Evidence from our studies on the effects of systemic injections of pharmacological agents in behaving monkeys has led to the proposal that the formation of stimulus memories depends on interaction between the cholinergic and glutamatergic systems. More specifically, our evidence suggests that the critical event for storage of the trace or representation of a stimulus is the potentiation exerted by activation of the cholinergic muscarinic receptor on activity mediated by the glutamatergic NMDA receptor. To test this hypothesis, we have been examining the effects of infusing pharmacological agents directly into the perirhinal cortex, which is known from lesion studies to be the most critical area in the temporal lobe for stimulus recognition. Our results thus far have shown that, like systemic injections of a cholinergic muscarinic receptor antagonist (scopolamine), microinjecting this drug into perirhinal cortex impairs recognition memory. By contrast, recognition memory is unaffected by either systemic or perirhinal injections of dopaminergic receptor antagonists (e.g. haloperidol). We have also demonstrated that, like systemic injections of an NMDA receptor antagonist (MK-801), perirhinal infusions of such an antagonist (D-AP5) impairs recognition memory. Again by contrast, recognition memory was unaffected by perirhinal injections of a kainate/AMPA receptor antagonist (CNQX). These results provide preliminary support not only for the hypothesis that stimulus memory depends on the interaction between muscarinic and NMDA receptor activation, but also for the notion that such interaction occurs within the neurons of the perirhinal cortex. The cholinergic/glutamatergic hypothesis of cognitive memory formation has been challenged on the basis of the finding that pretraining rats in the Morris water maze in one environment eliminates the impairing effects of muscarinic and NMDA receptor antagonists on learning the maze in another environment (for review, see Cain, Neuroscience and Biobehavioral Reviews 22: 181-193, 1998). In many of these studies, pretraining is used to familiarize animals with the procedural aspects of the task, and so it is commonly assumed that subsequent training in a novel environment should therefore be particularly sensitive to cognitive spatial processing. However, our recent behavioral findings in rats suggest that the pretraining promotes subsequent use in the novel environment of stimulus-response habits rather than cognitive memory. Therefore, corticostriatal habit circuits relying on intact nigrostriatal dopamine function may contribute to water maze performance following pretraining. Lesion and drug experiments will be conducted to test this hypothesis. Earlier findings in monkeys suggested that systemic injection of haloperidol, but not of scopolamine, retards the learning of a set of concurrent visual discriminations in which the stimulus pairs within the set are each presented just once every 24 hours. In a new study, using a version of this task in which the stimulus pairs of the set are each repeated a few times within each session, systemic injections of both drugs was found to retard learning. If confirmed, the differential results on the two versions of the task would support the notion that discrimination learning with pair-repetition just once every 24 hours can be mediated only by a dopaminergic-dependent corticostriatal habit system (and, hence, is susceptible to disruption only by haloperidol), whereas learning with pair-repetition within a session is mediated by both the latter system and a cholinergic-dependent cortico-limbic memory system (and, consequently, is susceptible to disruption by both pharmacological agents). The circuitry underlying the formation of stimulus memories is thought to involve a series of projections from the high-order sensory processing areas through structures in the medial temporal lobe, from there to the anterior group of thalamic nuclei and the magnocellular division of the medial dorsal nucleus (MDmc), and then to the ventral prefrontal and cingulate cortices. The parallel circuit underlying habit formation is thought to involve a series of projections from the neocortex through the basal ganglia, from there to thalamic nuclei within the ventral and intralaminar groups, and then to the premotor and supplementary motor areas. However, in the course of investigating medial thalamic efferents in macaques, we uncovered other thalamo-cortical routes that could contribute to stimulus memory and habit formation. The medial thalamic injection sites for anterograde tracers covered the midline nuclei, as well as MDmc, medial portions of the magnocellular ventral anterior nucleus (VAmc) and the intralaminar paracentral nucleus (Pc). These injections yielded terminal labeling in the outer half of layer I across an extremely large cortical expanse, sparing only the premotor and supplementary motor areas, precentral and postcentral gyri, and primary auditory cortex (the primary visual area in the occipital pole was not examined). In complementary studies, in which retrograde tracers were injected into various cortical areas, we searched for groups of neurons within the above medial thalamic region that were consistently labeled by the different injections and were therefore a potential source of the widespread projection to cortical layer I. Numerous retrogradely labeled neurons were seen in the midline group of thalamic nuclei after prefrontal, cingulate, and rhinal injections, suggesting that this particular thalamo-cortical projection could participate in the acquisition of stimulus memories. In addition, Pc and the medial portion of VAmc contained labeled cells from all the injected fields except rhinal cortex, suggesting that the widespread thalamo-cortical projections from these two nuclei, which belong to the ventral and intralaminar groups, might participate in habit formation.

National Institute of Health (NIH)
National Institute of Mental Health (NIMH)
Intramural Research (Z01)
Project #
Application #
Study Section
National Institute of Mental Health Initial Review Group (NIMH)
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
U.S. National Institute of Mental Health
United States
Zip Code
Turchi, Janita; Saunders, Richard C; Mishkin, Mortimer (2005) Effects of cholinergic deafferentation of the rhinal cortex on visual recognition memory in monkeys. Proc Natl Acad Sci U S A 102:2158-61