, 2010 and Wills et al., 2010), suggesting that these spatial circuits may be at least partly hard-wired. However, place cells appear to have more adult-like characteristics than grid cells, which raises the possibility that grid cells are dispensable for the formation of place cells in young animals. A recent

study with adult animals has shown that place cells can persist under conditions where the periodicity of grid fields is reduced as a result of medial septal inactivation (Koenig et al., 2011). However, because the grid cells and place cells were studied in different animals or, in one animal, in different hemispheres, it cannot yet be see more ruled out that a minimum of grid input was spared in those recordings that demonstrated intact place signals. It is not clear what alternative inputs could provide spatial signals to the hippocampus if no contribution is received from the grid cells; however, one possibility is that place

cells obtain the necessary spatial information from entorhinal border cells (Savelli et al., 2008 and Solstad et al., 2008), as proposed in early theoretical work (Hartley et al., 2000). Input from such cells may be sufficient to generate spatially localized activity. Another possibility is that grid patterns are present but difficult to visualize in time-averaged rate maps due to reduced spatial stability of neural activity in young and septum-inactivated animals. The jitter of firing may affect grid fields more than place fields, considering that the Inhibitor Library cost former are smaller. Finally, it is possible that the rudimentary periodicity of young grid cells, combined with Hebbian plasticity and phase precession, is sufficient to evoke localized firing in hippocampal target neurons. Following the discovery of grid cells in the MEC (Hafting et al., 2005),

no recent studies indicate the presence of a broader grid cell network in multiple parahippocampal structures. An abundant population of grid cells has now been reported in the pre- and parasubicular regions of the parahippocampal formation (Boccara et al., 2010). Compared to the MEC, pre- and parasubiculum have a higher percentage of grid cells conjunctive with a head direction preference, which may contribute to a slight reduction in the hexagonal periodicity of these grid cells compared to MEC grid cells. There are at least two possible mechanisms that could underlie the presence of grid cells in multiple parahippocampal cortices. First, the strong feed-forward projection from pre- and parasubiculum to MEC (van Groen and Wyss, 1990) gives rise to the suggestion that the MEC may inherit the grid signal from these input regions. This would require a complex wiring scheme based on minimal convergence between pre- and parasubicular cells with different grid phase, grid scale, or grid orientation.