Opposing effects of vasopressin on matching versus non-matching to position: further evidence for response, not memory, modulation.

Rats were trained on either of two related variants of an operant memory task. In the matching to position (MTP) task, one of two retractable response levers appeared, at random, as the sample. A response caused the lever to retract and this was followed by a delay (0-32 s) interval, during which the subjects had to approach and respond at the magazine tray. Both levers were then presented and the rat had to respond, for food reward, to the one which had appeared as the sample. A second group of rats learned non-matching to position (NMTP). This task was very similar to MTP, with one crucial difference: here, the subject had to respond to the lever which had not appeared as the sample. Both groups of rats learned their respective tasks rapidly, performance depending on the delay interval as expected. They were then injected, peripherally, with different doses of arginine-vasopressin (AVP: 0-25 micrograms/kg), a peptide which others have argued improves mnemonic performance. There was evidence to suggest that MTP performance was improved by AVP; on the other hand, NMTP performance appeared to be disrupted. It is suggested that AVP induces a bias towards responding on one side of the two lever test chamber. In other words, it affects motor or motivational, not mnemonic mechanisms.