The goal of this study was to compare different anthropometric measures in terms of their ability to predict T2DM and to determine whether predictive ability was modified by ethnicity. Anthropometrics were measured at baseline on 1073 non-Hispanic Whites (nHW), African Americans (AA) and Hispanics (HA), of which 146 developed T2DM after 5.2 years. Logistic regression models were used with areas under the receiver operator characteristic curve (AROC) comparing the prediction of models. Overall, there was no clear distinction between measures of overall and central obesity in terms of T2DM prediction. Waist-height ratio (AROC=0.678) was the most predictive measure, followed by BMI (AROC=0.674). Results were similar in nHW and HA, although, in AA, central adiposity measures best predicted T2DM. Measures of central and overall adiposity predicted T2DM to a similar degree, except in AA where central measures were most predictive.

HSV-2 may have adverse consequences in HIV. I evaluated the impact of HSV-2 co-infection on (highly active antiretroviral therapy)-untreated HIV infection in a systematic review of observational studies (study 1) and a retrospective cohort (study 2). I further evaluated whether HSV reactivation rates in co-infected persons differ by use of suppressive cART (study 3). Study 1 found modest evidence that HSV-2 seropositivity may be associated with accelerated progression to opportunistic infection or clinical AIDS, but not with increased HIV viral load. Some evidence suggests that HSV-2 disease activity is associated with increased HIV viral load and decreased CD4 counts. Study 2 compared rates of CD4 count change by HSV-2 status (Focus HerpeSelect ELISA) among 218 patients with a past period of ART-untreated follow-up using mixed linear regression models. No significant difference in the rate of CD4 count change was observed in HSV-2 seropositives at +13.6 cells/mm3/year (p=0.12) in univariate analysis, and -4.5 cells/mm3/year (p=0.68) in analysis adjusted for sex, HSV-1, oral and genital HSV symptoms, immigrant status, and immigrant*time interaction. These findings support the need for carefully designed and executed studies of HSV-2 suppression as an adjunctive management strategy for HIV disease, but raise questions regarding the exact mechanism of negative synergy between these viruses and the relative importance of HSV-2 latency and replication in driving these effects. In Study 3, 44 cART-naïve and 41 treated (HIV RNA<50 copies/mL) HIV+ adults with HSV-1 and/or 2 co-infection collected oral, genital and anal swabs daily for 28 days. Negative binomial models were used to quantify the relationship between cART and HSV shedding (Roche LightCycler HSV1/2). Overall HSV shedding was low, at a median (IQR) of 3.6% (0, 14.3%) of days. No relationship was seen between cART and HSV-1 or 2 shedding in univariate (RR=1.55, 95%CI=0.83,2.87) or multivariate analysis adjusted for sex, baseline CD4, recent immigrant status, and time since HIV diagnosis (aRR=1.05, 95%CI=0.43,2.58). Null results were also observed for HSV-1 and HSV-2 considered separately. That HSV shedding persists despite cART suggests that trials of anti-HSV drugs for improving HIV outcomes may be warranted in such patients.

The goal of this study was to compare different anthropometric measures in terms of their ability to predict T2DM and to determine whether predictive ability was modified by ethnicity. Anthropometrics were measured at baseline on 1073 non-Hispanic Whites (nHW), African Americans (AA) and Hispanics (HA), of which 146 developed T2DM after 5.2 years. Logistic regression models were used with areas under the receiver operator characteristic curve (AROC) comparing the prediction of models. Overall, there was no clear distinction between measures of overall and central obesity in terms of T2DM prediction. Waist-height ratio (AROC=0.678) was the most predictive measure, followed by BMI (AROC=0.674). Results were similar in nHW and HA, although, in AA, central adiposity measures best predicted T2DM. Measures of central and overall adiposity predicted T2DM to a similar degree, except in AA where central measures were most predictive.

