Recognition memory in rats--III. Neurochemical substrates.

In the first part of three overviews on recognition memory in the rat, we discussed the tasks employed to study recognition memory. In the second part, we discussed the neuroanatomical systems thought to be of importance for the mediation of recognition memory in the rat. In particular, we delineated two parallel-distributed neuronal networks, one that is essential for the processing of non-spatial/item recognition memory processes and incorporates the cortical association areas such as TE1, TE2 and TE3, the rhinal cortices, the mediodorsal thalamic nucleus and prefrontal cortical areas (Network 1), the other comprising of the hippocampus, mamillary bodies, anterior thalamic nuclei and medial prefrontal areas (Network 2), suggested to be pivotal for the processing of spatial recognition memory. The next step will progress to the level of the neurotransmitters thought to be involved. Current data suggest that the majority of drugs have non-specific, i.e. delay-independent effects in tasks measuring recognition memory. This may be due to attentional, motivational or motoric changes. Alternatively, delay-independent effects may result from altered acquisition/encoding rather than from altered retention. Furthermore, the neurotransmitter systems affected by these drugs could be important as modulators rather than as mediators of recognition memory per se. It could, of course, also be the case that systemic treatment induces non-specific effects which overshadow any specific, delay-dependent, effect. This possibility receives support from lesion experiments (for example, of the septohippocampal cholinergic system) or studies employing local intracerebral infusion techniques. However, it is evident that those delay-dependent effects are relatively subtle and more readily seen in delayed response paradigms, which tax spatial recognition memory. One interpretation of these results could be that some neurotransmitter systems are more involved in spatial than in item recognition memory processes. However, performance in delayed response tasks can be aided by mediating strategies. Drugs or lesions can alter those strategies, which could equally explain some of the (delay-dependent) drug effects on delayed responding. Thus, it is evident that neither of the neurotransmitter systems reviewed (glutamate, GABA, acetylcholine, serotonin, dopamine and noradrenaline) can be viewed as being directly and exclusively concerned with storage/retention. Rather, our model of recognition memory suggests that information about previously encountered items is differentially processed by distinct neural networks and is not mediated by a single neurotransmitter type.