FNWI building SILS-CNS part 1
FNWI building SILS-CNS part 2

SILS Swammerdam Institute for Life Sciences - CNS




Coding of spatial trajectories by CA1 neurons in
wildtype and CA1 NMDA-receptors



Amsterdam Science Park



Navigation in a complex environment can rely on several different strategies: for example, subjects may decide on the route to take based on a configuration of external landmarks (allocentric strategies), or memorize a well-known route as a sequence of body movements (egocentric strategy). The hippocampus is at the center of the network of brain structures supporting spatial memory and route computations, possibly contributing to a Cognitive Map function that has been proposed in the '40s by Tolman (O'Keefe and Nadel, 1978)
In rodents, interference with hippocampal function with lesion studies (see e.g. Morris et al., 1982), pharmacological (see e.g. Eichenbaum et al., 1990) and genomic means (Nakazawa et al., 2004), impairs the animal's ability to navigate to a goal. On the other hand, when a route can be computed in terms of a simple set of body movements (e.g. right/left turns), the role of hippocampus seems less important (Packard and McGaugh, 1996).

Interestingly, hippocampal subregions appear to have a differential involvement in the acquisition of spatial memories (Nakazawa et al., 2004), with CA1 disruption being more effective than intervention in CA3. For these reasons, the availability of specific genomic manipulations that affect only one of the hippocampal subfields have been particularly fruitful in dissecting the neural circuitry of spatial navigation. In particular, a series of knockout models have targeted the NR-1 subunit of the NMDA receptor specifically in CA1 (Tsien et al., 1996), CA3 (Nakazawa et al., 2002) and the dentate gyrus (McHugh et al., 2007), offering an unprecedented chance to explore how this receptor affects synaptic and systems plasticity in each of these structures.

The Starmaze Task.
Top : the starmaze, surrounded by cues hanging from a black curtain. The blue arrow indicates the most direct path to the rewarded arm in a standard trial; “dep arm” means departure arm.
Bottom : schematic view of a probe trial, where the animal departs from a different arm. According to the trajectory it takes, we can identify the strategy used: if the animal goes straight to the previously rewarded arm, we can assume that it is using the environmental cues (allocentric strategy, red arrow); on the other hand, if the animal just repeats the same sequence of body movements, as it was doing before, it will end up in a different arm (egocentric strategy; green arrow)


starmaze


In addition to these behavioral effects, the implication of the hippocampus in spatial processing is critically supported by the correlation between space and the activity of hippocampal cells: in all hippocampal substructures place cells have been found, cells that activate as the rat traverses a different place in the environment (Figure 2). Taken together, place cell activities form a map of the environment, which the animal can use to self-localize. Place cell activity is the result of a computation involving multiple inputs. Most importantly, the hippocampus has to combine inputs related to self-motion, or path integration (McNaughton et al., 2006) and external cues. To dissociate these different contributions, Rondi-Reig and co-workers (2006) tested wild-type and CA1 NR-1 KO (knockout) mice in a complex, star-shaped maze, where animals had to find the one out of five rewarded arms (Figure 1). In standard trials, mice always started from the same departure arm. NR-1 KO mice were found to be impaired in acquiring this task.
In probe trials, rats started from a different arm, so that, if they followed a body-turn based strategy, they should end up in an arm different from the goal arm, which was unchanged. If, however, they based their route on external cues, they should be able to reach the goal arm. Thus, NMDA receptors in CA1 are important for this type of route learning. However, nothing is known about how hippocampal cells encode routes computed according to different strategies.

Example of hippocampal place cell

Pyramidal neurons in area CA1 of mouse brain typically fire in a restricted area in the environment and are nearly silent elsewhere. They are therefore termed Place Cells. By precisely tracking the position of the mouse in the maze, using EthoVision XT (Noldus Information Technology)
we can synchronize the mouse's position with the firing activity of individual neurons. The result is this firing map, which was calculated over the most frequently used trajectory by the mouse in the session. Color scale is indicative of firing rate




Participants in this research line

Henrique Cabral (PhD)
Prof.Dr. Cyriel Pennartz
Dr. Francesco Battaglia

Internships

This project is open for internships of at least 6 months (Msc level)
for inquiries, contact F.P.Battaglia or C.M.A.Pennartz

Collaborations

Dr Laure Rondi-Reig, LPPA-CNRS, College de France
Dr Celine Fouquet, LPPA-CNRS, College de France
Fabrizio Grieco     Noldus Information Technology BV


Key publications

International literature

This page was last updated on 3 march 2011

Top