Cancer is the second leading cause of death in the United States. The majority of cases are caused by sporadic somatic mutation, which leads to cellular transformation over time. Therefore, cancer gene identification is a major focus of current research efforts. Understanding how key driver mutations result in cancer could lead to the design of better targeted therapies. The Sleeping Beauty (SB) transposon system can be used to identify driver mutations in a variety of tumor types. SB mutagenesis mimics the sporadic accumulation of somatic mutations found in spontaneous human cancers. This system also has an additional benefit over chemical carcinogenesis models in that key cancer gene candidates are easily identified with high-throughput sequencing and subsequent bioinformatic analysis. Using SB, our lab recently identified a novel oncogene involved in non-melanoma skin cancer called Zmiz1, the biological function of which is not well studied. A major focus of my thesis work was to characterize Zmiz1 and its role in skin cancer. This gene encodes a protein with predicted E3 SUMO ligase activity. My work has provided the first evidence firmly establishing an oncogenic role for Zmiz1 in cutaneous malignancy, thereby generating a novel transgenic mouse model of skin carcinogenesis. Importantly, we observed tumor-specific overexpression of an endogenous ZMIZ1 isoform in human squamous cell carcinomas. Non-melanoma skin cancer is the most common malignancy worldwide, and it disproportionally affects immunosuppressed patients. One proposed explanation for this is the concept of tumor immunosurveillance, whereby the immune system suppresses tumor growth. When the immune system is compromised, transformed cells can develop into tumors. However, immunocompetent people also develop cancer, despite an intact immune system. It is thought that while the immune system is keeping transformed cells from forming a tumor, it simultaneously influences the acquisition of new mutations that eventually allow escape from immune detection and clearance. This process, called immunoediting, is widely believed to be dependent upon the adaptive immune system and another focus of my research was studying immunoediting mechanisms using SB mutagenesis. Subtle differences were observed in the SB-induced mutation spectra of tumors generated in immunocompetent mice and immunocompromised mice, suggesting that some level of lymphocyte-dependent immunoediting of tumors had occurred. However, the adaptive immune system was not effective in suppressing tumor formation, which is in contrast with previously published data. My work represents an independent and original assessment of the immunoediting process, and cautions against reliance on a single animal model to study this area of cancer biology.

The neural crest (NC) is a vertebrate specific early embryonic stem-cell population that forms an array of derivatives in the developing embryo. These cells are induced from the developing ectoderm at the boundary of the neural plate and non-neural ectoderm (termed the neural plate border). After initial induction the NC migrate along predefined routes within the developing embryo and upon reaching their destination begin to differentiate into this myriad of derivatives. These derivatives include neurons, glia, cartilage, bone, connective tissue and pigmented melanocytes, among others. Through genetic studies a variety of genes important in these various stages of NC development have been identified. Together, these genes comprise the emerging NC gene-regulatory-network (NC-GRN). Although a variety of genetic components have been identified exactly how they interact with one another during NC development is not fully understood. One key family of genes important in the NC-GRN is the transcription factor activator protein 2 (TFAP2). Initial studies had identified an important role for TFAP2 in initial NC induction, as elimination of two family members (TFAP2A and TFAP2C) resulted in a complete absence of NC and all subsequent derivatives. Although TFAP2 was critical for NC development it was not clear exactly how TFAP2 fit into the NC-GRN and whether additional roles of TFAP2 existed within this network. Here in, we have addressed these questions by identifying a key transcriptional target of TFAP2, SOX10, during NC development. In addition, we have examined how TFAP2 is regulated by Wnt-signaling during NC development. Finally, we have identified an additional role for TFAP2 family members in regulating an additional step of NC development mainly the differentiation of NC-derived melanocytes and identified several targets that may mediate this effect. Together these findings provide a more clear understanding of TFAP2's role in the NC-GRN during development. Given disruption of this network contributes to disease of the NC these findings should provide insight into the etiology as well as provide potential therapeutic intervention for these diseases. In addition, understanding differences in the vertebrate NC-GRN to non-vertebrate GRN's should provide insight into the emergence of this important cell type during evolution.

