As we interact with the environment, our senses are constantly bombarded with information. Neurons in the visual cortex have to transform these complex inputs into robust and parsimonious neural codes that effectively guide behavior. The ability of neurons to efficiently convey information is, however, limited by intrinsic and shared variability. Despite this limitation, neurons in primary visual cortex (V1) are able to respond with high fidelity to relevant stimuli. My thesis proposes that high fidelity encoding can be achieved by dynamically increasing trial-to-trial response reliability. In particular, in this thesis, I use the mouse primary visual cortex (V1) as a model to understand how reliable coding arises, and why it is important for visual perception.  Using a combination of novel experimental and computational techniques, my thesis identifies three main factors that can modulate intrinsic variability.

First, using a method to perturb spatial statistics in natural movies, I demonstrate that broadband stimuli, with a power law distribution of spatial frequencies are more reliable processed than narrow-band stimuli. Using high-speed calcium imaging in vivo I demonstrate that these broadband stimuli recruit specific neuronal ensembles, which work together to reduce shared noise and increase reliability.

Next, I developed a novel intersectional genetic strategy to investigate the role of dendrite targeting somatotstatin-expressing (SST) and soma targeting parvalbumin-expressing (PV) inhibitory interneurons in reliable coding. Using a new dual color imaging method and optogenetic manipulations, I identified a novel circuit motif responsible for modulating the fidelity of pyramidal cell firing. Specifically, I demonstrate that SST neurons can reduce variability in pyramidal cells by transiently suppressing PV neurons, via the SSTàPV circuit. These results highlight the importance of finely tuned inhibitory microcircuits in sculpting the temporal fidelity of neural coding.

Finally, to demonstrate the importance of reliable coding in visual perception, I trained mice to perform a natural movie categorization task. In this task, mice had to correctly identify the movies that were more similar to a target movie to gain a water reward. Distractor movies were introduced by perturbing the statistics of the target movie. Optically activating SST neurons in V1 improved the performance of mice. This increase in behavioral performance correlated well with an increase in V1 coding reliability. Thus, increasing the reliability of the neural code increased the ability of mice to discriminate between complex visual stimuli.

Taken together, this work bridges the gap between cells, circuits and behavior, and provides mechanistic insight into how complex visual stimuli are encoded with high fidelity in the visual cortex.

EndFragment