Healthy brain development requires the proper establishment of synaptic connections between neurons. Many synaptogenic molecules have been identified in the mammalian brain, but few synapse-limiting molecules have been discovered. Of the latter, many have canonical functions in the immune system. One such family of molecules, called major histocompatibility class I (MHCI), plays a key role in the innate and adaptive immune systems, where it serves as a ligand for cytotoxic effector cells. MHCI is also found on multiple cell types in the central nervous system. Neuronal MHCI has been found to regulate synapse density, visual system plasticity and synapse removal. Despite the many functions mediated by MHCI, it remains unclear whether MHCI has different functions in regulating the synapses formed onto dendrite or formed by axons of individual neurons, or what binding partners MHCI signals through to enact these functions. In this thesis, I consider the role of MHCI in regulating the number of synapses that a neuron makes onto its targets, and develop a methodological pipeline for identifying binding patterns of MHCI in young cortical neurons.In Chapter 1, I review the current literature of synapse formation and the roles of MHCI in the CNS. In Chapter 2, I describe experiments aimed at testing the sufficiency of H2-Kb and H2-Db to regulate the number of synapses that an axon forms onto dendrites

of neighboring neurons. I find that neither MHCI molecule affects the number of synapses that axons form, but both decrease the synapse density on dendrites of overexpressing neurons. I also report the novel finding that synapse density is increased on dendrites of young cortical neurons in the absence of H2-Kb and H2-Db. In Chapter 3, I describe a proximity biotinylation based approach coupled with tandem mass spectrometry for finding binding partners of MHCI in young cortical neurons, including novel TurboID fusion constructs, functional validation and a protocol for scaled up protein collection from cultured neurons. Chapter 4 describes the implications of my findings from Chapter 2 and alternative ways to study MHCI in the intact brain. I also propose follow-up experiments for the proteomics results in Chapter 3 and their connection to presynaptic roles of MHCI. Finally, I discuss the challenges and exciting advances in the future of MHCI biology in brain development.

Since the early 1900s our understanding of the link between maternal sickness and offspring risk of neurodevelopmental disease has strengthened. Epidemiological findings challenged the idea that the neurological development was protected and separate from immune function and encouraged the development of the neuroimmunology field. It is now widely acknowledged that disorders including autism spectrum disorder (ASD) and schizophrenia (SZ) are characterized in part by peripheral immune dysregulation, often linked to both maternal and paternal alterations in peripheral immune function. Studies of these interactions using animal models have given rise to the theory that maternal immune activation (MIA) acts as a potent disease primer in some individuals. Animal models of MIA have long focused on two key cytokines as the major, if not sole, drivers of the development of neuropathology in exposed offspring—interleukin-6 (IL-6) and interleukin-17 (IL-17). Exposure to an immunogen that precipitates the release of these cytokines, or direct exposure to either alone, may alter the course of brain development when done during critical periods of neurogenesis, microglia development, axonal growth, and angiogenesis.Interestingly, not all at risk pregnancies exposed to various forms of maternal inflammation—infection, autoimmunity, toxicant exposure, or stress—produce offspring who go on to develop neurodevelopmental or neuropsychiatric disorders. Further complicating the issue, many distinct neuropsychiatric and neurodevelopmental disorders can result from ii maternal immune activation alone and despite the clear importance of IL-6, its administration does not produce the complete complement of possible behavioral alterations that MIA is known to elicit. When immunogens such as poly(I:C) and LPS are used, a wider variety of offspring’s behavioral sequelae follow, suggesting that more than one cytokine is involved in the precipitation of the various levels of susceptibility associated with MIA. Despite the explosion of MIA research and the many methods of MIA induction, the precipitating factors that predict resilience and susceptibility to MIA are still unknown. My thesis focusing on discovering these factors. There are six chapters in this dissertation. Chapter 1 is a literature review of maternal immune dysfunction in neurodevelopmental and neuropsychiatric disorders, the mechanisms of immune activity that precipitate alterations in the brain and the known impacts of immune signaling on immune function and development. In Chapter 2, I present experimental evidence in an animal model has shown that function of the maternal immune system before pregnancy is predictive of level of susceptibility and resilience in offspring to disparate clusters of altered behavioral functioning after exposure to maternal immune activation. These experiments were the first to show that IL-6 baseline immunoreactivity (BIR) varies widely in isogenic female mice, and even more surprisingly it is a predictive factor that can stratify the risk levels of offspring. Low or medium pre-pregnancy BIR, in combination with a 30mg/kg dose of poly(I:C), can determine before pregnancy which offspring will be most likely to develop alterations in repetitive behavior, alterations in immune signaling and changes in synapse development after exposure to MIA. These results suggest that utilizing BIR to predict offspring risk could provide a crucial method of increasing reproducibility in a challenging but highly useful environmental model, and in Chapter 3 I fully define the methods for using BIR in MIA studies. In Chapter 4, I describe experiments to fully characterize the behavioral alterations that cluster by maternal BIR and dose of poly(I:C), experiments that fully define the signaling associated with BIR beyond IL- 6, and finally explore the patterns in gestational immune signaling that lead to BIR x dose specific alterations in neuroimmune signaling in embryonic brains. Results of these experiments iii

