BME PhD Preliminary Exam Announcement for Om Kolhe (K. Jayant, advisor)

Everyone is invited to attend the public presentation beginning at 1:30 PM.

TITLE: INVESTIGATING LOCAL RECURRENT CIRCUIT MOTIFS DURING MOTOR DECISION MAKING

DATE: November 22, 2024

TIME: 1:30 PM

LOCATION: MRGN 129

COMMITTEE MEMBERS:

Dr. Krishna Jayant (Primary Advisor)

Dr. Edward Bartlett

Dr. Kevin Otto

Dr. Scott Pluta

Dr. Saeed Mohammadi

ABSTRACT:

Working memory is a fundamental feature in goal-directed decision-making, requiring both online maintenance and volitional manipulation of stored information. A number of studies have shown that the secondary motor cortex (M2), a pre-motor region in the prefrontal cortex, is a major hub for storing working memory. The prevailing view of M2 is that superficial layers L2/3 store working memory representations, evidenced by ramping of neural activity. However, it remains unclear if and how layer 5 contributes to this dynamical interaction. Our hypothesis is that stored information in L2/3 is modulated by deep layer 5 (L5) in a top-down fashion to extract task-relevant information. In this study, I aim to investigate how L5 enables efficient representation and manipulation of working memory in L2/3 through the emergence of neural ensembles in M2. I posit that L5 gates and spatiotemporally filters the information stored in L2/3 through propagating electrical activity known as traveling waves, allowing for the emergence of stable neural ensembles during working memory. Additionally, I postulate that reward reinforcement through dopaminergic inputs to this recurrent L2/3-L5 circuit selectively enhances task-relevant activity, thus increasing the dynamic range of M2. My research comprises two aims: investigating how TWs in the local recurrent circuit enable the emergence of stable neural ensembles (Aim 1) and examining dopamine's role in the maintenance of stable representations (Aim 2). To test these hypotheses, I have designed a novel recording platform combining high-resolution electrical recording in 3D volumes of brain tissue, two-photon imaging in head-fixed behaving mice, and electrochemical recordings of released neurotransmitters like dopamine. Successful completion of these aims will provide new insights into how local recurrent circuits store and compute information, and the role of dopamine in these processes. The outcomes will improve our understanding of cognitive processes like working memory and selective attention while potentially informing the design of large language transformer network models.