Cross -packing among retroviruses and characterization of the feline immunodeficiency virus (FIV) packaging signal: Implications for the development of FIV-based gene transfer systems

Feline immunodeficiency virus (FIV)-based gene transfer systems are being seriously considered for human gene therapy as an alternative to vectors based on primate lentiviruses, a genetically complex group of retroviruses capable of infecting non-dividing cells. The greater phylogenetic distance between the feline and primate lentiviruses is thought to reduce chances of the generation of recombinant viruses. However, safety of FIV-based vector systems has not been tested experimentally. Since primate lentiviruses such as human and simian immunodeficiency viruses (HIV/SIV) can cross-package each other's genomes, we tested this trait with respect to FIV. Unexpectedly, both feline and primate lentiviruses were reciprocally able to both cross-package and propagate each other's RNA genomes. This was largely due to the recognition of viral packaging signals by the heterologous proteins. However, a simple retrovirus such as Mason-Pfizer monkey virus (MPMV) was unable to package FIV RNA. Interestingly, FIV could package MPMV RNA, but not propagate it for further steps of replication. These findings suggest that upon co-infection of the same host, cross-packaging may allow distinct retroviruses to generate chimeric variants with unknown pathogenic potential. ^ In order to understand the packaging determinants in FIV, we conducted a detailed mutational analysis of the region thought to contain FIV packaging signal. We show that the first 90–120 nt of the 5′ untranslated region (UTR) and the first 90 nt of gag were simultaneously required for efficient FIV RNA packaging. These results suggest that the primary FIV packaging signal is multipartite and discontinuous, composed of two core elements separated by 150 nt of the 5 ′UTR. ^ The above studies are being used towards the development of safer FIV-based self-inactivating (SIN) vectors. These vectors are being designed to eliminate the ability of FIV transfer vector RNAs to be mobilized by primate lentiviral proteins that may be present in the target cells. Preliminary test of the first generation of these vectors has revealed that they are incapable of being propagated by feline proteins. The inability of FIV transfer vectors to express packageable vector RNA after integration should greatly increase the safety of FIV vectors for human gene therapy. ^

Targeting the Blood-brain Barrier with a Non-canonical Iron-mimicry Mechanism

Treatment of central nervous system (CNS) diseases is limited by the blood-brain barrier (BBB), a selective vascular interface restricting passage of most molecules from blood into brain. Specific transport systems have evolved allowing circulating polar molecules to cross the BBB and gain access to the brain parenchyma. However, to date, few ligands exploiting such systems have proven clinically viable in the setting of CNS diseases. We reasoned that combinatorial phage-display screenings in vivo would yield peptides capable of crossing the BBB and allow for the development of ligand-directed targeting strategies of the brain. Here we show the identification of a peptide mediating systemic targeting to the normal brain and to an orthotopic human glioma model. We demonstrate that this peptide functionally mimics iron through an allosteric mechanism and that a non-canonical association of (i) transferrin, (ii) the iron-mimic ligand motif, and (iii) transferrin receptor mediates binding and transport of particles across the BBB. We also show that in orthotopic human glioma xenografts, a combination of transferrin receptor over-expression plus extended vascular permeability and ligand retention result in remarkable brain tumor targeting. Moreover, such tumor targeting attributes enables Herpes simplex virus thymidine kinase-mediated gene therapy of intracranial tumors for molecular genetic imaging and suicide gene delivery with ganciclovir. Finally, we expand our data by analyzing a large panel of primary CNS tumors through comprehensive tissue microarrays. Together, our approach and results provide a translational avenue for the detection and treatment of brain tumors.

Long-term survival after multidisciplinary management of resected pancreatic adenocarcinoma.

