Cortical circuits are flexible and can change with experience and learning. However, the effects of experience on specific cell types including distinct inhibitory types are not well understood. Here we studied how excitatory and VIP inhibitory cells in layer 2/3 of mouse visual cortex were impacted by visual experience in the context of a behavioral task. Mice learned to perform an image change detection task with a set of eight natural scene images. Subsequently, during 2-photon imaging experiments, mice performed the task with these familiar images and three additional sets of novel images. Familiar images evoked less overall activity in both excitatory and VIP populations, and excitatory cells showed higher selectivity for familiar images. The temporal dynamics of VIP cells differed markedly between novel and familiar images: VIP cells were stimulus-driven for novel images but displayed ramping activity during the inter-stimulus interval for familiar images. Moreover, when a familiar stimulus was omitted, VIP cells showed extended ramping activity until the subsequent image flash. This prominent shift in response dynamics suggests that VIP cells may adopt different modes of processing during familiar versus novel conditions.
HIGHLIGHTS• Experience with natural images in a change detection task reduces overall activity of cortical excitatory and VIP inhibitory cells • Encoding of natural images is sharpened with experience in excitatory neurons • VIP cells are stimulus-driven by novel images but show pre-stimulus ramping for familiar images • VIP cells show strong ramping activity during the omission of an expected stimulus
RESULTS
Visual change detection task with familiar and novel imagesWe trained GCaMP6f-expressing transgenic mice to perform a go/no-go visual change detection task with natural scene stimuli ( Figure 1A). In this task, mice are presented with a continuous stream of flashing stimuli (250 ms image flash separated by 500 ms gray screen; Figure 1B). On 'go' trials, a change in image identity occurs at a time unknown to the animal. To earn water rewards, mice must report the image change by licking a reward spout within a 750 ms response window. False alarms are quantified during 'catch' trials in which the repeated image does not change, but licking behavior is measured in a similar time window ( Figure 1B; see Methods). Behavioral training proceeded through a series of stages: mice first learned task rules with simple full-field gratings of two different orientations with no intervening gray period, followed by introduction of a 500ms gray screen period between grating flashes, and lastly, training of natural scene stimuli ( Figure 1C). During the natural image stage, mice were trained with one set of eight images (image set A) for an extended number of sessions (range = 6-46 sessions with image set A, median = 17 sessions; Figure 1F, Supplemental Figure 1A). Mice viewed each of the eight scenes from the familiar image set an average of 10,350 times prior to the 2-photon imaging stage (range: ...