Search

The process of the brain developing from a single fertilized egg to the most sophisticated known organ requires precise spatial and temporal control to produce the necessary correct brain size and architecture. A particular region of interest is the cerebral cortex, responsible for higher functions such as language, reasoning and conscious thought. Its expansion in size and complexity from smaller mammals, such as mice, to humans is thought to contribute to our higher functions. However, a caveat of this increased complexity is the increased challenge of generating such a complex structure, and the potential for subtle changes during neurodevelopment to manifest in neurodevelopmental disorders such as Autism Spectrum Disorders (ASD). ASD is a spectrum disorder diagnosed early in childhood based on a range of diagnostic criteria. It is frequently characterised by impaired social interaction, repetitive behaviour and delayed development. While ASD patients share some symptoms, the genetic underpinnings of ASD are highly heterogeneous, with mutations to many single genes or larger genetic regions implicated as ASD risk factors. The 593kbp 16p11.2 locus encompasses 29 protein coding genes and its copy number variation (CNV) by heterozygous microduplication or microdeletion is implicated in around 1% of ASD cases, with many patients born with macrocephaly (deletion) or microcephaly (duplication), potentially indicating a possible problem with generating the correct number of brain cells during development. This suggests the hypothesis that some of the 16p11.2 region genes are involved in neural proliferation in early corticogenesis, and changes to the levels of these genes may affect proliferation, contributing to the 16p11.2 patient phenotype. This hypothesis is supported by a 16p11.2 deletion mouse model which exhibits ASD-like symptoms and altered proliferation in the cortex during embryonic development. Given that the 16p11.2 CNV's 1% autism incidence makes it the most frequent aetiology of ASD, this region is a promising area of study to understand how genetic dysregulation during critical prenatal cortical neurogenesis can contribute to the ASD phenotype. Despite its strong association with ASD, very little is known about the majority of the 16p11.2 genes, especially regarding brain development. In this thesis, focussing on the developing human neocortex, we aimed to identify which, if any, of these 29 genes were expressed in progenitor cells and describe their expression pattern during critical stages of cortical neurogenesis. We first used a bioinformatics approach to identify 16p11.2 genes expressed in progenitors to narrow down from the 29 genes in the region. We analysed a publicly available single-cell RNA sequencing (scRNA-seq) dataset from the proliferative zones, the ventricular zone and subventricular zone, of the 16-18 gestational week (GW) human fetal cortex. We identified six genes as being highly expressed in the cortical progenitor cells, and two as being significantly higher expressed in progenitors compared to post mitotic cells: KIF22 and ALDOA. We described their protein expression pattern in vivo at key stages of human fetal cortex development. We showed KIF22 protein to be expressed in the germinative zones, and its expression to be restricted to proliferating cells, suggesting a role for this protein in proliferation. We showed KIF22 protein levels to vary with the cell cycle, increasing from G1 through S and G2 phases to peak in mitosis. This suggests that changing the KIF22 protein level, as in the microduplication or microdeletion patients, will affect cell cycle and proliferation manifesting in changes to cortical size and architecture contributing to the 16p11.2 phenotype. ALDOA protein was shown to be present throughout the cortex, although higher in the proliferating regions. We demonstrated that ALDOA is predominantly localised to cytoplasm, and its protein levels or sub-cellular localisation do not change in proliferating or non-proliferating cells. ALDOA has a critical role in energy metabolism, and we can hypothesise that due to its expression throughout the cortex any changes to the ALDOA protein by the 16p11.2 CNV will induce a wide range of effects on brain development In conclusion we have identified two genes highly expressed in progenitors and expressed at much lower levels in post-mitotic cells from the 16p11.2 locus. These genes provide interesting targets for future studies to elucidate the mechanism by which they mediate proliferation and the effects of manipulating their protein levels. This is outwith the scope of this PhD thesis however a range of new techniques are emerging such as cerebral organoids, which can be easily manipulated. These will be a powerful tool to address the hypotheses produced by the descriptive work of this PhD thesis.

