Psychopharmacology

, Volume 193, Issue 1, pp 45–54

Putative antipsychotics with pronounced agonism at serotonin 5-HT1A and partial agonist activity at dopamine D2 receptors disrupt basal PPI of the startle reflex in rats

Authors

    • Division of Neurobiology 2Centre de Recherche Pierre Fabre
  • Alexandra Galinier
    • Division of Neurobiology 2Centre de Recherche Pierre Fabre
  • Joël Besnard
    • Division of Neurobiology 2Centre de Recherche Pierre Fabre
  • Adrian Newman-Tancredi
    • Division of Neurobiology 2Centre de Recherche Pierre Fabre
  • Ronan Depoortère
    • Division of Neurobiology 2Centre de Recherche Pierre Fabre
Original Investigation

DOI: 10.1007/s00213-007-0762-7

Cite this article as:
Auclair, A.L., Galinier, A., Besnard, J. et al. Psychopharmacology (2007) 193: 45. doi:10.1007/s00213-007-0762-7

Abstract

Introduction

Prepulse inhibition (PPI) of the startle reflex has been extensively studied because it is disrupted in several psychiatric diseases, most notably schizophrenia. In rats, and to a lesser extent, in humans, PPI can be diminished by dopamine (DA) D2/D3 and serotonin 5-HT1A receptor agonists. A novel class of potential antipsychotics (SSR181507, bifeprunox, and SLV313) possess partial agonist/antagonist properties at D2 receptors and various levels of 5-HT1A activation.

Materials and methods

It thus appeared warranted to assess, in Sprague-Dawley rats, the effects of these antipsychotics on basal PPI.

Results

SSR181507, sarizotan, and bifeprunox decreased PPI, with a near-complete abolition at 2.5–10 mg/kg; SLV313 had a significant effect at 0.16 mg/kg only. Co-treatment with the 5-HT1A receptor antagonist WAY100,635 (0.63 mg/kg) showed that the 5-HT1A agonist activity of SSR181507 was responsible for its effect. By contrast, antipsychotics with low affinity and/or efficacy at 5-HT1A receptors, such as aripiprazole (another DA D2/D3 and 5-HT1A ligand), and established typical and atypical antipsychotics (haloperidol, clozapine, risperidone, olanzapine, quetiapine, and ziprasidone) had no effect on basal PPI (0.01–2.5 to 2.5–40 mg/kg).

Discussion

The present data demonstrate that some putative antipsychotics with pronounced 5-HT1A agonist activity, coupled with partial agonist activity at DA D2 receptors, markedly diminish PPI of the startle reflex in rats.

Conclusions

These data raise the issue of the influence of such compounds on sensorimotor gating in humans.

Keywords

5-HT1A agonistAntipsychoticsDopamine D2 antagonistPrepulse inhibitionSensorimotor gatingStartle reflex

Introduction

Schizophrenia patients present information-processing abnormalities, and in particular, exhibit deficits in sensory processing such as the P50/P300 event-related potentials (Muller et al. 2001; Freedman et al. 2003), latent inhibition (Weiner et al. 1996; Moser et al. 2000; Weiner 2003), and prepulse inhibition (PPI) of the startle reflex (Braff et al. 1978; Swerdlow et al. 2000). PPI is defined as the attenuation of the response to a startling stimulus (pulse) when such a stimulus is preceded by a brief stimulus of subthreshold intensity (prepulse). Because PPI is relatively easy to implement and less time-consuming than models such as latent inhibition, it is widely used to study deficits of information processing, and has been the object of extensive pharmacological evaluation in rodents (see Geyer et al. 2001 for review). In particular, it has been reported that activation of serotonin 5-HT1A receptors by agonists such as 8-OH-DPAT and buspirone (Nanry and Tilson 1989; Rigdon and Weatherspoon 1992; Sipes and Geyer 1995) and agonists at dopamine (DA) D2 receptors (Mansbach et al. 1988; Varty and Higgins 1998) disrupt basal PPI in rats.

Schizophrenia patients are generally responsive to current antipsychotics, which have various levels of DA D2 receptor antagonism thought to be responsible—and even sufficient—to account for their clinical efficacy against positive symptoms (Kapur and Remington 2001). However, many antipsychotics are not adequately efficacious against negative symptomatology and deficits of the cognitive sphere, in particular, impairments of information processing.