showed that BIR is not an isolated IL-6 response but reflects an overall pattern of immune signaling that varies widely in isogenic female mice. Together, BIR and dose predicted three main groups of susceptibility to MIA, with the lowest BIR females exposed to a moderate dose of poly(I:C) producing male offspring with heightened fear responses and decreased sociability in both sexes. Medium BIR females exposed to a moderate dose of poly(I:C) produced male offspring with elevated repetitive behavior and decreased sociability and heightened fear responses but resilient female littermates. Both high BIR females exposed to a moderate dose of poly(I:C) and moderate BIR females exposed to a high dose of poly(I:C) produced litters with relatively lower susceptibility, with offspring exhibiting only moderate deficits. Far from being the response to a single cytokine—either IL-6 or IL-17—BIR negatively correlates with the overall immune responses in many cytokines, chemokines and growth factors of dams during pregnancy, with low and medium BIR dams exhibiting more inflammatory skewing than those with high BIRs. Results of these comprehensive experiments suggest that BIR is a measure of overall coherence of immune regulation, and those that respond most before pregnancy may be exhibiting a more balanced and homeostatic immune function that prevails during pregnancy, protecting the fetus from the deleterious effects of MIA. These immunoprotective factors seem to also be partially based on immunogen dose, as medium BIR dams exposed to a higher dose also produced offspring with higher relative resilience to the behavioral effects of MIA and showed less overall inflammatory activity during pregnancy despite higher IL-6 responses. The more susceptible low and medium BIR dams exposed to a moderate dose of poly(I:C) also produced the offspring with generally increased neonatal immune activity in brain tissue after MIA induction, further strengthening the idea that these BIR groups in combination with MIA magnitude are most likely to create a pathological environment for neural development. In Chapter 5 I present the results of pilot experiments aimed at determining whether endocrine factors underlie the development of BIR. These results showed that preliminarily, it appears that BIR may be partially environmentally determined, and these environmental influences may beiv

detectable in endocrine signaling in virgin female mice. Finally, in the Supplement, I present ongoing projects utilizing tissue I generated for three main sets of experiments. The first characterized the influence of BIR on immune signaling in response to ex-vivo stimulation before and during gestation, with a focus on the different cytokine responses and the expansion of cell populations that will soon be assessed with flow cytometry. The second set of experiments will investigate the genetic control of BIR by bulk sequencing RNA expression in virgin female mice from all three BIR groups with and without stimulation, as well as analyzing the differential methylation levels of the regulatory element FOXP3. Finally, the third set of experiments will investigate the differentially expressed genes in offspring exposed to MIA at multiple postnatal time points, with the opportunity to sequence any of six regions (frontal cortex, dorsal striatum, ventral striatum, hippocampus, cerebellum and brainstem) at P14, P70 and P150. These concurrent experiments will continue to develop the picture of how BIR arises, what it predicts in the peripheral immune system of the mothers, and how this leads to disparate responses in the brains of the offspring. Finally, in Chapter 6 I present concluding remarks on the implications and future directions of my research for understanding the role of maternal BIR in predicting risk and resilience to maternal immune activation, and how utilizing this predictive factor could lead to targeted interventions in individuals at risk of developing neurodevelopmental and neuropsychiatric disorders.