Stimulus coding over time in the hippocampus and prefrontal cortex
OA Version
Citation
Abstract
The ability to create temporal relationships between stimuli is a critical component for both working and long term memory. Research has shown that the prefrontal cortex and the hippocampus are critical regions supporting the ability to create these “what and when” associations. This dissertation will use two methods to explore the connections between these associations: an analysis of previously recorded data to investigate the properties of time cells, and the development of a novel task to investigate the impact of valence on sequence learning.
Previous work has identified sequences of time cells that fire across a delay in response to a given stimulus. It is unknown how the presentation of subsequent stimuli affects these sequences. To interrogate this question, I utilized previously recorded extracellular data, originally published in Warden and Miller (2010), from the prefrontal cortex in monkeys performing working memory tasks. In these tasks, monkeys were presented with a list of two stimuli that they needed to remember to obtain a reward. Decoding analyses showed that information about the identity of a stimulus persists past the presentation of a second list item. Additionally, information about the time of the stimulus presentation is carried at the population level. Utilizing model fits of individual cells, I identified subpopulations of cells that fire to time, stimulus and time, or list and time. I found that many “pure” time cells fire at similar times in both the first and second list position, but that some fire to what appears to be a conjunction of time and list position. Stimulus specific time cells often fire to a preferred stimulus regardless of which position it is presented. These results show that cells in the prefrontal cortex are able to encode a variety of task dimensions simultaneously.
Next, I developed a task to test the effect of valence on the encoding of “what and when”. Typically, previous work has focused on the ability to associate neutral stimuli with either positive or negative stimuli. However, I sought to simultaneously interrogate positive and negative valences in the same task. In our task, rats were trained to discriminate four tones, of which two predicted a positive outcome, and two predicted a negative outcome. I was able to show that rats are able to learn the task. Learning was initially defined as the ability to perform two sessions in a given day above 80% accuracy, and can also be seen as a significant increase in the proportion of trials that were correct. This task could be used in future work to include cellular recordings and more complex behavioral manipulations, such as interrogating the effect of stress on generalization.
Overall, this work furthers the understanding of how the brain is able to associate stimuli across time to determine “what happened when”.
Description
License
Attribution-NonCommercial-ShareAlike 4.0 International