Two procedures for training differential responses in alcohol and nondrug conditions

Drug Discrimination: Historical Origins, Important Concepts, and Principles

Research on the stimulus properties of drugs began with studies on state dependent learning during the first half of the twentieth century. From that research, an entirely new approach evolved called drug discrimination. Animals (including humans) could discriminate the presence or absence of a drug; once learned, the drug could serve as a discriminative stimulus, signaling the availability or nonavailability of reinforcement. Early drug discrimination research involved the use of a T-maze task, which evolved in the 1970s into a two-lever operant drug discrimination task that is still used today. A number of important concepts and principles of drug discrimination are discussed. (1) The discriminative stimulus properties of drugs are believed in large part to reflect the subjective effects of drugs. While it has been impossible to directly measure subjective effects in nonhuman animals, drug discrimination studies in human subjects have generally supported the belief that discriminative stimulus properties of drugs in nonhuman animals correlate highly with subjective effects of drugs in humans. In addition to the ability of the drug discrimination procedure to measure the subjective effects of drugs, it has a number of other strengths that help make it a valuable preclinical assay. (2) Drug discrimination can be used for classification of drugs based on shared discriminative stimulus properties. (3) The phenomena of tolerance and cross-tolerance can be studied with drug discrimination. (4) Discriminative stimulus properties of drugs typically have been found to be stereospecific, if a drug is comprised of enantiomers. (5) Discriminative stimulus properties of drugs reflect specific CNS activity at neurotransmitter receptors. (6) Both human and nonhuman subjects display individual differences in their sensitivity to discriminative stimuli and drugs. (7) The drug discrimination procedure has been used extensively as a preclinical assay in drug development. This chapter is the first in the volume The Behavioural Neuroscience of Drug Discrimination, which includes chapters concerning the discriminative stimulus properties of various classes of psychoactive drugs as well as sections on the applications and approaches for using this procedure.

5-HT1a Receptor Involvement in Temporal Memory and the Response to Temporal Ambiguity

It has previously been demonstrated that rats trained on the peak-interval procedure to associate two different cues with two different fixed interval schedules will generate a scalar peak function at an intermediate time when presented with the compound cue. This response pattern has been interpreted as resulting from the simultaneous retrieval of different temporal memories, and a consequential averaging process to resolve the ambiguity. In the present set of studies, we investigated the role that serotonin 1a receptors play in this process. In Experiment 1, rats were trained on a peak-interval procedure to associate the interoceptive states induced by saline and the 5-HT1a agonist, 8-OH-DPAT, with a 5 s or 20 s fixed-interval schedule signaled by the same tone cue (counter-balanced). While peak functions following administration of saline were centered at the appropriate time (5 s or 20 s), peak functions following administration of the agonist were centered around 7 s, irrespective of the reinforced time during training, suggesting agonist-induced disruption in selective temporal memory retrieval, resulting in increased ambiguity regarding the appropriate time at which to respond. In Experiment 2, rats were trained in a peak-interval procedure to associate a tone cue with a 10 s fixed interval and a light cue with a 20 s fixed interval. Administration of the 5-HT1a antagonist, WAY-100635, had no impact on timing when single cues were presented, but altered the intermediate, scalar, response to the stimulus compound, suggesting antagonist-induced disruption in the processes used to deal with temporal memory ambiguity. Together, these data suggest that manipulations of 5HT transmission at the 5-HT1a receptor cause changes in the temporal pattern of responding that are consistent with alterations in temporal memory processes and responses to temporal ambiguity.

Early preclinical studies of discriminable sedative and hallucinogenic drug effects

RATIONALE

One important technique in behavioral pharmacology is to train laboratory animals to discriminate between a psychoactive drug effect and a nondrug condition. Tests with different drugs have identified several categories of drugs that have different discriminable effects.

