Pathological ocular angiogenesis, or ocular neovascularization (NV) is a central feature of retinopathy of prematurity (ROP), proliferative diabetic retinopathy (PDR), and age-related macular degeneration (AMD). In the developed world, these diseases are the leading causes of blindness in infants, working-age individuals, and the elderly, respectively. As of today, there are two FDA-approved angiostatic agents being used to treat conditions characterized by ocular NV. Both angiostatic agents inhibit vascular endothelial growth factor (VEGF), the principle growth factor mediating ocular NV. Although VEGF-centric therapies reduce NV, they do not completely eliminate it. In order to more effectively prevent and/or treat these conditions, a more thorough understanding of the key players involved in the angiogenic cascade is needed. One enzyme that holds promise for therapeutic intervention is cyclooxygenase-2 (COX-2). We examined the involvement of COX-2 and COX-2-derived prostanoids in order to 1) understand their role in ocular angiogenic disease, and 2) develop more specific therapeutic targets for diseases comprised of an angiogenic component. We have shown that non-steroidal anti-inflammatory drugs (NSAIDs), which inhibit the activity of COX, significantly reduced the severity of NV in an animal model or oxygen-induced retinopathy (OIR). This anti-angiogenic effect was likely due to inhibition of VEGF-induced endothelial cell proliferation and tube formation. Furthermore, genetic deletion of COX-2 significantly reduced hypoxia-induced VEGF production by Müller cells. This effect was mediated by one of the five COX-derived prostanoids, PGE2. Additional experiments confirmed the role of PGE2 and, more specifically, the EP4 receptor, in angiogenic Müller cell and endothelial cell behaviors. Importantly, EP4 antagonism significantly reduced VEGF production by hypoxic Müller cells, VEGF-induced proliferation and tube formation in endothelial cells, and the severity of NV in rodent models of OIR and laser-induced choroidal NV (LCNV). Our findings are significant because they demonstrate that the EP4 receptor affects the ocular angiogenic cascade at more than one point, with the potential to be a powerful and effective therapeutic target for angiogenic diseases of the eye and other tissues.

Wound healing, an essential function to the survival of all multicellular organisms, requires the precise orchestration of multiple cell types in order to repair the damaged tissue. Wound healing involves three overlapping phases: inflammation, re-epithelialization, and scar formation. Matrix metalloproteinases (MMPs) are endopeptidases of the metzincin family, which have been shown to function throughout wound healing, but their precise functions and regulatory mechanisms are unclear <em>in vivo</em> largely due to the complications of redundancy. In mammals there are 24 MMPs (seventeen secreted type MMPs and seven membrane-anchored type MMPs). MMPs are inhibited by Tissue Inhibitors of Metalloproteinases (TIMP), of which there are four in mammals. To bypass the complications of redundancy, we utilized <em>Drosophila melanogaster</em> as a model system to study MMPs <em>in vivo</em> during wound healing. <em>Drosophila</em> serve as a simple system to not only study wound healing, but the <em>Drosophila</em> genome encodes only two MMPs, one secreted (<em>Mmp1</em>) and one membrane-anchored (<em>Mmp2</em>), providing a system to elucidate the functions of each MMP class <em>in vivo</em>. Our results indicate that <em>Mmp1</em> and <em>Mmp2</em>, as well as the inhibitor, <em>Timp</em> are required for re-epithelialization, where they may function together to promote cell migration. During re-epithelialization, <em>Mmp1</em>, under the control of the JNK signaling pathway, functions to promote cell migration by facilitating collagen IV remodeling, promoting actin cytoskeleton reorganization, and inducing RTK signaling. <em>Mmp1</em> from the hemocytes may function to limit the area of both JNK and RTK signaling to the immediate vicinity of the wound. Preliminary results suggest that <em>Mmp1</em> and <em>Mmp2</em> may regulate hemostasis and the melanization cascade, as we observe both clotting and melanization defects in <em>Mmp1</em> and <em>Mmp2</em> mutants. In unwounded tissue, both <em>Mmp1</em> and <em>Timp</em> are required for basement membrane maintenance, a function they do not share with <em>Mmp2</em>. The combination of both shared and independent phenotypes between <em>Mmp1</em>, <em>Mmp2</em>, and <em>Timp</em> mutants suggest that Mmp1, Mmp2, and Timp may form a complex to promote wound healing; however, if such a complex does form in vivo it is context-specific.

