Skip to main content
. 2020 Mar 13;9:e51972. doi: 10.7554/eLife.51972

Figure 5. Neocortical activation latency relative to SWRs spans a wide spectrum of negative to positive values.

(A) (i) Schematic of Asymmetry Index (AI) calculation. In this figure, AI was calculated for individual peri-SWR retrosplenial cortex (RSC) traces and called RSC AI. RSC AI values were used to quantify the latency of neocortical activation relative to SWR timestamps. (ii) The proportion of negative RSC AI values across animals (n = 14). Magenta dashed line indicates the chance level (50%). 55% (median, indicated by a red line) of peri-SWR RSC activity across animals have negative AI, meaning that on average, neocortical tendency to activate prior to hippocampal SWRs is greater than chance (n = 14; one-sided one-sample Wilcoxon signed-rank test; the median is greater than 0.5 with p=1.831 × 10−4). (iii) Representative peri-SWR RSC activity sorted by AI calculated for each individual peri-SWR RSC trace. Color bar represents z-scored iGluSnFR signal. The black trace shows the mean peri-SWR RSC iGluSnFR signal. Note that the chance of neocortical activation preceding SWRs is higher than following them. (B) (i) Distribution of RSC AI values for a representative animal. Dashed lines indicate the first (Q1) and third quartiles (Q3). The SWRs for which the associated RSC AI values are less and greater than Q1 and Q3 are called RSC QR1 and QR4, respectively. RSC QR2-3 consists of all other SWRs. (ii) Example plots of mean peri-SWR RSC iGluSnFR signal associated with RSC QR1 (blue) and QR4 (red). The thickness of the shading around each plot indicates SEM. Note that the activity associated with RSC QR1 and QR4 peak before and after SWR centers, respectively. (iii) Time course of exemplar mean peri-SWR hippocampal MUA associated with SWRs in RSC QR1 (blue) and RSC QR4 (red). Notice that both hippocampal MUA activity and RSC iGluSnFR activity are negatively (negative AI) and positively (positive AI) skewed for RSC QR1 and QR4, respectively. (iv) Distributions of hippocampal MUA AI values for SWRs in RSC QR1 (blue) and QR4 (red) in a representative animal. The vertical line represents the median of each distribution. (v) Summary of median values calculated in (iv) across all animals (n = 14, one-sided paired Wilcoxon signed-rank test; RSC QR1 versus RSC QR4 p=6.103 × 10−5). This figure has three figure supplement.

Figure 5—source data 1. Comparing HPC MUA AI median values across RSC QR1 and RSC QR4.

Figure 5.

Figure 5—figure supplement 1. Asymmetry Index is a more robust measure of activation latency compared to peak activation timestamp.

Figure 5—figure supplement 1.

(A) Peri-SWR RSC activity for individual SWRs sorted by peak timestamps detected in (−200 ms, 200 ms) interval. Peak timestamps were calculated for each individual peri-SWR RSC trace. Color bar represents the z-scored iGluSnFR signal. The black trace shows the mean peri-SWR RSC iGluSnFR trace. (B) (i-iii) Asymmetry index (AI), calculated for three representative traces of peri-SWR RSC activity patterns, is shown to demonstrate that the peak time may not accurately reflect the timing of activation relative to SWRs. In these examples, the peak timestamp (magenta lines) for the first two and for the third examples occur before and after the SWR time (0 s), respectively, suggesting that the corresponding SWRs have occurred in the late and early phases of RSC activation, respectively. However, the RSC AI values for these three traces do not match with this conclusion. Instead, the RSC AI values suggest that their corresponding SWRs have occurred around the middle phase of RSC activation for the first (AI = 0.0385) and third (AI = 0.0262) time traces because their corresponding AI values are close to zero, and at the early phase of RSC activation for the second (AI = 0.1067) time trace because its corresponding AI value is positive and away from zero.
Figure 5—figure supplement 2. Neocortical activation latency relative to SWRs is similar under sleep/urethane anesthesia and VSD/iGluSnFR imaging conditions.

Figure 5—figure supplement 2.

(A) There is no statistically significant difference in the proportion of negative RSC AI values in VSD versus iGluSnFR (i) nor in sleep versus urethane (ii) conditions. (B) The comparison of the distribution of medians of HPC MUA AI values across RSC QR1 and RSC QR4 in VSD versus iGluSnFR (i) and in sleep versus urethane (ii) conditions. A statistically significant difference was observed in comparing the HPC MUA AI median values for RSC QR4 in VSD versus iGluSnFR.
Figure 5—figure supplement 3. A delayed neocortical activation after SWRs in RSC QR1 was not observed.

Figure 5—figure supplement 3.

The mean of the peri-SWR RSC activity for SWRs in the RSC QR1 (blue curve shown in Figure 5Bii) were calculated for each animal and then the resultant signals were averaged across urethane-anesthetized (i) and sleeping (ii) animals separately.