OBJECTIVES

The two authors describe and discuss the early research on discriminable effects of sedative and hallucinogenic drugs and their acquaintance with each other at Yale University prior to their early and frequent publications on discriminable drug effects. Herb Barry studied sedative drugs primarily and Jim Appel studied hallucinogenic drugs.

RESULTS

Sedative drugs include ethyl alcohol, barbiturates, and benzodiazepines. Their discriminable effects are largely attributable to the activation of an inhibitory neurotransmitter, gamma-amino butyric acid. Alcohol has the most pervasive effect in accordance with the high dose required to alter behavior. Hallucinogenic drugs include lysergic acid diethylamide and mescaline. They increase the activity of the neurotransmitter 5-hydroxytryptamine and, perhaps, dopamine in the central nervous system (CNS). In spite of their relatively low concentrations in the brain, both of these neurotransmitters have many important behavioral effects.

CONCLUSIONS

Various sedative drugs cause a discriminable decrease in the function of the CNS. Different types of sedatives can be discriminated from each other. Indole and phenylethylamine hallucinogens have potent discriminative stimulus properties, which are related to the actions of biogenic amine neurotransmitters in the CNS.

Creation and first 20 years of the society for the stimulus properties of drugs (SSPD)

Clinical observations and novels in the 19th century recognized that memory of some events can be retrieved only under the influence of the same drug condition that was present during the event. This dissociative effect of drugs probably reflects the same drug effects that were later called the discriminative stimulus effects of drugs. The Society for Stimulus Properties of Drugs (SSPD) was founded in 1978 as a forum for communications and periodic meetings on this drug effect. During its early years many of its members were psychologists, but subsequent to that time the most frequent research application has been for the pharmacological purpose of identifying new drugs that have the same discriminative stimulus attributes as a prototype training drug. The majority of members have been in the United States, but several major international meetings have been in Europe. The methods used by the society's members involve both neuropharmacological and psychological processes, allowing them to make unique contributions to the study of both mind and brain.

Strategies for understanding the pharmacological effects of ethanol with drug discrimination procedures

Ethanol appears to produce a stimulus complex, or compound cue, composed of distinct components that are mediated by different receptor systems. In ethanol vs. water discriminations, it appears that ethanol produces a redundant stimulus complex such that separate, receptor-mediated activity can serve as the basis for the discrimination. These discriminations have been termed redundant, because multiple features of the cue could serve as the basis of the discrimination. In ethanol vs. water discriminations, one common feature is the asymmetrical generalizations between components of the ethanol cue and ethanol. There is also evidence for overshadowing of one component by other components of the ethanol stimulus complex. It appears possible to transfer the basis of the ethanol cue from a redundant cue to a conditional cue with specific training procedures. When the discriminative stimulus effects of ethanol are juxtoposed with those of one component of the ethanol complex, as in ethanol vs. water vs. pentobarbital discriminations, the ethanol discrimination shifts to a conditional basis. The ability to antagonize an ethanol discrimination may be dependent upon whether the discrimination is based on redundant component stimuli or conditional presence of all component stimuli.

The discrimination and durability of an ethanol cue in young and mid-aged female mice

Young adult (6 months) and mid-aged (12 months) C57BL/6 mice both learned to discriminate ethanol (ETOH, 1.0 g/kg) although criterion performance occurred later for mid-aged mice. ETOH discrimination improved with increasing dose (0.25-1.0 g/kg) and the dose-response function was unaffected by age. The ETOH cue had declined by 40 min postinjection for young mice not unlike a previous report for young rats. In contrast, the ETOH cue remained discriminable at 40 min for mid-aged mice, an effect perhaps due to their slower rate of ETOH metabolism and accountable for the previously reported reduction in ethanol consumption by mid-aged mice. Retention tests and reacquisition training both indicated that the ETOH cue can be retained by both age groups for at least 60 days without discrimination training or food deprivation. The present study suggests that the ethanol discriminative cue in mid-aged mice does not differ from that in young adult mice in potency but is more long lasting, the latter perhaps being related to their reduced ethanol consumption. Of significance from a therapeutic perspective, is that the ETOH cue remained discriminable for 2 months in both age groups (i.e., approximately 1/12 of their total life span).

