FIGURE 1. Procedures for inducing global remapping and rate remapping in hippocampal area CA3.
From the following article:
Hippocampal remapping and grid realignment in entorhinal cortex
Marianne Fyhn, Torkel Hafting, Alessandro Treves, May-Britt Moser and Edvard I. Moser
Nature 446, 190-194(8 March 2007)
doi:10.1038/nature05601

a–c, Firing fields of three simultaneously recorded CA3 cells during global remapping (different boxes in a constant location (a) or similar boxes in different rooms (b)) or rate remapping (different wall colours (c)). For complete cell samples see Supplementary Fig. 2a–f. Each row shows data from one active neuron (t, tetrode; c, cell). Each pair of columns shows, for each neuron, the location of spikes (red) on the trajectory of the rat (left) and a colour-coded map of the firing rate distribution (right; dark blue is zero, red is maximum; peak rate is indicated). d, Change in firing rates estimated by rate overlap (means
s.e.m.). Red lines indicate chance levels. Blue bars, same environment (A–A'); green bars, different environments (A–B). e, Spatial correlation for cells with activity in both environments (means
s.e.m.; best rotation). f–h, Rate overlap and spatial correlation in cells with activity in both environments. f, Different boxes (as in a); g, different rooms (as in b); h, different colours (as in c). Scatter plots show rate overlap as a function of spatial correlation for individual CA3 cells (open circles, A versus A'; filled circles, A versus B). Frequency distributions show spatial correlations (green bars) compared with distributions for shuffled cell pairs (purple bars) from the same experiments. In the different-boxes task, 2 out of 23 spatial correlations exceeded the upper limit of the 95% confidence interval of the shuffled distribution. In the different-rooms task, 3 out of 12 correlations were outside the 95% confidence interval. In the rate remapping task, 25 out of 26 cells were outside the 95% confidence interval. The rate overlap between A and B is significantly higher in f–h than in d because the scatter plots in f–h do not include the large number of cells with low rates in one of the environments.
