produced by drugs and concluded that a simple process of conditioned aversion could not account for all of the findings.
A variation of a one-taste model where preferences do emerge involves ► drug taste preference conditioning (see ► Alcohol Preference Tests). In the case of morphine taste preference conditioning, the development of a preference for the solution containing morphine can be made apparent against a baseline comprising a normal rejection of the aversive solution of morphine in subjects that do not have the pairing. However, the gradual consumption of a solution that is unpalatable does not necessarily evidence the development of a preference for several reasons. For example, repeated exposure to tastes lead to habituation, sensitization, contrast effects, changes in food intake, and so on. Thirst develops under conditions of forced drinking of an aversive solution and this would counteract the aversion. These processes and the drinking will also interact with the acute drug effects. The actual treatment drug may produce, for example, general malaise and hyperactivity. Drug effects would also affect taste sensitivity, alter general motivational states, and produce sensory impairment, overshadowing stimuli, etc. In such a one-taste paradigm like these, confounds can only be dealt with by applying additional groups and control conditions (Mucha 1992; Stolerman and D'Mello 1981).
An efficient solution to some of these problems is to offer the same test subject an alternate taste in a choice test. Much work on both conditioned aversion and conditioned preferences involve simply offering the test subjects a choice with water. For example, one can conclude that a preference is present when taste that was previously strongly avoided is now preferred relative to the water (see ► alcohol preference tests). The overall levels of fluid consumption would control for nonspecific effects of the training drug on fluid regulation. Care is nevertheless needed to avoid biases due to the left or right position of the experimental test on choice tests (Stolerman and D'Mello 1981). The use of choice tests with a water choice is still not fully adequate. In the case of drug taste preference conditioning using morphine-containing fluids, the preference produced by morphine could only be interpreted when the morphine solution was substituted for a taste similar to the bitter taste of morphine (Stolerman and D'Mello 1981).
The Two-Taste Test Paradigm and the Study of Experimentally Naive Rats
The two-taste method is a differential two-stimulus classical conditioning procedure. In this procedure, one taste is paired with the drug manipulation and a second taste is paired with a control manipulation. It provides robust within-subject control conditions and can be used to show subtle conditioned changes after only a few drug exposures (Mucha 1992). This method was first used to show a preference for a taste paired with a drug-related effect by Green and Garcia (1971). They paired a taste with the offset of the effect of a high dose of the ► dopa-mine agonist, ► apomorphine. The direct effect of the dosing induced illness: the same rats showed a taste aversion to a different taste presented just before the same injection. The conditioned preference was referred to as a "medicinal" effect of the offset of the emetic. Such a medicinal effect was also noted when the effects of nicotine infusions were blocked with a nicotine antagonist.
The use of two taste cues for discriminated conditioning procedures was largely developed over the course of work on the aversive effects of psychoactive substances (e.g., ► morphine, ► psychostimulants, and opioid agonists) during neuropharmacological studies at the University of Birmingham (UK) by David Booth, Charles Pilcher, Ian Stolerman, and others. However, the appetitive responses of tastes paired acutely with drugs only became apparent later when the procedure was applied to study low doses of sufentanil and fentanyl, two highly potent opioids (reviewed by Stolerman 1985).
In Fig. 1, there are taste preferences seen with adult rats in a two-taste conditioned preference procedure. There is a biphasic effect of dose with a preference effect seen at lower doses and aversion at higher doses. Thus, after a phase of adaptation to a water-deprivation schedule, rats were allowed to consume a target taste. This taste was paired shortly after using a subcutaneous injection of one of a range of doses of a test drug. The drug comprised one of three opioid analgesics (see Fig. 1). On another occasion, the same rats also received a second taste paired with a control manipulation (saline injection). The experimental and control pairings were repeated each day over 3 days. On a subsequent day each rat was then tested by offering the two tastes simultaneously for 24 h. Morphine shows the biphasic dose effect only with extended pairings (Fig. 1, open symbols).