Recently, activation at 5-HT1A receptors has been proposed to offer attractive perspectives as a possible additional (to DA D2 receptors) target for antipsychotic activity (Millan 2000; Bantick et al. 2001; Ichikawa et al. 2001). In fact, 5-HT1A receptor agonists prevent catalepsy (an animal model of extrapyramidal signs, EPS) produced by DA D2 receptors blockade (Broekkamp et al. 1988; Invernizzi et al. 1988; Wadenberg et al. 1994; Prinssen et al. 1999; Depoortere et al. 2003; Kleven et al. 2005) and reduce the incidence of EPS in schizophrenia patients treated with haloperidol (Goff et al. 1991; Moss et al. 1993; Yoshida et al. 1998). 5-HT1A receptor activation also increases DA release in prefrontal cortex (Rollema et al. 1997, 2000; Millan et al. 1998; Ichikawa et al. 2001), indicative of beneficial activity against negative symptomatology and cognitive disturbances of schizophrenia (Kapur and Remington 1996). These considerations have led to the development of compounds with preferential DA D2 receptor antagonist/partial agonist and 5-HT1A agonist activities such as bifeprunox (Feenstra et al. 2001; Wolf 2003), SSR181507 (Claustre et al. 2003; Depoortere et al. 2003; Boulay et al. 2004; Terranova et al. 2005), SLV313 (Feenstra et al. 2002; McCreary et al. 2002), and sarizotan (now developed as an anti-dyskinetic: Bibbiani et al. 2001; Rabiner et al. 2002; Bartoszyk et al. 2004).

In a preceding paper (Auclair et al. 2006), we showed that the level of efficacy at 5-HT1A receptors was crucial in the ability of antipsychotics to prevent apomorphine-induced PPI deficits, a preclinical test purported to detect antipsychotic activity (Geyer et al. 2001). In effect, antipsychotics with potent DA D2 receptor antagonist and no 5-HT1A receptor agonist properties (haloperidol, risperidone, olanzapine) were active in this model. Antipsychotics exhibiting moderate 5-HT1A receptor agonist properties (clozapine, nemonapride, ziprasidone, aripiprazole) reversed these deficits at some doses only. However, those with marked 5-HT1A receptor agonist and/or partial DA D2 receptor agonist activities (sarizotan, SLV313, and SSR181507) had no reversing effects; confirmation that 5-HT1A receptor activation was responsible for this was obtained in interaction studies using the 5-HT1A receptor antagonist WAY100,635.

Following those observations and taking into consideration the various reports that 5-HT1A and DA D2 receptor agonists decrease PPI in rats, the question arises as to the effects of these mixed DA D2/5-HT1A compounds on basal PPI in this species (considering that some of them, including SSR181507 and bifeprunox, are partial agonists at DA D2 receptors; Bruins Slot et al. 2006; Cosi et al. 2006). The question is relevant to the fact that schizophrenia patients present diminished basal PPI levels (Braff et al. 1978; for review, Swerdlow et al. 2000) and that deficient PPI is associated with impaired functional status in schizophrenia (Swerdlow et al. 2006). It would thus appear desirable, at least on theoretical considerations, for an antipsychotic not to further disrupt PPI. We therefore examined the effects of these mixed DA D2/5-HT1A compounds on basal PPI, as the influence of their 5-HT1A agonist properties on basal sensorimotor gating has not been characterized. Effects of these compounds on PPI were interpreted in the light of their in vitro affinities and efficacies at DA D2 and 5-HT1A receptors. Reference 5-HT1A receptor agonists and established antipsychotics were also included for comparison. Finally, interaction studies with the specific 5-HT1A receptor antagonist WAY100,635 were undertaken whenever appropriate.

Materials and methods

Animals

Male Sprague–Dawley rats (ICO: OFA SD, Iffa Credo, Les Oncins, France), weighing 180 ± 20 g upon arrival, were group-housed (n = 5 per cage) in an environmentally controlled room (temperature 21 ± 1°C and relative humidity 55 ± 5%) on a 12-h/12-h light/dark cycle (lights on at 7:00 a.m.) in stainless steel cages with grid flooring (internal dimensions, 26 × 42 × 18 cm; w × l × h). Animals were held in quarantine for 4 to 8 days, with free access to standard laboratory food (A04, Scientific Animal Food and Engineering, Epinay sur Orge, France) and filtered water (0.22-μm diameter pores; in bottles). A 5-day acclimatization period was allowed before animals were used in experiments. Twenty-four hours before testing, the animals were individually housed in an environmentally controlled test room in plastic hanging cages with a grid floor (internal dimensions, 11 × 31 × 18 cm; w × l × h) where they had free access to water. Animals were handled and cared for in accordance with the “Guide and Care and Use of Laboratory Animals (National Institutes of Health) and the European Directive 86/609 and was carried out in compliance with French regulations and the local ethical committee guidelines for animal research.

