Stewart AL, Roland JJ, Servatius RJ, Pang K (2011) Role of the medial septal-diagonal band nucleus in working memory: Effects of cholinergic or GABAergic lesions on memory demand and interference. Neuroscience 2011 Abstracts 513.09. Society for Neuroscience, Washington, DC.
Summary: The medial septum and diagonal band of Broca (MSDB), a major source of afferents to the hippocampal formation, is critical for learning and memory. The primary cells comprising the septohippocampal systems are cholinergic, GABAergic and glutamatergic. Selective damage of cholinergic MSDB neurons results in mild to no impairment of spatial working memory tasks, suggesting that non-cholinergic MSDB projections are important in learning and memory. Recently, we demonstrated that GABAergic MSDB lesions impair a delayed match to position task (DNMTP) with errors suggesting enhanced proactive interference. The current study assesses the effect of manipulating the intertrial interval (ITI) and retention interval (RI) on DNMTP performance in normal rats and those with cholinergic or GABAergic MSDB damage. In addition, activation of MSDB neurons on the last day of training will be assessed. Male Sprague Dawley rats receive sham, 192-IgG saporin (192-Sap) or GAT1-saporin (GAT1-Sap) administration into the MSDB before training on a DNMTP task using a T-maze. On the sample phase of each trial, rats are forced to one arm for reinforcement. Following an RI, a choice phase allows the rats to choose from both arms. Rats are reinforced for choosing the arm not entered during the sample phase. Following the choice phase, an ITI occurs before the sample phase of the next trial. Rats are trained one session per day, 12 trials per session, and 10 sessions. In a 2 x 2 experimental design, each rat is trained on either a 0 or 60s RI and a 0 or 60s ITI. Conditions with a long RI (60 s) are designed to tax working memory, whereas conditions with a similar RI and ITI are designed to increase interference. Immunocytochemistry for c-Fos is used to assess activation of cholinergic or GABAergic MSDB neurons following the last training session. As expected, our preliminary results show that sham rats performed better on 0 s than 60 s RI (0 s = 78% correct vs 60 s = 57%, both ITI’s pooled). Analysis of the 0 s RI demonstrates that performance in conditions with different RI and ITI was better than when RI and ITI were similar (0 s RI/60 s ITI = 74% vs 0 s RI/0 s ITI = 67%). Further analysis of the 60 s RI was difficult due to the near chance performance. Rats treated with either intraseptal 192-Sap or GAT1-Sap were impaired on the 0 s RI/60s ITI condition (Sham: 85%; 192-Sap: 65%; GAT1-Sap: 72%). However, only the 192-Sap rats were impaired in the 0 s RI/0 s ITI condition (Sham: 71%; 192-Sap: 62%; GAT1-Sap: 69%). Anatomical studies are currently underway. The results of this study will further elucidate the role of MSDB neurons in two aspects of working memory: memory demand and interference.