Primary cilia are sensory organelles present on most differentiated cells. Their assembly or function is impaired in several human diseases, including polycystic kidney disease (PKD). However, the mechanism by which ciliary dysfunction contributes to these pathologies is incompletely understood. PKD is characterized by altered trans-epithelial fluid transport, due to the mislocalization of ion pumps and changes in tight junctions. How ciliary dysregulation might lead to such changes in cell polarity and tight-junction function has not been shown. The overall hypothesis of this thesis is that ciliary dysfunction promotes specific changes in epithelial junctions and cell polarity that contribute to PKD. We developed an experimental model in which to test whether ciliary disruption indeed causes changes in cell polarity and tight junctions, and used it in conjunction with well-established methods to chemically induce ciliary shedding from cells, and with assays measuring the distribution of surface and secretory proteins in Madin-Darby canine kidney (MDCK) cells. This analysis revealed that several proteins, including Na/K-ATPase, were mislocalized following deciliation and that tight junctions became less permeable to paracellular ion flux. We infer that the combined effects of mislocalized ion transporters and increased tight-junction barrier function contribute to cyst formation in PKD. The functional specificity of tight junctions is determined by their expression of claudins, a protein family consisting of 24 members. Of these, claudin 2 expression has been shown to lead to an increase in junctional leakiness. We found that upon deciliation, claudin 2 expression was lost and trans-epithelial resistance increased 3-fold. The signaling pathway connecting deciliation to claudin 2 repression required ERK activation; phospho-ERK levels were elevated after exposure to deciliation stimuli, and an inhibitor of ERK activation blocked claudin 2 loss. The down-regulation of claudin 2 did not involve post-translational regulation, because protein half-life was not altered following deciliation. In contrast, the half-life of claudin 2 mRNA was greatly reduced in deciliated cells. Finally, when murine claudin 2 was ectopically expressed in MDCK cells, its expression was also suppressed upon deciliation. Collectively, these results support a model in which deciliation causes an ERK-dependent decrease in claudin 2 mRNA stability, and highlight a mechanism whereby ciliary dysregulation might contribute to altered trans-epithelial fluid transport associated with PKD.

Cell polarity is the asymmetric distribution of organelles that almost all cells use to separate individual processes and perform complex functions. Although the manner in which cells are polarized is very diverse, the processes necessary to assume polarized phenotypes are similar in many cell types. Epithelial cell polarization is of particular importance, as these cells serve form linings of organs and act as barriers distinguishing different compartments. Furthermore, loss of epithelial polarization occurs in some disease states and may result in cell invasion through underlying matrix. During initial polarization, vesicle trafficking is indispensible for assembly of structures, including apical junctional complex formation. Trafficking of new membrane and associated proteins to leading edges is also necessary for cell migration. RalA and RalB are members of the Ras superfamily of GTPases and have been implicated in several processes, including vesicle trafficking. Only 5 Ral effectors have been identified, two of which are members of the Exocyst complex, a hetero-octameric complex also involved with vesicle trafficking. I hypothesized that Ral GTPases were necessary for several aspects of cell polarization, and that they engage the Exocyst complex to mediate these processes. Initial investigation of tight junction assembly found that both RalA and RalB antagonistically affect paracellular permeability. Knockdown of RalA and RalB resulted in decreased and increased incorporation of components into assembling tight junctions, respectively. Furthermore, both RalA and RalB engaged the Exocyst in order to mediate tight junction assembly. I next examined the role of RalA-Sec5 and RalA-Exo84 interactions during tumor cell migration and invasion. Both interactions were necessary for invasion and single cell migration, although disruption of each interaction affected different aspects of migration. Furthermore, significant differences in cytoskeleton organization occurred in response to disruption of RalA-Sec5 and RalA-Exo84 interactions. Finally, I investigated the effects of RalA and RalB knockdown on growth of primary cilia and cyst formation. RalA decreased primary cilia growth and reduced average cilia length, while RalB increased cilia length. Knockdown of RalA and RalB also affected lumen formation during cystogenesis, as RalA knockdown prevented lumen formation and RalB knockdown caused formation of multiple lumens. Taken together, data presented here show that Ral engages the Exocyst to mediate distinct processes during tight junction assembly and cell migration, and implicates Ral GTPases in several different aspects of cell polarity.