Apparatus

Startle chambers (SR LAB, San Diego Instruments, San Diego, CA) were used. Each chamber consisted of a transparent acrylic cylinder (inside diameter, 8.8 cm; inside length, 18.4 cm) resting on a Plexiglas base in a sound-attenuated, ventilated enclosure illuminated by a 15-W lamp. Pulses and prepulses (white noise bursts) were presented via a loudspeaker mounted 28 cm above the animal. Startle reflexes within the cylinder were detected by a piezoelectric accelerometer attached to the base. Response sensitivities were calibrated using a standard calibrator tube (San Diego Instruments) and were adjusted to 150 units. Sound levels were calibrated (±1 dB) using the A scale of a sound level meter (Radio Shack, Tandy, Fort Worth, TX). All events were controlled and recorded on a PC using the San Diego Instruments Startle software.

Procedure

Pre-test and test sessions

Animals, which were used only once, were pre-tested in startle chambers 1 h and 45 min before the pharmacological challenge (test) session. The pre-test and the test sessions were performed during the light phase of the diurnal cycle of rats. Three different trial types were presented against a continuous 70-dB background noise: 118 dB pulse (pulse alone; PA) and 78 dB prepulse (pp) followed by a 118 dB pulse (prepulse–pulse; ppP) and no PA/no pp (NP). The PA duration was 40 ms, the pp duration 20 ms, and the interval between the onset of the pp and that of the PA was 100 ms. Sessions started with a 5-min adaptation period, after which, the animals were exposed to ten PA (included to induce habituation to startle such that habituation during the following PPI assessment would be minimized: These trials were not used for data analysis). These ten PA trials were followed by ten PA, ten ppP, and three NP trials presented in a pseudo-random order. The interval between trials was variable, but with a median of 15 s. The startle response was defined as the average of 100 consecutive samples of 1ms (amplitude) which were recorded from the onset of the pulse by means of a 12-bit AD acquisition card (range, 0 to 4,095 AU). Rats were returned to their home cages at the end of the pre-test session. This pre-test session, which lasted 13 min, was used to habituate rats to the procedure as well as to eliminate rats that showed any one of these four exclusion criteria at least once: (1) an amplitude of the startle reflex after presentation of the PA less than 10 AU (i.e., too low a startle response) or (2) more than four amplitudes equal to 4,095 (ceiling value of the 12-bits AD acquisition card) for the PA or the ppP condition (i.e., too high a startle response) or (3) a (PA–NP) amplitude/PA amplitude <80% (i.e., rat showing too high a level of motor activity in-between stimulus presentation periods) or (4) a pp amplitude > PA amplitude (i.e., a prepulse potentiation). Data from these pre-tests sessions are not reported.

Test sessions were in all respect similar to the pre-test sessions (vide supra). Rats that showed any one of the first three exclusion criteria (vide supra) at least once during the test session were not retained for data analysis (2.1% of rats were excluded based on this composite criterion).

Pharmacological treatments

At the end of the pre-test session, animals were injected s.c. or i.p. 60 min before the beginning of the test session with the test compound or its vehicle and 45 min later with saline s.c.. This second injection was performed so as to be under conditions similar to those used in the prevention of apomorphine deficits by antipsychotics, where vehicle was replaced by apomorphine (Auclair et al. 2006).

For the interaction of SSR181507 with the 5-HT1A antagonist WAY100,635 (0.63 mg/kg) the injection protocol was as follows:
https://static-content.springer.com/image/art%3A10.1007%2Fs00213-007-0762-7/MediaObjects/213_2007_762_Figa_HTML.gif

After each injection, the animals were put back in their home cages.

Data analysis

For each test session, the median of the amplitude of the startle responses for the last ten PA trials and for the ten ppP was calculated. The percentage PPI was calculated as follows:
$$ \frac{{{\left( {{\text{median}}\,{\text{PA}}\,{\text{amplitude}} - {\text{median}}\,{\text{ppP}}\,{\text{amplitude}}} \right)} \times 100}} {{{\text{median}}\,{\text{PA}}\,{\text{amplitude}}}} $$

Median was used instead of mean, as it is considered to reduce the contribution of “outliers” and is preferable to describe “small” populations (i.e., comprising seven to eight animals as is the case here; Myers and Well 1995). Data (percentage PPI) were analyzed with a one-way analysis of variance (ANOVA), with the treatment as the between-subjects factor, followed by a Dunnett’s post hoc test for comparison with the appropriate vehicle/vehicle group. An average of two rats were injected with vehicle/saline during each test session, and data from all these “control” rats were pooled at the end of the study. This was done to minimize the number of animals used.

