Elsevier

Neuroscience

Volume 150, Issue 4, 19 December 2007, Pages 743-753
Neuroscience

Behavioural neuroscience
Dynamic interplays between memory systems depend on practice: The hippocampus is not always the first to provide solution

https://doi.org/10.1016/j.neuroscience.2007.10.004Get rights and content

Abstract

Previous studies showed that the optimization of behavioral performance through extended training depends on a switch from hippocampus-based memory to striatum-based habit. Here we investigate whether the amount of training within one learning session influences the retention of memory for hippocampal versus striatal strategies. Mice were trained to search for a submerged cue-marked platform which remained in the same spatial location in the water-maze for each of three training regimens (4, 12 or 22 trials). Subsequently, they were either tested for retention of memory 1 h or 24 h later on a probe test or killed at different time points over a 7-h period to determine the kinetic of cAMP response element binding protein (CREB) phosphorylation in both memory systems. During the probe test mice had to choose between a submerged platform located in the same position as during the acquisition phase (spatial solution) and a platform marked by the cue but located in the opposite quadrant of the pool (cue-guided solution). Results showed that the animals first preferred the cue-marked platform, which represents a strategy that was selectively impaired by lesions of the dorsolateral caudate-putamen. With further practice, or context pre-exposure, animals transiently favored the hippocampus-dependent place solution but finally, both strategies became interchangeable and insensitive to either lesion. CREB phosphorylation increased in both memory systems following acquisition but training-dependent changes selectively occurred in the hippocampus wherein biphasic activation was initiated by the four-trial training and blocked by training for 22 trials. These findings indicate that learning in one session consists of three acquisition stages with parallel engagement of multiple memory systems at the beginning of learning. They suggest, however, that, in a later phase, dynamic interplays promote the use of the most adapted brain system depending on practice and this is accompanied by specific patterns of CREB phosphorylation in the hippocampus.

Section snippets

Subjects

Male mice (N=288) of the C57Bl/6 by JIco strain (Charles River, L’Arbresle, France) were used. Animals were received in the laboratory at 8 weeks of age and were housed collectively in an animal room equipped with air-conditioning (23 °C) and with an artificial 12-h light/dark cycle (from 7:00 AM on). At the age of 3–4 months (25–30 g), mice were housed in individual home cages and were given ad libitum access to food and water for 10 days before behavioral testing or surgery. During the

The memory strategy used differs as a function of the intensity of training and of the delay of retention

Firstly, intensities of training regimens were determined based on three learning levels as shown in the acquisition curve recorded for training over 22 trials (Fig. 2A). Results showed that the animals required 12 trials to display performance accuracy reaching an asymptotic level. Indeed, based on the swim latencies to escape the animals exhibited a significant (F(3,231)=8.91; P<0.001) but partial acquisition when given a shortened (four trials) training regimen whereas the performance level

Discussion

Using a variety of behavioral paradigms, many experiments have demonstrated that multiple memory systems are responsible for learning in rodents (McDonald and White 1995, Guillou et al 1999b, Kim and Baxter 2001, Packard and Cahill 2001, Gold 2004). The present results indicate that one learning session in a task that involves two competing solutions (i.e. either a spatial strategy or a cue-guided strategy) consists of three acquisition stages depending on the nature of the memory strategy that

Conclusion

These results are congruent with the theory according to which multiple memory systems are activated in parallel in a mammalian brain (White and McDonald, 2002) and support the view that post-training is a crucial phase for their interactions. They suggest that mechanisms of memory formation are initiated “on line” during learning and quickly involve dynamic interplays leading to “off line” reorganization between memory systems depending on practice. Moreover, in agreement with recent data

Acknowledgments

This work was supported by grants from the French Centre National de la Recherche Scientifique [C.N.R.S] and the Conseil Régional d’Aquitaine. The authors thank Dr. Gleb Shumyatsky for helpful comments on the manuscript.

References (49)

  • R.J. McDonald et al.

    Multiple memory systems: the power of interactions

    Neurobiol Learn Mem

    (2004)
  • D. Moncada et al.

    Phosphorylation state of CREB in the rat hippocampus: a molecular switch between spatial novelty and spatial familiarity?

    Neurobiol Learn Mem

    (2006)
  • R.G. Morris

    Developments of a water-maze procedure for studying spatial learning in the rat

    J Neurosci Methods

    (1984)
  • M.G. Packard et al.

    Affective modulation of multiple memory systems

    Curr Opin Neurobiol

    (2001)
  • M.G. Packard et al.