Flaviviruses are arthropod-transmitted positive-sense single stranded RNA viruses, capable of causing significant human disease. Their RNA genomes are translated by host machinery and produce a viral polypeptide of 10 viral proteins, 3 structural and 7 non-structural. The non-structural proteins are essential for mediating virus genome replication, host defense silencing, and cell remodeling. Much of this is accomplished through physical interactions between viral protein and host proteins, which interrupts or otherwise impacts their cellular function for virus benefit. Inhibition of host protein function during these interactions can be connected to viral pathogenesis and disease. In the case of Zika virus (ZIKV) the primary disease outcome of concern is birth defects or fetal demise that occur in fetuses that are infected in utero via vertical transmission from the infected mother. Birth defects arising from ZIKV infection are broad and collectively termed congenital Zika syndrome (CZS). Most infamous of these is microcephaly, a condition where brain and head are not fully grown at birth. Microcephaly is associated with a wide range of debilitating development problems that last throughout life. The molecular mechanisms that contribute to microcephaly observed in CZS cases are not fully understood and likely multifactorial. To understand if ZIKV-host protein-protein interactions are contributing to neurodevelopmental defect, global proteomics was previously performed on ZIKV proteins to identify host interactors. This revealed the interaction between ZIKV non-structural 4A (NS4A) and the host protein ankyrin repeat and LEM domain containing 2 (ANKLE2). ANKLE2 is involved in nuclear envelope dynamics during mitosis and asymmetric cell division during neurogenesis, a key step in brain development. Mutations in ANKLE2 are associated with primary congenital microcephaly in humans. Expression of ZIKV NS4A in vivo results in abnormal brain development in fruit flies. This phenotype is rescued by overexpression of human ANKLE2. Together this data supports the hypothesis that NS4A physically interacts with ANKLE2, inhibiting its function during neurodevelopment to cause microcephaly. However, the basis for the physical interaction and the specific ways NS4A inhibits ANKLE2 function through it are still mysterious. Beyond this is the question of why NS4A interacts with ANKLE2 to begin with? Is the interaction, and the pathogenesis that follows, purely coincidental, or does NS4A interact with ANKLE2 to serve in some aspect of ZIKV replication? This work serves to explore the questions revolving the NS4A-ANKLE2 interaction. First, we establish that ANKLE2 has the opportunity to impact ZIKV replication by showing colocalization of ANKLE2 with ZIKV factors during virus replication. To explore if ANKLE2 is actively participating in some aspect of virus replication we then genetically deplete ANKLE2 using CRISPRi knockdown or CRISPR mutagenesis. Cells with diminished or depleted ANKLE2 were infected with ZIKV and had reduced replication across multiple conditions and cell lines. This work provides substantial evidence that ANKLE2 supports ZIKV replication in human cells. During a flavivirus transmission cycle the virus must efficiently replicate in both human and mosquito cells. In collaboration with another group, we show that depletion of the Ankle2 ortholog in mosquito Aag2 cells also leads to reduced ZIKV replication, supporting that Ankle2 is beneficial to virus replication across hosts. Further, we show that physical interaction between ANKLE2 and NS4A is conserved across four other mosquito-borne flaviviruses and that ANKLE2 plays a role in the replication of some of these viruses. In order to investigate the physical determinants of the ANKLE2-NS4A interaction we developed a series of truncation mutants that serially express fewer domains of each protein. To test physical interaction between the proteins, we performed co-transfection and co-immunoprecipitation experiments. This revealed the N-terminal region of ANKLE2 interacts with the C-terminal region of NS4A. Further, we show that this non-interacting ANKLE2 mutant does not colocalize with ZIKV NS4A during infection. However, the data suggest that there are multiple contact sites between ANKLE2 and NS4A, across separate domains. ANKLE2 is a scaffolding protein that modulates the cell cycle through physical protein interactions. To explore how these interactions may be perturbed during infection and how new interactions may be driven to benefit virus replication, we performed affinity-purification and mass spectrometry on ANKLE2 with and without ZIKV infection. These revealed hundreds of candidate protein interactions which enlighten both how ZIKV inhibits normal ANKLE2 function and also potential pathways through which ANKLE2 promotes virus replication. Altogether, this work vastly expands on our understanding of the ZIKV NS4A-ANKLE2 interaction and supplies avenues for future exploration.

