Chemotaxis, the ability of a cell to change its motility as a response to environmental stimuli, belongs to the most important signal transduction mechanisms in life. Evolution has created a plethora of different molecular mechanisms to enable cells to react appropriately to extracellular changes. In this thesis, chemotactic signal transduction pathways in bacteria, in particular two component signalling in R. sphaeroides, and eukaryotes, in particular human axon guidance, are studied by means of X-ray crystallography complemented with biophysical, biochemical and cellular studies. Two-component signal transduction in bacteria is one of the most abundant signalling pathways in biology. Here the determinants for specificity for a crucial sensor kinase-response regulator interaction are presented and used to design novel, synthetic two-component pairs. The activation mechanism of response regulators has been extensively studied and a model for activation that crucially involves a Thr and a Tyr residue has been widely accepted in the field. In this thesis, two structures of a response regulator from R. sphaeroides are presented and together with biophysical and cellular assays suggest a novel paradigm for response regulator activation. Axon guidance is an essential process in human development and relies crucially on chemotaxis. Two signalling pathways, the plexin-semaphorin and the Rgm-Neogenin pathway are studied extensively in this work. Structures of the intracellular region of Plexin-B1 provide an elegant mechanism explaining how ligand binding events on the extracellular and intracellular side can be integrated into a single signalling output. The study of RgmB in complex with its receptor Neogenin provides the first structural insight into the important family of repulsive guidance molecules and explains their role in the disease juvenile hemachromatosis. In summary, this work provides insights into a plethora of chemotactic pathways and extends our current knowledge of these important mechanisms significantly.

Ceramide-1-phosphate (C1P) has proven to be a bioactive sphingolipid with diverse functions within the cell. At this time, ceramide kinase (CERK) is the only known enzyme known to generate C1P in mammalian cells, and this bioactive lipid is responsible for the activation and translocation of cytosolic phospholipase A2, the initial rate limiting step in eicosanoid synthesis. These studies investigate the regulation of ceramide kinase by calcium. While CERK activity has been shown to be calcium sensitive, little is known about how CERK is activated within the cell; one possibility is the interaction with calcium "sensor" proteins such as calmodulin. In this study, we develop two protocols to efficiently examine the interaction of CERK with calcium dependent proteins: V5 co-immunoprecipitation and Ni-NTA affinity purification. The methods utilize either adenoviral infection or Effectene© transfection of cells to ectopically express CERK with both a 6x His and V5 tag on its C terminus. Unlike the report of Igarashi and co-workers, our findings reveal that CERK does not specifically interact with the calcium sensor, calmodulin, in three different cell types. We also show that the calcium dependent membrane organizer, annexin A2, also does not bind to CERK. In light of these findings, we illustrate that while CERK may be sensitive to calcium, it does not, as previously reported by another laboratory, specifically bind to calmodulin. These studies eliminate possible calcium mediator proteins and are suggestive of another method for the calcium sensitive regulation of CERK lending to new avenues of investigation (i.e. CaMKII). This report also firmly established two successful protocols for investigating protein partners of CERK. Ultimately, through providing a clearer picture behind the calcium regulation of CERK, we can elucidate possible novel therapeutic targets within the inflammatory pathway.