Preterm birth (PTB) is a global public health problem that has significant adverse effects on neonatal mortality and morbidity. Progress in understanding the pathological mechanisms underlying PTB has been greatly hampered by the complex and polygenic nature of the disease. As a result, a multifaceted approach may hold promise for identifying true causal factors. The main objective of this thesis is to identify genes that play a role in the etiology of PTB using experimental data derived from different molecular levels (genome, transcriptome, and epigenome). To achieve this goal, we performed association studies using a candidate gene approach to identify genetic factors contributing to PTB. Our analysis of genetic variants in three OXT pathway genes (oxytocin (OXT), oxytocin receptor (OXTR), and leucyl/cystinyl aminopeptidase (LNPEP)) revealed several common polymorphisms in LNPEP that show significant association with prematurity. Large-scale sequence analysis of the OXTR gene identified several novel rare coding variants that might be of etiologic importance. Our results suggest that these variants, in aggregate, appear to make some contribution to susceptibility to PTB. We also examined the gene expression profiles in the human placenta to identify, at the transcriptomic level, candidate genes for PTB. Using splicing-sensitive microarray and deep sequencing technologies, we identified transcriptome signatures that differ between term (with and without labor) and preterm placental tissues and between placental and other human tissues. The transcriptome data were analyzed not only at the gene-level, but also at the exon-level, enabling the detection of alternative splicing events. The exon-level analysis revealed more frequent disruption of alternative splicing in preterm than term placental tissues, indicating that alternative splicing may represent one possible mechanism contributing to PTB. Our study at the epigenomic level was pursued through investigation of placental DNA methylation profiles. We, using a genome-wide approach, detected a panel of genes showing labor- and gestational age-associated methylation differences. Selected genes were validated using bisulfite sequencing and methylation-specific PCR. SLC30A3, a validated differentially methylated gene between term labor and preterm labor amnion tissues, for instance, may potentially play a role in the pathogenesis of PTB. Taken together, this thesis work provides a valuable source of novel candidate genes for PTB, and future research using integrative systems biology approaches may shed light on the molecular mechanisms underlying this complex, heterogeneous disease.

Interleukin-1-beta (IL-1β) is a potent proinflammatory cytokine. A primary outcome of IL-1β signaling is the activation of NFκB, a transcription factor that induces a large number of immune molecules, apoptotic factors, anti-apoptotic factors, and other transcription factors. Recent work has demonstrated that the activation of NFκB involves a multistep redox-signaling cascade that requires endocytosis of the interleukin receptor (IL-1R1)/ligand pair and superoxide production by NADPH oxidase 2 (Nox2) within the resulting newly formed early endosome. Hydrogen peroxide produced by the rapid dismutation of superoxide is necessary for the subsequent downstream recruitment of IL-1R1 effectors (TRAF6, IKK kinases) and ultimately the activation of NFκB. In this thesis, I have further dissected the spatial and temporal events that coordinate signaling processes of the IL-1β pathway. Using a combination of biophotonic imaging, immunofluorescence imaging, and lipid raft density gradient isolation, I demonstrate that both Nox2 and IL-1R1 are constitutively present in lipid raft microdomains on the plasma membrane. Stimulation by IL-1β induces endocytosis of Nox2 and IL-1R1 from the plasma membrane into caveolin-1, lipid raft positive early endosomes. Further, inhibition of lipid raft mediated endocytosis or deletion of caveolin-1 inhibits activation of NFκB, by IL-1β. We have also identified Vav1 as the Rac1 guanine exchange factor that is recruited to caveolin-1 positive lipid rafts following IL-1β stimulation, and demonstrated that dominant negative Vav1 inhibits NFκB activation by IL-1β. Following this work, I utilized assays for redox sensitivity and mass spectrometry to demonstrate that C70, C73, and C105 are hydrogen peroxide sensitive cysteines within the RING domain of TRAF6. I further demonstrate that hydrogen peroxide does not alter the E3 ubiquitin ligase activity associated with the TRAF6 RING domain. My findings suggest that the redox sensitivity of the RING domain mediates TRAF6 recruitment to the receptor complex. This is supported by the observation that hydrogen peroxide treatment of TRAF6, but not early signaling effectors (IL-1R1, IRAK1, IRAK4, MyD88) mediates TRAF6 recruitment to the IL-1 receptor complex. Further, mutation of the identified redox sensitive cysteines inhibits IL-1β signaling and NFκB activation. This research has helped to refine the understanding of the IL-1β signaling pathway, and may ultimately lead to new therapeutic targets for controlling inflammation.

