Fig. 5: Distinct patterns of temporal reorganization of CA3 spiking after DG and MEC lesions. | Nature Communications

Fig. 5: Distinct patterns of temporal reorganization of CA3 spiking after DG and MEC lesions.

From: Distinct roles of dentate gyrus and medial entorhinal cortex inputs for phase precession and temporal correlations in the hippocampal CA3 area

Fig. 5

a Phase-distance plot of spikes from an example CA3 cell from an MEC-lesion rat. All spikes of the cell’s trains are included. b Phase-distance plot of the same cell after replacing the spikes within each theta cycle with the mean phase of the spikes within each cycle. Using the cycle mean yielded a negative precession slope. An inconsistent distribution of spikes within theta-cycles after MEC lesions may thus have precluded the detection of phase precession. c Distribution (violin plot) of circular-linear regression slopes calculated from the cycle means. With regression slopes from cycle means there was no difference between cells from control and MEC-lesion rats (n = 101 and 158 cells, z-statistic = − 1.44, p = 0.151, MW test), while the difference between cells from control and the DG-lesion rats was retained (n = 84 and 68 cells, z-statistic = − 4.23, p = 2.4 × 10−5, MW test). However, as in analyses without within-cycle phase averaging (see Figs. 1 and 2), all groups showed some level of remaining phase precession (median slope less than zero: CTRL(DG): n = 84 cells, z-statistic = −6.49, p = 4.4 × 10−11, LESION(DG): n = 68 cells, z-statistic = − 2.33, p = 0.01; CTRL(MEC): n = 101 cells, z-statistic = − 7.42, p = 5.9 × 10−14, LESION(MEC): n = 158 cells, z-statistic = − 7.56, p = 2 × 10−14; one-sided sign tests). d Circular variance of theta phase with either all spikes (filled bars) or with each cycle’s spike mean (solid lines). In all groups, the circular variance decreased after replacing each cycle’s spikes with their mean, though the effect is least pronounced for CA3 cells of DG-lesion rats. The median circular variance of cycle mean spikes (dashed vertical lines) and all spikes (solid vertical lines), respectively [CTRL(DG), 0.557 and 0.707 (effect size = – 0.809; z-statistic = − 5.22, p = 1.8 × 10−7); LESION(DG), 0.661 and 0.733 (effect size = – 0.308; z-statistic = − 2.04, p = 0.041); CTRL(MEC), 0.603 and 0.782 (effect size = − 0.917; z-statistic = − 6.16, p = 7.5 × 10−10); LESION(MEC), 0.634 and 0.805 (effect size = − 0.779; z-statistic = − 7.26, p = 4.0 × 10–13; MW tests)]. Violin plots: Outline, distribution; error bars: 1.5 times the interquartile interval above the third and below the first quartile. n.s., not significant, * p < 0.05, *** p < 0.001. These results were confirmed when repeating the same analyses by using the cells’ place fields rather than the cells’ spike trains (Supplementary Fig. S7iii). Source data are provided as a Source Data file.

Back to article page