Carbamyl phosphate synthetase I (CPSI) is the liver-specific enzyme that catalyzes the first and rate-determining step of the urea cycle. Mutations in the CPSI gene cause CPSI Deficiency (CPSID), an autosomal recessive disease characterized by hyperammonemia. During the course of this project, many patient mutations likely causing RNA instability were identified. These genomic variants were present in patient genomic DNA but not detectable in patient RNA, indicating RNA degradation perhaps due to nonsense mediated decay (NMD) or splicing errors. This large number of suspected RNA instability mutations suggests CPSID is a good model to understand the prevalence of this mutation type in genetic disease. However, the ability to study a heterogeneous set of mutations in such a large gene has suffered from the lack of an appropriate model system. A goal of this project was to develop a novel model system for testing the functional consequences of these potential RNA instability mutations. Using two homologous recombination strategies, a BAC clone containing the full CPSI gene was modified to contain the pEHG vector and the CMV promoter upstream of CPSI. These modifications allow for stable transfection, selection and ubiquitous expression in eukaryotic cells. Patient mutations were then incorporated into the resulting BECC model system and these constructs were subsequently tested to assay for splicing changes as well as NMD. These mutations included the intronic substitutions c.654-3T>G and c.1210-1G>T, the exonic c.1893T>G mutation which directly creates a nonsense codon, and c.2388C>A which is a translationally silent exonic mutation with unknown function. Extensive control assays show that the BECC model system properly expressed wild-type CPSI, while each mutant shows decreased expression as measured by quantitative RT/PCR assays. To determine if this decrease in expression was due to NMD, it was inhibited by siRNA mediated knockdown of UPF2. Following NMD inhibition, relative expression levels of each mutant were increased, indicating these mutations cause degradation via the NMD pathway. In addition, cryptic splice sites activated by intronic mutations were identified. This project is significant because using CPSID as a model, it shows the prevalence of RNA instability mutations in a severe metabolic disease and provides a new model system to test any mutation, irrespective of type or location, for functional consequences including the activation of aberrant splicing and nonsense-mediated decay.

Adenovirus (Ad) vectors are promising vehicles for vaccine delivery and gene therapy, but vector development has been hampered by the lack of an atomic structure of the intact virion. In this work, cryoEM single particle reconstruction methods have been applied to study the structure of an adenovirus vector, Ad35F. The 9 Å resolution cryoEM structure of Ad35F was analyzed by docking the crystal structures of two major capsid proteins, hexon and penton base. A long α-helix of ~ 40 residues was observed on the capsid surface and assigned to the C-terminal domain of protein IX recently used as a platform for vector retargeting. In the 6 Å resolution structure of Ad35F α-helices of 10 or more residues are clearly visualized for the first time. This in combination with secondary structure prediction for proteins IIIa, VI, VIII, and IX has led to new and more precise assignments for these capsid proteins. The coiled coil of protein IX is assigned to a tetrameric coiled coil now observed on the outer capsid surface. Protein IIIa, which is predicted to be highly α-helical, is assigned to a cluster of helices below the penton base on the inner capsid surface. Protein VI, a membrane lytic factor, is sequestered within the hexon cavities on the inner capsid surface. The new protein assignments have important implications for vector design, virus assembly, and capsid disassembly during cell entry. First, since protein IIIa, VI, and VIII are all located on the inner capsid surface, our structural work indicates that the best candidate proteins for retargeting are hexon, penton base, fiber and protein IX. With the new interior position of protein IIIa, it is easier to envision how protein IIIa can facilitate DNA packaging during assembly. During cell entry in the low pH environment of the endosome, protein IIIa is well situated to allow rapid release of vertex proteins, including protein VI. The sequestration of protein VI is likely important to prevent membrane lysis until the appropriate time point in the viral life cycle.