Von Willebrand factor (VWF) is a plasma glycoprotein that can bind collagen at a wound site as well as circulating platelets. VWF forms high molecular weight multimers (>20,000 kDa). VWF can also form VWF strings that appear to be attached to the endothelial surface and are capable of binding platelets. These strings are only observed in vitro and in vivo in the absence of the VWF-cleaving protease ADAMTS13. Deficiency in ADAMTS13 results in thrombotic thrombocytopenic purpura (TTP), a clotting disorder characterized by thrombocytopenia, microangiopathic hemolytic anemia, renal dysfunction, neurological dysfunction and fever. Patients suffering from TTP demonstrate VWF-and platelet-rich thrombi in the microvasculature of numerous organ systems, but most notably in the kidneys, heart, and brain. While VWF strings have not been directly connected to TTP, their presence in vivo was only identified with the ADAMTS13 knockout mouse (a model of TTP), suggesting a possible relationship. Recently we identified glycerol as an agent, similar to histamine, that triggers the formation of VWF strings in vitro. We found that glycerol and histamine trigger TTP in an ADAMTS13-deficient mouse model. In addition, we determined conditions in vitro that promote the formation of dense VWF networks. These networks of VWF can be greater than 70 μm thick and appear to be able to form fibers as long as several millimeters in length. These networks have not been previously identified and may underlie a possible mechanism by which VWF-rich thrombi form in TTP. These networks were formed solely from cultured endothelial cells, leading us to believe that endothelial cells alone are capable of producing more VWF than perhaps previously appreciated. These data suggest that secretion of VWF from the endothelium may play an important role in the pathophysiology of TTP.

The most common cystic fibrosis transmembrane conductance regulator (CFTR) gene mutation is δF508 and this causes cystic fibrosis (CF). Animal models that recapitulate the human disease phenotype are critical to understanding pathophysiologic mechanisms in CF and developing therapies. New CF models in the pig and ferret have been generated that develop lung, pancreatic, liver, and intestinal pathologies that reflect disease in CF patients. Species-specific biology in the processing of CFTR has demonstrated that pig and mouse δF508-CFTR proteins are more effectively processed to the apical membrane of airway epithelia than human δF508-CFTR. The processing behavior of ferret wild-type (WT) and δF508-CFTR proteins remain unknown and such information is important to predicting the utility of a δF508-CFTR ferret. To this end, we sought to compare processing, membrane stability, and function of human and ferret WT- and δF508-CFTR proteins in a heterologous expression system using HT1080, HEK293T, BHK21, and Cos7 cells, as well as human and ferret CF polarized airway epithelia. Analysis of the protein processing and stability by metabolic pulse-chase and surface On-Cell Western blots revealed that WT-fCFTR half-life and membrane stability were increased relative to WT-hCFTR. Furthermore, in BHK21, Cos7, and CuFi cells, human and ferret δF508-CFTR processing was negligible, while low levels of processing of δF508-fCFTR could be seen in HT1080 and HEK293T cells. Only the WT-fCFTR, but not δF508-fCFTR, produced functional cAMP-inducible chloride currents in both CF human and ferret airway epithelia. Further elucidation of the mechanism responsible for elevated fCFTR protein stability may lead to new therapeutic approaches to augment CFTR function. These findings also suggest that generation of a ferret CFTRδF508/δF508 animal model may be useful. Furthermore, in the CFTR and CFTR+/+ ferret model we have characterized abnormalities in the bioelectric properties of the trachea, stomach, intestine and gallbladder of newborn CF ferrets. Short circuit current (ISC) analysis of CF and WT tracheas revealed the following similarities and differences: 1) amiloride sensitive sodium currents were similar between genotypes, 2) responses to 4,4'-diisothiocyano-2,2'-stilbene disulphonic acid (DIDS) were ~4-fold greater in CF animals, suggesting elevated baseline chloride transport through non-CFTR channels, and 3) as expected, there was a lack of IBMX/forskolin-stimulated and GlyH-101-inhibited currents in CF animals due to the lack of CFTR. CFTR mRNA and protein was present throughout all levels of the WT ferret and IBMX/forskolin-inducible ISC was only observed in WT animals. Interestingly, IBMX/forskolin-inducible intestinal ISC in WT animals was not inhibited by the CFTR inhibitor GlyH-101 or bumetanide. The luminal pH of the CF ferret stomach was significantly decreased relative to the controls, while both genotypes maintained near neutral pH along the length of the intestine. The WT stomach and gallbladder exhibited significantly enhanced IBMX/forskolin ISC responses and inhibition by GlyH-101 relative to CF samples. These findings demonstrate that multiple organs affected by disease in the CF ferret have bioelectric abnormalities consistent with the lack of cAMP-mediated chloride transport.

