The urea cycle enzymes play an important role in the processing of nitrogen to urea and in producing endogenous nitric oxide by the citrulline-nitric oxide cycle. This project characterizes the expression of urea and nitric oxide cycle enzymes and their intermediates at a cellular level. First, the expression of these enzymes in human tissues is examined. These findings reflect the expression of these genes in the human body at a baseline level. Second, the expression of these enzymes is determined in mouse tissue as well as concentrations of their intermediates (arginine and citrulline) and the end product nitric oxide. Here we discovered that citrulline is correlated with the production of nitric oxide metabolites in mouse tissues. Our third aim involved studying these enzymes in a cell culture environment and exposed to stimuli to measure changes in mRNA expression, protein expression and changes in the concentration of pathway intermediates (arginine and citrulline) as well as the end product nitric oxide. Finally, we examined protein interactions involved in the nitric oxide and urea cycle enzymes. This work highlights the functional co-localization and interactions of the components and intermediates of the urea and nitric oxide cycles, which has important implications for understanding the molecular biology underlying diseases associated with nitric oxide and urea cycle dysfunction.

Age-related macular degeneration (AMD) is a complex, late-onset disease that is the leading cause of blindness in the elderly. Age, smoking, and variants in complement factor H (CFH), LOC387715, complement factor B (CFB), and complement component 2 (CC2) are the strongest susceptibility factors for AMD. CFH inhibits activation of the alternative complement cascade, avoiding injury to self tissues by preventing excessive immune responses. Five CFH-like genes reside nearby on chromosome 1, and likely have similar roles in immune regulation. Deletion of CFHL1 and CFHL3 confers protection from AMD (Hughes et al. 2006). We have confirmed this protective effect (frequency=0.8%, frequency controls=2.6%, p=0.025). However, we also observed a protective haplotype without the deletion (haplotype frequency in cases=12%, controls=21%, p<0.001), suggesting that other protective variants have yet to be identified. As variants in CFH modify AMD susceptibility, exploration of other genes within the complement cascade is warranted. Variation in adjacent genes CFB and CC2 has been associated with decreased risk(Gold et al. 2006), though subsequent studies were unable to determine whether the effect of CC2 E318D was independent of polymorphisms in CFB or caused by LD between them(Maller et al. 2006). We dissected this region, and confirmed that both loci contribute to decreased risk of AMD, though the effect of CC2 is much smaller than CFB(CC2 E318D p=0.02, CFB R32Q p<0.0001). The search for novel AMD loci continues. A previous linkage screen identified chromosome 16p12 as a likely location for a novel AMD locus. After increasing SNP density across chr16p12, we combined linkage and association results, gene expression data, and the known biology of genes in the interval to select 4 candidates: CACNG3, HS3ST4, IL4R, and Q7Z6F8. CACNG3 is the strongest candidate (2pt LOD=3.3, haplotypes in family-based and case-control datasets p<0.01, ATA haplotype frequency cases=6.1%, frequency controls=10.3%, p=0.004). A common duplication covers the upstream sequence and first exon of CACNG3, and we hypothesize that this variation may have functional effects that influence susceptibility.

TOWARD AN UNDERSTANDING OF THE ROLE OF CHROMOSOME 15Q11-Q13 IN IDIOPATHIC AUTISM RYAN JAMES DELAHANTY Dissertation under the direction of Dr. James S. Sutcliffe The 15q11-q13 region is a genomic interval involved in a growing number of genomic disorders. The genes in the interval are subject to imprinting and parent-of-origin expression effects. Maternal duplication of the 15q11-q13 region is the most frequent chromosomal abnormality associated with autism. Extensive work has indicated that two genes in this interval, UBE3A and GABRB3, show very strong evidence for association with autism. To examine the extent to which these genes may contribute to autism, family-based association studies of UBE3A and GABRB3 were undertaken. Here we have investigated the role of common variants of UBE3A and GABRB3 in autism as well as the an intense investigation of the association of a rare variant, P11S in GABRB3 and its role in autism. In addition, we have investigated MECP2, a gene which when defective causes Rett syndrome, and potentially regulates gene expression of UBE3A and GABRB3. e have used genetic and biochemical methods to investigate two genes in the UBE3A network, ECT2 and GCH1. Finally, we used genotype data and multiplex ligation probe amplification (MLPA) to determine if copy number variation in the form of deletions and duplications in UBE3A and GABRB3 may play a role in the etiology of autism. Our findings indicated that a common allele of MECP2 is associated with autism, which was replicated by another group. We show association with UBE3A and its associated genes ECT2 and GCH1 as well as a relationship between UBE3A and GCH1 gene and protein expression, observed in a model system, and validated in our samples, which may provide guidance and support for a role of UBE3A and its action at the synapse and potential contribution to autism. We show modest association of GABRB3 with autism and epilepsy, but find a single coding variant, P11S, maternally overtransmitted and in such cases dramatically increasing autism risk. Finally, we found little evidence for microdeletions or microduplications in UBE3A and GABRB3 to contribute to autism pathology. The work presented in this thesis expands on earlier findings with regard to the role of GABRB3 and UBE3A in autism and represents an investigation of variants in these and their related genes spanning the spectrum from common variants of modest effect to rare variants of more profound effect. The availability of new technologies to evaluate copy number variation and next generation sequencing will likely uncover a wider role for 15q11-q13 and related loci in autism. The role of more highly penetrant private mutations of this nature is suggested as an avenue for further investigation.