Mouse models have been classically used to study human neurodevelopmental disorders. More recently, human stem cell models have advanced our ability to model unique aspects of human brain. While mice are able to recapitulate many aspects of human brain development, human cortex is vastly expanded compared to mice. This dissertation will show how mouse models and human iPSC models can be used in combination to generate increased understanding of the pathology and molecular mechanism of several human brain disorders.First, we studied Joubert syndrome. KIAA0753 is one of at least 35 Joubert causal genes. We describe four new patients with five novel KIAA0753 variants and study these variants in vitro. The variants produce truncated KIAA0753 protein products that are not competent to promote primary ciliogenesis. Patient fibroblasts and Kiaa0753 knockout 3T3 cells have a loss of primary cilia and subsequent deficit in SHH and WNT pathway signaling. Utilizing these complementary cell models allowed us to demonstrate the pathogenicity of specific human variants.Second, we identified a developmental brain disorder caused by human variants in CAMSAP1. Patients present with a characteristic craniofacial appearance and numerous structural brain abnormalities including microcephaly. We generated a Camsap1 null mouse allele which exhibits increased perinatal mortality but showed no severe structural brain abnormalities. We used human induced pluripotent stem cells (iPSCs) obtained from a CAMSAP1 patient to make neural rosette structures which display abnormal morphology, decreased proliferation, and premature neuronal differentiation. These cellular phenotypes suggest a neural progenitor defect. Mouse models showed us that Camsap1 is expressed in brain and required for postnatal survival, and adding a human iPSC model allowed us to demonstrate the effects of CAMSAP1 variants on early forebrain cell development.Finally, we studied the impact of sphingolipid biosynthesis on human brain development. SMPD4 variants cause a severe developmental brain disorder including microcephaly and cerebellar hypoplasia. We hypothesized that the role of SMPD4 in producing ceramide is important for making primary cilia. We found that Smpd4 null mice have cerebellar hypoplasia due to failure of Purkinje cell development. By using human iPSC models we were able to show that the microcephaly is likely due to a loss of neural progenitor cells, which exhibit cell death and decreased proliferation in vitro. We explored the role of primary cilia in these models and found that although Smpd4 null mice appear unaffected, human iPSC models have shortened and bulbous primary cilia which is rescued by adding exogenous ceramide. We found evidence for disrupted WNT pathway signaling. We suggest that during brain development, SMPD4 provides ceramide in an organelle-specific manner, regulating the cell cycle and allowing primary cilia to grow properly.In summary, common signatures such as primary cilia morphology and neural progenitor proliferation are impacted by patient variants in brain development genes. The power of combining mouse and human models lies in its potential to uncover the molecular roles of these genes. We have utilized this combination to advance our knowledge of the roles of KIAA0753, CAMSAP1, and SMPD4 genes for human brain development.

Prenatal exposure to Zika virus (ZIKV) can result in microencephaly and congenital Zika syndrome but why some brain cells and structures are initially spared by the virus is unknown. Here, a novel murine model of ZIKV infection incorporating in utero electroporation with cell type specific promotors was used to identify the time course of ZIKV infection and to determine which neural precursor cells are initially infected or spared. In vivo time course studies revealed early presence of ZIKV in apical radial glial cells (aRGCs) while infection of basal intermediate progenitor cells climbed after three days of virus exposure. ZIKV-exposed fetal brains exhibited microencephaly as early as 1 day post injection, caused by apoptosis and reduced proliferation, and this change in brain size persisted until birth regardless of developmental age at infection. During infection, 60% of aRGC basal fibers were perturbed while 40% retained normal morphology, indicating that aRGCs are not uniformly vulnerable to ZIKV infection. To evaluate this heterogeneous vulnerability, we generated cell type-specific fate mapping plasmid probes using a previously published single cell RNA-Seq dataset on the E15.5 mouse neocortical wall. The results indicate that one class of aRGC preferentially expresses the putative ZIKV entry receptor AXL, and that these cells are more vulnerable to ZIKV infection than the other aRGC subtypes with low AXL expression. Together, these data highlight important temporal and cellular details of ZIKV fetal brain infection and may be important for prevention strategies and for management of congenital Zika syndrome.