INTRODUCTION: Actual 5-year survival rates of 10-18% have been reported for patients with resected pancreatic adenocarcinoma (PC), but the use of multimodality therapy was uncommon in these series. We evaluated long-term survival and patterns of recurrence in patients treated for PC with contemporary staging and multimodality therapy. METHODS: We analyzed 329 consecutive patients with PC evaluated between 1990 and 2002 who underwent resection. Each received a multidisciplinary evaluation and a standard operative approach. Pre- or postoperative chemotherapy and/or chemoradiation were routine. Surgical specimens of 5-year survivors were re-reviewed. A multivariate model of factors associated with long-term survival was constructed. RESULTS: Patients underwent pancreaticoduodenectomy (n = 302; 92%), distal (n = 20; 6%), or total pancreatectomy (n = 7; 2%). A total of 108 patients (33%) underwent vascular reconstruction, 301 patients (91%) received neoadjuvant or adjuvant therapy, 157 specimens (48%) were node positive, and margins were microscopically positive in 52 patients (16%). Median overall survival and disease-specific survival was 23.9 and 26.5 months. Eighty-eight patients (27%) survived a minimum of 5 years and had a median overall survival of 11 years. Of these, 21 (24%) experienced recurrence, 7 (8%) after 5 years. Late recurrences occurred most frequently in the lungs, the latest at 6.7 years. Multivariate analysis identified disease-negative lymph nodes (P = .02) and no prior attempt at resection (P = 0.01) as associated with 5-year survival. CONCLUSIONS: Our 27% actual 5-year survival rate for patients with resected PC is superior to that previously reported, and it is influenced by our emphasis on detailed staging and patient selection, a standardized operative approach, and routine use of multimodality therapy.

Ribozyme-mediated reduction of wild-type and mutant cartilage oligomeric matrix protein (COMP) mRNA and protein.

Dominant-negative mutations in the homopentameric extracellular matrix glycoprotein cartilage oligomeric matrix protein (COMP) result in inappropriate intracellular retention of misfolded COMP in the rough endoplasmic reticulum of chondrocytes, causing chondrocyte cell death, which leads to two skeletal dysplasias: pseudoachondroplasia (PSACH) and multiple epiphyseal dysplasia (EDM1). COMP null mice show no adverse effects on normal bone development and growth, suggesting a possible therapy involving removal of COMP mRNA. The goal of this study was to assess the ability of a hammerhead ribozyme (Ribo56, designed against the D469del mutation) to reduce COMP mRNA expression. In COS7 cells transfected with plasmids that overexpress wild-type or mutant COMP mRNA and Ribo56, the ribozyme reduced overexpressed normal COMP mRNA by 46% and mutant COMP mRNA by 56% in a dose-dependent manner. Surprisingly, the use of recombinant adenoviruses to deliver wild-type or mutant COMP mRNA and Ribo56 simultaneously into COS7 cells proved problematic for the activity of the ribozyme to reduce COMP expression. However, in normal human costochondral cells (hCCCs) infected only with adenoviruses expressing Ribo56, expression of endogenous wild-type COMP mRNA was reduced in a dose-dependent manner by 50%. In chondrocytes that contain heterozygous COMP mutations (D469del, G427E and D511Y) that cause PSACH, Ribo56 was more effective at reducing COMP mRNA (up to 70%). These results indicate that Ribo56 is effective at reducing mutant and wild-type COMP levels in cells and suggests a possible mode of therapy to reduce the mutant protein load.

Variables affecting the stability of multiple cloning sites and hairpin structures in retroviral vectors

Retroviruses are RNA viruses that replicate through a double-stranded DNA intermediate. The viral enzyme reverse transcriptase copies the retroviral genomic RNA into this DNA intermediate through the process of reverse transcription. Many variables can affect the fidelity of reverse transcriptase during reverse transcription, including specific sequences within the retroviral genome. ^ Previous studies have observed that multiple cloning sites (MCS) and sequences predicted to form stable hairpin structures are hotspots for deletion during retroviral replication. The studies described in this dissertation were performed to elucidate the variables that affect the stability of MCS and hairpin structures in retroviral vectors. Two series of retroviral vectors were constructed and characterized in these studies. ^ Spleen necrosis virus-based vectors were constructed containing separate MCS insertions of varying length, orientation, and symmetry. The only MCS that was a hotspot for deletion formed a stable hairpin structure. Upon more detailed study, the MCS previously reported as a hotspot for deletion was found to contain a tandem linker insertion that formed a hairpin structure. Murine leukemia virus-based vectors were constructed containing separate sequence insertions of either inverted repeat symmetry (122IR) that could form a hairpin structure, or little symmetry (122c) that would form a less stable structure. These insertions were made into either the neomycin resistance marker ( neo) or the hygromycin resistance marker (hyg) of the vector. 122c was stable in both neo and hyg, while 122IR was preferentially deleted in neo and was remarkably unstable in hyg. ^ These results suggest that MCS are hotspots for deletion in retroviral vectors if they can form hairpin structures, and that hairpin structures can be highly unstable at certain locations in retroviral vectors. This information may contribute to improved design of retroviral vectors for such uses as human gene therapy, and will contribute to a greater understanding of the basic science of retroviral reverse transcription. ^

