Figure 3.

Approximately 50% of Active and Silent GCs Receive Spatially Tuned Synaptic Input

(A) Top, AP rate map of an active GC that fired only sparsely. The abscissa represents the position on the linear belt (2-cm spatial binning), and the ordinate denotes the lap number. The number represents the mean AP rate (top), and the color-code bar indicates the AP rate in spatial bins (right). Yellow arrowheads indicate APs. Bottom, plot of average AP frequency against the position across laps.

(B) Left top, V_m median (after AP removal) plotted against the position in the same cell. The color code indicates Z-scored V_m values. Left bottom, plot of V_m median against the position across laps. Right top, V_m variance (after AP removal) plotted against the position in the same cell. The color code indicates V_m variance. Right bottom, plot of V_m variance against the position across laps. Note more depolarized V_m and higher V_m variance at a specific location.

(C) Spatial tuning of subthreshold EPSP activity. Left, polar plot of EPSP event frequency. Spatial positions (0–180 cm) were converted into angles (0°–360°). Black circles represent EPSP event frequency in each bin (number of events divided by time spent in the respective bin), and the red arrow indicates the mean EPSP frequency tuning vector (multiplied by 10 for illustration purposes). Right, distribution of mean TVL from shuffled data. The red vertical line indicates the mean TVL of the original data. The mean TVL from the original data is significantly larger than the values obtained from the shuffled data. Although this cell fired APs ∼80–120 cm in only 2 of 21 laps, it showed a consistently higher V_m median, V_m variance, and EPSP rate at the same location.

(D–F) Similar analysis as in (A)–(C) but for a silent GC. This silent GC also received significant spatially tuned input.

(G–I) Similar analysis as in (A)–(F) but for another GC that showed no significant spatially tuned input. The gray vertical line indicates the mean TVL of the original data (not significantly different from the shuffled distribution in this cell).

(J) Proportion of GCs with spatially tuned synaptic input. Left, proportion of active nonplace GCs with spatially tuned EPSP input; 13 of 28 cells showed spatially tuned EPSPs. Right, proportion of silent GCs with spatially tuned EPSP input; 14 of 28 cells received spatially tuned EPSPs.

(K) Summary of EPSP frequency tuning vector directions of all significantly tuned GCs, plotted against the position. Cells were sorted in ascending order.

(L) Double-logarithmic plot of spatial information score per time against EPSP frequency TVL. Each data point represents a single GC recording. Spike information was significantly correlated with TVL. The red line represents the results from linear regression.