Variance Components based models are commonly used for linkage and association mapping of quantitative traits. Score Tests based on these models are generally more robust to various modeling assumptions than the corresponding likelihood ratio tests. They are also computationally much simpler than the likelihood ratio tests, making them the natural choice for whole genome scans, which have become increasingly common with the emergence of high-throughput genotyping technologies. However the popularity of score statistics have been limited, due to several practical issues, such as lack of availability of software and guidelines for choice of score statistic variants. In this dissertation, we develop novel score statistics for both linkage and association mapping, elucidate the theoretical properties of these and of the existing variants, and also compare some of the existing and proposed score variants using simulation. Analytical arguments and simulation results are used to develop guidelines for choice of appropriate score variants under different practical situations. In this dissertation, we are primarily concerned with identifying robust and powerful score statistics for detecting genetic susceptibility loci for complex diseases by mapping underlying quantitative phenotypes. Unlike Mendelian disorders, complex diseases in humans typically have a large number of modest genetic effects, which cumulatively have a significant impact on the disease. The work in this dissertation is aimed at maximizing the power of genome scans to detect more of these small genetic effects. This is of considerable public health significance, as the identified genetic variants can be followed up to gain important insights into the etiology of the disease, which can further lead to development of screening tests and preventive and therapeutic interventions for complex diseases.

Background: Schizophrenia is a disabling illness with unknown pathogenesis. Estimates of heritability suggest a substantial genetic contribution; however genetic studies to date have been equivocal. Uncovering liability loci may therefore require analyses of functionally related genes. Rooted in this assumption, this dissertation describes a series of studies investigating a genetic epidemiological foundation for the commonly cited hypothesis suggesting dopaminergic dysfunction in schizophrenia pathogenesis, i.e. the 'dopamine hypothesis'. Studies: The initial study investigated DRD3 and identified novel associations across the gene. The second study considered a larger network of dopaminergic genes in two independent Caucasian samples, detecting replicated associations and epistatic interactions. This study proposed a risk model for schizophrenia centered on the dopamine transporter. Study #3 investigated a dopamine precursor, phenylalanine hydroxylase, in four independent samples, identifying a single SNP (rs1522305) that was significantly replicated in two samples. Study #4 was motivated by the hypothesis of a shared genetic etiology for schizophrenia and bipolar disorder. This study comprehensively evaluated the dopaminergic network, selecting 431 'tag' SNPs from 40 genes among large schizophrenia and bipolar cohorts contrasted with adult controls. Across all genes 60% of nominally significant schizophrenia risk factors were also associated with bipolar disorder. The results supported DRD3 variations as risk factors for both disorders, confirmed several previously reported associations, and proposed new targets for future research. Conclusion: These results suggest dopaminergic gene variations could play an etiological role in the pathogenesis of schizophrenia and possibly bipolar 1 disorder. Additional replicate studies are warranted. Public Health Significance: Schizophrenia (SZ) is devastating. When the Global Burden of Disease study calculated disability adjusted life years, weighted for the severity of disability, they determined active psychosis seen in schizophrenia produces disability equal to quadriplegia. Schizophrenia has been estimated to be among the top ten causes of disability worldwide. As schizophrenia is common (roughly 1% point prevalence worldwide), the economic burden to society is substantial. Pathogenesis is unknown and treatment is palliative. Therefore understanding the genetic etiology could facilitate development of promising therapeutics.

MicroRNA genes are short, non-coding RNAs that function as post-transcriptional gene regulators. Although they have been implicated in organismal development as well as a variety of human diseases, there is still surprisingly little known about their transcriptional regulation. The understanding of microRNA transcription is very important for determining their regulators as well as the specific role they may play in signaling cascades. This dissertation focused on the comparison of mammalian microRNA promoters and upstream sequences to those of known protein coding genes. This dissertation is also focused on determining potential regulatory networks that microRNA genes may participate in, particularly those networks involved in the TGFβ / SMAD signaling pathway. The comparison of intergenic microRNA upstream sequences to those of protein coding genes revealed that the former are up to twice as conserved as the latter, except in the first 500 base pairs where the conservation is similar. Further investigation of the upstream sequences by RNA Polymerase II ChIP-chip revealed the transcription start site for 35 primary-microRNA transcripts. The identification of features capable of distinguishing core promoter regions from background sequences using a support vector machine approach revealed that the transcription start site of primary-microRNA genes share the same sequence features as protein coding genes. These results suggest that in fact microRNA genes are transcribed by the same mechanism by which protein coding genes are transcribed. This information allowed us to then identify the regulatory elements of microRNA genes in the same manner in which we use for protein coding genes. Identification of a SMAD family transcription factor binding site upstream of the human let-7d microRNA revealed a feed-forward regulatory circuit involved in epithelial mesenchymal transition. This provided the first evidence of a direct link between a growth factor and the expression of a microRNA gene. The understanding of microRNA transcriptional regulation has great public health significance. The ability to understand how these post-transcriptional gene regulators function in cellular networks may provide new molecular targets for cures or therapies to a variety of human diseases.