Bacterial resistance to 4,6-type aminoglycoside antibiotics, which target the 30S ribosomal subunit, has been traced to the arm/rmt family of rRNA methyltransferases. These plasmid-encoded enzymes transfer a methyl group from S-adenosylmethionine to N7 of the buried G1405 in the aminoglycoside binding site of 16S rRNA in the 30S ribosomal subunit. Neither 16S rRNA alone nor intact 70S ribosome is an efficient substrate for armA methyltransferase. To more fully characterize this family of enzymes, xiii we have investigated the substrate requirements of ArmA. We determined the Mg2+ dependence of ArmA activity and found that the enzyme could recognize both translationally active and translationally inactive forms of 30S subunits. To identify the site of interaction between ArmA and the 30S subunit, we used hydroxyl radical cleavage of 16S rRNA mediated by ferrous iron chelated to several sites on the ArmA molecule that were mutated to cysteine. This data suggests that significant conformational changes in 30S structure are involved in binding of ArmA. We hypothesized that a precursor intermediate in the biogenesis of the 30S subunit might be the optimal substrate for ArmA enzymes in vivo. To test this, we prepared 30S particles partially depleted of proteins by treatment with increasing concentrations of LiCl and assayed them for ArmA methylation. Even low concentrations of LiCl alter the 30S particles and greatly diminish their susceptibility to methylation. Additionally, a previously identified pre-30S particle isolated from an E. coli culture was assayed for its ability to support methylation by ArmA and found to be inferior to intact 30S particles as a methylation substrate. Thus, testing of immature particles prepared from in vitro and in vivo sources suggest that ArmA works very late in the 30S biogenesis pathway. Initiation factor 3 (IF3), a factor that only binds fully mature 30S particles, does not inhibit the ArmA methylation, while kasugamycin methyltransferase (KsgA) abolishes ArmA activity by sharing the same binding site with ArmA. From aforementioned experiments, we conclude that ArmA is most active toward 30S ribosomal subunits that are at or very near full maturation.

Our recent studies have shown a link between Adenovirus-5 (Ad-5) and elevated lipids, which prompted the hypothesis that Ad-5 infection might augment hepatic and/or adipose tissue lipid metabolism. To test our hypothesis, mice were infected with Ad-5 and screened for changes in lipogenesis and plasma markers associated with the metabolic syndrome. We observed increased expression of sterol regulatory element binding protein 1 (SREBP-1) in infected liver tissues, but not in adipose tissues and this correlated with elevated plasma and hepatic triglyceride levels. Elevated expression of adiponectin was seen in Ad-5 infected adipose tissues and this correlated with phosphorylated AMPK in infected liver tissues. These data suggested that the AMPK pathway was activated in livers of Ad-5 infected mice. Indeed, we observed reduced expression of PEPCK, a downstream target of AMPK in livers of Ad-5 infected mice. As PEPCK is an enzyme essential for gluconeogenesis, we hypothesized that Ad-5 infection would reduce blood sugar. Indeed, infected mice exhibited a transient decline in plasma glucose. The increase in SREBP1 levels in Ad-5 infected hepatic tissues was evaluated by looking at Ad-5 infected HepG2 cells. Ad-5 is thought to mimic insulin’s actions in which the PI3K pathway is activated. We hypothesized that Ad5-induced SREBP-1 expression levels are mediated through the induction of PKC λ/ζ/ι, and not through Akt because it has been shown that PKC λ/ζ/ι mediates insulin-dependent lipogenesis. To test our hypothesis, HepG2 cells were infected with Ad-5 and screened for downstream targets of the PI3K pathway. Through western blot analyses, we observed increased levels of phosphorylated PKC λ/ζ/ι. These results prompted the use of PKC pan inhibitor to see whether Ad-5 induced SREBP-1 levels would be down regulated. Indeed, with the presence of the PKC pan inhibitor, SREBP-1 expression levels were reduced. Together, these studies suggest that Ad-5 induces changes in gene expression, glucose, and lipid metabolism; which prompts the hypothesis that this common respiratory pathogen may be associated with the Metabolic Syndrome, and this may preclude its use as a vector for gene therapy.

