Limits...
Coupled variability in primary sensory areas and the hippocampus during spontaneous activity

View Article: PubMed Central - PubMed

ABSTRACT

The cerebral cortex is an anatomically divided and functionally specialized structure. It includes distinct areas, which work on different states over time. The structural features of spiking activity in sensory cortices have been characterized during spontaneous and evoked activity. However, the coordination among cortical and sub-cortical neurons during spontaneous activity across different states remains poorly characterized. We addressed this issue by studying the temporal coupling of spiking variability recorded from primary sensory cortices and hippocampus of anesthetized or freely behaving rats. During spontaneous activity, spiking variability was highly correlated across primary cortical sensory areas at both small and large spatial scales, whereas the cortico-hippocampal correlation was modest. This general pattern of spiking variability was observed under urethane anesthesia, as well as during waking, slow-wave sleep and rapid-eye-movement sleep, and was unchanged by novel stimulation. These results support the notion that primary sensory areas are strongly coupled during spontaneous activity.

No MeSH data available.


Related in: MedlinePlus

Coupling variability across primary sensory areas and HP.(a) Overall experimental design with freely behaving rats. (b) (top) Two seconds long samples of raster plots in time intervals with high and low cortical variability during spontaneous pre-exposure. (bottom) Corresponding 2 s long population firing rate plots of the raster plots found on (top) with the same color code after a 100 ms wide Gaussian kernel; the vertical bar corresponds to spikes/neuron/second. (c) Example of the CV of the population activity in each brain area: in S1 (green, n = 16), V1 (blue, n = 22) and HP (orange, n = 13) across different behavioral states; vertical black dashed lines delimit the exposure time period. The shading color code informs the behavioral states throughout the experiment, where each data point was calculated from 10s-long periods of population activity. (d) Scatter plots of pairs of CVs in each brain area shown in (a) segmented by selected behavioral states (WK,SWS and REM); also indicated is the Pearson correlation coefficient, r, between the CVs in each behavioral state. (e) Boxplot of group data for the Pearson correlation coefficient between CVs found in each pair of brain area during all experiments (8 datasets); all pairs had a significant correlation in CV with a significant difference only between S1V1 and S1HP. (f) Example of CV of the population activity segmented by behavioral states. (g) Samples of spiking correlations for 5 min long pre-experience time periods with different levels of cortical variability: (left) local spiking correlation histograms within S1, V1 and HP; (left-top) mean spike correlations for high levels of cortical variability; (left-bottom) mean spike correlations for low levels of cortical variability; (middle) inter-area spiking correlation histograms; (middle-top) mean spike correlations for low levels of cortical variability; (middle-bottom) mean spike correlations for low levels of cortical variability; (right, top and bottom) respective correlation matrices of the data.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC5385523&req=5

f2: Coupling variability across primary sensory areas and HP.(a) Overall experimental design with freely behaving rats. (b) (top) Two seconds long samples of raster plots in time intervals with high and low cortical variability during spontaneous pre-exposure. (bottom) Corresponding 2 s long population firing rate plots of the raster plots found on (top) with the same color code after a 100 ms wide Gaussian kernel; the vertical bar corresponds to spikes/neuron/second. (c) Example of the CV of the population activity in each brain area: in S1 (green, n = 16), V1 (blue, n = 22) and HP (orange, n = 13) across different behavioral states; vertical black dashed lines delimit the exposure time period. The shading color code informs the behavioral states throughout the experiment, where each data point was calculated from 10s-long periods of population activity. (d) Scatter plots of pairs of CVs in each brain area shown in (a) segmented by selected behavioral states (WK,SWS and REM); also indicated is the Pearson correlation coefficient, r, between the CVs in each behavioral state. (e) Boxplot of group data for the Pearson correlation coefficient between CVs found in each pair of brain area during all experiments (8 datasets); all pairs had a significant correlation in CV with a significant difference only between S1V1 and S1HP. (f) Example of CV of the population activity segmented by behavioral states. (g) Samples of spiking correlations for 5 min long pre-experience time periods with different levels of cortical variability: (left) local spiking correlation histograms within S1, V1 and HP; (left-top) mean spike correlations for high levels of cortical variability; (left-bottom) mean spike correlations for low levels of cortical variability; (middle) inter-area spiking correlation histograms; (middle-top) mean spike correlations for low levels of cortical variability; (middle-bottom) mean spike correlations for low levels of cortical variability; (right, top and bottom) respective correlation matrices of the data.

Mentions: Similar to what was observed in urethane-anesthetized rats (Fig. 1), freely behaving rats showed different levels of cortical variability in spontaneous activity associated with different population spiking modes (Fig. 2b, top). Namely, high levels of cortical variability were associated with a high density of populational silences, whereas low levels of cortical variability were associated with a sparse occurrence of populational silences. Additionally, there were large (small) fluctuations in population firing rate (Fig. 2b, bottom) during periods of high (low) levels of variability. Analysis of the CV for S1, V1 and HP during the experimental session reveals a strong Pearson correlation of the CV (normally distributed, p < 0.05) from distinct cortical areas (rS1V1 = 0.8776). Conversely, there was only a modest correlation between HP and either primary sensory areas S1 or V1 (rS1HP = 0.5270 and rV1HP = 0.4826; both p ≪ 0.01, Fig. 2c). Despite the high level of correlation between instantaneous local neocortical network states, levels were clearly different, which suggests that they do not necessarily share levels of variability and their spiking modes but rather share the timing and direction of changes of their local neuronal network states (Fig. 2c and Fig. S1).