    Inactivation of hippocampus or caudate nucleus with lidocaine differentially affects expression of place and response learning

    Neurobiol Learn Mem

    (1996)
  • R.A. Poldrack et al.

    Competition among multiple memory systems: converging evidence from animal and human brain studies

    Neuropsychologia

    (2003)
  • N.C. Voermans et al.

    Interaction between the human hippocampus and the caudate nucleus during route recognition

    Neuron

    (2004)
  • N.M. White et al.

    Multiple parallel memory systems in the brain of the rat

    Neurobiol Learn Mem

    (2002)
  • T. Ahmed et al.

    Long-term effects of brief acute stress on cellular signaling and hippocampal LTP

    J Neurosci

    (2006)
  • A. Barco et al.

    CREB, memory enhancement and the treatment of memory disorders: promises, pitfalls and prospects

    Expert Opin Ther Targets

    (2003)
  • R. Bernabeu et al.

    Involvement of hippocampal cAMP/cAMP-dependent protein kinase signaling pathways in a late memory consolidation phase of aversively motivated learning in rats

    Proc Natl Acad Sci U S A

    (1997)
  • B. Bontempi et al.

    Time-dependent reorganization of brain circuitry underlying long-term memory storage

    Nature

    (1999)
  • N.J. Broadbent et al.

    Rats depend on habit memory for discrimination learning and retention

    Learn Mem

    (2007)
  • Q. Chang et al.

    Switching memory systems during learning: changes in patterns of brain acetylcholine release in the hippocampus and striatum in rats

    J Neurosci

    (2003)
  • Cited by (40)

    • Dietary fish hydrolysate supplementation containing n-3 LC-PUFAs and peptides prevents short-term memory and stress response deficits in aged mice

      2021, Brain, Behavior, and Immunity
      Citation Excerpt :

      In our study, aged mice took more time to adopt spatial strategies to solve the task. This has been already suggested in several studies reporting that aging affects allocentric strategies, which are hippocampus-dependent, promotes the use of egocentric strategies, more striatum-dependent, and delays the evolution from non-spatial to spatial strategies in mice at 12 months of age and over (Bensalem et al., 2016; Fouquet et al., 2011; Gil-Mohapel et al., 2013; Kim et al., 2001; Martel et al., 2007). This has been also observed in an adapted version of the Morris water maze for humans in which allocentric spatial navigation learning, but not egocentric navigation, is particularly altered over 70 (Gazova et al., 2013).

    • Shifting between response and place strategies in maze navigation: Effects of training, cue availability and functional inactivation of striatum or hippocampus in rats

      2020, Neurobiology of Learning and Memory
      Citation Excerpt :

      Such correction is crucial, however, because it would demonstrate a capability to engage the hippocampal and striatal memory systems in dynamic, interactive and situation-adapted ways (e.g., Berke, Breck, & Eichenbaum, 2009; Chersi & Burgess, 2015; Eschenko & Mizumori, 2007; Mizumori et al., 2004; Regier, Amemiya, & Redish, 2015). Past research has suggested that the interaction between response and place memory systems is modulated by saliency of intramaze and/or extramaze cues (e.g., Packard & Goodman, 2013), task complexity (e.g., Cassel, Kelche, Lecourtier, & Cassel, 2012; Ruprecht, Taylor, Wolf, & Leising, 2014), practice (e.g., Martel et al., 2007), and a few other factors, including stress (e.g., Gardner et al., 2013; Packard & Goodman, 2013; Quaedflieg & Schwabe, 2018; Schwabe, 2013). However, general conclusions regarding hippocampal-striatal interactions cannot be drawn because the aforementioned factors have most often been tackled by using a diversity of experimental devices, protocols, training levels in different and separate experiments.

    • Increased training compensates for OX1R blockage-impairment of spatial memory and c-Fos expression in different cortical and subcortical areas

      2018, Behavioural Brain Research
      Citation Excerpt :

      Similar to results in the DG, the c-Fos activity reduction in the gRSC was only found in the weak training condition. The different impact of the level of training in the effects of OX1R blockade in dRSC and gRSC activity could be related to the fact that animals are more prone to using a response strategy to solve the task when the number of trials is low [60]. This type of strategy consists of proprioceptive and non-visual inputs, thus amplifying the importance of the gRSC, given that it is more connected to head-direction areas [61].

    View all citing articles on Scopus
    1

    Present address: Rutgers University, Life Sciences Building, Room 322, 145 Bevier Road, Piscataway, NJ 08854, USA.

    View full text