TFIID is a multi-subunit protein complex that regulates activation of gene transcription. The coactivator model proposes that a given transcription factor activates transcription of its target gene by recruiting TFIID to the gene promoter through direct, mutationally sensitive interactions with a subset of TFIID subunits. However, the coactivator model was developed from in vitro studies. To test the coactivator hypothesis in vivo, I examined the role of yeast TFIID in activating transcription through the cell-cycle regulated transcription factor, Swi6p. Through in vitro analysis, I found that Swi6p, in the context of the transcriptional complex SBF, interacts directly and selectively with a subset of TFIID subunits (Taf12p, Taf4p, and Taf5p). Through RNA blotting studies, I found that one of these interacting subunits, Taf12p, regulates the expression of several cell-cycle regulated genes. Finally, I found that Swi6p co-immunoprecipitates with Taf12p and Taf4p from yeast whole cell extract, and that this interaction is disrupted in yeast that harbor a mutation in another TFIID subunit, Taf9p (the taf9W133stop mutation). Analyses of protein expression and growth rates of various taf9W133stop mutant clones revealed an unexpected role for Taf9p in the regulation of cellular protein expression levels. I propose a model in which TFIID coordinately regulates protein expression and cell cycle progression by mediating transcription through the transcription factors, Rap1p and SBF, respectively, through a mechanism that involves a common set of TFIID subunits, Taf12p, Taf4p, Taf5p and Taf9p.

Glucoregulatory hormones tightly control glucose production and uptake by their direct actions on glucose metabolism and indirect actions on both lipid and protein metabolism. Our laboratory and others have shown that the main substrate responsible for the indirect actions of glucoregulatory hormones on hepatic glucose production (HGP) is free fatty acids (FFA). The antilipolytic effects of insulin mediate its ability to suppress HGP by reducing gluconeogenesis; whereas the lipolytic action of epinephrine influences its ability to stimulate HGP by limiting glycogenolysis and augmenting gluconeogenesis. Although glucagon has little to no effect on adipose tissue in vivo, elevated levels of FFA and glucagon are observed postprandially in individuals with type 2 diabetes. Part of this work determined how FFA interact with glucagon to acutely regulate HGP. Our findings indicate that elevated FFA inhibit the glycogenolytic effects of glucagon, as they do the glycogenolytic effect of epinephrine. However, this effect is short-lived compared to epinephrine. In contrast to epinephrine, an increase in FFA does not augment glucagon-stimulated hepatic gluconeogenesis. A dysregulation of lipid metabolism, or an elevation of lipids, could possibly interact with glucoregulatory hormones to cause harmful metabolic effects. There is strong evidence for elevated fasting and postprandial FFA levels in type 2 diabetes and those at risk for the disease. Both acute and chronic elevations of FFA have been shown to induce insulin resistance. Additionally, accumulation of intracellular triglycerides in liver and skeletal muscle have also been strongly implicated in the development of insulin resistance; however to date, their putative role is based solely upon correlative results. This work shows for the first time that repeated brief daily elevations of FFA induce insulin resistance in liver, skeletal muscle, and adipose tissue. This finding suggests that recurring elevations of FFA, such as might occur during the postprandial period in type 2 diabetes, could progressively lead to chronic insulin resistance. In addition, we show that accumulation of hepatic triglycerides does not worsen the effects of repeated elevations of FFA. Furthermore, this work shows that repeated brief elevations of FFA also alter mechanisms of basal HGP.

Pancreatic beta cells are the only cell type in the body that can produce insulin, a hormone required for maintaining glucose homeostasis. Type 1 and type 2 diabetes result from an absolute or relative reduction in functional beta cell mass, respectively. One approach to replacing lost beta cell mass is transplantation of islets from cadaveric donors; however, this approach is limited by lack of adequate amounts of donor tissue. Therefore, there has been much interest in identifying factors that enhance beta cell differentiation and proliferation in vivo or in vitro. Connective tissue growth factor (CTGF) is a secreted molecule expressed in pancreatic endothelial cells, beta cells, and ducts during pancreas development. CTGF is required for proper lineage allocation and islet morphogenesis during development and for beta cell proliferation at late gestation. The current study investigated the tissue interactions by which CTGF promotes normal pancreatic islet development. These results show that CTGF produced by both endothelial cells and beta cells is required for embryonic beta cell proliferation, making CTGF the first identified secreted endothelial-derived or beta cell-derived factor which regulates embryonic beta cell proliferation. Removing CTGF from any one source in the pancreas impairs embryonic beta cell proliferation, indicating that beta cell proliferation is particularly sensitive to the level of CTGF. In contrast, inactivating CTGF from one source in the pancreas does not affect lineage allocation or islet morphogenesis, indicating that these sources of CTGF act redundantly to promote these processes. This dissertation also demonstrated that over-expression of CTGF in beta cells during embryogenesis using an inducible transgenic system increases islet mass by promoting endocrine cell proliferation. Together, these findings demonstrate that CTGF acts in both an autocrine and paracrine manner during pancreas development and suggest that CTGF has the potential to enhance beta cell proliferation or differentiation in diabetic models.