Motivation: Areal bone mineral density (BMD) is one of the major risk factors for osteoporosis, a public health concern in the US and other countries. The goal of our study was to identify genes that influenced areal BMD in a population of older, originally healthy, African American and Caucasian American individuals. Methods: We used three genetic association methods (single SNP-single trait, single SNP-multiple trait, multiple SNP-single trait) to test for association between measures of BMD at three time points (2 years apart) and genotype data on 1439 single nucleotide polymorphisms (SNPs) in 138 candidate genes. We first developed a model to determine the genetic (SNP) coverage of our candidate genes, and then we assessed possible population within our populations, two factors that influence the power of association studies. We also investigated the effect of covariates on BMD traits using both cross-sectional as well as longitudinal methods, and used the BMD residuals from these analyses in our association studies. Results: The SNP coverage of our candidate genes was reasonable, 52.99% compared to the theroetical HAPMAP coverage of 55.8%. We also classified our Caucasian American and African Americans based on genetic ancestry and controlled for subtle substructure. We detected several associations between candidate genes and BMD traits in all the four groups, but the most significant and consistent result was obtained in Caucasian American males. Five SNPs in the GNRHR locus were significantly associated with hip BMD trait using both the single SNP association approaches, as well as the pathway based analysis. These results need to be followed up in additional populations. Public Health Significance: As the world population ages, the cost, rate of mortality and morbidity of osteoporosis is also increasing. Identification of genes that influence risk of developing osteoporosis may help identify people at risk, as well as facilitate development of drugs and other measures to mitigate the effects of this disease.

Sequence variation in gene promoters is often associated with disease risk. In this study, a 1,418 bp sequence of the 5'-flanking region of the APOH (encoding for β₂-glycoprotein I) has been functionally characterized by in vitro analysis. Associations of APOH promoter SNPs with systemic lupus erythematosus (SLE) risk and related phenotypes along with their effect on human plasma β₂GPI levels were examined. Transient transfections, dual-luciferase reporter gene assays were performed in COS-1 and HepG2 cells. HepG2 nuclear extracts was used for electrophoretic mobility shift assay (EMSA). A case-control design and genotyped 345 SLE women and 454 healthy control women for 12 APOH promoter single nucleotide polymorphisms (SNPs) (-1284C>G, -1219G>A, -1190G>C, -1076G>A, -1055T>G, -759A>G, - 700C>A, -643T>C, -627A>C, -581A>C, -363C>T -38G>A, and -32C>A). Haplotype analyses were performed using EH (Estimate Haplotype-frequencies) and Haploview programs. Deletion analysis localized the core promoter of APOH ∼160 bp upstream of ATG codon with the presence of critical cis-acting elements between -166 and -65. Functional relevance for three SNPs (-1219G>A, -643T>C and -32C>A) that resulted in lower promoter activity (51%, 40% and 37%, respectively) as compared to the wild-type alleles in COS-1 cells. EMSA demonstrated HepG2 nuclear protein(s) bind to the elements located in the regions of the three SNPs. Overall haplotype distribution of the APOH promoter SNPs was significantly different between cases and controls (P = 0.009). The -643C allele was found to be protective against carotid plaque formation (adjusted OR = 0.37, P = 0.013) among SLE patients. Three-site haplotype analysis revealed one haplotype carrying -32A to be significantly associated with decreased plasma β₂GPI levels (P < 0.001) and another haplotype harboring the minor allele for -1219A showed a significant albeit less pronounced association (P = 0.046). Our data indicate that APOH promoter variants may be involved in the etiology of SLE, especially the risk for autoimmune-mediated cardiovascular disease. Public Health Significance: Cardiovascular disease is the leading cause of death in the U.S. and other developed nations. Understanding its pathogenesis will help in formulating newer therapeutic strategies and treatment that may decrease the adverse consequences of accelerated atherosclerosis in SLE patients.