Canonical Wnt signaling regulates many fundamental developmental processes and is misregulated in a variety of disease states in humans. Ubiquitylation has been shown to play critical roles in the regulation of Wnt signal transduction, and many components of the Wnt pathway are known to be ubiquitylated. At the time I began my thesis work, however, little was known about how the ubiquitin system regulates Wnt signaling. Thus, I designed a RNA interference (RNAi) screen in Drosophila S2 cells to identify novel E3 ligases and deubiquitylases involved in Wg/Wnt signaling and identified XIAP (an E3 ligase) and USP47 (a deubiquitylase) as novel Wnt pathway components. <P>A key event in Wnt signaling is conversion of TCF/Lef from a transcriptional repressor to an activator, yet how this switch occurs is not well understood. Here, I describe an unanticipated role for X-linked Inhibitor of Apoptosis (XIAP) in regulating this critical Wnt signaling event that is independent of its anti-apoptotic function. I identified DIAP1 as a positive regulator of Wingless signaling in a Drosophila S2 cell-based RNAi screen and show that XIAP, its vertebrate homolog, is similarly required for Wnt signaling in cultured mammalian cells and in Xenopus embryos, indicating evolutionary conservation of function. I also demonstrate that upon Wnt pathway activation, XIAP is recruited to TCF/Lef where it mono-ubiquitylates Groucho/TLE: this modification decreases the affinity of Groucho/TLE for TCF/Lef. These data reveal a transcriptional switch involving XIAP-mediated ubiquitylation of Groucho/TLE that facilitates its removal from TCF/Lef, thus allowing assembly of &#946-catenin-TCF/Lef complexes and initiation of a Wnt-specific transcriptional program.</P> <P>In addition to the discovery of XIAP as a novel Wnt pathway component, I also identified the de-ubiquitylase Ubp64E as a positive regulator of Wingless signaling in the Drosophila S2 cell-based RNAi screen. USP47, its vertebrate homolog, is similarly required for Wnt signaling in cultured mammalian cells and in Xenopus embryos, indicating evolutionary conservation of function. My data indicate that USP47 likely functions at the level of transcription in the nucleus potentially through its interaction with the E3 ligases &#946-TRCP or XIAP or an as yet unidentified target.</P>

Colorectal cancer is the second most lethal, non-cutaneous epithelial cancer in the United States. Metastasis contributes to the majority of cancer-related deaths in this disease. Metastasis closely resembles the developmental process of epithelial-mesenchymal transition (EMT) and reliable models to recapitulate this process can be useful to shed light onto the biology and prognosis of patients with colorectal cancer. Herein we describe a mouse model that led to the discovery of a gene signature of metastasis using comparative functional genomics, which identified stage II and III colon cancer patients prone to recurrence and death from metastatic disease, in addition to a low-risk sub-group of stage III patients for whom adjuvant chemotherapy provided no additional survival benefit. These findings form the basis for substantive pre-clinical biomarker testing and eventual translational application to a clinical trial. On the other hand, loss of tumor suppressor genes affects the majority of colorectal cancer patients to permit progression of disease. Many pathways are defective in the pathogenesis of colorectal cancer; however, more than half of the patients have defects in the Wnt/β-catenin and Transforming Growth Factor-β (TGFβ) signaling pathways which are critical for development and intestinal homeostasis. This body of work also uses in vitro and in vivo models to describe a new role for the tumor suppressor Smad4 in the repression of β-catenin transcriptional activity in epithelial cells. This repression was associated with down-regulation of Wnt signaling and reversal of EMT. Clinical relevance of this effect was demonstrated by an epithelial cell-specific, Smad4-modulated gene expression profile associated with Wnt signaling suppression, which contributed prognostic information for colorectal cancer patients independently of pathological staging. These findings should facilitate hypothesis testing for biologically-targeted therapeutic interventions based on TGFβ/Smad and Wnt/β-catenin pathway activity levels. Overall, these results provide insight into the biology of metastasis and tumor suppression in colorectal cancer to promote seamless translation to care of the colorectal cancer patient at the bedside.

Sonic hedgehog (Shh) signaling regulates important biological processes during embryogenesis and adult homeostasis. Deregulation of this essential signaling pathway can lead to congenital defects and tumorigenesis. In this dissertation I explored the functional role of the cholesterol modification on Shh protein during central nervous system (CNS) patterning, determined previously unappreciated regulation of Shh pathway on hindbrain choroid plexus and cerebellum development, and established a novel mouse model for cerebellar tumor, medulloblastoma. These findings provide us new knowledge of how Shh signaling regulates various CNS organ development, as well as novel insights into the oncogenic processes in Shh-dependent brain tumor.

