Drug mixtures and ethanol as compound internal stimuli.

Drug discrimination methods that entail training with mixtures of drugs may shed light on polydrug abuse and on the actions of single drugs that interact with more than one receptor. In AND-discrimination procedures (drug A + drug B vs. vehicle), mixtures are discriminated primarily on the basis of their component drugs: these discriminations may be useful for testing interactions between component drugs in mixtures. The role of training dose, overshadowing and associative blocking in AND-discriminations have been investigated. For example, after prior training with midazolam, it was possible to demonstrate associative blocking of the nicotine element of the mixture stimulus, and vice versa. Using the AND-OR discriminations (drug A + drug B vs. drug A or drug B) increased pharmacological specificity considerably, and these procedures may be valuable for determining whether the effects of a novel mixture are similar to the combined effects of the training drugs. Ethanol is an example of a single drug that may produce a compound cue; rats trained to discriminate ethanol from water generalize (asymmetrically) to GABA(A) enhancers such as chlordiazepoxide (CDP) or pentobarbitone, to NMDA antagonists such as dizocilpine (MK-801), and to some serotonin agonists, such as trifluoromethylphenylpiperazine (5-HT(1B/2C)). In addition, rats trained to discriminate mixtures of either CDP or pentobarbitone plus MK-801 generalize to ethanol. A previous history of training with MK-801 or CDP (prior to ethanol discrimination training) enhanced the MK-801-like and CDP-like effects of ethanol respectively, but associative blocking of proposed elements in the ethanol stimulus was not seen. These studies provide some support for the multielement concept of ethanol discrimination but also suggest that rules governing three-component stimuli (such as those putatively produced by ethanol) may differ from those for the two-component mixtures of drugs studied previously.