Coordinated neuronal activity plays an important role in information processing and transmission in the brain. With current research predominantly focuses on understanding the properties and functions of neuronal coordination within cortical areas, however, whether coordinated neuronal ensembles (cNEs) are unique to cortical local networks or extend to neuronal populations in other brain regions, such as the auditory thalamus, remains unknown. Additionally, questions persist regarding whether information carried by cNEs is effectively transmitted to downstream areas compared to single neurons. In this study, we use single-unit recordings in female Sprague-Dawley rats to investigate the properties and functions of groups of neurons exhibiting coordinated activity in the auditory thalamus – the medial geniculate body (MGB). We reliably identify cNEs, which are groups of neurons that fire synchronously, in the MGB. We demonstrate that cNEs in the MGB have enhanced information encoding properties and are stable between spontaneous and evoked activity. These MGB cNE properties are similar to what is observed in the cNEs in the primary auditory cortex (A1), suggesting that ensembles serve as a ubiquitous mechanism for organizing local networks, playing a fundamental role in sensory processing within the brain. Furthermore, spikes from MGB neurons synchronized with other cNE members exhibit higher efficacy in driving A1 neuron firing compared to spikes unrelated to cNE activities. This increased efficacy of cNE spikes is target-specific and cell-type specific, rather than resulting in a general increase in the firing rate across all A1 neurons. These findings support the concept of a neuronal ensemble as a functional unit for encoding and transmitting information in the forebrain processing.

Interrogating the neurobiology underlying ASD pathology in traditional animal models is hampered by limited social behaviors in these species. Prairie voles display diverse, robust social behaviors and long-term relationships. Comparative studies suggest that the neural and genetic mechanisms mediating such attachments and complex social behaviors may be conserved across mammals. These prairie voles are therefore a powerful system to understand the biology of neuropsychiatric conditions characterized by the disruption of social behaviors, such as autism spectrum disorder (ASD). To capitalize on this model, we generated prairie voles bearing loss of functional alleles of Scn2a, a high-confidence ASD-associated gene, and characterized the social behaviors of adult heterozygous mutants. Scn2a encodes for the neuronal sodium channel NaV1.2 whose function has been studied well in mouse models. We find here that heterozygous loss of Scn2a in voles results in comparable changes in NaV1.2-dependent aspects of neuronal excitability. We assessed the social dynamics of heterozygous male and female Scn2a mutant prairie voles as compared to wildtype siblings through a battery of behavioral assays to determine how heterozygosity for Scn2a impacted social and attachment behaviors at multiple timepoints throughout pair bonding. We further investigated vocal dynamics of Scn2a-/+ males and females during the initial interactions with novel opposite sex animals. We found that mutants heterozygous for Scn2a display distinct, sex-specific, and context-specific differences in social dynamics across bonding. Scn2a-/+ adult male prairie voles display atypical affiliative behavior with a novel stranger as compared to wildtype siblings, regardless of bonding status. These experiments lay the groundwork to uncover potential biological mechanisms underlying social deficits in genetic models of ASD and determine if the neural or genetic differences observed in mutants persist throughout the lifespan.

What do we study in neuroscience? Recent advances in our understanding of both non-neuronal cells within the nervous system and contemporary machine learning models necessitate a renewed focus on answering this foundational question in sufficient generality to be able to capture a wide array of phenomena whose deep resonances may escape more restrictive starting points. Here, I present two investigations that both tug at tensions within traditional conceptions.

First, I explore calcium activity in the astrocyte network, where tantalizing results have suggested the possibility that these neglected cells carry wide-reaching consequences for the functioning of brains and organisms. The results of these investigations reveal not only that astrocytes' bidirectional interactions with neurons carry a remarkable amount of structure in relation to a core physiological state (sleep), but also suggest that astrocytes may perform sophisticated spatial and temporal integration of their proximal neuronal inputs, implementing a form of combinatorial logic.