Drug discrimination using a Pavlovian conditional discrimination paradigm in pigeons

Three pigeons were studied using a discriminated autoshaping procedure in which the presence or absence of methadone served as a conditional stimulus signalling which of two key light CSs would be followed by grain access. Drug sessions alternated randomly with no-drug sessions. Methadone (2.0 mg/kg) was administered prior to drug sessions in which a black vertical line on a white background served as CS+ and a diffuse white keylight served as CS- (reversed for bird 681). Saline or no injection was administered prior to no-drug sessions and the CS+/CS- contingencies were reversed. Discriminated performances emerged in which over 80% of the responding occurred to the appropriate stimulus. Stimulus control by methadone was assessed by presenting a range of methadone doses during 10-trial extinction sessions. A graded dose-effect curve was produced with low doses of methadone controlling saline-appropriate responding and higher doses controlling drug-appropriate responding. A range of doses of morphine, cocaine, and pentobarbital were also tested. Morphine produced methadone-appropriate responding while cocaine and pentobarbital did not.

Drug discrimination is a continuous rather than a quantal process following training on a VI-TO schedule of reinforcement

Debate continues as to whether drug discrimination in animals is an inherently quantal or continuous process. This issue is important in determining the appropriate interpretation of results from drug discrimination studies designed to assess the nature of drug-induced interoceptive cues. The quantal approach holds that subjects perceive a drug cue in an all-or-none manner, while the continuous view proposes that when appropriate training and testing procedures are used, subjects can discriminate along a continuum of interoceptive cues. Data consistent with the quantal view have consistently been generated by animals trained to respond on schedules of reinforcement having an FR component. Since quantal responding is a characteristic of these schedules, results from drug discrimination studies using training schedules with FR components are of little value in empirically determining whether drug discrimination reflects a quantal or continuous process. Use of variable schedules of reinforcement might be more appropriate because the pattern of responding generated does not preclude results consistent with either of the competing views. Data from the following studies that trained subjects using VI schedules with a concurrent TO for incorrect lever responding were analyzed: Barrett et al. (1982): L-5-hydroxytryptophan versus saline; Smith (1990): diazepam versus pentylenetetrazol; Barrett et al. (1992): amphetamine versus haloperidol; Barrett and Steranka (1983): amphetamine versus haloperidol. In every case, when experimental conditions produced a group mean intermediate to that for the training drugs, the distribution of scores for individual animals was normally rather than bimodally distributed.

Historical factors in the development of ETOH-conditioned place preference

Three groups of male Sprague-Dawley rats were conditioned with ethanol (ETOH) and water in a Conditioned Place Preference task. To assess the contribution of prior drug/behavioral history on the relative hedonic valence of ETOH, the three groups differed in the task demands and degree of prior ETOH exposure. One group was trained to self-administer 20% w/v ETOH in a home-cage self-administration task using a "Samson sucrose-fading" procedure prior to place conditioning trials (Group SA/CPP). A second group was initially trained in a two-choice ETOH-Saline drug discrimination (DD) task and subsequently conditioned in the place preference paradigm (Group DD/CPP). The third group of rats had no history of drug exposure and were experimentally naive prior to the place learning task (Group NH/CPP). Group SA/CPP developed a significant conditioned place preference; Group DD/CPP failed to demonstrate either a preference or aversion, whereas the Group NH/CPP demonstrated a significant place aversion. Data suggest that both present and past contingencies contribute to the algebraic summation of ETOH's hedonic valences.

