Selenium is an essential nutrient that is beneficial at daily intakes of 50-200μg/day and is detrimental at intake rates beyond 800μg/day. Selenium toxicity is an increasing environmental problem due to being a waste product of metal, coal, and oil refining. High selenium exposure causes developmental defects in wildlife, motor neuron degeneration in livestock and has been epidemiologically associated with the human motor neuron disease amyotrophic lateral sclerosis (ALS). In order to begin to define the cellular damage pathways activated by selenium, we have developed a genetic model of selenium toxicity using Caenorhabditis elegans. In this dissertation, we have begun to identify both environmental and genetic factors that affect selenium toxicity (like temperature, bacterial metabolism from the food source, calcium in the media), as well as a potential source of seleniums toxic effects, an increase in reactive oxygen species. We have also begun to determine which potential mechanism(s) of cell death are activated using strains with reduction-of-function mutations in cell death genes and pharmacologic treatments. In the second part of thesis, we characterize the neuronal damage caused by selenium because of its potential disease relevance. We demonstrate that selenium toxicity causes a decrease in cholinergic signaling that results in increased cytosolic protein catabolism in muscle which is predictably suppressed by nicotinic agonists and the acetylcholinesterase inhibitors. Combined, these data demonstrate that selenium toxicity causes muscle denervation, mimicking the pathology observed in ALS. We also demonstrate that selenium causes similar denervation of the muscles mediating egg-laying in C. elegans. Finally, we have investigated oxidative stress pathways regulated by DAF-2, a major stress response pathway in C. elegans, and identified a gene target of DAF-2 regulation, an iron-manganese superoxide dismutase (sod-2), that is a protective factor affecting sensitivity to selenium. This work demonstrates that selenium-related oxidative stress causes a progressive movement impairment due to motor neuron injury.

The human c-fes locus encodes a non-receptor tyrosine kinase (c-Fes) that is structurally and functionally unique. Originally, c-fes was isolated as the normal cellular homolog of sarcoma-inducing avian and feline retroviruses. However, unlike its viral oncoprotein counterparts that display constitutive tyrosine kinase activity, c-Fes exhibits restrained activity that is regulated by an undefined mechanism. Adding to its unique nature, recent studies have implicated c-Fes as a colorectal cancer-associated tumor suppressor despite its status as a proto-oncogene and tyrosine kinase. Previous work from our group has demonstrated that c-Fes forms high molecular weight oligomers in vitro, suggesting that c-Fes catalytic activity is governed by the interconversion of c-Fes between inactive monomeric and active oligomeric forms. However, this model was based largely on in vitro data and has not been assessed in living cells. To assess the involvement of oligomerization in regulating c-Fes activity in vivo, I employed a yellow fluorescence protein (YFP)-based bimolecular fluorescence complementation (BiFC) assay. Using BiFC, I demonstrated for the first time that c-Fes forms constitutive oligomers in vivo, regardless of its activation status. In addition, I determined that both coiled-coil domains mediate the oligomerization of c-Fes. Moreover, I established that c-Fes forms coiled-coil dependent oligomers in physiologically relevant cellular contexts, suggesting a new model for c-Fes regulation where conformational changes rather than oligomerization govern c-Fes kinase activity in cells. In colorectal cancers, loss of c-Fes expression is a common occurrence. This is not unusual, as tumorigenesis proceeds as oncogenes are activated and tumor suppressors are inactivated. To date, however, the mechanism responsible for c-fes gene repression has not been characterized. Upon determining that the absence of c-fes gene transcription was common among colorectal cancer cell lines, I used methylation inhibitor, bisulfite sequencing, and in vitro methylation analyses to establish that promoter methylation governs Fes gene and protein expression in colorectal cancers. Preliminary studies also suggest that promoter methylation governs c-Fes expression in human colon cancer surgical specimens. Taken together, the studies outlined in this thesis advance the field of c-Fes research by defining previously unknown regulatory mechanisms of both kinase activity and gene expression.