Second, I explore the structure of how language models respond when placed in relationship with highly-evocative naturalistic text. In the early stages of the explosion of interest around language models, a common argument levied against them was that, because they were trained to solve the task of next-token prediction (that is, of learning syntax), language models lacked actual understanding of the meaning of the tokens they were predicting (that is, of semantics). The results here—generating a rich and highly-structured portrait of human emotional life, through simple prompt-based impersonation and question-answering geared toward the practical task of hypothesis generation and symptom annotation in psychiatry—add to a growing body of work suggesting that, on the contrary, by solving the problem of generating language syntax, these models do (and, as implied by recent theoretical work, perhaps must) learn a latent structural image of semantics.

Finally, I draw the threads of these investigations together toward an answer of that foundational question: in neuroscience, we study the dynamics of interacting world models. I sketch a framework, the postmodern synthesis, of how the mechanics of these changing minds---when viewed as consisting solely of their relational structure---might be rigorously modeled using the mathematical language of double categories. This framework touches formal resonances with both quantum field theory and classical genetics. By framing our science the other way 'round, from the perspective of the mental content, rather than its material implementation, we may be able to ask new questions.

The prefrontal cortex (PFC) is comprised of a network of excitatory and inhibitory neurons that are critical for a number of cognitive processes including the ability to make decisions. PFC dysfunction causes deficits in these domains and major aspects of psychiatric disorders. Notably, layer V of the PFC is thought to play an important role in regulating these higher order processes and contains heterogeneous subpopulations of pyramidal neurons with distinct morphologies and projections. Layer V of PFC is also the major site of dopaminergic modulation. However, unlike the striatum, where D1 or D2 receptors are known to differentially express in separate projection subtypes of medium spiny neurons, the distribution of D1 and D2Rs in layer V projection neurons within PFC is unclear. Notably, Dopamine D2 receptors (D2Rs) play major roles in both normal and pathological PFC function. Thus, knowing their expression and effects in layer V neurons can provide helpful insights into important mechanisms that underlie prefrontal activity, and ultimately cognitive function. Chapter 1 of this thesis examines the expression of D1Rs and D2Rs in layer V projection subtypes. We also describe a novel mechanism by which D2R activation can drastically enhance the excitability of a D2R-expressing layer V pyramidal subtype. In Chapter 2, we explore differences in how long range excitation and feedforward inhibition is processed in D2R-expressing and D2R-lacking layer V pyramidal neurons. We propose that these differences might confer these subtypes with separate computational properties important for normal PFC function.

The thalamus, a subcortical brain structure involved in key aspects of sensation, perception, cognition, and consciousness, is endowed with intrinsic rhythmogenic properties and extensive reciprocal connections with the cerebral cortex. These reciprocal connections form thalamo-cortico-thalamic circuits which generate, synchronize, and propagate physiological brain rhythms such as sleep spindles. Disruptions in these circuits can cause sleep deficits, and generation of pathological rhythms like seizures. While the canonical thalamo-cortico-thalamic circuit diagram consists entirely of neurons and their synaptic connections, mounting evidence points to the profound impact of astrocytes in dynamically modulating circuit components. Astrocytes are non-neuronal glial cells that play diverse, complex roles in mediating normal neuronal function in the developing and mature brain, at the synaptic and circuit levels, in both health and disease (Chapter 1). Astrocytes are well-positioned to govern crucial aspects of thalamic circuit function, such as excitability and rhythmogenesis, which can be disrupted in neurological disorders such as epilepsy (Chapter 2), but the precise mechanisms by which astrocytes modulate thalamic synapse and circuit function in development and disease are poorly understood. In Chapter 3, we demonstrate the critical role astrocytes play in shaping the developing thalamus of rodents by providing a key immune signal—IL-33 (interleukin-1 family cytokine interleukin-33)—to control the number of excitatory synapses. Its absence can lead to runaway excitation in the thalamus, paving the path for epileptic activity in mice. In Chapter 4, we demonstrate that reactive astrocytes can trigger aberrant synaptic and circuit function in the thalamus, providing strong evidence that they can be drivers, rather than mere bystanders, of injury-related pathological sequelae such as post-traumatic epilepsy and sleep disruption. Furthermore, we identify an aspect of reactive astrocytes—GABA transporter GAT-3 loss of function—sufficient to confer pathological excitability in an otherwise normal thalamic circuit, and one which may underlie thalamic dysfunction observed in cases of brain injuries. In Chapter 5, we discuss our findings which highlight distinct forms of thalamic synapse modulation by astrocytes, as well as common pathways of astrocytic function in brain development and degeneration; we also speculate on the potential role of astrocytes in mediating the distinctive computational properties of the reticular thalamic nucleus. Altogether, our findings add to the growing evidence supporting the critical role of astrocytes in the modulation of thalamic circuits in physiological and pathological settings.