For the experiment on the interaction between WAY100,635 and SSR181507, a two-way ANOVA, with the pretreatment (WAY100,635 or vehicle) and the treatment (SSR181507 or vehicle) as the in-between factors, was used. It was followed by a Bonferroni’s post hoc test for multiple comparisons.

Drugs

The following compounds were obtained commercially: haloperidol and buspirone HCl, (Sigma RBI, St. Quentin Fallavier, France), (+)8-hydroxy-2-(di-n-propylamino)tetralin (+)8-OH-DPAT hydrobromide), and clozapine (Tocris, Illkirch, France). Ziprasidone HCl, aripiprazole, bifeprunox mesylate (DU-127090; N-[4-[(3-chloro-4-fluorophenyl) amino]-7-[3-(morpholin-4-yl)propoxy]quinazolin-6-yl]prop-2-enamide mesylate), SSR181507 HCl ((3-exo)-8-benzoyl-N-[[(2S)7-chloro-2,3-dihydro-1,4-benzodioxin-1-yl]methyl] -8-azabicyclo[3.2.1]octane- 3-methanamine monohydrochloride), sarizotan HCl (EMD-128130; (-)-3-[[[(R)-2-chromanylmethyl]amino]methyl]-5-(p-fluorophenyl) pyridine monohydrochloride), quetiapine, risperidone, olanzapine, and WAY100,635 (N-[2-[4-(2-methoxyphenyl)-1-piperazinyl]-N-(2-pyridinyl)cyclohexanecarboxamide) were synthesized by J.-L. Maurel from the Centre de Recherche Pierre Fabre (Castres, France). SLV313 HCl (piperazine, 1-(2,3-dihydro-1,4-benzodioxin-5-yl)-4-[[5-(4-fluorophenyl)-3-pyridinyl]methyl) was generously donated by Solvay Pharmaceuticals (Weesp, The Netherlands). SSR181507, buspirone, (+)8-OH-DPAT and WAY100,635 were prepared in distilled water and administered s.c.. SLV313, olanzapine, risperidone, quetiapine, clozapine, and haloperidol were prepared in distilled water with a drop of lactic acid, after which, the pH was adjusted to 5–7 with a 1 N solution of sodium hydroxide, and were injected s.c. (except clozapine, i.p.). Aripiprazole, bifeprunox, sarizotan, and ziprasidone were prepared as a suspension in aqueous Tween 80 (two drops per 10 ml distilled water) and administered i.p. Controls consisted of an injection of saline s.c. (for SSR181507, buspirone, (+)8-OH-DPAT, WAY100,635, SLV313, olanzapine, risperidone, quetiapine, and haloperidol) or i.p. (for clozapine) or saline with Tween 80 i.p. (for all other compounds). Distilled water alone, or with a mixture of lactic acid/NaOH or with Tween 80, were not administered, as these vehicles produce pain and discomfort in rats (most probably because of too low an osmolarity). An injection volume of 10 ml/kg was used throughout, and doses refer to the weight of the free base. For each compound, doses were administered in an unsystematic order.

Results

Full and partial 5-HT1A receptor agonists disrupt basal PPI level

As compared to control rats (circles in panels of Fig. 1), the selective 5-HT1A receptor agonist 8-OH-DPAT monotonically reduced basal PPI (Fig. 1a), with significant effects from 0.63 mg/kg s.c. [Dunnett’s test after significant ANOVA: F(4,63) = 13.54, p < 0.01]. Buspirone, a partial 5-HT1A receptor agonist, had a biphasic effect, with low doses diminishing PPI, whereas the percentage of PPI recorded after administration of 10 mg/kg was not significantly different from that of controls [Dunnett’s test after one-way ANOVA: F(4,65) = 2.44, p = 0.056].
https://static-content.springer.com/image/art%3A10.1007%2Fs00213-007-0762-7/MediaObjects/213_2007_762_Fig1_HTML.gif
Fig. 1

5-HT1A receptor agonists disrupt basal levels of PPI. Rats were injected 60 min before testing with vehicle or test compound, followed 45 min after by saline. The percentage PPI was calculated as: \( \frac{{{\left( {{\text{median}}\,{\text{PA}}\,{\text{amplitude}} - {\text{median}}\,{\text{ppP}}\,{\text{amplitude}}} \right)} \times 100}} {{{\text{median}}\,{\text{PA}}\,{\text{amplitude}}}} \). Symbols are means±SEM. *p < 0.05 and **p < 0.01, compared with the vehicle/vehicle (Veh/Veh) control group; Dunnett’s post hoc test after one-way ANOVA. N = 5–7 rats per dose, 42 for controls

Novel putative antipsychotics selectively targeting DA D2 and 5-HT1A receptors disrupt basal PPI level