Genome stability is essential for cells to survive. Consequently, cells have evolved intricate responses that include transcriptional changes, cell-cycle arrest, activation of DNA repair, and apoptosis. Such responses prevent permanent fixation of DNA damage induced by genotoxic agents into the genome, thus contributing to genome stability. A collection of proteins implicated in DNA damage responses are called Inhibition of Growth (ING) family, which are a group of small molecular weight proteins that regulate a variety of biological functions ranging from senescence, cell cycle arrest, apoptosis and DNA repair. ING proteins interact with Histone Acetyl-transferases (HAT) and Histone Deacetylases (HDAC) to alter the state of chromatin compaction and acetylation status of many proteins during DNA damage. A specific member of the ING family, ING2 has been implicated in modulating the tumor suppressor, p53 function through p300 HAT-mediated acetylation. Irradiation fails to upregulate ING2, but increases its association with transcription co-activator p300. That p300 HAT activation in the cells with ING2 knock down is hampered post-irradiation suggests that the interaction between ING2 and p300 is indispensable for the upregulation of the p300 HAT activity. Cells deficient in protein kinase Ataxia Telangiectasia Mutated (ATM) displays impaired p300 HAT activation and less association between p300 and ING2 following ionizing radiation, indicating that ATM function is also required. Alkylating agent, N-methyl-N-nitro-N-nitrosoguanidine (MNNG) upregulates ING2 level in both time- and dose-dependent manner. We further observed that ING2 regulates the cell death response induced by this alkylator through a mechanism involving acetylation and stabilization of p73. Induction/acetylation of p53, in response to MNNG, however, proceeds in an ING2-independent manner. Inhibition of c-Abl by STI571 treatment blocked ING2 upregulation and p73 acetylation induced by MNNG. Similarly, MLH1- suppressed or mutated cells displayed defective ING2 upregulation and p73 acetylation in response to MNNG, which suggests that Mlh1- and c-Abl-dependent upregulation of ING2 activates the cell death response to MNNG through p73 acetylation. Taken together, these findings demonstrate that ING2 plays an important role in the cellular responses to different DNA damage by regulating the acetylation of tumor suppressors.

The C2H2 zinc finger (ZF) transcription factors (TF) form the largest family of DNA binding proteins in eukaryotes. TFs are key proteins involved in gene regulation that bind to specific DNA sites. A major obstacle towards understanding the molecular basis of transcriptional regulation is the lack of a recognition code for protein-DNA interactions. We aim to understand molecular mechanisms of DNA recognition and to quantitatively estimate recognition rules for TF-DNA interactions. We identified key residues playing an important role in ZF-DNA interactions and found that they are prealigned to conformations observed in the bound state prior to binding. A binding site for Cl- ions corresponding to the pocket where DNA phosphates are found most buried in the complex of ZFs is identified. Bound ions constrain conformations of important residues consistent with observations of increased binding affinity with increased ionic strength in protein-DNA interactions. These results suggest a general mechanism where ZFs, through their key residues, rapidly form encounter complexes amenable for a fast readout of the DNA. We developed a novel experimentally-based approach using crystal structures and binding data on the C2H2 ZFs and decoded ten fundamental specific interactions for protein-DNA recognition. These are: Five hydrogen bonds, three desolvation penalties, a non-polar energy, and a novel water accessibility factor. The code is applied to three data sets with a total of 89 ZF mutants on three ZFs of EGR. Guided by simulations of individual ZFs, we mapped the interactions into homology models with all feasible intra- and inter- molecular bonds and selected the structure with the lowest free energy for each ZF. The interactions reproduce changes in affinity of 35 mutants of finger I (FI) (R2 = 0.99), 23 mutants of FII (R2 = 0.97) and 31 human ZFs on FIIII (R2 = 0.95). The method predicts bound ZF-DNA complexes for all mutants, decoding molecular basis of ZF-DNA specificity. These findings reveal recognition rules that depend on DNA sequence/structure, molecular water at the interface and induced fit of the C2H2 TFs. In summary, our method provides the first robust framework to decode the molecular basis of TFs binding to DNA.

