• chemogenetics
• mice
• top-down processing
• visual perceptual learning

Perceptual Learning (PL) is defined as an experience-dependent enhancement of discrimination abilities occurring after training to a specific stimulus. As an adaptive mechanism, PL improves the organism’s response to the environment by increasing the sensitivity to weak or ambiguous sensory stimuli. Among different types of PL, visual Perceptual Learning (vPL) is one of the most studied and it has been documented in response to a wide range of visual tasks, including discrimination of orientation, motion direction, depth differences, and even face recognition. Even though vPL is supposed to rely on functional rearrangements in brain circuitry occurring at early stages of sensory processing, with a pivotal role for the primary visual cortex (V1), top-down inputs from higher-order visual areas (HVAs) have been suggested to be critically involved. Indeed, HVAs can convey key information on attention, expectation and the precise nature of the perceptual task.
In the present Thesis, I aimed at directly assessing the possibility to modulate vPL by manipulating top-down activity in awake adult mice by using a combination of chemogenetics, behavioural analysis and multichannel electrophysiological assessments. I focused on the latero-medial secondary visual cortex (LM), the prime source for top-down feedback projections re-entering V1. I first evaluated the effects of a global chemogenetic blockade of LM activity, achieved via intracortical injections of a viral vector for the delivery of an inhibitory DREADD that could be selectively activated by an i.p. administration of Clozapine-N-Oxide (CNO). Then, I used a modified version of the visual water box task to train the animals in an operant vPL task. During the vPL protocol, mice were asked to discriminate two vertical gratings with different spatial frequencies that were made progressively closer to each other. Mice were divided into two groups of treatment (i.e., CNO or saline injection), and behaviourally tested 30 min after the i.p. injections. Statistical analysis of the minimum spatial frequency discrimination threshold (MDT) showed a significant impairment in vPL acquisition and retention in CNO-treated mice compared to controls.
Then, I focused on the effects of a selective blockade of top-down projections from LM to V1, using a double-injection strategy to induce the expression of a Cre-dependent inhibitory DREADD delivered into LM by injecting retro-Cre within V1 borders. Afterwards, mice were subjected to the vPL task 30 min after an i.p. administration of CNO. I found a marked deficit in vPL when LM>V1 projections were selectively suppressed. Moreover, V1 electrophysiological recordings after 30 min of CNO injections showed that LM inhibition did not affect the functionality of V1, ruling out the possibility that the observed vPL impairments were due to a decrease of visual acuity.
These results show a critical role of LM top-down projections projecting to V1 in vPL acquisition and retention.