SLE, a severe autoimmune disease is of major public health relevance since it predominantly affects women at child bearing age and even though immunosuppressives have increased the life span of SLE patients, lack of absolute cure is still troubling. Risk of premature coronary heart disease (CHD) is strikingly high in SLE women (35-44 years) than the general population. Low paraoxonase (PON) activity is associated with increased CHD as well as SLE risk. PON multigene (PON1, PON2 PON3) are anti-oxidants that cluster on chromosome 7q21-22 at 94.5- 94.6 Mb, in close vicinity to a linkage peak for SLE on 7q21.1 at 77.5Mb. PON1 (PON1/192, PON1/55) and PON3 (PON3/10340,PON3/2115) single nucleotide polymorphisms (SNPs) are the known significant modulators of PON/paraoxon activity. The purpose of this study was to determine the impact of PON2 tagSNPs with PON activity, SLE risk, lupus nephritis, parameters of LDL oxidation and subclinical carotid vascular disease measures. Nineteen PON2 tagSNPs were screened from HapMap and SeattleSNP databases in 489 SLE and 569 healthy control women from two recruitment sites (Pittsburgh and Chicago), using Pyrosequencing, RFLP or TaqMan allelic discrimination methods. Pairwise linkage disequilibrium (r2¡Ý 0.8) identified 15 tagSNPs that captured all the 19 PON2 variants in our sample. Although none of the PON2 tagSNPs revealed any obvious association with SLE risk, low PON/paraoxon activity was independently associated with SLE. Two PON2 variants [rs6954345(Ser311Cys) and rs987539] showed significant association with PON/paraoxon activity in Pittsburgh whites (cases+controls). Our data revealed few modest associations of PON2 variants with lupus nephritis (rs17876205, rs17876183, rs10261470, rs987539, rs9641164) in white (Pittsburgh+Chicago) SLE cases, parameters of LDL oxidation [PON2/rs11545941(Ala148Gly), rs13306702, rs2286233, rs10261470, rs17876205, rs4729189] in white (Pittsburgh) SLE cases and consistent association of PON2/rs11981433 and rs12704795 SNPs with carotid intima media thickness and plaque in white(Pittsburgh+Chicago) SLE cases. In conclusion, our data suggest that PON2 genetic variants have modest effect on serum PON activity, risk of lupus nephritis and subclinical carotid vascular disease measures in SLE patients.

Bipolar disorder (BPD) is a serious mood disorder that affects about 1% of the population of the United States. Twin, family, and adoption studies have shown evidence for a genetic component of BPD, but monozygotic twin concordance is less than one, indicating that BPD is a multifactorial disorder. First-degree relatives of an individual with BPD have approximately a 5-30% risk of developing BPD because of shared genes and environment. No strong susceptibility loci for BPD have been located, although some areas of interest are currently being evaluated. With increasing genetic information, demand for genetic counseling for BPD and other psychiatric disorders are increasing. This study used anonymous surveys for individuals with BPD and their first-degree relatives to assess knowledge, opinions, and acceptance of genetic counseling. The Health Belief Model was used to assess current health beliefs relating to BPD. Additionally, using a brief educational session, the effect of education on knowledge and health beliefs was assessed. A similar study was conducted by Dr. Triptish Bhatia in India and she has provided her unpublished results so that they could be included in this document for comparative purposes. Data show that the perceived severity of BPD and perceived barriers of testing were high in both populations. Data show that the perceived susceptibility, benefit, and knowledge of BPD in affected individuals were higher in the US population than in the Indian population. Ascertainment criteria and the evaluation procedures for the samples were different in both countries and they cannot be considered to be representative of the respective 'populations'. Results from this study can be applied to the creation of public health programs in which clients can learn more about their condition, how BPD is related to genomics and what the risk is to their offspring.

Cachexia is the most debilitating syndrome which manifests itself in several chronic, life-threatening diseases, especially in cancer. Cachexia is of major public health significance for the cancer population because it increases both morbidity and mortality and also reduces quality of life and survival time of cancer patients. Up to two-thirds of patients with advanced neoplasia develop signs and symptoms of cachexia, including anorexia, asthenia and severe unintentional weight loss leading to immobility and cardiac or respiratory failure. Cachexia accounts for more than 20% of all cancer-associated deaths. In the present study, we established a novel murine model for cancer cachexia induced by the human prostate cancer cell line PC-3. In a novel mouse model of cachexia induced by PC-3 cells, an established androgen-independent cell line derived from a bone metastasis of a human prostatic adenocarcinoma, in BALB/c nude mice we observed body weight loss, a 50% reduction in muscle weights and decreased muscle fiber diameters. Elevated levels of the phosphorylated p65 subunit of the nuclear factor of êB (NF-êB) were found in tibialis anterior and quadriceps muscles, but not in gastrocnemius muscle. Elevated levels of the muscle specific E3 ligase (MuRF1) confirmed activation of the ubiquitin-proteosome protein degradation pathway in these muscles. In contrast, elevated levels of the phosphorylated eukaryotic initiation factor 2 alpha (p-eIF2-á) in gastrocnemius muscle, but not in tibialis anterior and quadriceps muscles, suggested a greater component of cachexia due to decreased protein synthesis in this muscle. We also utilized a well-established murine model of cancer cachexia induced by murine colon adenocarcinoma cell line (C-26). C-26 tumor-bearing mice were treated with an intramuscular injection of an adeno-associated viral vector serotype 8 (AAV8) carrying the IêB super repressor (IêBSR) or cellular caspase-8-like inhibitory protein (cFLIP) gene driven by the cytomegalovirus (CMV) or muscle creatine kinase (MCK) promoter. We found that there was an improvement in body weight, individual muscle weight and muscle fiber diameter in mice receiving AAV8-IêBSR or AAV8-cFLIP. We also observed a reduction of MuRF1 protein expression, indicating that there was a reduction in muscle protein degradation via the ubiquitin-proteasome system. The result was confirmed by an increased level of myosin heavy chain protein expression. This study suggests the potential for AAV8 carrying IBSR gene mediated gene transfer to prevent or reverse cachectic symptoms in vivo.

