Effects of the selective dopamine D₁ antagonists NNC 01-0112 and SCH 39166 on latent inhibition in the rat

Abstract

Dopamine D₁ receptor blockade does not appear to be a prerequisite for antipsychotic activity since many clinically effective antipsychotics have little or no affinity for this receptor subtype. Clozapine, however, which has minimal liability for extrapyramidal symptoms, possesses affinities of similar order for D₁ and D₂ receptors. In earlier animal models used to predict antipsychotic effect, selective D₁ antagonists have shown effects similar to standard antipsychotics with preferential D₂ or mixed D₁/D₂ antagonism. We investigated the effects of haloperidol (0.1 mg/kg) and two selective D₁ antagonists, NNC 01-0112 (0.05, 0.1 and 0.2 mg/kg) and SCH 39166 (0.02, 0.2 and 2.0 mg/kg), on latent inhibition (LI) in rats. LI is a behavioural paradigm in which repeated nonreinforced preexposure to a stimulus retards subsequent associations to that stimulus. Disrupted LI has been suggested as a model for the attentional deficits in schizophrenia. Using preexposure to a flashing light stimulus, which subsequently served as a conditioned stimulus for suppression of water licking, we demonstrated a clear LI effect with haloperidol but with neither of the two D₁ antagonists. Since selective D₁ antagonists are not clinically effective, these results add further credibility for the relevance of LI as an animal model of psychosis.

Introduction

The clinical efficacy of classical antipsychotics has been demonstrated to be related to antagonism of the dopamine D₂ receptor subtype [17], [51]. D₁ receptor antagonism does not appear to be a prerequisite for antipsychotic activity, since many of these standard antipsychotics have little or no affinity for the D₁ receptor. However, in spite of being efficacious in the acute and maintenance treatment of schizophrenia, the clinical value of these agents is limited by several factors: they are ineffective in treating negative symptoms, they lack clinical efficacy in ‘treatment-resistant’ patients and they produce extensive side effects including extrapyramidal symptoms (EPS). By contrast, the atypical antipsychotic clozapine has minimal EPS liability in man [32] and is efficacious in patients with negative symptoms [26], [39] and in patients who do not respond to conventional treatment [26]. Furthermore, since in vitro studies on human brain tissue have shown that clozapine possesses affinities of comparable order for D₁ and D₂ receptors [18], it may be the case that D₁ receptor antagonism contributes to the antipsychotic effect of this compound.

SCH 23390, the prototype-selective D₁ receptor antagonist, was first described by Iorio [26] as a potential antipsychotic with an atypical pharmacological profile. However, it is unsuitable for clinical development because it has a very short biological half-life in the primate [2]. A number of analogues, which possess the same pharmacological profile as SCH 23390 but which have a longer duration of action, have since been synthesised and these include SCH 39166 as well as NNC 112, NNC 687 and NNC 756 (see Ref. [1] for a review). In vitro, specific D₁ antagonists bind in a saturable manner with high affinity to a single class of receptors, these receptors being selective for D₁ drugs [4]. Data from preclinical studies have demonstrated that selective D₁ antagonists reproduce the actions of standard antipsychotics, with preferential D₂ or mixed D₁ or D₂ antagonism, in animal models conventionally used to predict antipsychotic activity. These models include inhibition of conditioned responses, amphetamine cue, stereotyped behaviour [8], [9], [44], [45], inhibition of intracranial self-stimulation [43] and the paw test [48].

Studies with selective D₁ antagonists in nonhuman primates and rodents have produced data regarding their liability for negative symptoms and EPS. Acute administration of relatively high single doses of SCH 23390 has a comparable effect to haloperidol in that it induces dystonia in monkeys [5]. By contrast, NNC 01-0756 does not induce dystonia when administered in increasing doses, while the D₂ antagonist raclopride produces dystonia at low doses [22]. In monkeys sensitised to antipsychotic treatment, NNC 22-0215 and haloperidol both produce EPS when administered acutely, but there is rapid tolerance with the D₁ antagonist [6]. Furthermore, data from studies using rodents in the paw test imply that SCH 39166 has a lower potential for inducing EPS than traditional antipsychotics such as haloperidol. Such results indicate that the use of selective D₁ antagonists in the clinical situation may induce EPS for a short period only as opposed to the persistent symptoms associated with chronic treatment with standard antipsychotics.