SSR181507 and sarizotan induced marked, dose-dependent, and significant (ANOVA’s: both F > 6.36, both p < 0.01) deficits of PPI, with near complete disruption seen with the highest doses tested (Fig. 2). Bifeprunox also produced a monotonic but less robust effect, with the highest dose (10 mg/kg) producing a circa 42% reduction [Dunnett’s test after significant ANOVA: F(4,53) = 2.57, p < 0.05]. SLV313 induced a more moderate and biphasic disruption of basal PPI, with a significant effect at 0.16 mg/kg only [Dunnett’s test following significant ANOVA: F(5,71) = 2.49, p < 0.05].
https://static-content.springer.com/image/art%3A10.1007%2Fs00213-007-0762-7/MediaObjects/213_2007_762_Fig2_HTML.gif
Fig. 2

Novel putative antipsychotics selectively targeting D2 and 5-HT1A receptors differentially affect basal levels of PPI. Rats were injected 60 min before testing with vehicle or test compound, followed 45 min after by saline. See legend of Fig. 1 for details of PPI calculation. *p < 0.05 and **p < 0.01, compared with the vehicle/vehicle (Veh/Veh) control group; Dunnett’s post hoc test after one-way ANOVA. N = 7 rats per dose, 30–42 for controls

Pretreatment with the 5-HT1A receptor antagonist WAY100,635 blocks the disrupting effects of SSR181507

With respect to the control group (leftmost bar of Fig. 3), rats treated with 2.5 mg/kg i.p. of SSR181507 (second bar from the left) presented a profound reduction of basal PPI of a magnitude similar to that found in the preceding experiment (see Fig. 2). Administration of the 5-HT1A receptor antagonist WAY100,635 (0.63 mg/kg s.c.), which by itself had no effect on PPI (rightmost bar), fully reversed the disruptive effects of SSR181507 (second bar from right). This reversal was significant [Bonferroni’s post hoc test, after significant pretreatment, treatment effects, but non-significant interaction: F(1,37) = 11.51, p < 0.01, F(1,37) = 7.48, p = 0.01, and F(1,37) = 3.30, p > 0.05, respectively].
https://static-content.springer.com/image/art%3A10.1007%2Fs00213-007-0762-7/MediaObjects/213_2007_762_Fig3_HTML.gif
Fig. 3

Pretreatment with the 5-HT1A receptors antagonist WAY100,635 prevents the PPI-disrupting effects of SSR181507. Rats were injected with vehicle or WAY100,635 (0.63 mg/kg s.c.), vehicle or test compound, followed by saline, 75, 60, and 15 min, respectively, before testing. **p < 0.01; Bonferroni’s post hoc test after significant two-way ANOVA. See legend of Fig. 1 for details of PPI calculation. Amplitudes for PA were: 80.0 + 9.6 (n = 17), 128.6 + 10.2## (n = 9), 71.4 + 14.7 (n = 7), and 46.7 + 4.4# (n = 7), for the veh–veh–veh, veh–SSR–veh, WAY–veh–veh, and WAY–SSR–veh groups, respectively. #p < 0.05, ##p < 0.01, compared with the veh–veh–veh control group; Dunnett’s post-hoc test after significant one-way ANOVA; F(3,36) = 8.22, p < 0.01

Typical and atypical reference antipsychotics do not substantially modify basal PPI level

Over dose-ranges that encompassed pharmacologically active doses in classical behavioral tests, neither haloperidol, risperidone, clozapine, olanzapine, aripiprazole, nor quetiapine significantly affected basal levels of PPI (ANOVA’s: all F < 2.10, all p > 0.05; Fig. 4). However, a single dose of ziprasidone (0.16 mg/kg i.p.) gave a statistically higher level of PPI [F(4,53) = 3.81, p < 0.01].
https://static-content.springer.com/image/art%3A10.1007%2Fs00213-007-0762-7/MediaObjects/213_2007_762_Fig4_HTML.gif
Fig. 4

Established typical and atypical antipsychotics do not substantially modify basal levels of PPI. Rats were injected 60 min before testing with vehicle or test compound, followed 45 min after by saline. See legend of Fig. 1 for details of PPI calculation. **p < 0.01, compared with the vehicle/vehicle (Veh/Veh) control group; Dunnett’s post hoc test after one-way ANOVA. Note that there were two types of controls for i.p. treatments (with and without Tween 80). N = 7 rats per dose, 30–42 for controls