HIV-1 Nef is a small myristoylated protein capable of interaction with a diverse array of host cell signaling molecules. Multi-faceted in its function, Nef is a critical accessory factor, essential for high-titer viral replication and AIDS progression. Despite its essential role, the molecular mechanisms of Nef-mediated HIV pathogenicity are not fully understood. Previous biochemical and structural studies have suggested that Nef may form homodimers and higher order oligomers in HIV-infected cells. The studies summarized below investigated the oligomeric status of the HIV-1 nef gene product and its role relative to Nef-mediated function. We explored the formation of Nef oligomers in live cells by adapting a bimolecular fluorescence complementation (BiFC) assay, a well-defined system in which dimeric protein interactions are observed in live cells. Using this assay, we provided the first direct evidence for Nef oligomerization in vivo. We then assessed the generality of oligomerization by a group of Nef alleles broadly representative of all major HIV-1 subtypes and found oligomerization was highly conserved across all subtypes examined. We then used our BiFC system to define residues previously suggested via X-ray crystallographic studies to comprise the Nef dimerization interface. Using a systematic strategy for the mutagenic profiling of the oligomerization interface, we discovered two classes of residues were critical to Nef oligomerization. BiFC was completely abolished when either all four key hydrophobic interactions were simultaneously removed or when ionic interactions mediated by D123 and R105 were disrupted. Finally, we utilized Nef mutants identified in the mutagenic profiling of the oligomerization interface to explore the effects of oligomeric disruption on Nef function. Screening a panel of Nef mutants with varying degrees of oligomeric disruption we discovered, surprisingly, despite the varying effects on oligomerization, all of these mutants were shown to dramatically disrupt Nef-induced CD4 downregulation and viral replication. Taken together, the studies presented in this dissertation advance the field of HIV research by furthering our understanding of the regulation of Nef-mediated downregulation of CD4 and enhancement of HIV replication as well as validating the Nef oligomerization interface as a potential target for anti-retroviral drug design.

Extracellular signals generated during both feeding and fasting coordinately regulate transcription of metabolic enzyme genes that control glucose metabolism in thenÒ cell. A post-prandial rise in extracellular glucose levels promotes expression of various genes including the gene encoding the glycolytic enzyme L-type pyruvate kinase (L-PK). Conversely, under conditions of fasting, a rise in hormones that stimulate increased intracellular levels of cAMP results in suppression of glucose-activated genes such as L-PK. The L-PK gene is coordinately regulated by these two opposing stimuli. Therefore, we explored the mechanism of induction and repression of the L-PK gene by glucose and cAMP, respectively, using the 832/13 rat insulinoma cell line. Glucose mediates induction of the L-PK gene by stimulating the recruitment of two primary DNA binding transcription factors, the basic helix-loop-helix/leucine zipper protein Carbohydrate Response Element Binding Protein (ChREBP) and the orphan nuclear receptor, Hepatic Nuclear Factor 4Ñ (HNF4Ñ) to their respective response elements in the proximal L-PK promoter. In addition, glucose stimulates the recruitment of the coactivator CREB binding protein (CBP) to the L-PK gene promoter. Assembly of these three factors on the L-PK gene promoter facilitates alterations in the pattern of acetylation and methylation of histones associated with the promoter and coding region, respectively. These changes in histone modifications correlate with increased occupancy of the RNA Polymerase II (Pol II) holoenzyme on the L-PK promoter. Finally, glucose promotes changes in the phosphorylation state of the carboxyl-terminal domain (CTD) of Pol II at serines 5 and 2, which are necessary for the promoter clearance and elongation phases of transcription. cAMP represses the glucose-mediated induction of the L-PK gene by inhibiting the assembly of the ChREBP, HNF4Ñ and CBP-containing complex on the L-PK promoter. The cAMP-dependent decrease in complex assembly on the promoter is associated with alterations in the acetylation and methylation status of histones on both the promoter and coding region. Furthermore, cAMP inhibits the glucose-mediated recruitment and phosphorylation of Pol II CTD, ultimately blocking initiation and elongation of the L-PK gene by Pol II. In summary, these studies provide a detailed insight into the mechanism of regulation of the L-PK gene by glucose and cAMP in islet Ò cells.

