<p> The construction and refinement of early neuronal circuits is fundamentally relevant to adult brain function, developmental disorders, and learning and plasticity, but only a small part of this process is well understood (Cang and Feldheim, 2013; Ebert and Greenberg, 2013). A significant and long-standing debate in the field concerns the relative contributions of "hardwired" genetic and molecular determinants versus "plastic" environment and activity-driven alterations (Cline, 2003). While circuits are largely hardwired in most insect and invertebrate species (Hiesinger, 2006), the mammalian nervous system appears to rely on a combination of early molecular cues and later periods of activity-driven plasticity to refine circuits. An interesting middle ground in this process is a period during which circuits have begun to form and propagate activity, but do not yet function in an adult state (Huberman, 2008; Kirkby, 2013; Wong, 1999). At this time, multiple sensory systems are believed to experience "spontaneous" activity patterns that may help to refine circuits, but the form, relevance, and even existence of this activity is under debate (Cang and Feldheim, 2013). Similarly, relatively little is known about the molecules that drive these later stages of synapse and neuronal arbor formation, and the relation that they might have to available activity patterns (Feldheim and O'Leary, 2010). </p><p> In this thesis, I first describe experiments using <i>in vivo</i> calcium imaging in both retinal ganglion cell (RGC) axons and visual midbrain and cortex neuronal cell bodies that confirm the existence of patterned spontaneous activity ("retinal waves") throughout the early postnatal mouse visual system. In a second series of experiments with a genetic knockout believed to disrupt retinal waves <i>in vivo</i>, I find that both the frequency and pattern of waves are drastically disrupted in krrodcout mice relative to controls. Interestingly, I also find that downstream "spontaneous" activity patterns are "de-coupled" from retinal wave activity in this knockout. Conditional knockout experiments further revealed that retinal waves are required in a region-specific manner to drive circuit refinement, thus confirming the necessity of spontaneous activity to development in the visual system. Subsequent rescue experiments demonstrated that the properties of retinal waves are differentially relevant to separate visual circuits, implying that normal wave activity is likely optimized to refine multiple circuits concurrently. A final set of experiments was designed to investigate the role of Down syndrome cell adhesion molecule (DSCAM) in visual circuit development at ages when activity-dependent refinement is now believed to predominate. These results revealed that DSCAM knockout disrupts multiple visual circuits at a surprisingly late age, and in surprising ways. A striking "barrel" phenotype in the retinotopic map of germline and retina-specific conditional knockout mice strongly implies that loss of this cell-adhesion molecule can act on both axon-specific and non cell-autonomous levels during later ages when axon, synapse, and dendrite elaboration is currently believed to be primarily driven by spontaneous activity. </p><p> Together, these results depict visual system development as a process that initially relies on graded molecular cues to establish rough circuit guidelines, and then uses finely tuned patterns of spontaneous activity along with synaptic adhesion molecules to induce synapse and arbor elaboration and refinement. While there are likely a great deal of redundant and homeostatic mechanisms to ensure correct formation of such fundamental circuits, the defects induced by our manipulations were highly penetrant and often persisted into late adulthood, implying the existence of critical periods that will likely prove relevant to future studies of plasticity and developmental disorders. Overall, this work describes a system that relies on a complex synchronization of all available information to ensure the correct development of evolutionarily-relevant circuits within a short period of time.</p>

<p> Visual information is relayed from retina to the brain at first order synapses within the lateral geniculate nucleus (dLGN). During development, activity-dependent synaptic competition drives the segregation of retinal ganglion cell terminals into eye-specific zones. It has been assumed that the gain of synaptic transmission within these eye-specific zones is equivalent, providing uniform information transfer from the periphery to the CNS. Here, we revise this understanding. First, we demonstrate that anatomical segregation of retinal axons triggers a profound (200&ndash;300%) potentiation of neurotransmitter release selectively within the projection zone of the ipsilateral eye. Second, optogenetic recruitment of genetically defined axons within the ipsilateral projection zone provides evidence that functional synaptic connectivity is sub-stratified within the ipsilateral dLGN. Thus, we define a new functional organization within the dLGN and propose that synaptic competition acts as a developmental timer that triggers respecification of set point synaptic gain within the ipsilateral dLGN.</p><p>