Mechanisms of Ad-p53 induced apoptosis in human malignant glioma

Malignant brain tumors are one of the most challenging cancers affecting society today. In a recent survey, an estimated 17,000 annual cases were recorded with a staggering total of 13,300 deaths. A unique degree of heterogeneity typifies glial tumors and presents a challenge for solitary anti-neoplastic treatments. Tumors subsist as heterogeneous masses that progress through dysplasia to astrocytomas, mixed glioma and glioblastoma multiforme. Although traditional therapeutic approaches have provided increments of success, the median survival time remains 12 months. The urgency to improve upon current clinical protocols has encouraged alternative experimental strategies such as p53 adenoviral gene therapy (Ad-p53). This study addresses the efficacy of Ad-p53 for the treatment of glioma. Our model presents a tumor response that is unique among human cancers. Ad-p53 effectively induces apoptosis in mutant p53 expressing cells yet fails to do so in those with wildtype p53. In order to adopt Adp53 as a standard anti-cancer modality, we characterized the role of the tumor suppressor gene p53 in mediating apoptosis. We demonstrate that altering cellular p53 status through the introduction of a dominant negative mutant p53 (175H, 248W, 273H) sensitized cells to Ad-p53. We discovered that wild-type p53 expressing glioma cells retain the apoptotic machinery necessary to accomplish cell death, but have developed mechanisms that interfere with p53 signaling. Earlier studies have not addressed the mechanisms of Ad-p53 apoptosis nor the resistance exhibited by wild-type p53 glioma. To explain the divergent phenotypes, we identified apoptotic pathways activated and effectors of the response. We illustrated that modulation of the death receptor Fas/APO-1 is a principal means of Ad-p53 signaling that is impaired in wild-type p53 glioma. Moreover, the apoptotic response was found to be a multi-faceted process that engaged several caspases, most notably caspases -1, -3 and -8. Lastly, we assessed the ability of anti-apoptotic molecules Bcl-2 and CrmA to inhibit Ad-p53 apoptosis. These studies revealed that Ad-p53 is a powerful tool for inducing apoptosis that can be delayed but not inhibited by anti-apoptotic means. This work is critical for understanding the development of glioma and the phenotypic and genotypic alterations that account for tumor resistance. ^

Antitumor activity of an Ets protein, PEA3, in breast cancer cell lines MDA-MB-361DYT2 and BT474M1

Polyomavirus enhancer activator 3 (PEA3) is a member of the Ets family of transcription factors. We demonstrated in a previous study that, through down-regulating the HER-2/neu oncogene at the transcriptional level, PEA3 can inhibit the growth and tumor development of HER-2/neu-overexpressing ovarian cancer cells. Here, we established stable clones of the human breast cancer cell line MDA-MB-361DYT2 that express PEA3 under the control of a tetracycline-inducible promoter. The expression of PEA3 in this cell line inhibited cell growth and resulted in cell cycle delay in the G1 phase independently of the HER-2/neu down-regulation. In an orthotopic breast cancer model, we showed that expression of PEA3 inhibited tumor growth and prolonged the survival of tumor-bearing mice. In a parallel experiment in another breast cancer cell line, BT474M1, we were unable to obtain stable PEA3-inducible transfectants, which suggests that PEA3 possessed a strong growth inhibitory effect in this cell line. Indeed, PEA3 coupled with the liposome SN2 demonstrated therapeutic effects in mice bearing tumors induced by BT474M1. These results provide evidence that the PEA3 gene could function as an antitumor and gene therapy agent for human breast cancers. ^

