significantly attenuate the effects of PCP on PPI. This paradigm may therefore be useful for identifying other putative atypical antipsychotics.

Our results regarding the PPI disruptive effects of apomorphine and PCP in non-human primates, as well as the effects of typical and atypical antipsychotics on apomorphine or PCP induced disruption of PPI, are similar to what has generally been reported for rodent studies (reviewed in Geyer et al. 2001). In rodent and primate models, only atypical antipsychotics, such as clozapine or olanzapine restore PPI in animals treated with NMDA-antagonists such as PCP (Bakshi et al. 1994; Bakshi and Geyer 1995; Linn et al. 2001), suggesting that PCP-induced loss of PPI is a predictive measure for atypical antipsychotic properties. However, there are clinical and ethical reasons that preclude the use of PCP in humans. Ketamine is a non-competitive NMDA antagonist that has been used extensively in humans. Yet, the limited number of human studies of the effects of ketamine on PPI have had conflicting results, with some researchers reporting that ketamine induced deficits in PPI (Karper et al. 1994), while others found no change (van Berckel 1998) or an increase in PPI (prepulse facilitation) at the shortest interval between prepulse and startle stimulus (30 ms, Duncan et al. 2001). The differing results among human studies are probably due to dose or methodological differences. However, the discrepancy between results of studies in humans with ketamine and those in rodents with ketamine or PCP, or studies in non-human primates with PCP, may also reflect species differences in mechanisms regulating PPI. Alternatively, these differences may be due to specific properties of ketamine or PCP that make one of them more appropriate for models of schizophrenia. For example, there is also evidence in rodents that ketamine may increase PPI at short intervals between prepulse and startle stimulus (30 ms; Mansbach and Geyer 1991), indicating that ketamine, for reasons not yet clear, may differ from PCP in its effects on parameters of PPI.

Rodent studies indicate that chronic treatment with typical antipsychotics may reverse PCP-induced deficits in PPI (Martinez et al. 2000), though the reversal may be transient (Paabøl Andersen and Pouzet 2001). It is unclear whether this effect occurs in schizophrenia (Kumari et al. 1999; Weike et al. 2000). Future primate studies should evaluate the effects of chronic antipsychotic treatment on PCP-induced deficits in PPI.

Although prior exposure to typical antipsychotics did not affect PPI, we did observe dyskinesias when monkeys were re-exposed to a typical antipsychotic. We observed that three of four animals with prior long-term exposure to the typical antipsychotic fluphenazine decanoate displayed mild to moderate dystonias and dyskinesias upon their first exposure to an acute dose of haloperidol, even though their last exposure to fluphenazine occurred more than 10 years prior to the onset of the present study. We have previously reported that re-exposure to fluphenazine decanoate almost 2 years after initial exposure resulted in a more than 3-fold increase in observed dyskinesias (Linn et al. 2001) supporting the idea that intermittent treatment may lead to an increased incidence of extrapyramidal side effects (EPS) (van Harten et al. 1998). Our current results indicate that sensitization to typical antipsychotics may extend for more than 10 years in capuchin monkeys. Studies with typical antipsychotics in capuchin monkeys also indicate that the sedative effects of typical antipsychotics tend to decrease with repeated exposure (Casey 1988; Linn et al. 2001). In humans, sedative effects are postulated to account for the decreased startle response observed with antipsychotic treatment, though PPI is not affected (Graham et al. 2001). We observed a similar change in startle response amplitude, without affecting PPI, following clozapine treatment in our monkeys.

A recent review of rodent PPI models relevant to schizophrenia (Geyer et al. 2001) suggests that there are four distinct models based upon the manipulations used to disrupt PPI (i.e. DA agonists, serotonin agonists, NMDA or glutamate antagonists, and developmental perturbations). Our studies of PPI in non-human primates indicate that they constitute a valid model for DA agonist and NMDA antagonist pharmacological manipulations as models of schizophrenia.

In summary, our results indicate that PPI in monkeys may be an effective method for identifying clozapine-like atypical antipsychotics. Apomorphine induced deficits in PPI. Repeated exposure did not alter the deficit inducing effects of PCP on PPI. Clozapine reversed PCP-induced PPI deficits whereas haloperidol did not significantly attenuate PCP-induced PPI deficits even at doses that significantly attenuated apomorphine effects. Prior exposure to antipsychotics does not appear to be a confound for PPI studies. However, monkeys (Cebus apella) may remain sensitized to the effects of long-term treatment with typical antipsychotics for at least 10 years after the final dose. PPI in typical antipsychotic-sensitized monkeys may be a useful model for evaluating the potential for induction of EPS at effective treatment doses in new putative antipsychotics.

Acknowledgements

We thank Tammy McGinnis and Noelle O'Connell for help with data collection. This study was funded by R01 DA03383 and NIH 55620 to D.C.J.

Copyright information

© Springer-Verlag 2003