Coupled variability in primary sensory areas and the hippocampus during spontaneous activity
Coupling variability across primary sensory areas and HP.(a) Overall experimental design with freely behaving rats. (b) (top) Two seconds long samples of raster plots in time intervals with high and low cortical variability during spontaneous pre-exposure. (bottom) Corresponding 2 s long population firing rate plots of the raster plots found on (top) with the same color code after a 100 ms wide Gaussian kernel; the vertical bar corresponds to spikes/neuron/second. (c) Example of the CV of the population activity in each brain area: in S1 (green, n = 16), V1 (blue, n = 22) and HP (orange, n = 13) across different behavioral states; vertical black dashed lines delimit the exposure time period. The shading color code informs the behavioral states throughout the experiment, where each data point was calculated from 10s-long periods of population activity. (d) Scatter plots of pairs of CVs in each brain area shown in (a) segmented by selected behavioral states (WK,SWS and REM); also indicated is the Pearson correlation coefficient, r, between the CVs in each behavioral state. (e) Boxplot of group data for the Pearson correlation coefficient between CVs found in each pair of brain area during all experiments (8 datasets); all pairs had a significant correlation in CV with a significant difference only between S1V1 and S1HP. (f) Example of CV of the population activity segmented by behavioral states. (g) Samples of spiking correlations for 5 min long pre-experience time periods with different levels of cortical variability: (left) local spiking correlation histograms within S1, V1 and HP; (left-top) mean spike correlations for high levels of cortical variability; (left-bottom) mean spike correlations for low levels of cortical variability; (middle) inter-area spiking correlation histograms; (middle-top) mean spike correlations for low levels of cortical variability; (middle-bottom) mean spike correlations for low levels of cortical variability; (right, top and bottom) respective correlation matrices of the data.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC5385523&req=5

f2: Coupling variability across primary sensory areas and HP.(a) Overall experimental design with freely behaving rats. (b) (top) Two seconds long samples of raster plots in time intervals with high and low cortical variability during spontaneous pre-exposure. (bottom) Corresponding 2 s long population firing rate plots of the raster plots found on (top) with the same color code after a 100 ms wide Gaussian kernel; the vertical bar corresponds to spikes/neuron/second. (c) Example of the CV of the population activity in each brain area: in S1 (green, n = 16), V1 (blue, n = 22) and HP (orange, n = 13) across different behavioral states; vertical black dashed lines delimit the exposure time period. The shading color code informs the behavioral states throughout the experiment, where each data point was calculated from 10s-long periods of population activity. (d) Scatter plots of pairs of CVs in each brain area shown in (a) segmented by selected behavioral states (WK,SWS and REM); also indicated is the Pearson correlation coefficient, r, between the CVs in each behavioral state. (e) Boxplot of group data for the Pearson correlation coefficient between CVs found in each pair of brain area during all experiments (8 datasets); all pairs had a significant correlation in CV with a significant difference only between S1V1 and S1HP. (f) Example of CV of the population activity segmented by behavioral states. (g) Samples of spiking correlations for 5 min long pre-experience time periods with different levels of cortical variability: (left) local spiking correlation histograms within S1, V1 and HP; (left-top) mean spike correlations for high levels of cortical variability; (left-bottom) mean spike correlations for low levels of cortical variability; (middle) inter-area spiking correlation histograms; (middle-top) mean spike correlations for low levels of cortical variability; (middle-bottom) mean spike correlations for low levels of cortical variability; (right, top and bottom) respective correlation matrices of the data.
Mentions: Similar to what was observed in urethane-anesthetized rats (Fig. 1), freely behaving rats showed different levels of cortical variability in spontaneous activity associated with different population spiking modes (Fig. 2b, top). Namely, high levels of cortical variability were associated with a high density of populational silences, whereas low levels of cortical variability were associated with a sparse occurrence of populational silences. Additionally, there were large (small) fluctuations in population firing rate (Fig. 2b, bottom) during periods of high (low) levels of variability. Analysis of the CV for S1, V1 and HP during the experimental session reveals a strong Pearson correlation of the CV (normally distributed, p < 0.05) from distinct cortical areas (rS1V1 = 0.8776). Conversely, there was only a modest correlation between HP and either primary sensory areas S1 or V1 (rS1HP = 0.5270 and rV1HP = 0.4826; both p ≪ 0.01, Fig. 2c). Despite the high level of correlation between instantaneous local neocortical network states, levels were clearly different, which suggests that they do not necessarily share levels of variability and their spiking modes but rather share the timing and direction of changes of their local neuronal network states (Fig. 2c and Fig. S1).

View Article: PubMed Central - PubMed

ABSTRACT

The cerebral cortex is an anatomically divided and functionally specialized structure. It includes distinct areas, which work on different states over time. The structural features of spiking activity in sensory cortices have been characterized during spontaneous and evoked activity. However, the coordination among cortical and sub-cortical neurons during spontaneous activity across different states remains poorly characterized. We addressed this issue by studying the temporal coupling of spiking variability recorded from primary sensory cortices and hippocampus of anesthetized or freely behaving rats. During spontaneous activity, spiking variability was highly correlated across primary cortical sensory areas at both small and large spatial scales, whereas the cortico-hippocampal correlation was modest. This general pattern of spiking variability was observed under urethane anesthesia, as well as during waking, slow-wave sleep and rapid-eye-movement sleep, and was unchanged by novel stimulation. These results support the notion that primary sensory areas are strongly coupled during spontaneous activity.

No MeSH data available.


Related in: MedlinePlus