Mechanism of anti-tumor activity of adenovirus type 5 E1A in ovarian cancer

The adenovirus type 5 E1A gene was originally developed as a gene therapy to inhibit tumorigenicity of HER-2-overexpressing cells by transcriptional downregulation of HER-2. Our goal is to improve the overall efficacy of E1A gene therapy. To achieve this goal, we have conducted two preclinical experiments. ^ First, we hypothesized that Bcl-2 overexpressing ovarian cancer is resistant to E1A gene therapy. This hypothesis is based on that the 19 kDa protein product of the adenoviral E1B gene which is homologous to Bcl-2 inhibits E1A-induced apoptosis. Treating high Bcl-2-xpressing cells with E1A in combination with an antisense oligonucleotide to Bcl-2 (Bcl-2-ASO) resulted in a significant decrease in cell viability due to an increased rate of apoptosis relative to cells treated with E1A alone. In an ovarian cancer xenograft model, mice implanted with low HER-2, high Bcl-2 cells, treated with E1A plus Bcl-2-ASO led to prolonged survival. Bcl-2 thus may serve as a predictive molecular marker enabling us to select patients with ovarian cancer who will benefit significantly from E1A gene therapy. ^ Second, we elucidated the molecular mechanism governing the anti-tumor effect of E1A in ovarian cancer to identify a more potent tumor suppressor gene. We identified PEA-15 (phospho-protein enriched in astrocytes) upregulated in E1A transfected low HER-2-expressing OVCAR-3 ovarian cancer cell, which showed decreased cell proliferation. PEA-15 moved ERK from the nucleus to the cytoplasm and inhibited ERK-dependent transcription and proliferation. Using small interfering RNA to knock down PEA-15 expression in OVCAR-3 cells made to constitutively express E1A resulted in accumulation of phosphoERK in the nucleus, an increase in Elk-1 activity, DNA synthesis, and anchorage-independent growth. PEA-15 also independently suppressed colony formation in some breast and ovarian cancer cell lines in which E1A is known to have anti-tumor activity. We conclude that the anti-tumor activity of E1A depends on PEA-15. ^ In summary, (1) Bcl-2 may serve as a predictive molecular marker of E1A gene therapy, allowing us to select patients and improve efficacy of E1A gene therapy. (2) PEA-15 was identified as a component of the molecular mechanism governing the anti-tumor activity of E1A in ovarian cancer, (3) PEA-15 may be developed as a novel therapeutic gene. ^

Neurocognitive impairment in relation to glutathione S-transferase enzyme polymorphisms among medulloblastoma patients

Background. Medulloblastoma is a type of brain cancer that accounts for approximately 7-8% of all intracranial tumors and 20-30% of pediatric brain tumors. It is the most common type of malignant brain tumor in childhood. It was reported that majority of survivors with medulloblastoma have social problems, endocrine deficits, and neurological complications. Furthermore, all had significant deficits in neurocognitive functioning. Glutathione S-transferases belong to a family of isoenzymes that catalyze the glutathione conjugation of a variety of electrophilic compounds. ^ Objective. We aimed to determine whether the development of neurocognitive impairment is associated with GST polymorphisms among children and adolescents diagnosed with medulloblastoma (MB) after radiation therapy. ^ Methods. A pilot study composing of 16 children and adolescents diagnosed with MB at Texas Children's Cancer Center was conducted. The t-test was used to determine if the GST polymorphisms were related to neurocognitive impairment and logistic regression was performed to explore association between GST polymorphisms and gender, age at diagnosis, race/ethnicity, and risk group. ^ Results. An association was observed between GSTT1 polymorphism and cognitive impairment one year after radiation and GSTM1 polymorphism two years after radiation. It was observed that patients with GSTT1 null genotype have lower performance IQ (p=0.03) and full scale IQ (p=0.02) one year after radiation and patients with GSTM1 null genotype have lower verbal IQ (p=0.02) two years after radiation. Patients under age 8 have a statistically non-significant higher risk of having not null genotypes compared to those older than age 8 (OR= 7.5, 95%CI: 0.62-90.65 and OR= 2.63, 95%CI: 0.30-23.00 for GSTT1 and GSTM1 respectively). ^ Conclusion. There was a significant association between GSTT1 polymorphism and cognitive impairment one year after radiation and between GSTM1 polymorphism and cognitive impairment two years after radiation. Further large scale studies may be needed to confirm this finding and to examine the underlying mechanism of neurocognitive impairments after treatment of medulloblastoma patients.^

Simultaneous induction of NKG2D and NKG2D ligand for promoting immune surveillance by IL-12 plus doxorubicin treatment