The mouse model is a well-established and powerful tool for genetic research, with countless transgenes and optimized protocols available. However, knockout of many genes is lethal at very early stages of embryogenesis, making it hard to predict the pathogenic effect of hypomorphic mutations and precluding more in-depth studies of organogenesis. Mouse models of hypomorphic mutations can allow in vivo study of genes required for pre-organogenesis survival and provide more specialized study of specific mutations identified clinically. We use forward genetics in the mouse and powerful diagnostic genotyping tools to identify and model novel mutations affecting craniofacial and cerebral development, demonstrating novel roles for two essential genes in these processes and studying the underlying pathogenic mechanisms.We recovered the dorothy mutation in an N-ethyl-N-nitrosourea (ENU) forward genetic screen designed to produce recessive mutations affecting craniofacial development in the mouse. Dorothy embryos die prenatally and exhibit many striking phenotypes commonly associated with ciliopathies, including a severe midfacial clefting phenotype. We used exome sequencing to discover a missense mutation in nucleotide binding protein 2 (Nubp2) as a candidate variant. This finding was confirmed by a complementation assay with the dorothy allele and an independent Nubp2 null allele (Nubp2null). We demonstrated that Nubp2 is indispensable for embryogenesis. NUBP2 is implicated in both the cytosolic iron/sulfur cluster assembly pathway and in negative regulation of ciliogenesis. Conditional ablation of Nubp2 in the neural crest lineage with Wnt1-cre recapitulated the dorothy craniofacial phenotype. Using this model, we found that the proportion of ciliated cells in the craniofacial mesenchyme was unchanged, and that markers of important signaling pathways were unaltered. Finally, we show evidence that the phenotype is the result of a marked increase in apoptosis within the craniofacial mesenchyme.Another method of discovering novel genetic contributions to developmental processes is the sequencing and modeling of patient mutations, which provides the added benefit of a translational context. Primary microcephaly is a congenital brain malformation characterized by a head circumference less than three standard deviations below the mean, and results in moderate to severe mental deficiencies. We studied two children with primary microcephaly in an otherwise unaffected family. Exome sequencing identified an autosomal recessive mutation leading to an amino acid substitution in a WDR domain of the highly conserved Coatomer protein complex, subunit beta 2 (COPB2). To study the role of Copb2 in neural development, we utilized genome-editing technology to generate an allelic series in the mouse. Two independent null alleles revealed that Copb2 is essential for early embryogenesis. Mice homozygous for the patient variant (Copb2R254C/R254C) appeared to have a grossly normal phenotype. Strikingly, mice heterozygous for the patient mutation and a null allele (Copb2R254C/Zfn) showed a severe perinatal phenotype including low neonatal weight, significantly increased apoptosis in the brain, and death within the first week of life. Immunostaining of the Copb2R254C/Zfn brain revealed a reduction in CTIP+ Layer V neurons, while the overall cell density of the cortex was unchanged. These results identified a general requirement for COPB2 in embryogenesis and a specific role in corticogenesis.