Effects of schedule of reinforcement on a pentobarbital discrimination in rats

The purpose of this study was to determine the effects of the schedule of reinforcement on a pentobarbital discrimination in rats. Five rats were trained to discriminate 10 mg/kg pentobarbital from saline under a multiple fixed-interval 180-s fixed-ratio 20 schedule of reinforcement. During both saline and pentobarbital training sessions, subjects emitted a higher percentage of correct responses under the fixed-ratio component as compared to the fixed-interval component of the multiple schedule. Determination of the pentobarbital dose-response curve under the fixed-ratio component resulted in a steep curve characterized by responding on the saline lever at low doses and on the drug lever at higher doses. Under the fixed-interval component, a graded dose-effect curve was produced, with considerable responding on both levers after intermediate doses of pentobarbital. The administration of phencyclidine and MK-801 resulted in an intermediate level of drug-lever responding for some subjects. Administration of d-amphetamine resulted in saline (nondrug) appropriate responding. The results of this study demonstrate that the schedule of reinforcement is a determinant of drug stimulus control, just as it is a determinant of other drug effects.

Quantal vs. graded generalization in drug discrimination: measuring a graded response

In drug discrimination research, detection of drug stimuli by animals is used for investigating various properties of psychoactive drugs. The major issue addressed by this paper is whether detection of drug stimuli by animals is a quantal or graded event. Some data suggest that detection of a drug stimulus by animals is quantal in nature. Thus, variations in drug stimulus substitution may only reflect variations in threshold for detecting the training stimulus rather than the current concept of these data reflecting graded responding to stimulus intensity. Therefore, drug discrimination procedures may have limited utility for detecting quantitative differences in the subjective effects of varying drug doses. In order to examine this problem, a method for measuring continuous response gradients in individual animals is needed. Tests for quantal responding generally use the distribution of responses on two manipulanda as the dependent measure. However, this variable may be inadequate for detecting a graded response, and may actually reflect loss of stimulus control or a deterioration in performance, rather than changes in response magnitude. Most alternative measures utilize response rate. Unfortunately, these measures are influenced by the direct rate-altering properties of some drugs. One possible alternative method is conditioned taste aversion as the discriminative task. This paradigm provides a means for not only ascertaining if graded discriminative responses occur in individual animals, but also more rapidly training a drug discrimination. Thus, using conditioned taste aversion techniques for measuring a drug discrimination may provide better indices for detecting response gradations.

Effects of ethanol on mature offspring of mice given ethanol during pregnancy

Male offspring of mice maintained on isocaloric liquid diets containing either 20% ethanol or sucrose derived calories during pregnancy were tested on their ability to discriminate different ethanol doses as adults. They were trained to lever-press for a food reward on each of two levers in an operant chamber, and were then maintained on an FR20 reinforcement schedule. After response rates stabilized, ethanol discrimination training was initiated by reinforcing only responses made on the drug appropriate-lever after i.p. injections of ethanol (1.0 g/kg) or water. After learning to discriminate the 1.0 g dose, the animals' ability to discriminate doses of 0.25, 0.50, 0.75, 1.0, and 1.5 g/kg was assessed by determining the percentage of responses made on the drug lever during a 2-min test period. Compared to sucrose controls, mice exposed to alcohol in utero learned the lever response more slowly and were less responsive to injected ethanol as evidenced by a reduced effect of the drug on response rates and by a reduction in their ability to discriminate the presence of injected ethanol. The results indicate that prenatal ethanol exposure can have long term consequences which reduce the effects of ethanol in fully mature animals.

Effects of variations of reinforcement magnitude on alcohol discrimination using intracranial stimulation as the reinforcer

Male Wistar rats were trained to discriminate intraperitoneal injections of 2 g/kg of alcohol from 0.5 g/kg alcohol injections. Intracranial stimulation in the posterolateral area of the hypothalamus served as the reinforcer according to a FR 10 schedule of bar pressing. After discrimination between the two alcohol doses had been established, the response-reinforcement contingency was modified by progressively increasing the intensity of the intracranial stimulation. The results indicate that the generalization gradient varies with the number of reinforcements. The more the lever associated with a training dose was reinforced, the more the ED50 had the tendency to approach this dose. Our paradigm, by using electrical brain reinforcer, allowed to deliver in an equal manner the reinforcement schedule in order to obtain equal performances and equal number of reward received.

