Abstract
Integrating episodic experiences to drive efficient behavior stems from the ability of the brain to encode and retrieve mnemonic representations of the surrounding environment. The capacity of an organism to maintain and flexibly manipulate these internal representations is crucial for adapting behavior to changing environmental conditions, and there are multiple brain regions that coordinate activity to facilitate the physiological processes that underlie the neural substrates that control behavior. Notably, the hippocampus and prefrontal cortex are two areas that are often associated with memory and have been extensively studied for their role in cognition. The inextricable relationship between these structures positions this circuit to support a wide range of memory-related cognitive processes.
During learning, episodic memories are are thought to be encoded in a labile population of hippocampal neurons that integrate over space and time, and these representations are essential for learning. Similarly, the prefrontal cortex is critical for the consolidation and retrieval of these hippocampally encoded memories, and neural dynamics within this circuit that occur at multiple timescales are hypothesized to serve as a framework for how memory-dependent actions are stored and executed. The focus of this thesis is twofold. Firstly, we present a phenomenological account of how hippocampal patterns of activity, namely theta oscillations and sharp-wave ripples (SWRs), broadly organize activity in the hippocampus and prefrontal cortex during different behavioral states. Secondly, using these findings as a basis, we investigate whether coordinated neural activity at multiple timescales is related to relevant behavioral demands to support ongoing decision making.
We first give an account of how hippocampal theta oscillations mediate processing of spatial information. We find that, in addition to the spatial encoding in hippocampal area CA1, neuronal activity in the prefrontal cortex is organized by hippocampal theta and encodes spatial location that can be further refined by incorporating theta phase information. Furthermore, this spatial encoding is coherent between the hippocampus and prefrontal cortex, indicating a theta-mediated mechanism through which shared spatial information is processed. In the context of hippocampal SWR events, we show that coordinated hippocampal-prefrontal reactivation during the awake state is a more accurate representation of ongoing experience than sleep reactivation, potentially reflecting a process through which memories can be stored and retrieved to guide ongoing behavior. Furthermore, awake reactivation is strongest during the early stages of learning, suggesting a crucial role for this coordinated reactivation during initial learning.
The aforementioned results raise the possibility that these coordination mechanisms may organize underlying neural activity to support ongoing behavior. Indeed, we find that reverse and forward hippocampal replay events during pauses in exploration represent possible past and future trajectories and exhibit distinct gradients over learning, supporting the proposed roles of reverse and forward replay in retrospection and prospection, respectively. In addition, coordinated hippocampal-prefrontal replay distinguished correct past and future paths and could be used to accurately predict correct and incorrect behavioral choices. We further examined hippocampal-prefrontal activity at multiple timescales and have uncovered a relationship between behavioral and fast timescale sequences. Behavioral-timescale sequences during navigation in both the hippocampus and prefrontal cortex maintained representations of the current goal. Importantly, we also show the existence of prefrontal theta sequences and demonstrate that, unlike hippocampal theta sequences, they are able to predict actual choices. During incorrect trials, there was an impaired interaction between behavioral and fast timescale processes, indicating that cooperative interactions at multiple cognitive timescales support memory-guided decision making.
Together, the contents of this thesis investigate the neural mechanisms through which hippocampal-prefrontal coordination may support encoding, consolidation, and retrieval of memories to support decision making. Furthermore, these works provide an extension to what is known about the roles of the hippocampus and prefrontal cortex in cognition, leading to a broader understanding of how interregional coordination may engender actions that lead to effective behavior.