It should be noted that effects of compounds on PPI and on pulse alone (PA) amplitude did not appear to be correlated (Table 1). For example, some drug treatments did not reduce PA amplitude, but could disrupt PPI to a great extent (SSR181507) or minimally (SLV313), whereas other treatments could either diminish PA amplitude or preserve it and, at the same time, leave PPI intact (olanzapine) or even had a tendency to potentiate it (aripiprazole).
Table 1

Effects of compounds on pulse alone amplitude

Dose (mg/kg)

0

0.01

0.04

0.16

0.63

2.5

10

40

8-OH-DPAT

122.4 ± 12.5a

NT

144.6 ± 35.7

121.9 ± 27.8

128.8 ± 20.1

110.4 ± 24.6

NT

NT

Buspirone

122.4 ± 12.5a

NT

NT

76.6 ± 13.6

94.1 ± 16.1

177.0 ± 46.0

152.7 ± 36.9

NT

SSR181507

122.4 ± 12.5a

NT

100.3 ± 18.4

230.0 ± 80.4

149.1 ± 38.6

252.3 ± 82.5*

103.6 ± 7.2

NT

Sarizotan

96.6 ± 10.7b

NT

NT

122.1 ± 20.8

93.7 ± 17.1

84.8 ± 9.7

145.7 ± 42.4

NT

Bifeprunox

96.6 ± 10.7b

NT

NT

71.6 ± 20.8

96.4 ± 24.1

154.1 ± 49.1

174.8 ± 42.7

21.9 ± 5.6

SLV313

122.4 ± 12. 5a

180.0 ± 45.7

80.5 ± 9.2

180.3 ± 67.2

181.1 ± 41.4

157.8 ± 31.8

NT

NT

Haloperidol

122.4 ± 12.5a

164.6 ± 52.0

76.9 ± 13.5

106.9 ± 43.5

43.6 ± 5.2

29.2 ± 8.4*

NT

NT

Risperidone

122.4 ± 12.5a

NT

NT

44.1 ± 4.9*

45.9 ± 8.8*

56.9 ± 9.6

NT

NT

Clozapine

86.8 ± 12.3c

NT

NT

NT

113.9 ± 42.5

94.3 ± 29.2

36.4 ± 8.5

NT

Olanzapine

122.4 ±  12.5a

NT

NT

NT

97.8 ± 11.9

85.6 ± 20.5

24.0 ± 7.2**

NT

Aripiprazole

96.6 ± 10.7b

NT

NT

117.9 ± 32.1

75.4 ± 11.4

67.5 ± 9.5

86.6 ± 27.8

89.3 ± 27.7

Ziprasidone

96.6 ± 10.7b

NT

NT

141.1 ± 55.5

81.4 ± 12.2

70.1 ± 18.2

71.2 ± 8.5

NT

Quetiapine

96.6 ± 10.7 b

NT

NT

NT

NT

73.6 ± 14.4

50.9 ± 12.4*

54.7 ± 12.1

Each number is the mean±SEM amplitude of the startle reflex (AU) after presentation of the pulse alone.

aSaline administered s.c.

bSaline with Tween 80 administered i.p.

cSaline administered i.p.

*p < 0.05, **p < 0.01, compared with the vehicle/vehicle (Veh/Veh) control group; Dunnett’s post hoc test after significant one-way ANOVA

NT Not tested

Discussion

The influence on basal PPI of the “new generation” antipsychotics exhibiting DA D2/5-HT1A properties appears closely related to their different levels of interaction/activation of these receptors. In this context, it is interesting to compare data from the present study with results from recent in vitro measures of affinity and efficacy at DA D2 and 5-HT1A receptors (Bruins Slot et al. 2006).

Novel mixed DA D2/5-HT1A antipsychotics disrupt PPI to various extents

Sarizotan, SSR181507, bifeprunox and—to a lesser extent SLV313—disrupted basal PPI in rats. Sarizotan has high affinity for, and more importantly, very high efficacy at 5-HT1A receptors (Table 2, second and third columns), and furthermore, has higher affinity for 5-HT1A than for DA D2 receptors (Ki ratio of 0.3, sixth column). As a result, from 0.63 mg/kg, its activation of 5-HT1A receptors starts diminishing PPI (last column), with limited influence of its DA D2 receptor antagonistic property (17% of activity of dopamine in an activation of ERK phosphorylation assay: fifth column).
Table 2

Comparison of in-vitro affinities and efficacies at dopamine D2 and 5-HT1A receptors with effects on PPI

 

5-HT1A

DA D2

 