Behavioral analogues of anxiety. Animal models

In the absence of fully characterized biological indexes, anxiety is at present measured as unpleasant effects reported verbally by patients. Because of the subjective nature of the syndrome, animal analogues have been difficult to design, but quests for new anxiolytics and a deeper understanding of the neurobiology of anxiety have fostered the development of several animal models. Usually, animals are exposed to exteroceptive or interoceptive stimuli which can be interpreted as capable of causing anxiety in humans. Then, the animals are observed for responses or behavioral deficits resulting from those stimuli in order to provide an index of anxiety. Behavioral responses that are reliably produced by those stimuli and that are also antagonized by anxiolytic drugs are accepted as analogues of anxiety. Exteroceptive stimuli, useful in this respect, consist of a variety of noxious treatments such as exposure to conflict-situations or unavoidable electric shock, whereas interoceptive stimuli consist of treatment with anxiogenic drugs or electrical stimulation of selected brain areas. Elicitation of unconditioned behavior or changes in the rate of conditioned (learned) responding have been employed as measures of anxiety responses following application of either exteroceptive or interoceptive stimuli. These measures, although useful in detecting anxiolytic drugs, possess several weaknesses. They suffer from difficulties in obtaining quantitative and objective data, they do not differentiate between anxiety and stress or fear, they are unable to measure further deterioration of behavior expected to occur when more potent anxiogenic stimuli are tested and they often present difficulty in differentiating direct motor effects of a number of stimuli are not related to anxiety. More recently, interest in the development of other analogues of anxiety has led to the use of drug-discrimination paradigms. In this approach, interoceptive discriminative stimuli, produced by anxiogenic drugs, are used as analogues of anxiety in animals. As an example of this approach, data are reviewed showing that pentylenetetrazol, an anxiogenic drug in humans, produces interoceptive stimuli which can be readily discriminated by rats. Further, these stimuli can be easily quantified through dose-response analysis. All known anxiogenic drugs generalize to pentylenetetrazol-induced interoceptive discriminative stimuli. Similarly, other anxiety-provoking situations in humans, such as withdrawal from dependence on benzodiazepines, also generalize to the pentylenetetrazol-induced stimuli. Alternatively, all known anxiolytic drugs antagonize these stimuli with a relative potency similar to

Extended schedule transfer of ethanol discrimination

Stimulus control was established in rats with ethanol (600 mg/kg) and saline by employing a two-lever response choice task and an FR10 schedule of food reinforcement. Subjects were then tested with an extended schedule procedure in which lever selection and its perseverance were measured under the training conditions and after the administration of pentobarbital at doses of 2 to 12 mg/kg. With decreasing doses of pentobarbital, drug-lever selection was observed to decline. The dose at which initial lever selection was evenly distributed between the two levers (ED50) was determined to be 4 mg/kg. However, at this dose the perseverance on the ethanol-appropriate lever was not significantly different than that observed after the training dose of ethanol. In addition, the perseverance of saline-lever selection produced by saline was observed to be greater than that produced by the training dose of ethanol on the ethanol-lever. The advantages inherent in employing the extended schedule performance procedure in transfer experiments are discussed.

Drug discrimination training with progressively lowered doses

Rats were trained to discriminate drug from no-drug conditions in a two-lever operant task. Moderately high dosages were used initially. Whenever the discrimination was learned, training was continued with progressively reduced dosages. Eventually the rats discriminated extremely low doses of phenobarbital, chlordiazepoxide, cyclazocine, and fentanyl.

