Display Settings:

Items per page

Results: 8

1.
Figure 2

Figure 2. L4 neurons spike with precise latencies during object location discrimination. From: Neural coding during active somatosensation revealed using illusory touch.

a. Spike rasters and peri-touch spike histogram for one L4 neuron aligned to first touch (same session as in ).
b. Peri-touch spike histogram averaged across all rapidly touch excited L4 neurons < 250μm from C2 center (13 neurons) for the first touch per trial.

Daniel H. O’Connor, et al. Nat Neurosci. ;16(7):958-965.
2.
Figure 6

Figure 6. Millisecond time scale precise spike latencies are not required for detecting an object at a particular location. From: Neural coding during active somatosensation revealed using illusory touch.

a. Top, ‘Delayed’ photostimulation of L4 neurons was triggered by whisker crossings with varying delays (Δt). Bottom, whisker movements with whisker crossings (red circles) and corresponding photostimuli (cyan circles) for Δt = 50 ms.
b. Fooling Index as a function of the delay between whisker crossing and photostimulation.
c. Fooling Index as a function of azimuthal angle at the time of stimulation.

Daniel H. O’Connor, et al. Nat Neurosci. ;16(7):958-965.
3.
Figure 8

Figure 8. Illusory object location can only be evoked during periods of tactile exploration marked by whisking bouts. From: Neural coding during active somatosensation revealed using illusory touch.

a. Example of whisking bouts in relation to trial start and pole motion. Whisker movement (θ, grey) and whisking amplitude (θamp, black).
b. Eight example trials, showing the time course of whisking (θamp, black) and the corresponding photostimulation pattern (cyan circles). Left, trials in which photostimulation occurred during periods without whisking. Right, trials in which photostimulation occurred during whisking.
c. Fooling Index for ‘whisking’ and ‘not whisking’ trials. Also plotted are interleaved standard virtual pole trials (Δt = 5 ms; as in ). Error bars, SEM.

Daniel H. O’Connor, et al. Nat Neurosci. ;16(7):958-965.
4.
Figure 4

Figure 4. Optogenetic stimulation of L4 neurons mimics touch-evoked spiking. From: Neural coding during active somatosensation revealed using illusory touch.

a. Targeting ChR2 to L4 neurons. Left, ChR2-expression (magenta) in one barrel. Right, genetic scheme.
b. Single example neuron responding to different light intensities. Cyan, photostimulus.
c. Population peri-stimulus time histograms recorded in different cortical layers (n=85 neurons total), following a ChR2 stimulus. Responses are averaged across light intensities (Methods).
d. Overlay of population peri-stimulus time histogram (grey, touch; same data as ; magenta, photostimulation, delayed by 5 ms from stim).
e. Comparison of L4 activity evoked by touch and photostimulation.

Daniel H. O’Connor, et al. Nat Neurosci. ;16(7):958-965.
5.
Figure 5

Figure 5. Closed-loop photostimulation causes illusory perception of object location. From: Neural coding during active somatosensation revealed using illusory touch.

a. Four trial types during a photostimulation behavior session, depending on pole location and photostimulation (cyan lightning bolts). The virtual pole (magenta) was in the θROI. Mice reported object location by licking or not licking.
b. Photostimulation (blue circles) coupled to whisker movement (grey, θ) during object location discrimination. Asterisk, answer lick.
c. Responses in the four trial types across one behavioral session. Green, ‘yes’ responses; yellow, ‘no’ responses.
d. Photostimulation on NO trials (red) in the C2 barrel increases the fraction of yes responses. Blue, YES trials. Error bars, SEM. Lines show individual mice.
e. Fooling Index (defined in d). Black circles, individual mice. Grey circle, first session, averaged across all mice. Error bars, SEM. Grey bar, mean maximum possible Fooling Index.
f. Same experiment as in a–e, without ChR2 expression.
g. Same experiment as in a–e, with ChR2 expression and photostimulation in the E3 barrel.
h. Symmetric response task; both object locations were indicated by licking at one of two lickports (lick left/lick right). Black circles, individual mice. Grey bar, mean maximum possible Fooling Index. Performance of each mouse is different from zero by one-tailed permutation test.