The brain computes and uses uncertainty to guide decision-making. While this is well established for information sensed externally in the form of perceptions, it is less established whether information retrieved from internal storage, in the form of episodic memory, is also treated probabilistically. To test this question, we developed a spatial episodic memory task in which rats gamble their time on a memory choice in each trial, indicating their confidence in its accuracy. We found that rats express higher confidence on correct trials than errors, indicating a degree of self-reflective consciousness thought previously to exist only in humans. We introduce a generative model for episodic memory confidence that predicts the observed patterns of memory confidence.

To investigate the neural correlates of memory confidence, we implanted four rats with triple-site, local field potential (LFP) and single-unit recording devices targeting the orbitofrontal cortex (OFC), nucleus accumbens (NAc), and dorsal hippocampus. To perform these surgeries, we developed a novel method for the implantation of thin-film polymer electrode arrays through the dura mater. We demonstrate that this technology can yield long-term, high quality single unit and LFP recordings.

To investigate the neural activity in these three regions as it may relate to memory confidence, we took as a starting point the decades-old observation that the hippocampus is required for memory, and the more recent finding that hippocampal neurons store and send information about past experience to the rest of the brain. In particular, a hippocampal neural activity pattern known as the sharp wave-ripple (SWR) is an LFP event associated with highly synchronous neural firing in the hippocampus and modulation of neural activity in distributed brain regions. A growing body of evidence indicates that SWRs support both memory consolidation and memory retrieval. This work is summarized in a synthetic review that introduces the perspective that the SWR may mediate the retrieval of stored representations that can be utilized immediately by downstream circuits in decision-making, planning, recollection, or a confidence evaluation, while simultaneously initiating memory consolidation processes. Finally, a proof-of-concept study of SWR function in the episodic memory task is presented.

The prefrontal cortex (PFC) functions to integrate information from the internal and external world in order to flexibly guide behavior. As the most recently evolved brain region, the PFC is implicated in a range of psychiatric disorders such as depression, anxiety, and schizophrenia. The circuitry within the PFC consists of a diversity of excitatory and inhibitory neurons and each neuronal subtype represents a unique locus for action by neuromodulators, such as dopamine and serotonin. The serotonergic system is heavily implicated in mood and emotion and is the target of many modern psychiatric drugs. Despite this, we have a limited understanding of the actions of serotonin on the diverse cell types in the PFC. In this dissertation, I describe the actions of serotonin (5HT) on interneurons and excitatory inputs arriving from discrete brain regions and suggest how this may relate to behavior and disease. I find that 5HT increases the excitability of fast-spiking interneurons but not somatostatin-expressing interneurons by reducing conductance through potassium channels, leading to enhanced summation of gamma frequency inputs. Furthermore, this causes an increase in gamma-frequency inhibitory events in downstream pyramidal cells, a finding which may contribute to my observation that serotonergic signaling reduces low frequency oscillatory power without changing in gamma power in vivo. I also describe how serotonin may act at multiple loci within the prefrontal circuit to reduce anxiety behavior. By both presynaptically suppressing ventral hippocampal inputs and increasing inhibition, 5HT may reduce theta oscillations and attenuate anxiety behavior. These findings may have wide implications for our understanding of psychiatric disorders.