Spatial representations of place cells in darkness are supported by path integration and border information

Abstract

Effective spatial navigation is enabled by reliable reference cues that derive from sensory information from the external environment, as well as from internal sources such as the vestibular system. The integration of information from these sources enables dead reckoning in the form of path integration. Navigation in the dark is associated with the accumulation of errors in terms of perception of allocentric position and this may relate to error accumulation in path integration. We assessed this by recording from place cells in the dark under circumstances where spatial sensory cues were suppressed. Spatial information content, spatial coherence, place field size, and peak and infield firing rates decreased whereas sparsity increased following exploration in the dark compared to the light. Nonetheless it was observed that place field stability in darkness was sustained by border information in a subset of place cells. To examine the impact of encountering the environment's border on navigation, we analyzed the trajectory and spiking data gathered during navigation in the dark. Our data suggest that although error accumulation in path integration drives place field drift in darkness, under circumstances where border contact is possible, this information is integrated to enable retention of spatial representations.

Keywords: CA1; hippocampus; place cells; sensory.

Figure 1

Overview of the experimental paradigm. Recording trials for the spatial drift paradigm are illustrated. Animals were habituated to the familiar circular arena one day before testing to make sure that stable place fields were established based on the visually perceived environment. On the experimental day, animals were first allowed to explore the circular arena under illuminated conditions. Recordings were initiated 5 min after entry and lasted for 15 min in the light (S1). In the second trial, animals re-entered in the arena in under illuminated conditions. The lights were turned off after 5 min and recordings were conducted for 15 min (S2). The condition for the last trial was the same as the first trial (S3). After each recording trial, animals were removed from the arena and the floor was cleaned.

Figure 2

Firing rate maps of 51 place cells in the spatial drift paradigm reveal changes in place cell firing. Firing rate maps of 51 place cells across the spatial drift paradigm are illustrated. Recordings were conducted in a circular arena. Animals were allowed to explore the arena in the sequence of light-dark-light. Three firing rate maps were shown for each cell and were marked with different trial numbers. Trials S1 were recorded under illuminated conditions. Trials S2 were recorded in the dark. Trials S3 were recorded under illuminated conditions again. Before each recording trial started, animals always re-entered into the arena under illuminated conditions and were allowed to explore the arena for 5 min. Firing rate maps in S3 are not shown for five cells (C20, C21, C22, C23, and C51) due to insufficient movement of the animals during the recording period.

Figure 3

A change from light to darkness reveals distinct firing rates in place cells. Firing rate maps of 3 cells in the spatial drift paradigm are illustrated. Recordings were conducted in a circular arena. For each cell, three firing rate maps are shown with regard to three recording conditions: light-dark-light, respectively. Peak firing rates are indicated at the upper right region of each rate map. Analogs of spike waveforms of corresponding cells are illustrated for a time window of 2 ms. Cell C06 showed a strong reduction in firing when the lights were turned off, including reduced peak firing rates and increased firing fields. Cell C31 revealed minor changes upon the same environment change. The changes were shown only by a small number of cells, whereas the pattern of activity shown by C08 was expressed in the majority of place cells recorded. This cell showed small but notable increases in place field sizes and decreases in the peak firing rate.

Figure 4

Characteristics of place fields. (A) Spatial correlations were significantly higher between S1 and S3 (two trials conducted in light conditions) than the other two trial-pairs (Friedman ANOVA + post hoc tests; * p < 0.05). No significant difference was observed between S1-S2 and S2-S3. This suggests that the stability of place fields decreased in the dark. (B) Spatial information was significantly lower in trial S2 (conducted in the dark) than in the other two trials (conducted in the light) (Friedman ANOVA + post hoc tests; * p < 0.05). No significant difference was observed between S1 and S3 (both trials conducted in the light). (C) Sparsity was significantly higher in trial S2 (conducted in the dark) than in the other two trials (conducted in the light) (Friedman ANOVA + post hoc tests; * p < 0.05). No significant difference was observed between S1 and S3. (D) Place field size was significantly higher in trial S2 (conducted in the dark) compared to the other two trials (conducted in the light) (Friedman ANOVA + post hoc tests; * p < 0.05). No significant difference was observed between S1 and S3 (both trials conducted in the light). (E) The average velocity of the animals in each trial did not vary significantly (One Way Repeated Measures ANOVA tests). (F) Spatial coherence was significantly lower in trial S2 (conducted in the dark) than that in the other two trials (conducted in the light). No significant difference was observed between S1 and S3 (Friedman ANOVA + post hoc tests; * p < 0.05). (G–I) Histograms are shown illustrating the distributions of correlation coefficients between trial S1 and S2 (G, light–dark) between trial S2 and S3 (H, dark– light); and between S1 and S3 (I, light–light).

Figure 5

Measures on firing frequency in conditions of light (S1), dark (S2) and light again (S3). (A) Peak firing rate was significantly lower in trial S2 (conducted in the dark) than in the other two trials (conducted in the light). No significant difference was observed between S1 and S3 (Friedman ANOVA + post hoc tests; * p < 0.05). (B) The average firing rates of place cells in each trial are illustrated. No significant difference was observed (Friedman ANOVA test). (C) Infield firing rates of place cells in each trial are illustrated. Infield firing rate was significantly lower in trial S2 (conducted in the dark) than in the other two trials (conducted in the light). No significant difference was observed between S1 and S3 (Friedman ANOVA + post hoc tests; * p < 0.05). (D) Outfield firing rates of place cells in each trial are illustrated. No significant difference was observed (Friedman ANOVA test).

Figure 6

Cross-correlations under conditions of light (S1), dark (S2) and re-exposure to light (S3). Cross-correlation coefficients were significantly lower in the dark (S2) than those conducted in the light, i.e., S1 and S3 (Friedman ANOVA + post hoc tests; * p < 0.05). No significance was observed between two trial conducted in the light. This suggests that the intra-trial stability of place fields was impaired in the dark due to drifts in place cell firing.

Figure 7

Example of place cells illustrating firing patterns from from-border and towards-border segments. Each row depicts the activity of one cell during under illuminated conditions (two leftmost columns) and in darkness (two rightmost columns). The firing activity is further split into from-border (columns no. 1 and 3) and towards-border segments. Activity in both from and towards cases was expected to stay largely the same with lights on and differ once lights are switched off. The second row depicts a cell where this was not the case and the rat was able to successfully detect its position during darkness. In contrast, the third row depicts a cell which features successful self-localization after just having encountered a wall but failing to do so when no wall was encountered for some time.

Figure 8

Statistical analysis of place fields generated from from-border and towards-border trajectory segments. Histograms of the differences between from-border and towards-border labeled trajectory segments during light and darkness (upper row). Note that differences between firing activity maps are measured as the cosine of the angle between two firing maps, and thus lower values indicate a larger difference between maps. The lower row shows the histogram of the same difference value being based on a large set of semi-randomly generated surrogate data instead of the original experimental recordings. The surrogate data was used to find the threshold value (dashed red line in all plots) determining an actual significant difference between two from-border/towards-border activity maps. The lower right plot shows this surrogate data in a higher resolution, while the lower left depicts the same data in the resolution as the histograms of the original data above– where it can be seen that 13 of 51 (25%) cells feature a significant different in their from-border/towards-border firing behavior in darkness, while the same effect seems to be restricted to mere outliers during the lights on condition.