Hippocampal CA3-region is crucial for acquisition and memory consolidation in Morris water maze task in mice

Abstract

This experiment investigated the involvement of the dorsal hippocampal CA3-region in the different phases of learning and memory in spatial and non-spatial tasks. To do so, we temporarily inactivated the CA3-subfield by a focal injection of diethyldithiocarbamate (DDC) which chelates most of the heavy metals present in this region.

The effects of temporary inactivation of the CA3-region were examined in an associative task, the Morris water maze (MWM). To study the different phase of memory we used a new behavioural massed-procedure founded on four massed training sessions in the spatial and the non-spatial (cue) version of this task. In the spatial version, we showed that a bilateral injection of DDC into the CA3-region impairs the acquisition but not the recall of spatial information. The main result of this study is that the same injection performed immediately after the training session also perturbed memory consolidation. In the cue version of the MWM, we found no difference between the DDC-injected mice and their controls in acquisition or memory consolidation of non-spatial information.

These results suggest that the hippocampal CA3-region is essential for spatial memory processes and specifically in memory consolidation of spatial information.

Introduction

Many neurobehavioral experiments have revealed the involvement of the dorsal hippocampus in spatial learning and memory [4], [6], [24], [25], [37], [43]. The hippocampus is a heterogeneous structure that consists of a largely unidirectional loop of excitatory pathways through different subfields, the dentate gyrus, CA3, CA1 and subiculum [16]. In this present study, we investigated the role of one of the hippocampal subfields, the CA3-region, in three different phases of spatial memory (i.e. acquisition, memory consolidation and recall). This CA3-region holds a strategic position in the hippocampus because it receives sensorial information from the external and internal environment via two main pathways: the mossy fibres and the perforant path [1], [9]. Moreover, as CA3 pyramidal cells have numerous recurrent collaterals, this subfield is considered, in different computational models, as an auto association network for storage and retrieval of information particularly in working memory processes [33], [40]. Recently, different authors have demonstrated that this CA3-region is involved specially in spatial working memory [11], [20], [21]. This CA3 subfield seems particularly implicated during the initial acquisition phase of a spatial learning [12], [19] but not during recall. To our knowledge, nobody investigated, in the same study, the role of this subfield in the three memory phases of an associative spatial task, and particularly during the initial memory consolidation phase. Nevertheless, in a previous study, which have showed the involvement of this CA3-subfield in acquisition but also in memory consolidation in a particular non-associative spatial task, the spatial open-field task [38].

To study the role of the CA3-region in different memory phases in spatial and non-spatial associative tasks, it was necessary to develop new behavioural procedures and to block the CA3 subfield only during the learning phase studied. For the first point, we developed a behavioural procedure in the Morris water maze (MWM), where animals, are required to locate a hidden platform in a swimming pool in relation to environmental cues in less than 80 min. This massed-procedure is crucial to study the memory phases for several reasons. First, to study the acquisition phase, it was essential to avoid systematic drugs injections at the beginning of sessions which could possibly damage the neural tissue. Second, it was important to find a massed learning procedure brief enough to be sure the influence of the drug was maximal during the whole of a learning session [38]. Especially, this procedure was realized in order to allow working only on the first information memory consolidation. This point is very important because when learning sessions are done during several days, memory trace is reactivated and reconsolidated successively during each learning session [27], [35]. Implicated structures or molecular mechanisms can be different during first memory consolidation or during reconsolidation after reactivation of the memory trace [27], [28], [31], [35].

To study the different phases of memory, it is also necessary to temporarily block the hippocampal CA3 subfield only during one of these phases. That, we used the diethyldithiocarbamate (DDC) which present a maximal duration action into the dorsal hippocampus about 1 h 45 min [38]. This drug chelates most of the intra- and extracellular heavy metals [7], [15], [39]. DDC can interfere with synaptic transmission [7], with synaptic plasticity via the blockage of post-synaptic enzymes [14] and blocks, in vitro, the induction of mossy fiber LTP [22]. This chelator has been used before to block temporarily the dorsal hippocampus [11], the hippocampal mossy fibres [19] and in our group, the whole CA3-region [38].

So, this short duration of action of the DDC and our massed-behavioral procedure in the MWM, allowed us to work without ambiguity on the role of the CA3-region in the three different phases of spatial memory.