Successful bacterial pathogens manipulate crucial intracellular host processes as a virulence strategy. One particular potent mechanism utilized by bacterial phytopathogens is to inject virulence factors (effectors) directly into the host cell. While many effectors have been identified and shown to suppress plant immune responses, very few have well-characterized enzymatic activities or host targets. To overcome the challenges of functional analysis of effectors, I designed two heterologous screens to characterize effector proteins of the bacterial phytopathogen Pseudomonas syringae. Specifically, my objective was to identify those P. syringae effectors that target evolutionarily conserved host proteins or processes and to subsequently elucidate the molecular mechanisms of these effectors. The first heterologous screen that I performed was to utilize tandem-affinity-purification (TAP)-tagged effectors in human cells to identify potential interacting host proteins. The second heterologous screen iii utilized a high-throughput genomics approach in yeast, known as the pathogenic genetic array (PGA), to characterize P. syringae effectors. Using the first heterologous approach, I have identified HopZ1a as the first bacterial phytopathogen effector that binds tubulin. I have shown that HopZ1a is an acetyltransferase activated by the eukaryotic co-factor, phytic acid. In vitro, activated HopZ1a acetylates itself and tubulin. In Arabidopsis thaliana, activated HopZ1a causes microtubule destruction, disrupts the secretory pathway and suppresses cell wall-mediated defense. The acetyltransferase activity of HopZ1a is dependent on the conserved catalytic cysteine residue (C216) and a conserved lysine residue (K289). Using the second heterologous screen in yeast, I have shown that HopZ1a may target the mitogen-activated protein kinase (MAPK) signaling cascades. Together, my work has identified novel eukaryotic targets and elucidated the virulence functions of HopZ1a.

Purpose: The prevalence of diabetes (DM) is increasing globally with type 2 diabetes (T2DM) being primarily responsible for this alarming trend. Age and DM have been associated with an increased prevalence of AR cataract in earlier studies but T2DM has not been considered separately from type 1 diabetes. Furthermore, study results have been inconsistent in terms of whether nuclear sclerosis (NS), cortical cataract (CC) or posterior subcapsular (PSC) are specifically associated with DM. The purpose of this thesis was to provide Canadian data on these issues while considering the limitations found in earlier studies in terms of variable age group selection and cataract definition. Logistic regression analysis was extended beyond risk analysis to model the prevalence of AR cataract across the human age range. Finally, as statins are commonly prescribed for patients with T2DM, the impact of using this pharmaceutical on AR cataract prevalence was investigated. Methods: A file review of over 6397 clinic files was performed to create the Waterloo Eye Study (WatES) database. Abstracted data included patient age and sex, the presence of early to late AR cataract (NS, CC, PSC or related lens extraction-LE), systemic health diagnoses including a diagnosis of T2DM or type 1 diabetes, and any medication used. Data quality was looked at through repeatability with double-entry of files and calculation of missing data rates. Comparisons were done between the study population demographics (age and sex) and those available on the general population and representative Canadian optometric patients. Prevalence of AR cataract was determined for the entire study group and for yearly age-groups. The probability of AR cataract generated from logistic regression analysis was used to model the prevalence of AR cataract over the entire age range of patients. Similar functions were determined for T2DM and non-diabetic (ND) subgroups and then again after further subdividing them into patients who did and did not use statins. The age of 50% prevalence of AR cataract were determined for each of these functions. Distribution rates of mixed and uniform cataract were calculated and compared for the T2DM and ND subgroups. Age of first lens extraction and differences in LE rates were also determined for these groups. Multivariable logistic regression analysis was done to determine odds ratios (OR) for associations between variables (patient age, being female, having a diagnosis of T2DM, smoking, systemic hypertension, and statin use) and the outcome of AR cataracts or its subtypes. Results: Data abstraction repeatability was found to be high and missing data rates were found to be low. While significant differences existed between the demographics of the general population and this clinic population, the sex and age distributions were comparable to optometric practices in Canada. The overall prevalence of AR cataract, NS, CC,and PSC in this population was 35.3%, 28.8%, 9.9%, and 3.6% respectively. The yearly prevalence of AR cataract in this population was found to increase in a sigmoid trend over the course of the human age span that began to rise after 38 years of age and approached 100% by 75 years of age. When modelled into a probability of cataract function, 50% prevalence of AR cataract occurred at 56.6 years of age. T2DM was reported in 452 WatES patients; 97% of whom were over 38 years of age. The probability of 50% AR cataract, NS, and CC prevalence occurred almost four years earlier in the T2DM subgroup compared to those without diabetes. PSC was much less prevalent and did not reach 50% levels, but the age of 10% prevalence was eight year earlier in the T2DM group compared to the ND group. Patients with T2DM had more mixed cataract, a higher rate of LE and an earlier age of first LE than non-diabetics. Statin use was reported in 761 patients; 96% who were over 38 years of age. Statin use was 3.5 times more common in patients with T2DM compared to non-diabetics. When the diabetic subgroups were further subdivided by those who do and do not use statins, the age of 50% probability of AR cataracts was now almost eight years earlier in the T2DM patients using statins compared to the ND patients who did not. The probability functions were similar between T2DM patients not using statins and ND patients who did report statin use. Having a diagnosis of T2DM was significantly associated with early to late NS and CC when controlling for statin use, whereas statin use was significantly associated with NS and PSC when controlling for a diagnosis of T2DM. Conclusions: AR cataract, T2DM and statin use were prevalent conditions in this clinic population, especially over 38 years of age. Modelling the prevalence of AR cataract over a broad age range could assist predicting cataract in Canadian optometric patients. A diagnosis of T2DM resulted in an earlier development of all three cataract subtypes, resulting in increased rates of LE and mixed cataract. However, the association was only significant for NS and CC when controlling for statin use. Given the frequent use of statins in patients with T2DM, the significant association found between statin use and increased risk of AR cataract warrants further study.

