Influence of training history on ethanol discrimination in rats

The compound stimulus hypothesis of ethanol discrimination predicts that a history of training to discriminate drugs that mimic individual elements of the ethanol stimulus should attenuate stimulus control by other stimulus elements (associative blocking). Rats were trained initially to discriminate either chlordiazepoxide (5 mg / kg s.c., n = 10) or dizocilpine (0.08 mg / kg i.p., n = 10) from vehicle in two-lever procedures with food reinforcers presented on a tandem variable-interval fixed ratio schedule. Control rats received 'sham training' (vehicle injections only, n = 9). All subjects were then trained to discriminate ethanol (1.5 g / kg intragastrically (i.g.)) until discrimination accuracy reached 95%. Chlordiazepoxide (1.25-10.0 mg / kg s.c.) produced more drug-appropriate responding in rats with a previous history of training to discriminate chlordiazepoxide than in either of the other two groups, but stimulus control by dizocilpine was not attenuated. Equivalent results were obtained in rats with a previous history of training to discriminate dizocilpine. Ethanol (0.375-3.0 g / kg i.g.) produced similar dose-related increases in drug-appropriate responding in all three groups. Thus, previous discrimination training modified the characteristics of ethanol discrimination in a way that may be explained by persistence of the original discriminations. The lack of evidence for associative blocking contrasts with results of previous experiments on the discrimination of compound stimuli produced by administering drug mixtures. The findings provide limited support for the hypothesis that ethanol produces a compound stimulus that includes elements of positive modulation of gamma-aminobutyric acid (GABA(A)) receptors and of N-methyl-D-aspartate (NMDA) antagonism.