Research reportThe match/mismatch of visuo-spatial cues between acquisition and retrieval contexts influences the expression of response vs. place memory in rats
Highlights
► Further exploration of the novel response vs. spatial memory double-H maze test. ► Response and place memories coexist after response training in the double-H maze test. ► Competition vs. cooperation between memory systems for response and place cues. ► More stops in place vs. response cues users on the 1st swim track in a probe trial. ► Cue match/mismatch- and error-driven shifts from response to place memory.
Introduction
The fact that evolved animals like laboratory rodents possess distinct, though not necessarily independent, memory systems relying upon different brain structures is supported by evidence from maze learning tasks coupled to measures of brain activity, reversible inactivation or/and permanent lesion studies (e.g. [1], [2], [3], [4], [5], [6], [7], [8]). In a variety of such tasks (e.g., water, Barnes, star… mazes), rodents may learn by using place cues and elaborating cognitive maps, as demonstrated in Tolman's pioneering work (e.g. [23]). Thereby, they acquire the capability to navigate to their goal by reference to a memorized configuration of allocentric room cues. This learning process is often termed place learning in the literature. Generally, the resulting capabilities are acquired over a relatively low number of trials. They rely on a highly flexible declarative-like memory system usually presented as requiring the hippocampus (e.g. [8], [9]).
Rodents can also acquire displacement habits by the way of a response learning process. For instance, in a maze, they may learn the only correct route connecting a start box with a goal box or memorize simple (e.g., turn systematically right in a cross maze) or more complex (e.g., make successive right, left, right, right, left turns in e.g., a serial alternation maze) response patterns. Such capabilities are acquired gradually, over a larger number of trials than needed for place learning. They are based on a rigid procedural memory system usually presented as requiring the dorsal striatum [9], [8].
These declarative-like and procedural memory systems have sometimes been presented as competing with each other for the control of a learning process or the expression of a memory (e.g. [1], [2]). The engagement of one or the other memory systems also depends on aspects of training and cues in dual solution tasks, rather than on genetic or other biological factors (e.g. [10], [11]). Based on a cross-maze task, pioneering data reported by Packard and McGaugh [8] suggested that the initial approach of a response-learning task may in fact engage a hippocampus-dependent, declarative-like memory system. Over additional training, the rats gradually shift to the use of response cues requiring the striatum. Most interestingly, when animals have shifted to the response cues, the place representation remains intact (e.g. [7], [8] for a review). This observation is in line with a view considering that the striatum-dependent system would be able to take a functional advantage over the hippocampus-dependent system. Packard and McGaugh [8] nevertheless regarded their data as suggesting that the “two memory systems are functioning independently” (p. 70). They also emphasized that it is “the nature of the environment [which] appears to interact with separate learning systems in determining the expression of place or response learning” (p. 70), as proposed by Restle [10]. Evidence for the last assertion, however, was not provided by the authors.
Using our recently described double-H maze task (Pol-Bodetto et al. [17]), we confirmed that rats keep an allocentric representation of the goal's location despite their use of procedural memory based on response cues to approach a goal. We also observed that the behavior of the rats having acquired a place response was less affected by the habit when the view on room cues from the start point (termed “visuo-spatial context” hereafter) was markedly different from that of training trials, conversely to what happened when the visuo-spatial contexts of training and probe trials were close. The immediacy of the place response indicated that the procedural approach of the task can be switched off very quickly. This issue is challenging as it is not fully compatible with several ideas, including that of a genuine competition between procedural and declarative-like memory systems (e.g. [1], [2]), that of independent memory systems [8], or that of an inhibition of declarative-like mechanisms once habits are established [8]. We hypothesized that the expression of a memory for response vs. place in rats trained until habits have formed is the consequence of a very rapid comparison between training and testing contexts at the start of a trial. When these contexts are matching, or at least when they are not too different from each other, rats preferentially use a response memory. In case of a detected mismatch, a cognitive map is preferred. The current study aimed at furthering the exploration of this issue by the means of a non-invasive approach in the double-H maze test. Thus, conversely to our former experiments in which drugs acting at muscarinic or NMDA receptors were used or control injections made, the rats of the current study were trained and tested free of any such treatment in a variety of protocols. Furthermore, in our former study, we always used the same start location to train the rats, namely an arm located at a room side opposite to where the rat cages were kept. It turns out that this aspect of the protocol does not seem to be anecdotal, as Restle [10] provided solid meta-analytic evidence showing that the location of the rat cages relative to the maze goal constitutes a strong, if not the strongest place cue for the rat. Therefore, in the current study, we used a protocol balancing this bias by training equal proportions of rats from the side of the cages or from the opposite. Finally, based on whether rats first used response or place cues in the probe trial, detailed a posteriori analyses of the behavioral patterns were made, an approach which the use of a variety of drug treatments had hindered in our former experiment.
Section snippets
Subjects
We used 52 adult, male Long–Evans rats weighing between 240 and 268 g at the start of the experiment. Rats from the Long–Evans strain shift from place to response learning over training in a repetitive orientation task [12]. They were provided by the Centre d’Elevage R. Janvier, Le Genest St-Isle, France. All rats were kept in individual transparent Makrolon cages (42 cm × 26 cm × 15 cm) in temperature-controlled (23 ± 1 °C) rooms that were maintained on a 12:12 h dark–light cycle (light on at 7:00 am).
