Functional modularity in entorhinal grid cells

The aim of this project was to determine the organization of high-end cerebral cortices in mammals. During the past decades, neuroscientists have made great advances in the understanding of isolated parts of the cortex, such as in the visual cortex, and we are beginning to appreciate how sensory signals are converted to activity patterns in the brain. However, these advances are generally limited to early stages of sensory processing. We still know little about the later stages of cortical processing, in high-end multisensory association cortices. In the present project, we sought to identify neural codes near the apex of the cortical hierarchy ? in the entorhinal cortex. We have measured activity from grid cells ? a functional cell type that we discovered in this area in 2005. We took as a starting point the fact that these cells are organized in modules, where each module consists of grid cells with a given scale and orientation as well as a partially independent functional organization. In the present project we have shown that grid cells distribute widely across the mediolateral axis of the medial entorhinal cortex, in a bandlike organization. We have shown that grid modules have a similar scale relationship across modules in rats and mice, with a scale factor of approximately 1.4, and we have demonstrated that this organization of grid cells develops even in animals that are raised in extreme geometric environments, deprived of salient visual references such as vertical borders. Experience has a temporary effect on grid cell development but deviations disappear with added exposure, suggesting that the circuit is to a large extend hard-wired. The final part of the project showed that the organization of the medial entorhinal circuit has implications for the formation of multiple maps of place cells in the hippocampus.

The medial entorhinal cortex (MEC) is part of the brain's circuit for dynamic representation of self-location. The metric of this representation is provided by grid cells, cells with spatial firing fields that tile the environment in a periodic hexagonal pattern. Much has been learned about the properties of individual grid cells but the population dynamics of the grid map is not well understood, mainly because of limited sampling within animals. Preliminary data from my lab suggests that grid cells are o rganized into a conglomerate of modules with distinct geometric features and differences in functional capacity. The aim of the present project is to determine, with data from more than a hundred grid cells in the same animal, how grid modules with distin ct geometric properties are mapped onto the neuronal sheet of the MEC. We shall identify the number and distribution of grid modules and relate these parameters to variations in the expression of immunocytochemical markers. Using a combination of computat ional and experimental approaches, we shall establish whether fractionation follows differences in input patterns or underlying structure of the network or whether discretization is an emergent property of self-organizing inhibitory recurrent networks suc h as the MEC. The developmental origin of grid modules, and the capacity of grid networks before the appearance of modularity, will be addressed by recording large numbers of MEC cells from juvenile animals when they explore open environments for the firs t time. The endeavour will pioneer the functional analysis of complex neural circuits and may, as one of the first initiatives in neuroscience, provide us with mechanistic insight into a non-sensory cognitive function in the mammalian cortex.