Locomotion in vertebrates depends on proper formation and maintenance of neuronal networks in the hind-brain and spinal cord. Malformation or loss of factors required for proper maintenance of these networks can lead to severe neurodegenerative diseases limiting or preventing locomotion. A powerful tool to investigate the genetic and cellular requirements for development and/or maintenance of these networks is a collection of zebrafish mutants with defects in motility. The zebrafish mutant quetschkommode ( que) harbors a previously unknown gene defect leading to abnormal locomotor behavior. Here I show that the que mutants display a seizure-like behavior starting around four days post fertilization (dpf) that is characterized by a lack of an initial high amplitude body bend (C-bend) and simultaneous contra-lateral contractions leading to a seizure-like phenotype and paralysis. Peripheral nerve recordings show a significant increase in the number of initiated swimming bouts and overlap between left and right motor neuron activity. These data suggest that the que mutation leads to defects in nervous system function, at the level of motor neurons or central control of motor neurons. I have genetically mapped the que locus to a 0.36cM interval on chromosome 22 using meiotic mapping. I identified a splice mutation in the gene `dihydrolipoamide branched-chain transacylase E2' (dbt) as defective in que mutants. An orthologous mutation in humans lead to Maple Syrup Urine Disease (MSUD), a devastating metabolic disorder leading to seizures, mental retardation, coma and neonatal death if untreated. In zebrafish, dbt is expressed throughout early development and dbt transcripts become enriched in the hind-brain as well as in the gut and liver by 96 hpf. In MSUD patients levels of branched chain amino acids (BCAA) and their keto acids are significantly increased due to the essential role of the dbt enzyme for the BCAA metabolic pathway. The que mutation causes a significant increase of branched chain amino acids in the zebrafish mutant and a strong decrease of neurotransmitters such as glutamate and GABA as well as precursors like glutamine. I hypothesize that reduced neurotransmitter levels in que lead to the observed motility phenotype. Consistent with this hypothesis, I show a tissue specific reduction of glutamate in the hind-brain and spinal cord of que mutants. To evaluate the que mutant's potential as a vertebrate model for MSUD I performed a pilot drug screen using a selection of metabolites of the pathway as well as diet additives currently evaluated in clinical trials. Conversely, application of phenylbutyrate, one of the diet additives, had a beneficial influence on swimming abilities of que mutant embryos, while the keto acid α-ketoisocaproate (KIC), one of the elevated keto acids in human patients, decreased the percentage of larvae capable of swimming. These results help establish the zebrafish que mutant as a new model for MSUD disease that can be used to further the understanding of this disorder and to help identify therapeutic agents.

<p> Mitochondrial metabolism of reactive oxygen species (ROS) is tightly regulated during brain development. Imbalance has been correlated to neuropsychiatric disorders. Nevertheless, the contribution of ROS accumulation to aberrant cortical circuit organization and function remains unknown. Individuals with 22q11 deletion syndrome (22q11DS) are highly susceptible to psychiatric disorders; therefore, 22q11DS has been suggested as a model for studying the neurodevelopmental origins of these disorders. Six genes &ndash;<i>Mrpl40, Tango2, Prodh, Zdhhc8, Txnrd2</i> and <i>Scl25a1</i>&ndash; located in the 22q11DS commonly deleted region encode proteins that localize to mitochondria. This project aimed to characterize the effects of altered mitochondrial function, due to diminished dosage of these genes, on cortical projection neuron development, using the <i>LgDel</i> mouse model of 22q11DS. I found growth deficits in <i>LgDel</i> neurons that are due to increased mitochondrial ROS and are <i>Txnrd2</i>-dependent. Antioxidant treatment, by n-acetyl cysteine (NAC), rescues neuronal morphogenesis in <i>LgDel</i> and <i> Txnrd2</i>-depleted neurons <i>in vitro</i> and <i>in vivo.</i> Electroporation of <i>Txnrd2</i> restores ROS levels and normal dendritic and axonal growth. <i>Txnrd2</i>-dependent redox regulation underlies a key aspect of cortical circuit differentiation in a mouse model of 22q11DS. These studies define the effects of mitochondrial accumulation of ROS on neuronal integrity, and establish the role of altered pyramidal neuron differentiation in the formation of circuits in 22q11DS. These data provide novel insight into the role of redox imbalance in aberrant development of cortical circuits.</p><p>