NKG2D (natural killer group 2, member D) and its ligands interaction in tumor microenvironment directs tumor infiltrating immune cells to recognize tumor cells, stimulate cytotoxic effector immune cells, and therefore eradicate tumor cells. IL-12, a cytokine produced by antigen presenting cells, has remarkable antitumor effect by activating innate and adaptive immunity. Doxorubicin, a commonly used chemotherapeutic agent also boosts the host antitumor immune response to cause tumor cell death. Our previous publication suggests that IL-12 plus doxorubicin enhances NKG2D function-dependent inhibition of tumor progression and promotes CD8+T cells infiltrating into tumors. The purpose of this study is to determine the underlying mechanism. Our study reveals a novel function of doxorubicin, which is to augment IL-12–induced NKG2D expression in CD8+T cells but not in NK or CD4+T cells. This observation was further validated by NK and CD8+T cell-depletion studies, in which only depletion of CD8+T cells abolished the expression of NKG2D in lymphocytes. The induced NKG2D expression in CD8+T cells is tightly associated with tumor-specific localization of CD8+T cells and improved antitumor efficacy. The IL-12 plus doxorubicin treatment-induced antitumor efficacy is also due to NKG2D ligand Rae-1 induction in tumors. Rae-1 induction in tumors is a long term effect in multiple tumor models, but not in normal tissues. A novel CD8+T cell direct contact dependent mechanism accounts for Rae-1 induction in vivo and in vitro, and CD80 is the receptor through which CD8+T cells interplay with tumor cells to upregulate Rae-1 on tumor cells. In summary, increased NKG2D expression in CD8+T cells in response to IL-12 plus doxorubicin was closely associated with tumor-specific localization of CD8+T cells and greater antitumor efficacy of the combined regimen than either agent alone. NKG2D ligand Rae-1 induction is triggered by the interaction of CD80 on tumor cells with tumor infiltrating CD+8 T cells.

Potential roles for bcl-2 in the pathogenesis, progression, and treatment of prostate cancer

The most common molecular alterations observed in prostate cancer are increased bcl-2 protein expression and mutations in p53. Understanding the molecular alterations associated with prostate cancer are critical for successful treatment and designing new therapeutic interventions. Hormone-ablation therapy remains the most effective nonsurgical treatment; however, most patients will relapse with hormone-independent, refractory disease. This study addresses how hormone-ablation therapy may increase bcl-2, develops a transgenic model to elucidate the role of bcl-2 multistep prostate carcinogenesis, and assesses how bcl-2 may confer resistance to cell death induction using adenoviral wild-type p53 gene therapy. ^ Two potential androgen response elements were identified in the bcl-2 promoter. Bcl-2 promoter luciferase constructs were transfected into the hormone- sensitive LNCaP prostate cell line. In the presence of dihydrotestosterone, the activity of one bcl-2 promoter luciferase construct was repressed 40% compared to control cells grown in charcoal-stripped serum. Additionally, it was demonstrated that both bcl-2 mRNA and protein were downregulated in the LNCaP cells grown in the presence DHT. This suggests that DHT represses bcl-2 expression through possible direct and indirect mechanisms and that hormone-ablation therapy may actually increases bcl-2 protein. ^ To determine the role of bcl-2 in prostate cancer progression in vivo, probasin-bcl-2 mice were generated where human bcl-2 was targeted to the prostate. Increased bcl-2 expression rendered the ventral prostate more resistant to apoptosis induction following castration. When the probasin-bcl-2 mice were crossed with TRAMP mice, the latency to tumor formation was decreased. The expression of bcl-2 in the double transgenic mice did not affect the incidence of metastases. The double transgenic model will facilitate the study of in vivo effects of specific genetic lesions during the pathogenesis of prostate cancer. ^ The effects of increased bcl-2 protein on wild-type adenoviral p53-mediated cell death were determined in prostatic cell lines. Increased bcl-2 protected PC3 and DU145 cell lines, which possess mutant p53, from p53-mediated cell death and reductions in cell viability. Bcl-2 did not provide the same protective effect in LNCaP cell line, which expresses wild-type p53. This suggests that the ability of bcl-2 to protect against p53-mediated cell death is dependent upon the endogenous status of p53. ^

Effects of exogeneous p53 on growth suppression, apoptosis, and differentiation in oral cancer cell lines