The primary cilium is an organelle consisting of a specialized microtubule-based extension of the cell membrane. These non-motile cilia extend nearly ubiquitously across cell types and are necessary for the regulation of an array of signaling pathways including many critical for development. Ciliopathies are a class of diseases resulting from impaired ciliary function, affecting several different organ systems. Cognitive impairment is a common phenotype present in ciliopathy patients. Additionally, multiple mouse models of ciliary dysfunction exhibit defects in brain development. These disruptions of ciliary function in both humans and mice and the resulting phenotypic consequences display the importance of primary cilia for proper central nervous system health. To gain a better understanding of the role of ciliary proteins in cortical development we performed a series of genetic experiments to ablate three intraflagellar transport (IFT) genes in discrete spatiotemporal domains as described in Chapter 2. Kif3a and Ift88 are two components of the IFT-B complex responsible for anterograde trafficking to the tip of cilia, and Ttc21b is part of the IFT-A complex, necessary for proper retrograde trafficking. This study led to three primary findings: cilia mediate essential tissue-tissue interactions during embryonic head development, these genes have unique and specific spatiotemporal requirements in cortical development, and cortical phenotypes associated with Ttc21b are due to events outside brain tissue itself and prior to definitive forebrain development. The understanding that events prior to a definitive forebrain were responsible for causing microcephaly and cortical malformations in the Ttc21b-alien(aln) mouse model helped generate a number of hypotheses about the earliest mechanisms leading to these phenotypes, which are explored in Chapter 3. Examining the expression of multiple genes in the prospective forebrain led to the conclusion that Ttc21baln is a phenocopy of Gli3-null mouse forebrain phenotypes. Ttc21b is known to be critical for regulating the Sonic Hedgehog (SHH) pathway via processing of GLI3 proteins into repressor and activator forms. Impaired processing of GLI3 in Ttc21b-aln would explain these similarities to Gli3-nulls. The full pathway by which Gli3 disruption causes loss of forebrain tissue has not been completely delineated. By using Ttc21b-aln as a model we have implicated two regulators of WNT/BMP signaling in the early embryonic head that are regulated by GLI3 repressor activity, Dkk1 and Gsc. Finally, Chapter 4 describes a quantitative trait locus (QTL) experiment to detect genetic modifiers of the reduced forebrain size in Ttc21b-aln. The severity of the microcephaly phenotype of Ttc21baln is dependent on the genetic background of the mouse strain carrying the mutation. This QTL screen identified one statistically significant locus on chromosome 4 and a locus suggestive of significance on chromosome 6. Both regions were validated by generating a congenic line for each QTL. Additionally, a candidate gene, Gpr63, in the chromosome 4 region was tested using CRISPR-Cas9 technology and found to be a genetic interactor with Ttc21b which affects forebrain size. These studies increase our understanding of the role of primary cilia and SHH signaling in forebrain development as well as the complex genetics impacting ciliary function.

Human genetics has identified essential roles for many centriole- and cilia-related proteins during human development. Mutations in centrosome-associated genes commonly cause microcephaly, or "small brain," and mutations in cilia-associated genes cause a diverse spectrum of diseases termed "ciliopathies." However, the functional relationships between these two crucial organelles are less well studied. The activities of centrosome-related proteins during mitosis and cytoskeletal remodeling are well-characterized, but their in vivo functions are incompletely understood. Here, we identify novel human mutations in a centrosomal gene which encodes a regulatory subunit of a microtubule interacting protein, and uncover unexpected pathways during vertebrate development. Human mutations cause severe microlissencephaly, reflecting defects in cerebral cortical neurogenesis, and loss of function in mice and zebrafish confirm essential roles in embryonic development, neurogenesis, and cell survival. Surprisingly, null mutant embryos display hallmarks of aberrant Sonic hedgehog signaling, including holoprosencephaly. Deficient induced pluripotent stem cells and lymphoblasts show defective proliferation and spindle structure, while deficient fibroblasts also demonstrate a remarkable excess of centrioles, including excessive maternal centrioles, with supernumerary cilia but deficient Hedgehog signaling. Our results reveal novel roles for this protein in regulating overall centriole number, mother centriole and cilia number, and as an essential gene for normal Hedgehog signaling during neocortical development.