THE ROLE OF FOXM1 IN GROWTH FACTOR-MEDIATED PANCREATIC BETA-CELL PROLIFERATION JIA ZHANG Dissertation under the direction of Professor Maureen A. Gannon Both type I and type II diabetes are due to an either absolute or relative loss of â-cell mass. A thorough understanding of â-cell mass regulation is needed for treatment of diabetes by restoring â-cell mass and glucose homeostasis. In adults, â-cell mass is replenished mainly through â-cell proliferation. Growth factors such as placental lactogen (PL), insulin-like growth factor-1 (IGF-1) and hepatocyte growth factor (HGF) are potent â-cell mitogens. However, little is known about the intrinsic factors within a â-cell that mediate the mitogenic effects of growth factors. In searching for those intrinsic factors, previous studies from our laboratory have demonstrated that Foxm1, a pro-proliferation transcription factor, is required for the maintenance and expansion of adult â-cell mass under basal conditions and in injury models. This dissertation examined the hypothesis that Foxm1 plays a central role downstream of one or multiple growth factors to stimulate â-cell proliferation. The pregnancy hormone PL is mainly responsible for increased â-cell proliferation in pregnancy. We found that Foxm1 was up-regulated in maternal islets during pregnancy and PL induced Foxm1 expression in cultured islets. Additionally, Foxm1 was required for PL-mediated â-cell proliferation in vivo. Our preliminary data further identified Foxm1 as a novel direct target of the transcription factor Stat5, a downstream effector of PL signaling. To determine whether Foxm1 is essential for increases in â-cell proliferation and mass during pregnancy, we evaluated female mice with a pancreas-wide Foxm1 inactivation (Foxm1Äpanc). These mice developed gestational diabetes. Compared to controls, Foxm1Äpanc mice exhibited blunted â-cell proliferation and mass in response to pregnancy, which is likely due to the elevated islet expression or activity of two cell cycle inhibitors, Menin and p27. Foxm1-mediated â-cell proliferation is thus crucial for â-cell mass expansion and glucose homeostasis during pregnancy. Our ongoing endeavors include examining whether Foxm1 is required for HGF and IGF-1-stimulated â-cell proliferation. In conclusion, the central role of Foxm1 in â-cell maintenance and growth factor stimulated-growth makes it a promising therapeutic candidate for enhancing â-cell mass in vivo.

The COPII machinery is the primary mediator of ER-to-Golgi transport and mutations in different COPII (Coat Protein II complex) paralogs are associated with diverse human diseases. Each tissue has vastly different requirements for the secretory machinery in terms of cargo type and volume, and COPII must accommodate these requirements. The mechanisms by which COPII function may be regulated, however, are only beginning to be elucidated. Because skeletal tissues are highly secretory and traffic large cargoes such as collagen, we used forward genetics approaches in zebrafish coupled with a positional cloning strategy to identify novel genes important for protein trafficking in skeletal tissues. Here we identify a mutation in the transcription factor feelgood/creb3l2, which directly regulates the expression of genes encoding select COPII proteins, including sec23a and sec24d. We found that these COPII components, but not their close paralogs sec23b or sec24c, are required for collagen secretion in chondrocytes and the notochord, but not for secretion of other cargos such as peptidoglycans, laminin, and N-cadherin. Loss of function of sec23a and sec23b paralogs leads to highly divergent phenotypes. We found that collagen deposition in the notochord is dependent on Sec23a while fibronectin deposition is dependent on Sec23b. To evaluate whether the unique phenotypes are due to differences in expression or function of sec23 paralogs, we have performed rescue experiments in crusher/sec23a mutants. sec23a overexpression in crusher was able to rescue morphology of the pectoral fin, and collagen trafficking within the fin, significantly better than sec23b overexpression. These data suggest a functional difference between the paralogs. Because Sec23a and Sec23b amino acid sequence is 93% similar except for a highly divergent 18 amino acid loop, we hypothesized that this loop may be essential for their functional difference. We therefore build a sec23b construct containing the 18AA loop from sec23a, as well as the reciprocal sec23a with the sec23b loop. In contrast to WT sec23b, overexpression of sec23b with the sec23a loop was able to efficiently rescue collagen secretion and fin morphology in crusher/sec23a mutants, while sec23a with the sec23b loop failed to do so. These data indicate that functional differences between sec23a and sec23b exist and that the highly divergent 18 amino acid loop plays a role in collagen secretion. Our work has established that Creb3l2/Sec23a/Sec24D are critical for collagen/ECM secretion during the maturation stage of chondrogenesis and suggests that this secretory axis is central to a broad transcriptional regulatory network that optimizes cellular machinery to facilitate trafficking in highly secretory cell types.