The p53 gene is known to be one of the most commonly mutated genes in human cancers. Many squamous cell carcinomas of the head and neck (SCCHNs) have been shown to contain nonfunctional p53 as well. The use of p53-mediated gene therapy to treat such cancers has become an intensive area of research. Although there have been varied treatment responses to p53 gene therapy, the role that endogenous p53 status plays in this response has not been thoroughly examined. Because of this, the hypothesis of this study examined the role that the endogenous p53 status of cells plays in their response to p53 gene therapy. To test this, an adenoviral vector containing p53 (p53FAd) was administered to three squamous cell carcinoma lines with varied endogenous p53. The SCC9 cell line demonstrates no p53 protein expression, the SCC4 cell line displays overexpression of a mutant p53 protein, and the 1986LN cell line displays low to no expression of wild-type p53 protein as a consequence of human papillomavirus infection. After treatment with p53FAd, the cells were examined for evidence of exogenous p53 expression, growth suppression, alterations in cellular proteins, G1 growth arrest, apoptosis, and differentiation state. Each cell line exhibited exogenous p53 protein. Growth suppression was seen most prominently in the SCC9 cells, to some extent in the 1986LN cells, and little was seen with the SCC4 cells. WAF1/p21 protein was induced in all three cell lines, while PCNA, bcl-2, and bax expression was not significantly affected in any of the lines. Apoptosis developed first in SCC9 cells, next in 1986LN cells, with little seen in the SCC4 cells. The SCC9 line was the only line to show significant GI growth arrest. No significant differences were observed in the overall expression of differentiation markers, aside from increased keratin 13 mRNA levels in all three lines indicating a possible tendency toward differentiation. This study indicates that the endogenous p53 status of squamous cell carcinomas appears to play a critical role in determining the response to p53 adenoviral gene therapy. ^

POLG DNA testing as an emerging standard of care before instituting valproic acid therapy for pediatric seizure disorders.

PURPOSE: To review our clinical experience and determine if there are appropriate signs and symptoms to consider POLG sequencing prior to valproic acid (VPA) dosing in patients with seizures. METHODS: Four patients who developed VPA-induced hepatotoxicity were examined for POLG sequence variations. A subsequent chart review was used to describe clinical course prior to and after VPA dosing. RESULTS: Four patients of multiple different ethnicities, age 3-18 years, developed VPA-induced hepatotoxicity. All were given VPA due to intractable partial seizures. Three of the patients had developed epilepsia partialis continua. The time from VPA exposure to liver failure was between 2 and 3 months. Liver failure was reversible in one patient. Molecular studies revealed homozygous p.R597W or p.A467T mutations in two patients. The other two patients showed compound heterozygous mutations, p.A467T/p.Q68X and p.L83P/p.G888S. Clinical findings and POLG mutations were diagnostic of Alpers-Huttenlocher syndrome. CONCLUSION: Our cases underscore several important findings: POLG mutations have been observed in every ethnic group studied to date; early predominance of epileptiform discharges over the occipital region is common in POLG-induced epilepsy; the EEG and MRI findings varying between patients and stages of the disease; and VPA dosing at any stage of Alpers-Huttenlocher syndrome can precipitate liver failure. Our data support an emerging proposal that POLG gene testing should be considered in any child or adolescent who presents or develops intractable seizures with or without status epilepticus or epilepsia partialis continua, particularly when there is a history of psychomotor regression.

Comparison of multiple comparison methods for identifying differential gene expression in simulated and real papillary thyroid cancer microarray data

The difficulty of detecting differential gene expression in microarray data has existed for many years. Several correction procedures try to avoid the family-wise error rate in multiple comparison process, including the Bonferroni and Sidak single-step p-value adjustments, Holm's step-down correction method, and Benjamini and Hochberg's false discovery rate (FDR) correction procedure. Each multiple comparison technique has its advantages and weaknesses. We studied each multiple comparison method through numerical studies (simulations) and applied the methods to the real exploratory DNA microarray data, which detect of molecular signatures in papillary thyroid cancer (PTC) patients. According to our results of simulation studies, Benjamini and Hochberg step-up FDR controlling procedure is the best process among these multiple comparison methods and we discovered 1277 potential biomarkers among 54675 probe sets after applying the Benjamini and Hochberg's method to PTC microarray data.^