Neither genome size nor gene number is indicative of organism complexity. Complex regulation of genes within a genome likely contributes to organism complexity. Coordinate regulation of genes through shared cis regulatory elements is one such mechanism. Peptide tyrosine tyrosine (PYY) and N-acetylglutamate synthase (NAGS) are divergently transcribed and thus share a 5 flanking region, raising the possibility these genes are coordinately regulated through common cis regulatory elements. These genes are separated by less than 1000 base pairs and may be regulated by a bidirectional promoter. The purpose of this project is to investigate the molecular and physiologic relationship between PYY and NAGS. To locate the promoter region, the transcriptional start sites (TSSs) for these genes were identified in multiple tissues. Results of these experiments demonstrate the major TSSs for PYY and NAGS are separated by greater than 1000 base pairs indicating they are not under control of a bidirectional promoter by the strict definition. However, they may share cis regulatory elements in the intergenic region. Expression of PYY and NAGS was determined in a panel of human tissues, revealing distinct expression patterns for these genes. Importantly, PYY and NAGS are similarly expressed in the ileum, indicating that if the genes are coordinately regulated it is in a tissue-specific manner. To better understand the physiologic connection between these two genes, plasma PYY levels were measured in patients with diminished capacity for nitrogen processing. Plasma PYY levels were increased in these patients and therefore may contribute to some of the clinical symptoms in this group. Finally, results from in vivo studies in mouse point toward possible coordinate expression of PYY and NAGS in response to different feeding conditions. Taken together the data are consistent with the coordinate regulation of PYY and NAGS, and lead to a model in which these genes are coordinately regulated as a means to prevent overconsumption of dietary protein. This mechanism may explain, in part, why protein is more satiating than other macronutrients and may also help elucidate the mechanisms of nitrogen balance.

Hirschsprung disease (HSCR) is characterized by a lack of enteric ganglia in a variable portion of the distal bowel. The complex inheritance pattern of this disorder has led researchers to focus on genetic effects other than the putative disease mutation. Mouse models provide a controlled background for these types of studies. <i>Sox10</i> is an essential gene for the development of the enteric nervous system (ENS). <i>Sox10^Dom</i> mice on a mixed genetic background exhibit the variable aganglionosis seen in HSCR cases. Congenic lines of <i>Sox10^Dom</i> mice on distinct inbred genetic backgrounds, C57BL/6J (B6) and C3HeB/FeJ (C3Fe), differ in penetrance and extent of aganglionosis. A linkage screen for modifiers of <i>Sox10^Dom</i> aganglionosis was undertaken in a large B6 X C3Fe F₂ population. Several potential modifier regions were identified, with the most significant located on chromosome five (<i>Sox10m3</i>). The most relevant candidate gene in this region was <i>Phox2B</i>, an essential factor in autonomic neurogenesis.<p> This goal of this dissertation was to understand the developmental differences between the congenic <i>Sox10^Dom</i> lines and how <i>Phox2B</i> impacts those differences. This study involved exploring differential expression of <i>Phox2B</i>, genetic variation at this locus between inbred strains, and the effects of different alleles of <i>Phox2B</i> on ENS development. The results suggest that <i>Phox2B</i> is differentially expressed between the B6 and C3Fe strains. The putative differences in expression are most likely due to genetic variation at key regulatory regions that are associated with aganglionic severity across multiple inbred strains. Strain-specific alleles at <i>Phox2B</i> modulate the effect of the <i>Sox10^Dom</i> mutation on ENS development.<p> The work presented in this dissertation confirms the identification of <i>Phox2B</i> as the <i>Sox10m3</i> modifier and demonstrates the effects of genetic background on complex phenotypes such as HSCR-related aganglionosis. By extending the principles used in this work to other genes and inbred strains, a greater understanding of the pathogenesis of HSCR can be achieved.

Abnormal regulation of mitochondrial deoxyribonucleoside triphosphate pools can lead to mitochondrial DNA depletion syndromes, a set of genetic diseases associated with depletion of mitochondrial DNA. Besides mitochondrial DNA depletion syndromes, improper maintenance of mitochondrial deoxyribonucleoside triphosphate pools and mitochondrial DNA have also been implicated in a host of other human pathologies. The unifying objective of this dissertation was to enhance our knowledge of the regulation of mitochondrial deoxyribonucleoside triphosphate pools. The first step was to investigate the characteristics of mitochondrial and cytoplasmic deoxyribonucleoside triphosphate levels. I calculated mitochondrial and cytoplasmic deoxyribonucleoside triphosphate concentrations from previously published data. Cytoplasmic and mitochondrial deoxyribonucleoside triphosphates are strongly correlated in normal cells but not in transformed cells. Following up this discovery with analysis of gene expression, I discovered that, consistent with the trends in deoxyribonucleoside triphosphate concentrations in cells, genes coding for enzymes that maintain cytoplasmic and mitochondrial deoxyribonucleoside triphosphates have correlated expression across normal tissues but not across transformed tissues. To further understand the influence of cytoplasmic deoxyribonucleoside triphosphates on mitochondrial deoxyribonucleoside triphosphates, I simulated the metabolism of mitochondrial deoxyribonucleosides with a computational model. Cytoplasmic deoxyribonucleotides have a substantial and indispensable contribution to mitochondrial deoxyribonucleoside triphosphates in most circumstances. My results further show that import from the cytoplasm would need to occur at either deoxyribonucleoside diphosphate or triphosphate levels.

