The impact of bilateral ongoing activity on evoked responses in mouse cortex

  1. Daisuke Shimaoka
  2. Nicholas A Steinmetz
  3. Kenneth D Harris
  4. Matteo Carandini  Is a corresponding author
  1. University College London, United Kingdom
10 figures and 1 additional file

Figures

Figure 1 with 2 supplements
Ongoing cortical activity is largely bilateral.

(A) Location of somatosensory and motor areas based on the anatomical database by Allen Brain Atlas. The background image indicate a raw fluorescence signal obtained from voltage imaging. …

https://doi.org/10.7554/eLife.43533.002
Figure 1—figure supplement 1
Pre-processing of the imaging signals has no effect on bilaterality.

(A) Singular Value Decomposition efficiently compresses calcium-imaging data. Example GCaMP traces in V1, using 5 numbers of SVD components, from 50 to 2000 components. The resulting five traces …

https://doi.org/10.7554/eLife.43533.003
Figure 1—video 1
Ongoing cortical activity is largely bilateral.
https://doi.org/10.7554/eLife.43533.004
Figure 2 with 1 supplement
Bilateral fluctuations differ across areas.

(A) Seed-area maps of correlation from an example animal expressing a voltage indicator. Correlation was computed against the average signal within the red square to signals of the all the other …

https://doi.org/10.7554/eLife.43533.005
Figure 2—figure supplement 1
Bilateral correlation during the task and during the passive condition.

This analysis was performed on six mice where we had measured both active and passive condition, and where all six visual areas were identified functionally. Two of these mice expressed VSFP (orange)…

https://doi.org/10.7554/eLife.43533.006
Figure 3 with 1 supplement
Bilateral ongoing activity continues during unilateral visual responses.

(A) Three example sequences of voltage signal during the period after presentation of 50% contrast to the left monocular visual field in an example mouse expressing VSFP. In each examples, the …

https://doi.org/10.7554/eLife.43533.007
Figure 3—video 1
Bilateral ongoing activity continues during unilateral visual responses.
https://doi.org/10.7554/eLife.43533.008
Task stimuli evoke unilateral responses and do not quench bilateral fluctuations.

(A) Mean calcium signal across trials from seven animals in the right stimulated V1 (red) and in the left unstimulated V1 (black). Time period of 50–150 ms after stimulus onset (shaded area) was …

https://doi.org/10.7554/eLife.43533.009
Figure 5 with 4 supplements
Bilateral ongoing activity and visually evoked responses add linearly.

(A) Single trial traces aligned by the onset of visual stimulation of 100% contrast to the left visual field in the left unstimulated PM and right stimulated PM. The time period (50–150 ms) used for …

https://doi.org/10.7554/eLife.43533.010
Figure 5—figure supplement 1
Bilateral ongoing activity and visual responses interact additively regardless of the similarity of their patterns.

(A) Template obtained as average visual response to left and right 100% contrast stimuli obtained from one animal. (B) Example single-trial responses to after presentation of 100% contrast the the …

https://doi.org/10.7554/eLife.43533.011
Figure 5—figure supplement 2
The effects of bilateral ongoing activity were due to fast changes in cortical activation, not to slow changes in cortical state.

(A) Instantaneous amplitude at unstimulated PM (vertical axis) and pre-stimulus fluctuation amplitude between 2–7 Hz at stimulated PM (horizontal axis) of single trials in area PM. Each dot …

https://doi.org/10.7554/eLife.43533.012
Figure 5—figure supplement 3
The effects of bilateral ongoing activity were due to fast changes in cortical activation, not to slow changes in pupil diameter.

Same as Figure 5—figure supplement 2, except that trials are divided by prestimulus pupil diameter instead of cortical state.

https://doi.org/10.7554/eLife.43533.013
Figure 5—figure supplement 4
Ongoing activity sampled as prestimulus activity.

(A) Cross-correlation between activity in left and right PM, in individual mice (colored traces) and across all seven mice (black trace). (B) Activities 50–150 ms after visual stimulation of 100% …

https://doi.org/10.7554/eLife.43533.014
Figure 6 with 1 supplement
Bilateral ongoing activity does not affect perceptual reports.

(A) Probability of turning the wheel to the correct direction as a function of stimulus contrast in one animal. The probability is computed separately in trial groups of low (blue) and high (cyan) …

https://doi.org/10.7554/eLife.43533.015
Figure 6—figure supplement 1
Bilateral ongoing motif does not affect perceptual reports.

(A) Probability of turning the wheel to the correct direction as a function of stimulus contrast in one animal. The probability is computed separately in trial groups of negative (dark green) and …

https://doi.org/10.7554/eLife.43533.016
Figure 7 with 1 supplement
Additive model accounts most trial-by-trial variations in visual responses.

(A) Schematic representation of the additive model (Equation 1). (B) Single-trial PM traces predicted by the additive model. (B1) bilateral activity, estimated from left PM. (B2) visually evoked …

https://doi.org/10.7554/eLife.43533.017
Figure 7—figure supplement 1
Variance explained by the additive model in different visual areas.

(A) Variance of activity in each visual area explained by the additive model, as a function of distance from location of the corresponding area in the other hemisphere. Gray curves indicate results …

https://doi.org/10.7554/eLife.43533.018
Author response image 1
Singular Value Decomposition efficiently compresses calcium-imaging data.

A few tens of components are sufficient to capture close to 100% of the variance, both in the calcium-sensitive channel (blue) and in the calcium-independent channel (red). In the paper, we use 500 …

https://doi.org/10.7554/eLife.43533.021
Author response image 2
Example GCaMP traces in V1, using 5 numbers of SVD components, from 50 to 2,000 components.

The resulting five traces look identical. The number of components used in the paper is 500.

https://doi.org/10.7554/eLife.43533.022
Author response image 3
Bandpass filtering the signals does not affect the estimate of bilateral activity.

Left: Results obtained after taking the derivative and lowpass filtering (as used in our paper). Right: Results obtained without any filtering. By definition, filtering affects the power spectral …

https://doi.org/10.7554/eLife.43533.023

Additional files

Download links