High-risk human papillomaviruses (HPVs), such as HPV-16, are the etiological agents of squamous cell carcinomas (SCCs) of the anogenital tract and a subset of oropharyngeal cancers. High-risk HPVs encode two oncoproteins, E6 and E7, which promote unscheduled host cell proliferation by targeting the p53 and pRB tumor suppressor proteins, respectively. HPV-16 E7 has been shown to stimulate structural chromosomal instability and DNA breakage. These findings raise several important questions. First, how does HPV-16 E7 induce DNA damage? Second, what are the precise consequences of HPV-16 E7-induced DNA damage for host cell genomic integrity, and lastly, how do HPV-16 E7-expressing cells maintain proliferation despite activated DNA damage checkpoints? Here, we show that HPV-16 E7 activates the Fanconi Anemia (FA) pathway, a branch of the host cell DNA damage response that primarily responds to stalled DNA replication forks. Importantly, we show that HPV-16 E7 expression in FA-deficient cells accelerates the formation of structural chromosomal alterations, which may help to explain the heightened susceptibility of FA patients to HPV-associated tumors. However, we also provide evidence that HPV-16 E7-induced FA pathway activation in FA-proficient cells may contribute to evasion of anti-proliferative host cell barriers by promoting alternative lengthening of telomeres (ALT). Finally, we demonstrate that HPV-16 E7 circumvents DNA damage checkpoint control and promotes aberrant mitotic entry by increasing the proteolytic turnover of claspin, which plays a role in the ATR/CHK1-mediated replication stress response. Collectively, our results underscore that HPV-16 E7 interferes with host cell genome integrity by inducing DNA replication stress. The detrimental effects of HPV-16 E7 on the genomic integrity of host cells with a deficient FA pathway support the notion that this DNA damage response pathway is crucial to prevent HPV-16 E7-induced genomic instability and malignant progression. However, we also provide evidence that HPV-16 E7 can exploit the FA pathway to promote cellular immortalization. Future experiments to explore these events for cancer therapy and/or prevention are warranted.

When the cationic lipid DOTAP (1,2-dioleyltrimethylammoniumpropane) is used to encapsulate an antigenic peptide from the human papillomavirus E7 oncogene (E7), the resultant DOTAP/E7 particles act as a therapeutic vaccine to cause tumor regression through an antigen-specific immune response when the vaccine is injected into mice bearing E7-positive TC-1 tumors. Of critical importance, the DOTAP works as both a delivery vehicle and an adjuvant without induction of a pro-inflammatory cytokine response in vivo. It is hypothesized the antigen specific immune response is mediated by dendritic cells in vivo. To that end, the interaction of murine bone marrow-derived dendritic cells (BMDC) with the vaccine in vitro was investigated. When BMDC were incubated in the presence of DOTAP or the DOTAP/E7 vaccine, there was a dose and time-dependent upregulation of co-stimulatory molecule expression, indicating that BMDC were activated by the cationic lipid DOTAP. Further experiments indicated that BMDC were capable of internalizing DOTAP liposomes through many endocytic routes and the vaccine trafficked through the vacuolar pathway. An indirect method was used validate antigen presentation by BMDC, wherein the generation of antigen specific effector cells after incubation of CD8+ T lymphocytes, purified from the spleens of naïve mice, with fixed dendritic cells that had been activated by the vaccine, was examined. Not surprisingly, the DOTAP/E7 therapeutic vaccine was capable of initiating the generation of effector CD8+ T lymphocytes in vitro through a conventional mechanism, which requires dendritic cell activation and presentation of the peptide antigen. Interestingly, it was found the same simple vaccine was capable of generating antigen specific effectors in vitro in the absence of antigen presenting cells in a novel pathway of effector generation, based on the expression of CD8 on the cell surface, T cell receptors that recognize a specific peptide antigen/MHC complex as well as an antigen-specific increase in IFN-γ production. Further studies revealed a possible mechanism for this novel pathway. Taken as a whole, these observations may lead to additional applications of DOTAP both in vitro and in vivo to modulate the immune response toward the correction of a variety of diseases.