DETERMINING THE EFFECT OF DOUBLE-STRANDED RNA TREATMENT IN OVARIAN CANCER By Charlotte Faye Roberts, B.A. A thesis submitted in partial fulfillment of the requirements for the degree of Masters of Science in Biochemistry at Virginia Commonwealth University.Virginia Commonwealth University, 2011Major Director: Jessica K. Bell, Ph.D.Assistant Professor, Department of Biochemistry & Molecular Biology. Epithelial ovarian cancer is a lethal gynecological malignancy. Due to its asymptomatic nature it is typically detected in the latter metastatic stages. Standard treatment protocol involves surgical cytoreduction, followed by a combination of taxane and platinum-based chemotherapeutics. Initially this treatment is successful however, most patients face recurring tumors that over time become resistant to current drug regimens. Thus, novel chemotherapeutic development is necessary. Cancer cells express receptors of the innate immune system, pattern recognition receptors (PRRs) that function to alert the host of invading pathogens. PRRs such as toll-like receptor 3 (TLR3), retinoic acid-inducible gene I (RIG-I), melanoma differentiation-associated gene 5 (MDA5), and dsRNA-dependent protein kinase receptor (PKR) recognize double-stranded RNA (dsRNA), a viral replication intermediate, and trigger apoptosis. Numerous studies have been conducted on the four dsRNA receptors in cancer. The findings have shown that stimulation of individual or a group of these receptors have led to a multitude of responses such as activation of apoptosis, inhibition of tumorigenic growth, and inhibition of metastasis in several cancer types (prostate, breast, nasopharyngeal, and melanoma cancer). Previous work in the Bell lab has shown that within a panel of ovarian cancer cell lines, one subset upregulates dsRNA receptors upon stimulation with polyinosinic-polyuridylic acid (pI:pC) and leads to apoptosis. A second subset of ovarian cancer cells do not upregulate dsRNA receptors and their survival is not affected by dsRNA treatment. We hypothesize that all or a subset of dsRNA receptors are required to elicit a dsRNA-induced apoptotic response. To test this hypothesis we examined the dsRNA-induced apoptotic response the responding cell lines (CAOV3 and OVCAR3) via three methods: selective ligand assays, transient knockdowns with siRNA, and stable lentiviral knockdowns with shRNA. Then we examined the dsRNA-induced apoptotic response in the non-responders (DOV13 and SKOV3). The first objective was to determine if all or a subset of these four dsRNA receptors were required for the dsRNA-induced apoptotic response. The second objective of this thesis was to examine if dsRNA receptor expression in cell lines resistant to dsRNA-induced apoptosis could restore dsRNA responsiveness. To execute the first objective, we first examined receptor contribution to the dsRNA-induced apoptotic response via a selective antagonist (2- aminopurine) to PKR and a selective agonist (polyadenylic-polyuridylic acid, pA:pU) to TLR3. Inhibition of PKR did not blunt the apoptosis levels in the responders and was determined to be inessential for the dsRNA-induced apoptosis. Selective ligation of TLR3 with pA:pU showed an increase in apoptosis, but not to levels seen with pI:pC. Objective one was also carried out via transient knockdown using siRNA. Knockdowns via this method were less than 70% and the lipid vehicle of one of the transfection reagents was found to be sensitizing to the cells. Stable lentiviral knockdowns with shRNA were utilized to conduct the knockdown assays. By qPCR, lentiviral knockdown of TLR3 showed an 85% decrease and showed a great decrease in the dsRNA-induced apoptotic response in the cell death assay. The lentiviral knockdown of RIG-I showed a 54% decrease via qPCR and did not alter dsRNA-induced apoptotic responses. The lentiviral knockdown of MDA5 could only be assessed via the TLR3/MDA5 double knockdown, and it showed a 53% decrease via qPCR analysis. The cell death assay of the TLR3/MDA5 double knockdown showed a great decrease in the dsRNA-induced response. The work presented in this thesis is the first to address the contribution of all four dsRNA receptors to the dsRNA-induced apoptotic response in one study. In this work, we have found that PKR is not needed for the dsRNA-induced apoptosis. Loss of TLR3 in the responders reduces death, but not back to basal levels. This may be due to the delivery method of pI:pC such that it goes directly to the endosome. Forced expression of the dsRNA receptors (TLR3, MDA5, and RIG-I) can all induce apoptosis to similar levels indicating redundancy. The importance of this work reveals that any of the three dsRNA receptors, TLR3, MDA5, and RIG-I, could be possible targets for individualized chemotherapeutic regimens for women with ovarian cancer expressing these receptors.

