Extended Data Fig. 1: Behavioural and physiological responses in mice undergoing task learning and optogenetic manipulation.
From: Volatile working memory representations crystallize with practice
(a) Performance of the animals while they learned the task over eight days assessed with discriminability index (D’). (b) Normalized licking distribution of novice and expert mice during the 20 second trial period. (c) Odour intensity measured by photoionization detector (PID). (d) Effect of airflow (left), delay period duration (middle), and odour concentration (right) on animal’s task performance. (e) Licking distribution for different trial types in expert animals. (f) Schematic from Paxinos and Franklin44 illustrating the muscimol injection target and a histological section after the fluorescent dye (DiI) injection marking the injection location. (g) Muscimol’s impact on expert mice’s performance and licking, indicating significant performance disruption (paired t-test; 95% CI; performance p = 0.0001; licking p = 0.159). Mice received muscimol or saline randomly across consecutive sessions. Each colour indicates one mouse with two mice being experimented with twice. (h) Schematics and effect of muscimol inhibition on mice in a non-working memory task. (i) Schematic from Paxinos and Franklin44 illustrating the injection target and a histological section after the stGtACR2 showing injection localization. (j) Blue light illumination silences stGtACR2-expressing neurons, as demonstrated by a neuron’s activity raster plot and its averaged firing rates during photostimulation, alongside the collective activity of 20 recorded units before, during, and after inhibition, showing significant changes (paired t-test; before-during p = 0.026; during-after p = 0.020; before-after p = 0.235). The right panel is another example of the effect of blue light on modulating the activity of stGtACR2-expressing neurons showing the activity of 102 electrophysiologically recorded units before, during, and after photoinhibition (paired t-test; before-during p = 1.28 × 10−7; during-after p = 1.01 × 10−6; before-after p = 0.70). (k) Illustration of bilateral optogenetic setup and histology showing stGtACR2-expressing neurons. (l) Optogenetic inhibition during different task epochs reduces expert mice’s task performance (paired t-tests: delay 4, p = 0.0088; delay 5, p = 0.005; second odour, p = 0.0004; choice 1, p < 0.0001). (m) Effect of blue light on the performance (top) and licking (bottom) of expert mice expressing EGFP. The plot titles indicate the epochs at which the illumination was done. Each colour indicates one mouse. (n) Normalized pupil diameter for various trial types in three distinct animals exhibiting expert-level performance. (o) Normalized paw position across different trial types in three distinct animals demonstrating expert-level performance. (p) Distribution of forelimb speed in three individual expert animals across various odour combinations and temporal segments. (q) Photoinhibition of the M2 during the last 2 seconds of the delay period across the first 7 days of training impairs task performance. Performances are calculated over twenty-trial blocks in each session (n = 4 stGtACR2-expressing mice; n = 4 mCherry-expressing mice). (a), (b), (d), (e), (g)–(j), (l)–(o), and (q) Mean ± s.e.m. *p≤ 0.05, **p≤ 0.01, ***p≤ 0.001, ****p≤ 0.0001, ns, not significant. See Methods for details of statistical analyses.