Daniel H. O’Connor, et al. Nat Neurosci. ;16(7):958-965.
6.
Figure 3

Figure 3. Decoding object location and behavioral choice based on L4 spikes. From: Neural coding during active somatosensation revealed using illusory touch.

a. Neural coding of object location. Top, whisker position (θ, grey) and the two pole locations (blue, YES; red, NO). Bottom, schematic spike probability for the two object locations.
b. Spike-triggered θ (for every spike in the exploration window, adjusted by spike latency 9 ms) (YES trials, blue, n = 73; NO trials, red, n = 42; same as in ).
c. Spike count during the exploration window (same data as in ). Dotted line, mean reaction time.
d. L4 neurons discriminate object location equally based on spike count and spike-triggered θ (black circles, individual neurons; black cross, population mean and standard error; p = 0.57 paired two tailed t-test). ‘Discrimination performance’ is the area under the receiver operating characteristic curve for a linear classifier. Dotted lines, chance discrimination performance and equal discrimination performance.
e. L4 neurons discriminate behavioral choice based on spike count better than spike-triggered θ (p = 0.0085).

Daniel H. O’Connor, et al. Nat Neurosci. ;16(7):958-965.
7.
Figure 1

Figure 1. Overview of the experimental system and whisking strategy during object location discrimination. From: Neural coding during active somatosensation revealed using illusory touch.

a. Recordings were made from L4 neurons while mice localized objects with the C2 whisker. Whisker movements were measured with high-speed video. The mouse indicated its decision by licking for a water reward. Whiskers were detected as they crossed a virtual pole (an infrared laser). A real-time system controlled photostimulation of ChR2-positive L4 neurons based on whisker position.
b. Schematic of the object location discrimination task. θ is the azimuthal angle of the whisker at the base.
c. Whisking during object location discrimination (data from one representative session). Left, whisker θ at touch onset (408 touches; YES trials, blue; NO trials, red). Right, distribution of whisker positions during task-related whisking (θamp > 2.5 degrees). Occupancy (in seconds) is the time spent at particular θ (bin size, one degree).
d. Whisker movements (grey; θ) in two example behavioral trials (top, NO trial; bottom, YES trial). Black trace segments correspond to contact periods. Pole entry (grey) and availability (black) indicated by uppermost line. Protraction corresponds to increasing θ. Ticks, spikes; asterisk, lick.
e. The number of contacts per trial for all sessions (36,910 trials; YES trials, blue; NO trials, red). Dots, means.

Daniel H. O’Connor, et al. Nat Neurosci. ;16(7):958-965.
8.
Figure 7

Figure 7. Optogenetic silencing of the C2 column biases behavioral choice toward ‘no’ responses, consistent with spike count coding. From: Neural coding during active somatosensation revealed using illusory touch.

a. Left, silencing the C2 cortical column using ChR2-based stimulation of GABAergic neurons. Right, recordings from putative excitatory neurons under control (black) and photostimulation (yellow) conditions (peristimulus time histogram aligned to first touch; bin size, 2 ms; n = 6 neurons from 4 mice; Methods). The photostimulus (1.4 mW) began approximately 200 ms before first touch.
b. Reducing spike count in the C2 column reduces performance on YES trials, and improves performance on NO trials. Lines, 3 individual mice.
c. Same as b, for a symmetric response version of the object location discrimination task. Lines show data from 3 individual mice, for two photostimulation conditions with 2 mW average power (continuous illumination, dashed lines; illumination with 1 ms pulses at 80 Hz, same average power, solid lines).
d. We speculate that mice monitor spike count within the ensemble of L4 neurons in the C2 column normally activated by contact on YES trials (blue ‘neurons’, upper left corner of the table). The table shows a schematic of the L4 ensemble under the conditions tested in this study (–7). For example, on NO trials, a distinct but overlapping ensemble is activated (red ‘neurons’; the YES trial ensemble is indicated with blue outlines). On photostimulated NO (i.e. virtual pole; cyan) trials, activity is evoked in a subset of the YES ensemble, fooling mice into making ‘yes’ responses. Bottom, hypothetical distribution of decision variable (spike count in YES ensemble) used by mice to decide between a ‘yes’ and a ‘no’ response. Red, NO trials; blue, YES; cyan, NO with virtual pole and photostimulation (, ); yellow dashed lines, silencing. If spike count in the YES ensemble of neurons exceeds a threshold value (the ‘decision boundary’), the mouse makes a ‘yes’ response; otherwise the mouse makes a ‘no’ response.

Daniel H. O’Connor, et al. Nat Neurosci. ;16(7):958-965.

Display Settings:

Items per page

Supplemental Content

Recent activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...
Write to the Help Desk