Sterile inflammation is a key component of a variety of diseases including, gout, arthritis, type 1 diabetes, Alzheimer’s disease and multiple sclerosis (MS). Sterile inflammation induces the recruitment of immune cells via chemokines, such as CCL5 and CXCL10. Expression of these chemokines is dependent on IRF-1. Recently the FDA has approved the use of a pro-drug, FTY720 that after phosphorylation becomes a S1P mimetic for the treatment of MS. This report describes two novel and opposing mechanisms of S1P action in sterile inflammation. First, intracellular S1P acts as a cofactor of cIAP2 that inducesIL-1-dependent K63-polyubiquitination of IRF-1, which leads to the recruitment of immune cells to the site of inflammation. Conversely, extracellular S1P provides a feedback loop that inhibits CXCL10 and CCL5 expression through S1PR2 signaling. Accordingly, immune cell infiltration to sites of sterile inflammation is increased in S1PR2-/- production via calcium-dependent, but cAMP- and PKA-independent mechanisms that likely involve c-Fos expression and unconventional PKC activation. Elevated c-Fos could competitively inhibit CCL5 expression directly or indirectly via blocking IFN production. These two novel pathways highlight unexpected aspects of S1P signaling, and provide potential mechanisms that can be exploited for the improvement of therapeutics for the treatment of MS.

Vasoconstrictor tone in the splanchnic circulation redistributes blood flow during hemorrhage and resuscitation. A metabolic sensor, 5’adenosine monophosphate-activated protein kinase (AMPK), has been proposed to relax arteries by inhibiting myosin light chain (MLC) kinase (MLCK) and rho kinase (ROCK) activities. Because AMPK activation might be beneficial in re-establishing splanchnic blood flow during resuscitation, we sought to explore the relative ability of AMPK activators (AICAR, A769662, berberine (BBR) and simvastatin (SIMV)) to relax mesenteric artery (MA) contraction. Our data revealed that these drugs caused vasorelaxation when tissues were stimulated either with KCl (producing primarily a Ca2+ dependent contraction) or phenylephrine (PE; producing a primarily Ca2+ independent contraction). We further investigated the potential mechanisms by which BBR induced mesenteric artery relaxation. We found that BBR did not inhibit MLC phosphorylation, nor did it phosphorylate AMPK, and therefore is likely working through another mechanism to cause vasorelaxation. Notably, PE induced an increase in AMPK phosphorylation and, of all the AMPK activators examined, only AICAR phosphorylated AMPK in rabbit mesenteric artery, which provides a guide for future studies.

In the presented study, we demonstrate that the interaction of group IVA cytosolic phospholipase A2 and ceramide-1-phosphate is crucial for production of eicosanoid synthesis in inflammation. Inflammation is a critical component of many disease states including anaphylaxis, cancer, cardiovascular disease, rheumatoid arthritis, diabetes and asthma. Eicosanoids are well established mediators of inflammation, and the initial rate limiting step in the production of eicosanoids is the liberation of arachidonic acid (AA) from membrane phospholipids by a phospholipase A2 (PLA2). The major phospholipase involved in this liberation of AA during the inflammatory response is group IVA cytosolic phospholipase A2 (cPLA2α). Previous studies from our laboratory demonstrated that the bioactive sphingolipid, ceramide-1-phosphate (C1P), binds cPLA2α at a three amino acid sequence, which is located in the cationic β-groove of the C2 domain of cPLA2α. In this study we examined the effects of the genetic ablation of ceramide kinase (CERK) on eicosanoid synthesis, as CERK is the only known enzyme to produce C1P in mammalian systems. We utilized primary mouse fibroblasts (MEFs) and macrophages isolated from CERK-/- and +/+ mice. The ceramide-1-phosphate and eicosanoid profiles were investigated, and both ceramide-1-phosphate and eicosanoid levels in CERK-/- MEFs were found to be dysregulated. This study also presents the development of a global eicosanoid method to analyze eicosanoids via LC-ESI-MS/MS. Using this new analysis method, we demonstrated that there are significant differences in eicosanoid levels in ex vivo CERK-/- cells when compared to wild type counterparts, but the effect of the genetic ablation of CERK on eicosanoid synthesis and the serum levels of C1P was not apparent in vivo.