Breast cancer is currently the second most common form of cancer and the second leading cause of death due to cancer in the United States. Breast cancer itself is subdivided into at least four subtypes, luminal A, luminal B, HER2-enriched, and basal-like, based on genomewide molecular expression patterns. Luminal A is the most common form and typically characterized by high levels of estrogen receptor (ER). HER2-enriched cancers usually, but not always, harbor amplified copies of the HER2 oncogene. Luminal B cancers share characteristics with the luminal A and HER2-enriched subtypes. Finally, basal-like cancers are more oftentimes defined by their lack of any markers or molecular targets. Thus, they are often called triple-negative breast cancer. Recent evidence suggests that there are a number transcription factors that play critical roles in the cancer progression of these malignancies. Indeed, TFAP2C has been clearly shown to positively regulate ER in luminal A cancers. Alternatively, TFAP2A appears to play an interesting, but as of yet incompletely, understood role in basal-like cancer. There has been additional evidence that suggests TFAP2C regulates multiple members of the ErbB family of receptor tyrosine kinases. Thus, we hypothesize that the TFAP2 family of transcription factors play a critical role in breast cancer progression. More specifically, we will show that TFAP2A and TFAP2C not only regulate a few critical genes in luminal and basal-like cancer, but instead are responsible for the genomewide expression pattern of these two breast cancer subtypes. Moreover, we argue that TFAP2C's regulation of certain receptor tyrosine kinases in luminal A cancers indicates promising therapeutic targets, particularly with small molecule inhibitors that are already FDA-approved. In addition, we provide data suggesting that TFAP2C likely plays an oncogenic role in HER2-positive breast cancer, possibly through the regulation of certain members of the ErbB family of receptor tyrosine kinases, such as EGFR. To address these points, we use a combination of genetically engineered mouse models, xenografts, siRNA mediated knockdown technology, western blot, qPCR, and number of additional molecular biological techniques. These results will not only establish the family of TFAP2 family of proteins as critical regulators of cancer progression, but our findings will specify how and to what extent each subtype of breast cancer is affected by individual members of the TFAP2 family of transcription factors.

Maspin (Mammary Serine Protease Inhibitor) was first reported in 1994 as a serpin with tumor suppressive properties. Maspin was initially isolated through subtractive hybridization and differential display analysis as a 42kDa protein that is expressed in normal mammary epithelial cells but reduced or absent in breast carcinomas. Further research led to maspin's characterization as a class II tumor suppressor based on its ability to inhibit cell invasion, promote apoptosis and inhibit angiogenesis. Since then, efforts have been made to characterize maspin's tumor suppressive mechanisms. In particular, researchers have studied maspin localization, the regulation of maspin expression, and more recently, maspin protein interactions. We employed a maspin-baited yeast two-hybrid system and subsequently identified Interferon Regulatory Factor 6 (IRF6) as a maspin-binding protein. Whereas many of the IRF family members have been well characterized, IRF6 remains poorly understood. In this dissertation, we elucidate some of the complex mechanisms involved in maspin activity, specifically relating to IRF6 regulation and function. We have examined the expression of IRF6 in breast cancer cells and we show that, similar to maspin, IRF6 is reduced or absent in breast carcinomas. We further show that the re-expression of IRF6 in breast cancer cells results in genotypic and phenotypic changes which can be abrogated in the presence of maspin. We identify ERK1/2 as a kinase involved in IRF6 phosphorylation, and we demonstrate a possible role for toll-like receptor signaling in the activation of IRF6. We also evaluate the differential expression of maspin and IRF6 during murine mammary gland development and we show that both maspin and IRF6 are maximally expressed during lactation. These studies have increased our understanding of the complex, pleiotropic nature of maspin and provide an avenue to develop maspin's potential as a diagnostic marker for cancer progression and as a potentially powerful therapeutic agent in the fight against breast cancer.