Pretraining
On the day of pre-training, rats were given four successive trials in which they had to swim from one arm extremity to the opposite one, having no access to the other parts of the maze. In average, all groups swam this distance in less than 16 s (overall mean = 14.3 ± 0.9 s), and we found significant Group effects neither on latency (F3/47 = 0.27, ns; data not illustrated) nor on swim velocity (overall mean = 12.9 ± 0.6 cm/s; F3/47 = 0.66, ns; data not illustrated).
Training
The data are illustrated in Fig. 2. We
Discussion
In our previous study (Pol-Bodetto et al. [17]), we found that a protocol fostering response learning contributed to a preferential expression of response memory in a probe trial. We also observed that despite the use of a training protocol adapted to response learning, memories for a response or a place co-existed, as rats were able to shift from the former to the latter when a navigation based on response memory missed the goal. In this study, however, there were two major and three minor
Conclusion
The current findings suggest that, in a goal-directed maze task, place and response learning systems, which are driven by the hippocampus and the striatum, respectively, do not necessarily operate in a competing manner. In fact, they may instead cooperate according to immediate match-mismatch computation and subsequent error-driven adjustment principles. This view could now be challenged on the basis of e.g., lesion, reversible inactivation and immediate early gene expression studies.
Conflict of interest
The authors of the current work have no conflict of interest to declare. The funding sources had no role in the study design, the collection, analysis and interpretation of data, the writing of the report, and the decision to submit the article for publication or the choice of the journal.
Acknowledgments
This work was supported by the University of Strasbourg and the CNRS. The authors are grateful to O. Bildstein, D. Egesi and G Edomwonyi for their assistance in animal care. They also thank Dr A. Pereira de Vasconcelos and Dr R. Goutagny for helpful comments on an almost final draft of the manuscript.
References (23)
- et al.
Intra-hippocampal lidocaine injections impair acquisition of a place task and facilitate acquisition of a response task in rats
Behav Brain Res
(2003) - et al.
Parallel processing across neural systems: implications for a multiple memory system hypothesis
Neurobiol Learn Mem
(2004) Anxiety, cognition, and habit: a multiple memory systems perspective
Exp Brain Res
(2009)- et al.
Inactivation of hippocampus or caudate nucleus with lidocaine differentially affects expression of place and response learning
Neurobiol Learn Mem
(1996) - et al.
Transfection of mutant CREB in the striatum, but not the hippocampus, impairs long-term memory for response learning
Neurobiol Learn Mem
(2008) Behavior strategy learning in rat: effects of lesions of the dorsal striatum or dorsal hippocampus
Behav Process
(2004)- et al.
The double-H maze test, a novel, simple, water-escape memory task: acquisition, recall of recent and remote memory, and effects of systemic muscarinic or NMDA receptor blockade during training
Behav Brain Res
(2011) - et al.
Memory influences on hippocampal and striatal neural codes: effects of a shift between task rules
Neurobiol Learn Mem
(2007) - et al.
Immediate early gene activation in hippocampus and dorsal striatum: effects of explicit place and response training
Neurobiol Learn Mem
(2007) - et al.
Switching memory systems during learning: changes in patterns of brain acetylcholine release in the hippocampus and striatum in rats
J Neurosci
(2003)
A double dissociation between the rat hippocampus and medial caudoputamen in processing two forms of knowledge
Behav Neurosci
Cited by (14)
Altered activity-regulated H3K9 acetylation at TGF-beta signaling genes during egocentric memory in Huntington's disease
2022, Progress in NeurobiologyVentral midline thalamus activation is correlated with memory performance in a delayed spatial matching-to-sample task: A c-Fos imaging approach in the rat
2022, Behavioural Brain ResearchCitation Excerpt :Before being trained with ITI of increasing durations, all rats had been pre-trained using a repetitive egocentric navigation task, which does not prevent the formation of a cognitive map, as the room was illuminated and allothetic cues were visible [22]. Consequently, as found in a former study [25], after a 180° rotation of the start arm for the probe trial, a large majority of rats displayed an immediate place memory-based behavior. This was reflected in the first arm to which they swam after being released in the maze, namely that of the platform location during training.
Subtle alterations in memory systems and normal visual attention in the GAERS model of absence epilepsy
2016, NeuroscienceCitation Excerpt :However, both (correct latency and magazine latency) can be affected by motor dysfunctions. This test was used to explore procedural memory (D’Hooge and De Deyn, 2001; Pol-Bodetto et al., 2011; Cassel et al., 2012). Apparatus: The apparatus was composed of a transparent acrylic glass maze formed by six parallel arms (160 9 20 cm) that are joined in their middle by a perpendicular central path.
The reuniens and rhomboid nuclei: Neuroanatomy, electrophysiological characteristics and behavioral implications
2013, Progress in NeurobiologyCitation Excerpt :Dolleman-van der Weel et al. (2009) discussed their finding by reference to changes in strategy shifting, speculating that rats with Re lesions exhibited deficits of inhibitory response control mechanisms. We used our recently developed double-H task (Cassel et al., 2012; Pol-Bodetto et al., 2011; see Fig. 9) to test failure in response inhibition. Rats had to learn two escape response sequences (right-left or left-left turns) which relied on where in the maze they were positioned at the start of each trial (see Fig. 10).