Head and neck squamous cell carcinoma (HNSCC) is the sixth most common malignancy worldwide. The 5-year survival rate when the cancer remains as a primary tumor is 81% but when it metastasizes to distant sites, defined as a metastatic cancer, it decreases dramatically to 26%. Approaches to prevent these cancers from undergoing these metastatic changes can greatly improve the survival and outcome of these cancer victims. This current study is examining the expression profiles of chemokines and VEGFs in HNSCC. By investigation the underlying pathways involved in the expressions of chemokines and VEGFS we hope to sort out the transcriptional regulation of these molecules. We used pharmological inhibitors of several important kinase pathways and the receptors involved in the transcription of chemokines and VEGFs. This study specifically looked at the proangiogenetic chemokines, CXCL5 and CXCL8, and their receptor CXCR2, and their possible impact on VEGFs, specifically VEGF-C and VEGF-A. From experimentation we concluded that HNSCC uses the MAPK pathway for regulation of the chemokines CXCL5 and CXCL8, but not for its downregulation. VEGF-A showed to be positively controlled by the MAPK pathway. The Akt pathway was found to downregulate VEGF-C, possibly from CXCR2. VEGF-C was not under control of the chemokines’ expression, VEGF-C and VEGF-A were also differentially regulated. The current study has begun to sort out the expression and regulation of chemokines and VEGFs in HNSCC. There are still many unanswered questions about the role these molecules play in HNSCC, but hopefully these conclusions will aid in finding improved treatments for patients diagnosed with head and neck cancer.

Platelet-Activating Factor Acetylhydrolase (PAFAH) is a monocyte-derived phospholipase A2 that catalyzes the hydrolysis of platelet-activating factor (PAF) and has been implicated in atherosclerosis. Although PAF and other proinflammatory stimuli are postulated to induce the enzyme, mechanisms controlling PAFAH expression are largely unknown at present. We have shown that PAFAH induction in monocytes is increased in response to oxidized phospholipids. The PAFAH 5' flanking region has at least 10 putative Stat elements, and IL-6 has been shown to be downstream from the prostaglandin receptor, EP2, which has been shown to bind oxidized phospholipids, prompting the hypothesis that Stat proteins might regulate its expression. To test this hypothesis, we treated human monocytes with IL-6, a monocyte-derived cytokine that activates Stat3, IL-8, a monocyte-derived cytokine induced by Stat3, and oxidized 1-palmitoyl-2-arachidonoyl-sn-3-phosphocholine (oxPAPC), a major component of the oxidized LDL particle. Two monocyte-derived cell types, macrophages (MO) and dendritic cells (DC) were prepared from primary human monocytes. The cells were treated with various doses of IL-6, IL-8, or oxPAPC for various time frames in the absence of serum. Culture supernatants from the cytokine-treated cells were harvested and screened for PAFAH protein and activity and cell monolayers were assessed for PAFAH mRNA by quantitative real time PCR (qPCR). Cells treated with oxPAPC were further analyzed for secreted IL-6 using ELISA and activation of Stat3 using Western Blot. Both IL-6 and IL-8 induced PAFAH expression in a dose-dependent manner. Although both MO and DC responded to the cytokines, preliminary experiments suggested that induction of PAFAH is more robust in DC than MO. Cytokine-treated cells exhibited increased PAFAH activity in their culture supernatants that correlated with increased PAFAH protein levels. Treatment with oxPAPC induced IL-6 secretion and subsequent Stat3 activation in DC. Together, these data support the hypothesis that PAFAH expression is regulated by oxidized phospholipids and proinflammatory cytokines in developing atheromas.