Effects of cholinergic manipulation on operant delayed non-matching to position performance in two inbred strains of mice.

With the increasing demand on phenotyping of mouse mutants there is a clear need to develop novel paradigms for testing mice. Mice are able to learn a non-matching to position rule to high accuracy in a variety of maze paradigms, but an operant version of this task is desirable. In the present study, mice of the C57BL/6 and DBA/2 strains were trained and tested on an operant delayed non-matching to position (DNMTP) paradigm. Data were analysed according to the methods of signal detection theory (SDT), which allows conclusions as to whether strain differences in DNMTP performance are more related to changes in accuracy or in motivational factors. Mice can learn to respond on an operant DNMTP paradigm with high accuracy, and accurate performance depends on the duration of the delay-period, i.e. forgetting curves can be generated. Comparison between the two strains of mice revealed that DBA/2 mice learned faster than C57BL/6 mice to associate the lever press with food during initial shaping, but no further strain differences were observed in accurate responding during later stages of the experiment. However, differences in biased responding and, in particular, responsivity were observed between the two strains. Muscarinic blockade with scopolamine (0.1--1.0 mg/kg) failed to affect accuracy in the two strains, but altered responsivity. This task should be of great value for a more in-depth analysis of cognitive function in mutant mice as it allows a better dissociation between mnemonic and non-mnemonic factors. In particular, such paradigm may be of interest for testing conditional mutants, which allow time-sensitive induction or inhibition of gene expression, i.e. where animals can be trained while non-impaired to stable baseline and then tested once the gene is activated or inhibited.