Affinity (pKi)a

Efficacy (Emax)b

Affinity (pKi)a

Efficacy (Emax)b

Ki ratio 5-HT1A/D2a

PPI disruption

Sarizotan

8.65

118

8.18

17

0.3

++

SSR181507

8.53

73

18.41

66

0.8

++

Bifeprunox

7.19

74

8.83

76

44

+

SLV313

8.64

72

8.58

0

0.9

±

Aripiprazole

7.42

60

8.59

59

15

0

Clozapine

6.31

33

6.38

0

3.3

0

Haloperidol

5.77

NT

9.01

NT

1738

0

Olanzapine

5.38

0

7.82

NT

275

0

Risperidone

6.03

NT

8.70

NT

468

0

Ziprasidone

8.30

56

7.90

0

0.4

0

(+)8-OH-DPAT

8.85

100

NT

NT

++

Buspirone

7.65

66

NT

NT

±

aInhibition of [3H]8-OH-DPAT binding, rat cortex or inhibition of [3H]nemonapride binding, rat striatum; Newman-Tancredi et al. 2005

bActivation of ERK phosphorylation at human 5-HT1A or D2s receptors; Bruins Slot et al. 2006

NT Not tested

– Cannot be determined

Both SSR181507 and bifeprunox exhibit substantial efficacy at 5-HT1A receptors (albeit lower than that of sarizotan). The former of the two compounds has slightly higher affinity for 5-HT1A than for DA D2 receptors (Ki ratio of 0.8). Accordingly, even at fairly low doses (i.e., 0.16 and 0.63 mg/kg), effects on PPI are already observable. Furthermore, like bifeprunox, SSR181507 behaves as a partial DA D2 receptor agonist (66% activity of DA). This may accentuate further the deficit of basal PPI so that at higher doses (2.5 and 10 mg/kg), there is almost complete abolition of PPI. Bifeprunox has higher affinity for DA D2 than for 5-HT1A receptors (Ki ratio of 44). Hence, the partial agonist activity at DA D2 receptors (76% that of dopamine) would be expected to predominate at low doses (i.e., <1 mg/kg). This partial D2 receptor agonist activity may be the cause of the limited PPI disruption observed under these conditions. At higher doses (10 and 40 mg/kg), the efficacious 5-HT1A agonist activity would be recruited, exerting an additional decreasing effects on PPI.

The case for SLV313 is slightly different: Although it has substantial efficacy at 5-HT1A receptors, its Ki ratio is close to unity (0.9), and most importantly, it behaves as a DA D2 receptor antagonist (Table 2). As a result, at low doses, there is limited PPI deficit. As the dose increases, blockade of D2 receptors may exert a protective effect and will tend to cancel out the negative impact of 5-HT1A receptors activation on PPI. In fact, a similar explanation can be made for the biphasic effect of buspirone: At higher doses, D2 receptor antagonism (pKi, 7.49; in-house data) opposes the diminishing effects of 5-HT1A receptor activation, resulting, overall, in an absence of effect on PPI at higher doses. This assumption that a DA D2 receptor antagonist property protects from the negative effect of 5-HT1A receptors activation on PPI is further reinforced by previous findings with F15063 (an efficacious agonist at 5-HT1A receptors, but an antagonist at DA D2 receptors; Newman-Tancredi et al. 2006, 2007). This compound has no detrimental effect on basal PPI over a wide range of doses (Depoortere et al. 2007).

On the whole, these data with DA D2/5-HT1A novel antipsychotics are consistent with the disruptive influence of 5-HT1A receptor agonists (present results, Nanry and Tilson 1989; Rigdon and Weatherspoon 1992; Sipes and Geyer 1995), including the selective 5-HT1A receptor agonist, F13714 (Auclair et al. 2006). Recently, it was reported that the weak 5-HT1A receptor agonist buspirone slightly disrupted PPI in female volunteers (Gogos et al. 2006), in line with what had been previously observed in rodents. When co-administered with SSR181507, the 5-HT1A receptor antagonist WAY100,635 completely abolished the disruptive effects of SSR181507 on basal PPI. This demonstrates that 5-HT1A receptor activation is the major contributor to the disrupting activity of SSR181507 on basal PPI and probably of sarizotan and bifeprunox. Nonetheless, one cannot rule out a marginal effect of partial agonist activity at DA D2 receptors as participating to these PPI-disrupting effects. The predominant implication of 5-HT1A receptors activation is also in line with the lack of activity of sarizotan, SSR181507, and SLV313 against apomorphine-induced PPI disruption; indeed, a similar co-treatment experiment with WAY100,635 and SLV313 or SSR181507 clearly showed that the 5-HT1A agonist activity of these compounds prevented reversion of apomorphine-induced PPI deficit (Auclair et al. 2006).