MOLECULAR PHYSIOLOGY AND BIOPHYSICS DOPAMINE DEPLETION ALTERS THE BALANCE BETWEEN CA2+/CALMODULIN-DEPENDENT PROTEIN KINASE II AND PROTEIN PHOSPHATASE I ABIGAIL M. BROWN Dissertation under the direction of Professor Roger J. Colbran Parkinsons disease is characterized by the preferential loss of the nigrostriatal dopaminergic neurons, whose axons terminate within the striatum and release dopamine onto medium spiny neurons. Striatal dopamine loss results in several changes in striatal medium spiny neuron morphology and disrupts corticostriatal synaptic plasticity. However, the molecular and biochemical mechanisms underlying the consequences of dopamine loss in Parkinsons disease are not well understood. This dissertation presents evidence linking both short-term and chronic dopamine loss to the misregulation of two key synaptic signaling enzymes in the striatum, CaMKII and PP1. Although it is unclear what causes Parkinsons disease in most human patients, the possibility that such striatal calcium-sensitive signaling pathways are misregulated suggests that targeting these pathways may be a useful strategy in developing future treatments for Parkinsons disease.

The transcription factor Pdx1 is expressed throughout the pancreas early, but selectively in insulin-producing cells postnatally. Previous studies showed that conserved regulatory regions, Areas I and II (Pdx1PB), directed pancreatic endocrine cell expression, while an adjacent region (Pdx1XB) containing the conserved Area III directed transient Ò cell expression. We generated transgenic mice in order to lineage trace cells in which these promoter fragments were activated: Pdx1PBCre mediated only endocrine cell recombination, while Pdx1XBCre directed broad and early recombination throughout the developing pancreas. An Areas I-II-III reporter transgene was expressed throughout the E10.5 pancreas, gradually becoming Ò cell selective, similar to endogenous Pdx1. These data suggested that sequences within Area III mediate early pancreas-wide Pdx1 expression. Area III contains a binding site for PTF1, a transcription factor complex essential for development of all pancreatic cell types. This site contributed to Area III-dependent reporter gene expression in an acinar cell line, while PTF1a specifically trans-activated an Area III-containing reporter in a non-pancreatic cell line. Importantly, Ptf1a occupied Area III of the endogenous Pdx1 promoter in E11.5 pancreatic buds. These data strongly suggest that PTF1 is an early activator of Pdx1 in pancreatic progenitor cells. Ptf1a expression in the early pancreatic primordium requires signaling from the adjacent vascular endothelial cells, which are reciprocally sensitive to VEGF signaling from the pancreas. To elucidate the role of the VEGF signaling pathway in pancreas differentiation, we analyzed transgenic mice that express agonists selective for VEGF receptor isoforms (VEGFR1 or VEGFR2) throughout the pancreas. Morphometric analysis of antibody labeled cross-sections revealed that at postnatal day one, mice expressing the VEGFR2 agonist had a 49% increase in endothelial to endocrine ratio, but islet size was similar to wild type. In contrast, VEGFR1 agonist expression caused a decrease in endothelial to endocrine ratio, but surprisingly, islets were smaller than in wild type. Taken together, these results suggest that VEGFR1 and VEGFR2 signaling have distinct roles in pancreatic islet development, and the expansion of endocrine mass mediated by increased VEGF may require increased signaling through both receptors.