The increase in waste heat after consuming energy or burning fossil fuels is an issue environmentally and economically. Thermoelectric (TE) materials are developed to use in various applications because of their ability in converting waste heat into electricity. However, the applications are limited due to a low efficiency of materials, and research on thermoelectric materials is an on-going project for future use. Type-I clathrates are one of the TE materials which are studied in depth since the proposal of Slack’s PGEC (Phonon-Glass-Electron-Crystal) concept in 1995 due to their excellent thermoelectric properties. In this study, development and optimization of quaternary type-I clathrates will be the focus because double substitution often leads to better figure-of-merit, ZT, but it hasn’t really been studied. Higher ZT value is necessary because the energy conversion efficiency of TE materials is depending on the ZT value along with a larger temperature difference. Addition of lanthanoid elements as 2nd guest atoms to the main type-I clathrate structure, realized in Ba8Ga16Ge30, will be attempted to form quaternary compounds. The formation of the quaternary clathrates will be analyzed through powder X-ray diffraction, single crystal analysis and energy dispersive X-ray analysis. Also, as the performance of TE materials is examined through the figure of merit, ZT = TS²σ/κ, various techniques will be used to determine the Seebeck coefficient, the electrical conductivity and the thermal conductivity. The quaternary clathrates, Ba8-xLnxGa16Ge30 and Ba8-xLnxGa16+xGe30-x, where Ln = La, Ce and Eu were synthesized from the pure elements in stoichiometric ratios at 1000°C with slow cooling to room temperature. The products were then annealed at 600°C to acquire homogeneous samples for analyses. The various compositions of lanthanoid were intercalated into the structure of clathrates, which resulted in the quaternary clathrates with homogeneity. The crystal structure of quaternary clathrates with the space group of Pm-3n exhibited the same structure type as the ternary clathrates. The successfully formed products were refined with Rietveld refinements to understand their structures. The Eu containing clathrates crystallized with a lattice parameter a = 10.78251(6) Å, V = 1253.60(2) ų, for x = 0.3. The Ce containing clathrates also adopted the same space group with a lattice parameter a = 10.77331(6) Å, V = 1250.40(2) ų, for x = 0.3. The La containing clathrates formed with a lattice parameter a = 10.78494(6) Å, V = 1254.45(2) ų, for x = 0.3. Between 0.2 and 1.0 lanthanoid elements per formula unit were substituted with decreasing amount of barium where the actual amount of Ln in clathrates was lower than nominal amount. All these quaternary clathrates were found to be n-type semiconductors as determined through the Seebeck coefficient and electrical conductivity measurements.