Such preclinical findings suggest that selective D₁ antagonists are potential antipsychotics with less EPS than traditional antipsychotics. Unfortunately, this expectation of antipsychotic efficacy has not been fulfilled. When administered to patients, SCH 39166, the first selective D₁ receptor antagonist developed for clinical trials in schizophrenia, did not induce any apparent antipsychotic effects [11], [13], [29]. No beneficial changes were found in positive symptoms with this selective D₁ antagonist [11], [13], [29], although an apparent effect on negative symptoms was found by the latter study group. A preliminary study with NNC 01-0687 in chronic schizophrenic patients has reported a modest improvement in psychotic symptoms [28].

Thus, the preclinical tests produced false positives. This casts doubt upon the validity of such preclinical models used for predicting the antipsychotic activity of drugs. The aim of the present study was to investigate the effects of two selective D₁ antagonists, SCH 39166 and NNC 01-0112, on the animal latent inhibition (LI) model of schizophrenia. LI has superiority over other models of schizophrenia such as amphetamine-induced hyperactivity, which have been criticised for possessing little face or construct validity [16]. These models do not bear an obvious relation to the symptoms of the disease and, in addition, do not model the psychopathological construct hypothesised to underlie the disease process. The LI model simulates an attentional deficit that is a putative central characteristic of certain forms of schizophrenia. This deficit has often been described as an inability to distinguish between relevant and irrelevant stimuli. In the LI paradigm, repeated nonreinforced preexposure to a stimulus retards subsequent associations to that stimulus. LI has been described as reflecting the process of learning to ignore irrelevant stimuli [33], [34], [35], [36] and has been demonstrated behaviourally in both animals and humans.

LI is disrupted in rats by systemic administration of amphetamine [12], [31], [52], [53], [63], [64], [65], [67]. This disruption has been suggested as a model for the attentional deficit in schizophrenia [21], [23], [42], [52], [58], [63], [64]. In human studies, acute schizophrenics show a deficit in LI, whereas chronic schizophrenics exhibit no such deficit [3], [24]. Nonschizophrenic volunteers given amphetamine also fail to show LI [25].

The validity of the LI paradigm as a model of schizophrenic attentional deficit is strengthened by the fact that antipsychotics, including both typical and atypical compounds, reverse amphetamine-induced disruption of LI and demonstrate an emergence of LI when administered alone [10], [14], [20], [21], [30], [40], [42], [47], [54], [55], [57], [59], [60], [61], [62], [66]. Significantly, in a human study employing an auditory LI task, the typical antipsychotic chlorpromazine has been demonstrated to produce an LI effect in healthy volunteers [37].

The effects of the selective D₁ antagonist SCH 39166 have previously been assessed on LI, but only when this drug was administered centrally. Using the taste aversion paradigm, Ellenbroek et al. [15] demonstrated that administration of SCH 39166 into the medial prefrontal cortex had no influence on LI. The present study investigated the effects of peripheral administration of SCH 39166 (0.02, 0.2 and 2.0 mg/kg), NNC 01-0112 (0.05, 0.1 and 0.2 mg/kg) and haloperidol (0.1 mg/kg) on LI in rats. Haloperidol served as a comparator to ensure that any apparent LI effect with either of the selective D₁ antagonists could be taken as reliable. The enhancing effects of haloperidol on the LI model are well established [10], [14], [21], [48], [60], [62]. Pilot experiments conducted by us have shown 0.1 mg/kg haloperidol to be the most effective dose in enhancing LI.

A three stage off-baseline conditioned emotional response procedure modelled on that of Feldon and Weiner [20] and Weiner et al. [62], [68] was employed: (1) preexposure to the to-be-conditioned stimulus (flashing houselight); (2) conditioning, during which the flashing houselight stimulus was paired with a mild footshock; and (3) test, in which LI was measured by the degree of suppression of water licking elicited by the flashing houselight stimulus. Ten preexposures were used since it has been established that using a low number of preexposures produces a small LI effect and hence a more accurate evaluation of drug influences [14], [21], [41], [59].