Drug discrimination and generalization in pigeons

In a three-key operant conditioning situation six pigeons were trained to select the response key which was associated with each of three drug treatment conditions: d-amphetamine (2 mg/kg), pentobarbital (5 mg/kg), and saline. Thus, the drug state served as a discriminative stimulus for food reinforcement. After 20 sessions of discrimination training in each of the three conditions, more than 90% of the responses were correctly emitted in the presence of the appropriate drug or saline stimulus. Acquisition of the discrimination progressed at approximately equal rates for the three treatments. Subsequent to discrimination training, generalization gradients were obtained for several doses of the training drugs and for dose ranges of cocaine, morphine, and methocarbamol. The pigeons responded to morphine by choosing the key paired with pentobarbital during training; further, cocaine administration resulted in choice of the amphetamine key. However, metocarbamol, over the doses used, produced responding more characteristic of saline than of the other training drugs. The data suggest that a three-key operant discrimination procedure using pigeons provides a sensitive method for investigating the stimulus properties of relatively low doses of behaviorally relevant drugs.

Temporal parameters of d-amphetamine as a discriminative stimulus in the rat

Three groups of rats were trained on a two-lever operant discrimination using d-amphetamine (0.8, 1.6 or 2.4 mg/kg) and saline as cues. Reinforcement of responding on one lever was associated with the drug and reinforcement on the other lever was associated with saline. Following acquisition, behavioral control was evaluated during 10-min extinction tests at 0, 15, 30, 60, 90, 120, 180 or 240 min after injection of drug or saline. Onset of the stimulus properties of d-amphetamine occurred within 10 min and maximal effects were attained by 15-30 min post-injection. Drug effects were minimal or absent 2-4 h after administration. The time course for all doses of d-amphetamine was identical when each group was tested iwth its respective training dose. When subjects were tested with doses other than their training dose, the time course of the drug effect varied as a function of training dose, test dose and time after injection.

The stimulus properties of morphine and ethanol

The present investigation sought (a) to establish the efficacy of morphine and ethanol as discriminative stimuli when each is paired with the administration of saline and (b) to compare, in a qualitative sense, the stimulus properties of the two drugs. Additional experiments examined the effects of treatment with naloxone or l-propranolol upon morphine and ethanol-mediated discriminated responding. Finally, the stereospecificity of the stimuli produced by morphine was determined by a comparison, in morphine-trained rats, of levorphanol and dextrorphan. Discriminated responding developed rapidly in both the morphine and ethanol groups. In tests in which ethanol was administered to morphine-trained animals and vice versa, no similarity to stimulus properties was apparent. Antagonism of discriminated responding induced by morphine and ethanol was attempted using naloxone and l-propranolol. Naloxone blocked the actions of morphine but was without effect upon ethanol. No evidence of antagonism of either drug by propranolol was found. When a range of doses of levorphanol (0.1-3 mg/kg) and dextrorphan (3-100 mg/kg) was tested in morphine trained animals, only levorphanol was able to substitute for morphine. The present results suggest that the stimulus properties of morphine represent typical opiate effects.

Role of dopamine in d-amphetamine-induced discriminative responding

In two-lever operant chambers, rats were trained in a food-reinforced discrimination task. Reward was contingent upon correct lever choices to the induced differential cue conditions of d-amphetamine (0.8 mg/kg) or saline throughout training on a differential reinforcement of low response rate (DRL-15 sec) schedule. Upon acquisition of discriminative response control the animals were pretreated with various neurochemical agents. Pretreatment with atropine, phentolamine, propranolol, methysergide or cinanserin did not block production of d-amphetamine lever responding. Nicotine and oxotremorine did not produce the amphetamine-like cueing effect. However, pimozide blocked the ability of animals to discriminate d-amphetamine and L-DOPA, in combination with Ro 4-4602 with or without amantadine, generated amphetamine-like responses. These results indicate a role of dopamine in the production of the amphetamine state and from the failure of apomorphine to exhibit the stimulus property of d-amphetamine and the antagonism by alpha-methyltyrosine of the d-amphetamine responding it is further suggested that this amphetamine state is produced via the releasing of newly synthesized dopamine.