Autism spectrum disorders (ASDs) are a heterogeneous class of neurodevelopmental disorders that share the three core behavioral symptoms of impaired social interactions, communication deficits, and restrictive and repetitive behaviors. Our previous studies identified the novel X-linked ASD gene NEXMIF (KIAA2022, KIDLIA). Mutations in NEXMIF leading to loss of its protein product are responsible for the development of autistic features and intellectual disability in humans. However, the role NEXMIF plays in brain development and ASD remains largely unknown. Therefore, I investigated the behavioral impairments and cellular and molecular dysregulation that result from loss of NEXMIF in a transgenic mouse model. I found that male NEXMIF-/y hemizygous knockout (KO) mice replicate the behavioral alterations reported in affected humans and that cultured neurons from NEXMIF-/y KO brains show a significant decrease in neurite outgrowth, synaptic protein expression, and spine and synapse density. Loss of NEXMIF in cultured neurons also leads to altered expression of many genes including several involved in synaptic development and function. Reintroduction of some of the downregulated genes in cultured neurons rescued the decreased spine density and synaptic AMPAR levels observed from loss of NEXMIF. Several clinical reports have indicated that in females, haploinsufficiency of the X-linked NEXMIF gene causes symptoms similar to those observed in males lacking NEXMIF. Therefore, I examined the behavioral and molecular phenotypes in a transgenic mouse model of NEXMIF haploinsufficiency, female NEXMIF+/- mice. These animals displayed ASD-like behaviors, including impaired social interactions, repetitive self-grooming, and memory deficits. NEXMIF haploinsufficiency results in mosaic expression of the protein, resulting in two populations of neurons in the brain, those that express NEXMIF and those that do not. Interestingly, I found that both types of neurons demonstrated impairments in dendritic outgrowth, synaptic density, and the expression of important synaptic proteins. Together, these findings provide new insights into the cellular and molecular mechanisms of NEXMIF-dependent ASD and the role of NEXMIF in neurodevelopment, in both males and females.

The hippocampal CA1 is crucial for myriad types of learning and memory. It is theorized to provide a spatiotemporal framework for the encoding of relevant information during learning, allowing an individual to create a cognitive map of its environment and experiences. To probe CA1 network dynamics that underlie such complex cognitive function, in this work we used recently developed cellular optical imaging techniques that provide high spatial and temporal resolutions. Genetically-encoded calcium indicators offer the ability to record intracellular calcium dynamics, a proxy of neural activity, from hundreds of cells in behaving animals with single cell resolution in genetically-defined cell types. In complement, recently developed genetically-encoded voltage indicators have enabled direct recording of transmembrane voltage of individual genetically-defined cells in behaving animals. The work presented here uses the genetically-encoded calcium indicator GCaMP6f and the genetically-encoded voltage indicator SomArchon to interrogate the activities of individual hippocampal CA1 neurons and their relationship to the dynamics of the broader network during behavior. First, we provide the first in vivo, real-time evidence that two unique populations of CA1 cells encode trace conditioning and extinction learning, two distinct phases of hippocampal-dependent learning. The population of cells responsible for the representation of extinction learning emerges within one session of extinction training. Second, we perform calcium imaging in a mouse model containing a total knockout of NEXMIF, a gene causative of autism spectrum disorder. We reveal that loss of NEXMIF causes over-synchronization of the CA1 circuit, particularly during locomotion, impairing the information encoding capacity of the network. Finally, we conduct voltage imaging of CA1 pyramidal cells and parvalbumin (PV)-positive interneurons, with simultaneous recording of local field potential (LFP), to characterize how cellular-level membrane dynamics and spiking relate to network-level LFP. We demonstrate that in PV neurons, membrane potential oscillations in the theta frequency range show consistent synchrony with LFP theta oscillations and organize spike timing of the PV population relative to LFP theta, indicating that PV interneurons orchestrate theta rhythmicity in the CA1 network. In summary, this dissertation utilizes genetically-encoded optical reporters of neural activity, providing critical insights into the function of the CA1 as a flexible, diverse network of individual neurons.