Real time processing of hyperspectral imaging data of the kind acquired by Computed Tomography Imaging Spectroscopy (CTIS) presents some unique challenges in computational power and data storage. The general approach pursued in this work is a direct application of numerical solution methods implemented in the computing cluster environment of Vanderbilt Universitys Advanced Computing Cluster Resource (ACCRE). The examination of four reconstruction algorithms, Simultaneous Algebraic Reconstruction, Cimmino Component Averaging, Expectation Maximization-Maximum Likelihood, and Multiplicative Algebraic Reconstruction Technique were explored. The Multiplicative Algebraic Reconstruction Technique proved robust in quality and speed as implemented with the matrix preconditioning method as employed.

Glucagon-like peptide-one (GLP-1) secreted from the endocrine L cell in the gut after a meal results in elevations of the peptide in arterial blood, the hepatic portal vein (~2x arterial levels) and vasculature in the gut directly surrounding the site of secretion (~4x arterial levels). An intraportal physiological bolus of GLP-1 increases afferent vagal signaling from the hepatoportal region, which could initiate the gluco-regulatory effects of GLP-1. In Specific Aim I, hyperglycemia was induced with a combination of peripheral and intraportal glucose infusions in the 42 h fasted conscious dog. GLP-1 was given intraportally or into the hepatic artery in physiologic amounts. Intraportal, but not hepatic artery, GLP-1 delivery significantly increased nonhepatic glucose uptake relative to that observed in saline infused control dogs, without altering pancreatic hormone levels. In Specific Aim II, GLP-1 or saline was infused intraportally in 42 h fasted conscious dogs, and hyperglycemia was induced by peripheral glucose infusion, alone. Under these conditions, GLP-1 did not alter pancreatic hormone levels or glucose utilization. In Specific Aim III, dogs were fasted for 18 h to increase Ò-cell responsiveness and insulin sensitivity. GLP-1 or saline was infused intraportally, and hyperglycemia was induced by a combination of intraportal and peripheral glucose infusion. GLP-1 had no effect despite the presence of a portal vein glucose infusion. In Specific Aim IV, dogs were administered a mixed meal in the presence or absence of exendin (9-39), a GLP-1R antagonist. Blocking postprandial GLP-1 action resulted in expedited gastric emptying for about 2 h which ultimately increased peripheral glycemia in both normal and insulin resistant dogs. Blocking postprandial GLP-1 action did not enhance the incretin effect in either group of animals. In conclusion we found that in the dog 1) GLP-1R activation is not responsible for the incretin effect, 2) the dominant effect of postprandial GLP-1 secretion is slowed gastric emptying, 3) GLP-1 has a small direct effect on the liver, and 4) intraportally delivered GLP-1 induces increased nonhepatic glucose uptake under conditions that mimic a meal.

The melanocortin 4 receptor (MC4R) is a G-protein coupled receptor, which is widely, yet weakly, expressed throughout the CNS. Although the sites of MC4R expression are known, the field still does not understand how the gene is regulated temporally. <p> To determine key regulatory regions of the MC4R promoter, I have established in vitro and in vivo models for studying MC4R promoter activity. By transiently transfecting cell lines with mouse and human promoter-luciferase constructs, I identified regions within the proximal 900bp that correlate with positive expression in endogenous MC4R cell lines. <p> A 4.8kb promoter tauEGFP transgene transcript expression was demonstrated by Real Time RT-PCR to be restricted to the CNS, and similar to endogenous MC4R transcript expression in three of four independent lines. Using a Beta-Galactosidase reporter, 3.3kb 5-flanking and 650bp 3-flanking murine MC4R sequence also directed expression in a CNS specific manner. Interestingly, fetal tissues stained for Beta-Galactosidase activity showed non-CNS expression nearly identical to endogenous extra-neural MC4R expression in fetal rats. The luciferase transgenic mice, from the largest construct (identical in sequence to the LacZ construct) to the smallest (only containing 65bp of the 100% conserved region CR-8) were found to express the transgene in the CNS. <p> These results suggest that the 5-flanking proximal sequence of the MC4R gene is sufficient for brain preferential expression in vivo. Furthermore, the highly conserved CR-8 region is sufficient to target brain preferential expression in a heterologous promoter in vivo in a pattern similar to the larger constructs.