The major goals of my research were to characterize the hypothalamic neural progenitors and to understand how Hh/Gli signaling plays a role in regulating cell proliferation in the hypothalamic neurogenic zone. In contrast to mammals, the zebrafish brain has a life-long potential to grow continuously. Thus, for comparative neurogenesis studies, zebrafish become an indispensible model organism to understand adult neurogenesis and regulatory signaling pathways. Identification of the regulatory mechanisms underlying the controlled cell proliferation in adult zebrafish brain will pave the way to manipulate the healing potential of the mammalian brain. Using immunohistochemistry and in situ hybridization techniques to label known markers for neural stem/ and progenitor cells I have identified three different populations of cells with radial glia (RG) like morphology in the adult zebrafish hypothalamic ventricular zone. In adult zebrafish, cells with RG-like morphology in the ventricular regions are thought to be the neurogenic population. The first population of cells I identified was positive for the neural stem cell marker NESTIN and showed additional characteristics of neural stem cells. Using a label retention assay we showed that Nestin(+) cells are slow cycling. The second population of RG-like cells was Hh responsive, and expressed markers of neural progenitor/transit amplifier cells. Double labeling experiments reveal that the Hh responsive cells were distinct from the Nestin(+) cells These cells were proliferative and cycled faster compared to nestin(+) neural stem cells. The third population of cells with RG morphology in the hypothalamic ventricular zone expressed shh ligand, indicating a regulatory role for Hh signaling in the hypothalamic ventricular zone. Down-regulation of Hh signaling at larval and adult stages reduced proliferation in the hypothalamic ventricle, indicating that Hh acts as a positive regulator of proliferation, as in the dorsal brain. According to our working model, nestin(+) cells are slow cycling, and/or quiescent neural stem cell population in the hypothalamic ventricular zone, whereas Hh responsive cells are the fast cycling transit amplifier cells which proliferate and give rise to new neurons and glia in the adult. My comprehensive analysis of the neural stem/progenitors in the adult zebrafish hypothalamic ventricular zone provides a starting point for the continued study of the mammalian hypothalamic ventricular zone. This study also demonstrates Hh signaling functions as a positive regulator of cell proliferation in the post-embryonic zebrafish hypothalamus consistent with its role in the dorsal brain. (Abstract shortened by UMI.)

The subventricular zone (SVZ) of the mammalian brain serves as one of only two sources of adult born neurons. Adult born neural progenitors - known as neuroblasts - acquire the ability to migrate and travel large distances to their destination in the olfactory bulb. Disruption of neuroblast migration is associated with learning and memory deficiencies and, following injury, neuroblasts are re-routed to promote neurodegeneration. While a lot of research has attended to augmenting the production and survival of neuroblasts, the body of evidence for pharmacological targets or compounds that promote migration is comparatively sparse. This thesis set out to identify novel strategies to modulate neuroblast migration for brain repair by studying proteins known to modulate migration and identifying new ones through compound screening. Firstly, an explant migration assay from the mouse SVZ was used to investigate the potential to use growth factors to stimulate neuroblast migration. This confirmed that that insulin-like growth factor 1 (IGF1) and IGF2 regulate neuroblast migration, as previously reported by other research groups. The role of IGF2 is investigated further using a mouse model in which the binding of IGF2 to IGF2R is disrupted, resulting in increased proliferation in the embryonic cortical SVZ, brain overgrowth and perinatal lethality. In the second half of this thesis I try to tackle one of the major bottlenecks limiting the search in for pharmaceutical interventions targeting neuroblast recruitment: the lack of high-fidelity in vitro migration assays. Drawing concepts from existing in vitro migration assays and cerebral organoid models, I developed a novel neuroblast spheroid migration assays that permits the investigation of large numbers of interventions, concurrently, in 3 dimensions. Using the spheroid assay I successfully screened 1012 small molecule kinase inhibitors for their effects on neuroblast migration. Several compounds were identified that significantly increased or decreased neuroblast migration. Two genes: MUSK and PIK3CB were selected from the screen as putative biological targets and genetic knockdown of these genes validated that interruption of their activity increased neuroblast migration. In the future these compounds could be studied further to explore their potential for augmenting the recruitments of new neurons to sites of injury so support neuroregeneration, or for decreasing invasion of brain malignancies.