The hippocampal formation (HF) is the core of a memory system that is crucial for the formation of new episodic (unique event) memories in humans and lab animals such as rodents. These functions are thought to rely on a variety of specialized neural network computations: It is known, for example, that the hippocampus keeps track of the animal’s position in space, via the activation of place cells. A place cell is a neuron which increases its firing rate when the animal is in a specific location in space. Episodic memories are then formed by creating a joint representation of the animal’s current location and the events that occur at that location. Since the discovery of place cells in the 1970s, it is believed that place cell firing is generated by input from an upstream navigational system. The presence of such a navigational system was confirmed upon the recent discovery of multiple highly specialized cell types in the medial entorhinal cortex (MEC), the most prominent being grid cells. Grid cells are similar to place cells in that they form spatially selective firing fields (place fields). However, while place cell firing is confined to one to two place fields, grid cells form multiple firing fields that are arranged in equilateral triangular lattices. These firing patterns are thought to create a map like representation of the environment. In addition, the MEC contains head direction cells and speed cells which keep track of the animal’s running direction and speed, respectively. Together, these three cell types have all that is requires in order to accurately navigate through space. While the potential importance of place cells and grid cells was recognized by the Nobel Prize 2013, their function could not be experimentally evaluated due to the lack of tools that allow to manipulate individual functional cell types within the HF. We were able to overcome this obstacle by disrupting the normal development of grid cells in very young mice. This manipulation allowed us to experimentally test the long-standing theory that grid cells enable the animal to navigate through space through the generation of place cells in the hippocampus. To our surprise, we found that hippocampal place cells were able to keep track of the animal’s position in space, even when grid cells were abolished. The animal's ability to create shortcuts in order to navigate to a goal location, however, was nevertheless disrupted. Our results suggest that information about the current location of the animal can reach the hippocampus via routes that do not involve gird cells, and that this information is sufficient to keep track of the animal’s position in space. Behavioural tasks with high navigational demand, including the calculation of short cuts, in contrast, require computations performed by grid cells.
The formation of episodic memories does not only require the association of information about events and their places, but also that events are encoded in the correct temporal order. The temporal organization of information relies on the precise interplay between brain oscillations and the firing or individual neurons. The medial septum (MS) is a structure known to be a key generator of prominent brain oscillations referred to as the theta rhythm. How the individual cell types within the MS contribute to this function, however, remains unknown. Using patch clamp recordings and neuronal tracing techniques, the Monyer laboratory recently identified two distinct long-range-projecting GABAergic neuron (LRGN) types in the MS, which can be distinguished by their selective expression of parvalbumin (PV) and calbindin (CB) (Fuchs et al., 2016). These two cell types target specific cell types in the MEC and are ideally suited to coordinate distinct temporal computations in the MEC. We probed the function of these two LRGNs, using optogenetics and electrophysiological in vivo recordings in behaving mice. We found that the two cell types have distinct, and yet complementary functions: PV expressing LRGNs are important for the generation of the theta rhythm in the MEC, and thus for the generation of a clock-like signal for the coordination of memory components. CB expressing LRGNs, in turn, are not required to generate a clock-like signal, but instead organize the sequential firing of MEC neurons.
The results were presented at the Spring Hippocampal Research Conference 2017 in Taormina (Italy) as well as at the 11th Federation of European Neurocience Scoieties (FENS) 2018 in Berlin (Germany).