Smoking is one of the leading risk factors that contribute to the development of head and neck squamous cell carcinoma (HNSCC). Smoking induces chromosome breaks. One of the most frequent and earliest alterations in HNSCC is the segmental loss of the short arm of chromosome 3 (3p). Another frequent observation is the activation of one of the most common fragile sites in HNSCC, FRA3B, which spans the FHIT gene, considered to be a tumor suppressor. Distal to FHIT is FANCD2, which encodes a protein required for the proper function of the Fanconi anemia/BRCA pathway and eventual repair of DNA damage through homologous recombination repair. Recently, it has been observed that FANCD2 gene and protein expression is decreased in HNSCC cells. Therefore, we hypothesize that 3p loss, resulting from the activation of the fragile site FRA3B, leads to FANCD2 gene copy loss and a defective DNA damage response. Western blots showed decreased expression of FHIT protein in HNSCC cell lines. Fluorescence in situ hybridization revealed partial loss of FANCD2 in HNSCC cell lines studied. Western blotting and quantitative RT-PCR showed decreased FANCD2 protein and gene expression in HNSCC cell lines with FANCD2 gene copy loss. HNSCC cell lines with decreased FHIT protein expression, FANCD2 gene copy loss, and decreased FANCD2 gene and protein expression exhibited decreased FANCD2 and RAD51 focus formation. Our results suggest that smoking could induce breakage of the fragile site, FRA3B, leading to FANCD2 loss and resulting in defective DNA damage repair. The use of targeted therapy on cancer cells with deficient DNA damage repair, like CHEK1 small molecule inhibitors, could improve the currently available cancer treatment schemes. The public health relevance of our studies involves the use of an abnormal FA/BRCA pathway as a marker for selective use of targeted cancer therapy for HNSCC.

Triploidy is the presence of 69 chromosomes instead of the normal diploid number of 46 and can occur in a complete form or in a mixoploid state in which there are populations of diploid and triploid cells in the same individual. The extra haploid set can be of paternal or maternal origin. Triploidy is one of the most common chromosome aberrations seen in 1-2% of all recognized pregnancies and can lead to partial mole which can in turn lead to serious complications for the mother and fetus. Given the high incidence of chromosome abnormalities including triploidy and its impact on individuals with chromosomally abnormal pregnancies, a greater understanding of their etiology has a potential to contribute greatly to public health by enhancing the management and possible future prevention. Though complete triploidy is not compatible with postnatal survival, mixoploid individuals are capable of surviving into adulthood. Both syndromes have a broad phenotypic spectrum though it is generally less severe in mixoploids. Though much has been learned in the nearly half century since the first case report of diploid/triploid mixoploidy was published, many questions still remain. A major issue is a large between study difference in the ratio of diandric to digynic triploidy and the prevalence of partial hydatidiform mole. Additionally, there is a clear parent-of-origin effect on fetal and placental morphology as well as developmental age that is believed to be related to genomic imprinting. The goals of this paper include summarizing the current body of knowledge on triploidy and diploid/triploid mixoploidy, examining the remaining questions, and a side-by-side comparison of the two syndromes. An exhaustive literature search was undertaken which produced many case reports of triploidy and diploid/triploid mixoploidy as well as studies on the mechanisms leading to triploidy, phenotypic characteristics, and the characteristics of triploid cells. It appears that the complex pattern surrounding parental origin of the extra haploid set of chromosomes may have contributed to between study ascertainment bias. More complex studies with careful attention to detail must be undertaken to fully understand the etiology and pathophysiology of triploidy.