Like conditioned place preferences (see ► conditioned place preference and aversion), taste preferences index simple approach or appetitive responses produced by a previously neutral stimulus paired with a drug reinforcer. They extend to gustatory and flavor stimuli experimental work based on environmental or place cues. However, until the 1980s, considerable attention was given to the idea that the initial application of self-administered
Conditioned Taste Preferences. Fig. 1. Taste preference scores of rats trained with different doses of three mu-opioid receptor agonists; the data for each point were collected using independent groups of subjects. Abscissa: the doses of the pairing drug. Ordinate: mean of the difference between the volume of consumption of the drug- and the vehicle-paired tastes expressed as a percent of total fluid consumed. The open circles indicate data from rats trained with morphine using six pairings instead of three pairings. The asterisks indicate significant effects. (redrawn from Mucha and Herz (1985) Psychopharmacology 86:274-280.)
substances only produced taste aversions and that taste preferences were not produced in these situations. The fact that low doses of opioids in drug-naive animals also show conditioned taste preferences refute this notion. Some of the paradoxical effects of self-administered drugs are indeed due to the toxicity of high dosing typically used in the study of conditioned taste aversions. However, there are important differences between classes of addictive drugs, such as between opioids and psychostimulants. Indirect dopamine agonists such as ► amphetamine, for example, do not appear to evoke conditioned taste preferences, despite the fact that they act in the same general area of the brain as the opioids (Stolerman 1985). Therefore, conclusions about the opioids in the drug-naive animal probably do not generalize to drugs from other pharmacological classes.
The pattern of conditioned taste preferences in Fig. 1 likely reflects the activity of these training drugs on the mu-opioid receptors, as ► kappa-opioid agonists produced only aversions. The peak increases in intake of the drug-paired taste in the figure also correspond to the known relative activities of the test substances on the mu-opioid receptor. The conditioned taste preference also only occurs with ► stereoisomers active on the mu receptor, suggesting that, indeed, opioid receptors are involved. However, it is still not known whether opioid antagonists block the formation of these preferences (Stolerman 1985). Opioid antagonists used as training drugs also produce conditioned taste aversion in morphine-naive rats and this response is potentiated in animals treated chronically with opioids, (Mucha 1992). The taste aversions produced by opioid antagonists in drug-naive rodents are also limited to the stereoisomer active on the opioid receptors. Therefore, taste preference and taste aversions conditioning appear to be modulated by increases and decreases in activity on the opioid receptors, respectively (Stolerman 1985).
The conditioned taste preferences have not been studied using local brain injections, which would be necessary to show that they occur in the reinforcement system in the brain. The pattern of effects seen for the opioids in Fig. 1 parallel conditioned place preferences produced by these substances, which are known to be centrally mediated. Consistent with this, the potent opioid analgesics, sufen-tanil and ► fentanyl, which have the strongest effects in Fig. 1, have physicochemical properties allowing them to penetrate into the brain much more readily than morphine. However, there are a number of open questions, particularly regarding other substances active on the reinforcement system that fail to produce conditioned taste preferences under acute treatment conditions (e.g., amphetamine).
An important issue is that opioids give rise to physical dependence and their removal results in ► withdrawal syndromes. These states are highly aversive and can be alleviated by the reapplication of opioids (Mucha 1992). This was a major explanation of the motivation for the oral intake of drugs in the early studies of oral self-administration of drug. Clear evidence for the hypothesis is actually rare partly because incentive effects of a training drug confound tests for conditioned taste preferences produced by withdrawal. The effects in Fig. 1 are also seen at doses that are too low and infrequent to produce dependence. There are data from one-taste conditioning work and on morphine taste preference conditioning where animals were given supplementary injections of morphine independent of the taste preference conditioning. There was a suggestion for a facilitation of the preference conditioning in injected animals; however, the conclusions were still not fully in line with a withdrawal alleviation hypothesis. It was possible that the injections sensitized the incentive effects of the morphine or produced a habituation of the initial aversive effects (Gaiardi et al. 1991; Stolerman and D'Mello 1981). However, this
Was this article helpful?