Although it is known that the response of endothelial cells to atherogenic disturbed flow is distinct from that elicited by non-atherogenic laminar flow, the mechanisms involved are poorly understood. Observing that expression of the endothelial receptor tyrosine kinase, Tie1, is evident only at regions of atherogenic flow in mature animals, we hypothesized that Tie1 plays a role in the endothelial response to atherogenic shear stress. Because deletion of Tie1 results in embryonic lethality secondary to vascular dysfunction, we utilized conditional and inducible mutagenesis (Tie1-/flox:SCL-ERT-Cre) to study the effect of Tie1 attenuation on atherogenesis in apolipoprotein E deficient (ApoE-/-) mice, and found a dose dependent decrease (70%) in atherosclerotic lesions. To test our hypothesis, we isolated primary aortic endothelial cells from Tie1flox/flox:SCL-ERT-Cre immorto mice and found that atheroprotective laminar flow decreased Tie1 expression in vitro. Attenuation of Tie1 was associated with an increase in eNOS expression and Tie2 phosphorylation, In addition, Tie1 attenuation increased IkB-α expression while attenuating. In summary, we found that shear stress conditions that modulate atherogenic events also regulate Tie1 expression and Tie1 may play a novel pro-inflammatory role in atherosclerosis.

Mesothelium is a simple squamous epithelium that forms the surface layer of all coelomic cavities and organs. Despite its widespread localization, mesothelial development has been studied almost exclusively in the heart. In this organ, mesothelial progenitors originate external to the initial heart tube and migrate across the pericardial cavity to contact and envelop the heart forming the cardiac mesothelium (epicardium). Newly formed epicardial cells then delaminate from the epithelium, invade the myocardium, and differentiate into fibroblasts, vascular smooth muscle and endothelial cells. Thus, the initial heart tube, composed of splanchnic mesoderm and endocardial cells, is devoid of both mesothelial and vascular progenitor populations and must recruit these cells from an exogenous source. Studies of the liver, lungs, and intestine demonstrated a similar capacity for mesothelial cells to invade each respective organ and contribute to the vasculature indicating mesothelial cells may have a conserved role in coelomic vasculogenesis. The aim of the research presented here was to identify the origin of mesothelial and vascular cells of the intestine to determine if the extensively studied cardiac model of mesotheliogenesis and vasculogenesis was widely applicable among coelomic organs. A comprehensive examination of mesodermal development in the avian embryo revealed a potential mesothelial progenitor population present from the onset of intestinal formation. Additionally, large blood vessels appeared to be derived via angiogenesis from the vitelline artery. A detailed lineage tracing study revealed mesothelial progenitors are indeed intrinsic to the splanchnic mesoderm of the intestinal primordium. A chick-quail chimera study demonstrated that both the vascular smooth muscle and endothelial cells at all levels of the intestine are derived from progenitors resident to the splanchnopleure through remodeling of a primordial endothelial plexus. Thus, the intestine displays a novel method of mesotheliogenesis and vascular formation compared to the heart.

Cdk1 kinase dephosphorylation and activation by Cdc25 phosphatase is essential for mitotic entry. Activated Cdk1 phosphorylates Cdc25 and other substrates, further activating Cdc25 to form a positive feedback loop that drives the abrupt G2/mitosis switch. Conversely, mitotic exit requires Cdk1 inactivation and reversal of Cdk1 substrate phosphorylation. This is mediated in part by Clp1/Cdc14, a Cdk1 antagonizing phosphatase, which reverses Cdk1 phosphorylation of itself, Cdc25, and other Cdk1 substrates. Thus, Cdc25 phosphoregulation is essential for proper G2-M transition and its contributions to cell cycle control have been modeled based on studies using Xenopus and human cell extracts. Since cell extract systems only approximate in vivo conditions where proteins interact within dynamic cellular environments, here we use Schizosaccharomyces pombe to characterize experimentally and mathematically, the in vivo contributions of Cdk1-mediated phosphorylation of Cdc25 to the mitotic transition. Through comprehensive mapping of Cdk1 phosphosites on Cdc25 and characterization of phosphomutants, we show that Cdc25 hyperphosphorylation by Cdk1 governs Cdc25 catalytic activation, the precision of mitotic entry, and unvarying cell length, but not Cdc25 localization or abundance. We propose a mathematical model that explains Cdc25 regulation by Cdk1 through a distributive and disordered phosphorylation mechanism that ultrasensitively activates Cdc25. We also show that Clp1/Cdc14 dephosphorylation of Cdk1 sites on Cdc25 controls the proper timing of cell division, a mechanism that is likely due to the double negative feedback loop between Clp1/Cdc14 and Cdc25 that controls the abruptness of the mitotic exit switch.