Article Figures & Data
Figures
- Figure 1.
Sustained attention task in freely-moving ferrets. A, Illustration of trial sequences of the 5-CSRTT. Each trial begins with illumination of the water spout, which is centrally placed on the back wall of the chamber. The ferret initiates the trial by approaching the water spout, which is equipped with an IR proximity sensor. Then, the spout light is extinguished and the 5-s delay period starts during which the animal is required to sustain attention to the five windows on the front wall of the chamber. A white solid square (stimulus) will randomly present in one of five windows after the delay ends. Nose-poke to the stimulus window during stimulus presentation (2 s) or in the first 2 s after stimulus offset (HP) triggers a tone and delivery of water reward at the spout (correct trial, left). Touch by the ferret is indicated with an asterisk. If the ferret responds before the stimulus (PreTouch), or touches an incorrect window (MissTouch), or fails to respond before the end of the HP (omission), a 6-s time-out (TO) period is introduced where the house light is on and no water is delivered (right). After collecting the reward (8-s time window) or at the end of the TO period, a new trial can be started. B, Representative photographs of one animal during a single trial: a, initiation; b, turn to the stimulus windows (numbered 1 through 5 in subpanel a); c–d, paws on the platform and checking for occurrence of stimulus; e, stimulus on; f, find the stimulus; g, nose poke the stimulus window; h, turn back to collect the reward; i, collect the reward; j, complete trajectory for a single trial obtained from video tracking; k, heatmap of animal locations during the session. Time stamps are shown in the corner of each frame. C, Behavioral performance. Mean accuracy and omission rates across sessions. Error bars: SEM across sessions. D, Mean distance between animal location and stimulus location as a function of time for correct trials. The shorter distances to middle windows (W2 – W4) indicate that animals were centered relative to the stimulus windows before stimulus onset. E, Distribution of touch reaction times (RT) for correct trials. In most trials, the correct window was touched during stimulus presentation (RT < 2 s).
- Figure 2.
Single-unit responses during task performance. A, Representative coronal section; stars indicate the estimated locations of the electrode tips. Electrodes outside of LP/pulvinar (LP/Pulv) were excluded. LGN, lateral geniculate nucleus; MGN, medial geniculate nucleus; PO, nucleus of the posterior group; SGN, suprageniculate nucleus. B, top, Example of high-pass filtered raw trace from representative recording session. Bottom, Action potentials of two representative single units. C, Peri-event raster plots (top) and peri-event histograms of the corresponding firing rates (bottom) of two representative neurons. The unit on the left exhibited an increasing firing during the delay period, whereas the unit on the right did not show such modulation but rather increased its firing rate after stimulus onset. D, Z-score normalized population firing rate of attention-modulated (red, n = 130) and non-attention-modulated units (black, n = 129) in correct trials (solid lines) and in incorrect trials (dotted lines). Attention-modulated units gradually increased their firing rate during the delay period. Both neuron types displayed a transient increase in firing rate after stimulus onset, with stronger responses in attention-modulated neurons. Shaded areas indicate SEM. E, Peri-event raster plots from all the attention-modulated (red, n = 130) and non-attention-modulated units (black, n = 129) in all correct trials. Units from each category are sorted by their mean firing rate in the delay period.
- Figure 3.
Interaction of spiking activity and the LFP in the θ band. A–D, SFC across sessions and animals for different task epochs for attention-modulated (red) and non-attention modulated units (black). A, Before initiation. B, Delay period (from initiation to stimulus onset). C, Before touch (from stimulus onset to touch response). D, After touch. The inset in B shows the STA of the LFP averaged across the attention-modulated (red) and non-attention-modulated units (black), and the shuffle control (green) in the 5-s delay period for all correct trials. The vertical dashed line indicates the spike time. Shaded areas both in the SFC plots and in the STA plot represent SEM. E, SFC of attention-modulated and non-attention-modulated units in θ band (4–6 Hz) for different task epochs. θ Coherence was differentially modulated by task performance. Attention-modulated units showed significantly higher coherence than non-attention-modulated units at all task epochs, except after touch. For attention-modulated units, θ coherence increased during the delay period compared to before trial initiation and reached the maximum before touch. No change in coherence occurred for non-attention-modulated units during the task; *p < 0.05; NS indicates p > 0.05. Error bars represent SEM.
- Figure 4.
Task-modulated θ/γ PAC. A, Modulation of θ power during task performance. θ Power gradually increased and was maintained during the delay period and peaked between stimulus onset and screen touch. Shaded area represents SEM. An example LFP trace (bandpass in 1–20 Hz) is displayed on the top and shows the θ oscillation throughout a trial. B, Heatmap of PAC during task performance shows selective coupling of θ and γ oscillations. An example LFP trace (bandpass in 1–100 Hz) is superimposed on the heatmap and shows γ-θ oscillation nesting. C, Task-dependent modulation of time-resolved θ/γ PAC. θ/γ PAC was elevated throughout the delay period and decreased after screen touch. Shaded area shows SEM. D, Modulation of γ power during task performance. γ Power gradually decreased during the delay period. Shaded area represents SEM.
In this issue