Osteoclastogenesis and angiogenesis are correlated in bone during physiological and pathological processes including development, fracture healing, bone metastases and inflammatory bone disease. However, it is unclear if and how these processes are linked. This dissertation investigates a possible causative role for osteoclasts in bone angiogenesis. First, changes in osteoclast formation and activity affected angiogenesis in a parallel fashion. Osteoclast inhibition decreased angiogenesis, while osteoclast stimulation increased angiogenesis in fetal mouse metatarsal explants. Likewise, osteoclast stimulation also increased angiogenesis in mouse calvaria in vivo, thus showing that osteoclasts and angiogenesis are linked. Further studies were conducted to determine the mechanism by which osteoclasts may increase angiogenesis. Angiogenic factor expression by osteoclasts was analyzed by reverse-trancriptase PCR and Q-PCR angiogenesis arrays of human bone marrow osteoclasts. MMP-9 was the most highly expressed osteoclast angiogenic factor at the mRNA level. Because MMP-9 is important for osteoclast and blood vessel invasion of the growth plate and fracture calluses, the role of MMP-9 in osteoclast stimulated angiogenesis was studied in depth. Osteoclast stimulation with RANKL or PTHrP failed to stimulate angiogenesis in MMP-9-/- mouse calvaria or metatarsal explants. Surprisingly, osteoclast stimulation was dramatically blunted in MMP-9-/- calvaria or metatarsal explants. However, the number of vessels per osteoclast was not different between WT and MMP-9-/- mice, indicating that osteoclasts lacking MMP-9 do not have an intrinsic angiogenic defect. Further, bone marrow cultures from WT and MMP-9-/- mice formed similar numbers of osteoclasts, demonstrating that osteoclast differentiation or precursor number is not responsible for the inability of PTHrP or RANKL to increase osteoclastogenesis in MMP-9-/- mice. These results suggest that MMP-9 is important for osteoclast-stimulated angiogenesis by affecting the number of osteoclasts at the angiogenic site due to its previously reported effects on osteoclast migration. These studies greatly increase our understanding of angiogenesis in bone and suggest an important role for osteoclasts in angiogenesis during bone development, fracture healing, bone metastasis, inflammatory bone diseases and the potential effects of osteoclast inhibitory agents on angiogenesis.

Duchenne muscular dystrophy (DMD) is a devastating primary muscle disease with pathological changes in skeletal muscle that are ongoing at the time of birth. Progressive deterioration in striated muscle function in affected individuals ultimately results in early death due to cardio-pulmonary failure. Since affected individuals can be identified prior to birth by prenatal genetic testing for DMD, gene replacement treatment can be started in utero. This approach offers the possibility of preventing pathological changes in muscle that begin early in life. Previous studies with systemic in utero adenoassociated viral (AAV) vector serotype 1 gene delivery to embryonic day 16 (E-16) pups resulted in high levels of transduction in diaphragm and intercostal muscles, but no detectable levels in limb muscle. Recently newer AAV serotypes such as AAV8 have demonstrated widespread and high transgene expression in skeletal muscles and diaphragm by systemic delivery in adults and neonatal mice. In this study I tested AAV8 vector gene delivery by intraperitoneal administration in E-16 mice in utero. Using an AAV8 vector carrying a lacZ transgene, I observed high level transduction of diaphragm and more moderate transduction of multiple limb muscles and heart. Encouraged with these results I tested in utero gene transfer in the mdx mouse model of DMD, a minidystrophin gene driven by the human cytomegalovirus promoter was delivered systemically by an intraperitoneal injection to the fetus at embryonic day 16. Treated mdx mice studied at 9 weeks after birth demonstrated widespread expression of recombinant dystrophin in skeletal muscle, restoration of the dystrophin associated glycoprotein complex in dystrophin-expressing muscle fibers, improved muscle pathology, and functional benefit to the transduced diaphragm compared to untreated littermate controls. In order to further extend these studies, AAV9 carrying a minidytsrophin gene was also tested. Robust expression in heart and muscles were seen at 4 weeks post treatment by in utero gene delivery. Furthermore robust heart expression persisted as long as 3 months post treatment. These results support the potential of AAV8 and AAV9 vectors to efficiently cross the blood vessel barrier to achieve systemic gene transfer to skeletal muscle in utero in a mouse model of muscular dystrophy, to significantly improve the dystrophic phenotype and to ameliorate the processes that lead to exhaustion of the skeletal muscle regenerative capacity.