The activity of polymerase ã (pol ã) is complicated. To understand how its kinetics values affect the final function of the pol ã, I created a stochastic model of pol ã replication on the single nucleotide incorporation level. Using this model, I analyzed replication pauses of both wild-type and pathogenic mutated pol ã and discovered that the pausing time is proportional to the number of disassociations occurring in each forward step of the pol ã, and studied mitochondrial toxicity caused by nucleoside analogs in antiretroviral treatment. To enrich the yield of rare disease alleles, a probability-based approach, SampleSeq, has been developed to select samples for a targeted resequencing experiment that outperforms over sampling based on genotypes at associated SNPs from GWAS data. To detect somatic mutations, novel algorithms have been developed to detect base substitution and loss of heterozygosity, using next-generation sequencing data for normal-tumor sample pairs.

HUMAN GENETICS <p> DETERMINING THE ROLE OF GROWTH DIFFERENTIATION FACTOR-6 (GDF6) IN THE DEVELOPMENT OF THE CORONAL SUTURE <p> DAWN ELIZABETH CLENDENNING <p> Dissertation under the direction of Professor Douglas P. Mortlock <p> Growth differentiation factor-6 (Gdf6) is a member of the Bone Morphogenetic Protein (BMP) family of secreted signaling proteins. The Gdf6 mutant mouse presents with fusions in the bones of the wrist and ankle, hypoplasia of the thyroid cartilage, abnormalities of the bones of the middle ear, and craniosynostosis of the coronal suture. Craniosynostosis is the premature fusion of one or more of the cranial sutures, the joints that separate the flat bones of the skull. <p> The primary objective of this work was to determine the developmental timing of the coronal suture defect, when and where Gdf6 is expressed during cranial development, what mechanism leads to fusion of the coronal suture, and how Gdf6 interacts with other members of the BMP family in coronal suture development. We found that Gdf6 is expressed in the rudiment of the frontal bone early in suture development. Loss of Gdf6 leads to the osteogenic differentiation of the coronal suture mesenchyme, which must remain undifferentiated to remain a site for cranial growth. Furthermore, Gdf6 interacts with fellow BMP family member Bmp4 not only in the development of the coronal suture, but in multiple skeletal elements. Additionally, we found that the BMP antagonist Noggin does not play an important role in embryonic suture development. The findings of this project not only further the the study of craniosynostosis but also broadens the knowledge of the roles of BMPs.

Late-onset Alzheimer disease (LOAD) is a complex neurodegenerative disorder with a strong genetic component. APOE is a well-established risk gene for LOAD, and several other genes have also been identified. However, at least half of the genetic component for LOAD remains unexplained. Genetic heterogeneity complicates identifying the remaining risk genes. To overcome this problem the objective of my thesis work was to study the Amish communities of Ohio and Indiana, a genetically isolated founder population to identify a novel LOAD risk gene(s). I first performed quality control procedures on a set of genome-wide single nucleotide polymorphisms (SNPs) genotyped in the Amish. I then performed genome-wide linkage and association analyses in the cleaned dataset. Lastly, I performed a candidate gene sequence analysis.

Many common diseases such as obesity and type-2 diabetes have a significant genetic component that contributes to susceptibility. Peroxisome proliferator activated receptors (PPARs) are nuclear receptors that heterodimerize with the retinoid X receptors (RXRs) to influence the expression of many genes involved in adipocyte differentiation and lipid metabolism such as the fatty acid binding proteins (FABPs) and the uncoupling proteins (UCPs). Genetic variation in any of these gene families could potentially alter metabolic traits related to obesity and type-2 diabetes. The goal of this project is to identify genetic variation in the PPARs and RXRs and then to determine if this variation is associated to quantitative traits related to obesity and type-2 diabetes using a multi-locus analysis approach. In this study, three sets of regression models were constructed: the first containing polymorphisms in just the PPARs or RXRs; the second with variants from all four gene families; and the third using polymorphisms from the gene isoforms showing the highest level of expression in each of three tissues. Some of the models were only able to account for small portions of the particular trait variation; however, many of the models accounted for a large amount of variation in the trait, up to 23.4% in the Hispanic female model for fasting free fatty acids. Multi-locus genotypes, as opposed to single locus effects, were found to be the best predictors of variation in almost all of the final models. These analyses confirmed the importance of gene-gene interactions on traits related to obesity and type-2 diabetes such as fasting free fatty acids and cholesterol; therefore, multiple polymorphisms should be considered together to fully understand their influence on a quantitative trait.