One limitation of the present study is that the impact of novel D2/5-HT1A antipsychotics was assessed in Sprague–Dawley rats that presented a “normal” level of PPI. It would be informative to investigate the activity of these novel antipsychotics in strains of rats with innate decreased level of PPI (i.e., Brattleboro), or in Sprague–Dawley rats screened for low levels of PPI. Furthermore, chronic dosing studies may determine if there is tachyphylaxia to the PPI-disrupting effects of some of these compounds.

Antipsychotics with DA D2 antagonist activity and with little or no 5-HT1A agonism, spare basal PPI

Over a broad range of doses (0.01 to 40 mg/kg), none of the seven reference antipsychotics tested dose-dependently and robustly disrupted PPI. All antipsychotics, with the exception of aripiprazole, behave as DA D2 receptor antagonists (Table 2). Furthermore, the Ki ratio at 5-HT1A versus DA D2 receptors were 3, 15, 275, 468, and 1738, for clozapine, aripiprazole, olanzapine, risperidone, and haloperidol, respectively. As a consequence, DA D2 antagonist activity will largely prevail over the moderate (ziprasidone), low (clozapine, quetiapine), or negligible (olanzapine, risperidone, and haloperidol) 5-HT1A receptor activity. The lowest dose of ziprasidone (0.125 mg/kg) significantly increased PPI. While further studies are warranted to explore more thoroughly this issue, the present data clearly indicate that the increase is not due to agonist properties of ziprasidone at 5-HT1A or D2 receptors. The increase may be related to the multiple additional interactions of ziprasidone with adrenergic α1, 5-HT2A/2c, 5-HT7, or histaminergic H1 receptors (Leysen 2000).

Aripiprazole is a recently marketed antipsychotic, with some agonist activity at, but moderate affinity for 5-HT1A receptors, and it behaves as a DA D2 receptor partial agonist. It had no particular impact on basal PPI (and even non-significantly tended to potentiate PPI at higher doses tested). At low doses, the partial agonist activity at DA D2 receptors should slightly diminish PPI, with its 5-HT1A receptor agonist activity further participating to diminishing PPI as the dose increases. One possible explanation for a lack of PPI-disruption at higher doses might be that in rats, a major metabolite of aripiprazole is a pure DA D2 receptor antagonist (Lawler et al. 1999). This metabolite may oppose the partial agonist activity of the parent compound at DA D2 receptors, hence, the lack of effect on PPI. It should also be noted that aripiprazole has a particularly “rich” binding profile (Shapiro et al. 2003) interacting at numerous other receptors (e.g., 5-HT2A receptors), which might also explain its pharmacological activity.

In this context, other antipsychotics (risperidone, clozapine, olanzapine, ziprasidone, quetiapine) also have marked affinity for the 5-HT2A receptor. However, activation of these receptors has been reported to produce deficit of PPI (Geyer et al. 2001), and these antipsychotics act as antagonists at these sites. The lack of disruptive effects by these seven antipsychotics is, on the whole, in line with what has been reported by others (see Appendix 1 in Geyer et al. 2001).

Summary and conclusions

The ability of novel putative mixed D2/5-HT1A antipsychotics to disrupt basal PPI in rats is quite diverse: There seems to be a narrow window between efficacy and affinity balance at these receptors to avoid interference with basal PPI. Hence, too much 5-HT1A activation, especially when coupled to partial agonism at DA D2 receptors, produces marked decrease of basal PPI. However, 5-HT1A agonism is not necessarily coupled to PPI deficit; hence, F15063, a DA D2 receptor antagonist and 5-HT1A receptor agonist, with higher affinity at the former type of receptor, does not modify basal PPI over a wide dose-range (Depoortere et al. 2007).

Additionally, one should remain cautious about directly transposing findings on the pharmacological sensitivity of PPI from rodents to humans. Indeed, 5-HT1A receptor agonists can have opposite effects in rats (see above) and in mice (Dulawa et al. 2000). Second, ketamine, a phencyclidine-like compound, or the serotonin releaser MDMA, robustly disrupt PPI in rats (see review by Geyer et al. 2001), but can increase it in volunteers (Vollenweider et al. 1999; Abel et al. 2003). Similarly, opposing effects between the two species have been reported for the hallucinogenic drug psilocybin (Gouzoulis-Mayfrank et al. 1998; Geyer et al. 2001). However, the possibility remains that some new D2/5-HT1A putative antipsychotics might diminish basal PPI in human volunteers or schizophrenia patients. Studies would, therefore, be desirable to clarify the impact of these compounds on basal PPI in humans.

Acknowledgments

The authors thank Solvay Pharmaceuticals for supplying SLV313. All authors of this manuscript are employees of the Pierre Fabre Research Center.

Copyright information

© Springer-Verlag 2007