Assessing the Neural Circuits and Behavioral Manifestations of Working Memory in Rats
Date
2023-06-14
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
ORCID
0000-0001-6002-3982
Type
Thesis
Degree Level
Masters
Abstract
Understanding the brain requires us to answer both what the brain does, and how it does it (Niv, 2021). Working memory (WM) is an essential cognitive ability that maintains, and actively manipulates, a finite amount of information to support goal directed behavior. My thesis project consists of a series of experiments that aim to increase our understanding of what brain circuits underly WM, and how behavior can be measured to infer WM in rats. The Nucleus Accumbens (NAc) core is a striatal brain region that integrates cognitive and limbic information to direct attention and behavioral action sequences towards reward-predicting stimuli. Previous work from the Howland group describes both cortical and subcortical brain regions that underly performance on the Odor Span Task (OST), a measure of olfactory working memory capacity (WMC) in rodents; however, the role of the NAc core in mediating performance of the OST is currently unknown. To investigate this role, I assessed how chemical inactivation of the NAc core, and infusion of dopamine (DA) receptor-specific antagonists, impact odor span and foraging behavior. We found that chemical inactivation of the NAc core caused a profound reduction in odor span, indicating impaired WMC. We also observed a profound reduction in odor span when D2, but not D1, receptors were selectively antagonized. Chemical inactivation and DA receptor-specific antagonism had little effect on behavioural measures used to infer locomotion. Together, these results point to a substantial role of D2 receptors of the NAc core in mediating performance of the OST. Recent advances in automated behavioral analysis enable researchers to obtain a detailed and objective record of a diversity of behaviors across species. As automated analysis of a behaviorally complex task such as the OST poses significant technical challenges, I collaboratively worked to quantify behaviors used to infer WM within a spontaneous novelty recognition task. We found that supervised machine learning (SML)-generated behavioural predictions were accurate on much of our dataset, but required some human intervention to correct for sub-optimal predictions. SML-based analysis enabled quantification of rat-stimulus interaction duration, total distance travelled, interaction bout count, and novel stimulus approach latency. This analysis was used to assess the effects of Cannabis smoke exposure on novelty preference to infer WM, where we found that Cannabis smoke exposure impacts novelty preference in a load-dependent, and stimuli- specific manner. To our knowledge, this study is the first demonstration of SML-based behavioral analysis in the context of a spontaneous interaction-based test.
Description
Keywords
Working Memory, Nucleus Accumbens, Cannabis, Machine Learning, Behavior
Citation
Degree
Master of Science (M.Sc.)
Department
Anatomy, Physiology, and Pharmacology
Program
Anatomy, Physiology, and Pharmacology