Functional dissociations between the medial septal area (MSA) and the nucleus basalis magnocellularis (NBM) were examined using the concepts and experimental procedures developed by scalar timing theory. Rats were tested in variations of a signalled discrete-trial peak-interval schedule of reinforcement in which the response rate functions identified the time when the rats expected reinforcement. The variations assessed aspects of both reference and working memory for information obtained from prior trials and from the current trial. A double dissociation was found in reference memory. Rats with NBM lesions, like those with frontal cortex (FC) lesions, remembered the time of reinforcement as having occurred later than it actually did; rats with MSA lesions, like those with fimbria-fornix (FF) lesions, remembered the time of reinforcement as having occurred earlier than it did. A single dissociation was found in working memory. MSA lesions and FF lesions impaired working memory, while NBM and FC lesions had no effect on it. These data begin to identify the brain mechanisms underlying temporal memory; they indicate that the frontal and hippocampal systems are both involved, but in complementary ways; and they provide information that helps specify more clearly the functions of the frontal and hippocampal systems.

The functional contribution of nucleus basalis magnocellularis (NBM) and the medial septal area (MSA) to memory was evaluated in two different spatial discriminations. Preoperatively, rats were trained to a criterion level of performance in a simultaneous left/right discrimination on the stem of a T-maze (a trial-independent memory) and a discretetrial, rewarded alternation discrimination on the arms of the T-maze (a trial-dependent memory). Bilateral lesions were made by injecting ibotenic acid (IBO) into the NBM, MSA, both NBM and MSA, or dorsal globus pallidus (DGP), and by radiofrequency current (RF) in the NBM and MSA. Control rats received operations in which either no current was passed or no neurotoxin was injected. Lesions in the NBM, MSA, or both the NBM and MSA produced a similar pattern of behavioral changes relative to the performance of controls; postoperative reacquisition of the arm discrimination was initially impaired but showed recovery to normal levels, whereas postoperative reacquisition and reversal of the stem discrimination was not impaired (except following the combined NBM and MSA lesion). Lesions of the DGP had no effect on choice accuracy in any discrimination.When the discrimination on the arms was made more difficult by increasing the delay interval during which the information had to be remembered, rats with combined NBM and MSA lesions were again impaired relative to controls and showed no signs of recovery of function. These results provide information about the behavioral functions of the basal forebrain cholinergic system and suggest that pathological changes in certain components of this system can cause disorders of memoryWithin the basal forebrain of the rat is a core of acetylcholinergic (ACh) neurons (Sofroniew et al., 1982) (NBM), diagonal band of Broca, and the medial septal area (MSA) Lehmann et al., 1980;Johnston et al., 1981;Bigl et al., 1982;Hartgraves et al., 1982; Pearson et al., 1983; Saper, 1984). The NBM projects ACh afferents primarily to the neocortex, and the MSA and vertical limb of the diagonal band project ACh afferents primarily to the hippocampus. The frontal neocortex and hippocampus, and the cholinergic systems afferent to these areas, have an important role in memory (Bartus, 1980; Bartus et al., 1982; Mishkin, 1982). Recently, the behavioral effects of lesions produced in either the NBM (LoConte et al., 1982; Flicker et al., 1983) or MSA (Mitchell et al., 1982) have been evaluated separately in individual tasks, but the behavioral effects of these two lesions have not been compared on the same tasks.In the present study, both ibotenic acid (IBO) and radiofrequency (RF) current were used to produce lesions. IBO causes degeneration of neuronal perikarya at the site of injection, without damaging axons of passage or neurons distant from the injection site (Coyle and Schwartz, 1983). RF lesions destroy both neuronal perikarya and fibers of passage. The comparison of RF and IBO lesions allows for the assessment of the relative importance of d...

No abstract