A Go/No-go delayed nonmatching-to-sample procedure to measure object-recognition memory in rats.

The novel-object preference (NOP) test is widely used to assess object-recognition memory in rats. When interpreting behaviour on the test, a common assumption is that the magnitude of a rat's novel-object preference reflects the persistence or accuracy of its memory for the previously encountered object. However, some concerns have been raised regarding the latter interpretation, and hence, the internal validity of the NOP test as a gauge of object-recognition abilities. Given the concerns, we developed a new Go/No-go delayed nonmatching-to-sample (DNMS) procedure to measure object-recognition memory, which circumvents the interpretational problems associated with the NOP test. Rats were trained to displace an unfamiliar object (sample) from over a food well to obtain a food reward. Then on a choice phase, rats were presented with novel objects ("Go" trial) or copies of the sample object ("No-go" trial). On Go trials rats received a reward for displacing the novel object and on No-go trials no reward was provided for displacing the sample object. Rats required an average 54 sessions to reach a performance criterion of at least 80 % correct choices on five consecutive sessions (16 correct choices out of 20). Afterwards, rats were tested on the NOP test, and we found that scores on both tasks were not significantly correlated, indicating performance on the Go/No-go DNMS task did not predict novelty preference scores. The findings from this experiment reveal the benefits of an alternative approach to assess object-recognition memory in rats.