BAYESIAN STATISTICAL METHODS IN GENE-ENVIRONMENT AND GENE-GENE INTERACTION STUDIES

Complex diseases such as cancer result from multiple genetic changes and environmental exposures. Due to the rapid development of genotyping and sequencing technologies, we are now able to more accurately assess causal effects of many genetic and environmental factors. Genome-wide association studies have been able to localize many causal genetic variants predisposing to certain diseases. However, these studies only explain a small portion of variations in the heritability of diseases. More advanced statistical models are urgently needed to identify and characterize some additional genetic and environmental factors and their interactions, which will enable us to better understand the causes of complex diseases. In the past decade, thanks to the increasing computational capabilities and novel statistical developments, Bayesian methods have been widely applied in the genetics/genomics researches and demonstrating superiority over some regular approaches in certain research areas. Gene-environment and gene-gene interaction studies are among the areas where Bayesian methods may fully exert its functionalities and advantages. This dissertation focuses on developing new Bayesian statistical methods for data analysis with complex gene-environment and gene-gene interactions, as well as extending some existing methods for gene-environment interactions to other related areas. It includes three sections: (1) Deriving the Bayesian variable selection framework for the hierarchical gene-environment and gene-gene interactions; (2) Developing the Bayesian Natural and Orthogonal Interaction (NOIA) models for gene-environment interactions; and (3) extending the applications of two Bayesian statistical methods which were developed for gene-environment interaction studies, to other related types of studies such as adaptive borrowing historical data. We propose a Bayesian hierarchical mixture model framework that allows us to investigate the genetic and environmental effects, gene by gene interactions (epistasis) and gene by environment interactions in the same model. It is well known that, in many practical situations, there exists a natural hierarchical structure between the main effects and interactions in the linear model. Here we propose a model that incorporates this hierarchical structure into the Bayesian mixture model, such that the irrelevant interaction effects can be removed more efficiently, resulting in more robust, parsimonious and powerful models. We evaluate both of the 'strong hierarchical' and 'weak hierarchical' models, which specify that both or one of the main effects between interacting factors must be present for the interactions to be included in the model. The extensive simulation results show that the proposed strong and weak hierarchical mixture models control the proportion of false positive discoveries and yield a powerful approach to identify the predisposing main effects and interactions in the studies with complex gene-environment and gene-gene interactions. We also compare these two models with the 'independent' model that does not impose this hierarchical constraint and observe their superior performances in most of the considered situations. The proposed models are implemented in the real data analysis of gene and environment interactions in the cases of lung cancer and cutaneous melanoma case-control studies. The Bayesian statistical models enjoy the properties of being allowed to incorporate useful prior information in the modeling process. Moreover, the Bayesian mixture model outperforms the multivariate logistic model in terms of the performances on the parameter estimation and variable selection in most cases. Our proposed models hold the hierarchical constraints, that further improve the Bayesian mixture model by reducing the proportion of false positive findings among the identified interactions and successfully identifying the reported associations. This is practically appealing for the study of investigating the causal factors from a moderate number of candidate genetic and environmental factors along with a relatively large number of interactions. The natural and orthogonal interaction (NOIA) models of genetic effects have previously been developed to provide an analysis framework, by which the estimates of effects for a quantitative trait are statistically orthogonal regardless of the existence of Hardy-Weinberg Equilibrium (HWE) within loci. Ma et al. (2012) recently developed a NOIA model for the gene-environment interaction studies and have shown the advantages of using the model for detecting the true main effects and interactions, compared with the usual functional model. In this project, we propose a novel Bayesian statistical model that combines the Bayesian hierarchical mixture model with the NOIA statistical model and the usual functional model. The proposed Bayesian NOIA model demonstrates more power at detecting the non-null effects with higher marginal posterior probabilities. Also, we review two Bayesian statistical models (Bayesian empirical shrinkage-type estimator and Bayesian model averaging), which were developed for the gene-environment interaction studies. Inspired by these Bayesian models, we develop two novel statistical methods that are able to handle the related problems such as borrowing data from historical studies. The proposed methods are analogous to the methods for the gene-environment interactions on behalf of the success on balancing the statistical efficiency and bias in a unified model. By extensive simulation studies, we compare the operating characteristics of the proposed models with the existing models including the hierarchical meta-analysis model. The results show that the proposed approaches adaptively borrow the historical data in a data-driven way. These novel models may have a broad range of statistical applications in both of genetic/genomic and clinical studies.