In a series of studies in one of our laboratories, we examined whether an acute disruption of the septo-hippocampal pathway would produce a profile of errors in the DNMTS task that was different from that produced by a chronic disruption of this system. In these studies, rats were injected with chlordiazepoxide (CDP) into the medial septum at different times relative to training or testing. Intraseptal injections of benzodiazepines such as CDP, or GABA-A or GABA-B agonists, decrease high affinity choline transport (HAChT) in the hippocampus, and reduce acetylcho-line turnover and release.34-36
These studies examined the time-dependence, site-specificity, and potential mechanism of action of CDP. Rats injected with CDP into the medial septum, but not the lateral septum, amygdala, or nucleus basalis, exhibited dose-dependent deficits in DNMTS performance. Furthermore, there was a distinct time-dependent profile with deficits observed when CDP was injected immediately, but not 15, 30, or 45 min after the pre-delay session. Rats injected with CDP immediately after training were able to acquire and use information during the post-delay session, but they were unable to utilize information from the pre-delay session to guide performance. They exhibited a significant decrease in CC and an increase in RE, but no alterations in PE. These rats exhibited impaired memory of the arms entered in the pre-delay session but they could acquire and retain new information during the post-delay session. CDP disrupted working memory (i.e., retrograde amnesia) but did not produce an anterograde amnesia or performance deficit. The lack of PE is important since it demonstrates that working memory is intact during the post-delay session in this group.
Deficits were also observed when CDP was injected immediately prior to the post-delay session but there was a qualitatively different pattern of errors, with a selective increase in PE, but not RE. Rats were able to use information acquired prior to the delay but their acquisition of information during the test session was impaired. These data demonstrate that intraseptal CDP impairs working memory, and the degree to which different phases of working memory are disrupted depends upon the time of injection. Post-training injection of CDP impairs the encoding or maintenance of working memory, while pretest infusion impairs the ability to use or encode information to guide performance during the test session. In contrast, rats injected with CDP immediately before the post-delay test session exhibited a significant increase only in PE. An increase in PE suggests an impaired ability to store or maintain current arm choices into working memory.
Was this article helpful?