This thesis investigates the internal structure and the computational representation of the lexical entries in Combinatory Categorial Grammar (CCG). A restricted form of typed feature structures is proposed for representing CCG categories. This proposal is combined with a constraint-based modality system for basic categories of CCG. We present some linguistic evidence to explain why both a unication-based feature system and a constraint-based modality system are needed for a lexicalist framework. An implementation of our system is also presented.

The purpose of this study is to investigate the effect of rock types and morphologic classes on the locations of underground cities existing in Cappadocia. To achieve this purpose four databases are created that contain related information of underground cities, present settlements, rock types and morphologic classes. Four main analyses are carried out using the data created fort the study. These analyses are: 1) Distance analysis to determine the distances between underground cities and present settlements, 2) Density analysis to inspect the areas where the underground cities are concentrated, 3) Distribution analysis to explore the spatial distribution of underground cities within the rock types and morphologic classes, and 4) Neighbourhood analysis to examine whether the underground cities within rock types and morphologic classes are located along or far inside the marginsof the polygons. The conclusions reached after the analyses are as follows: 1) The mean distance between two underground cities is about 4 km. 2) The mean distance between an underground city and the nearest present settlement is about 700 m. 3) Underground cities are concentrated in Derinkuyu-NevSehir-&Ouml / zkonak belt. Present settlements, on the other hand, are concentrated along Aksaray-Ortak&ouml / y-HacibektaS. 4) For the underground cities, pyroclastic dominant Neogene sequences are preferred whereas all other units are avoided. 5) In terms of morphology, the class defined as &ldquo / mesa&rdquo / is strongly preferred for underground cities. 6) Neither lithology nor morphology played a role in the site selection for present settlements. 7) Both for rock types and morphologic classes the underground cities are located along margins of the polygons.

Neisseria meningitidis is a causative agent of meningitis and septicaemia. Pili are one of the major virulence factors that contribute to the pathogenicity of N. meningitidis. Pili of Neisseria are type IV fimbriae composed primarily of thousands of identical pilin subunits. Pilin of N. meningitidis is post-translationally modified by trisaccharide, phosphorylcholine and -glycerophosphate. The genes involved in pilin expression, pilin glycosylation and phosphorylcholine modification are phase variable (high frequency ON/OFF switching of expression). The function of pilin post-translational modifications and their phase variable expression in host:pathogen interactions is unknown. The phase variable expression of glycosylation in bacteria has been proposed to function in bacterial adherence and immune avoidance. However, the function of pilin glycosylation in N. meningitidis is unclear. Phosphorylcholine is expressed in a number of respiratory organisms including P. aeruginosa (on teichoic acid), S. pneumoniae (on lipoteichoic acid) and H. influenzae (on LPS). Phosphorylcholine in these organisms is important in colonisation of the nasopharynx and invasion of the epithelium. Studies on N. meningitidis pilin post-translational modifications have been restricted by difficulties in purification of pilin protein. In this thesis, we evaluated current pilin purification methods and established an efficient method of purifying pilin from N. meningitidis by Flag-tag purification system. Flag-tag purified pilin is post-translationally modified. The LC-ESI/MS/MS analysis performed in this thesis using Flag-tag purified pilin successfully determined the phosphorylcholine post-translational modification sites. Based on the MS data and the mutagenesis analysis, phosphorylcholine is covalently linked to serine 157 and serine 160 of pilin. The colony immunoblot of a serine 157/160 to alanine mutant revealed that phosphorylcholine modifications of these sites on pilin are the only surface exposed phosphorylcholine and is responsible for binding to TEPC-15 (the monoclonal antibody which binds to phosphorylcholine). In this thesis, molecular modelling demonstrated that surface exposure of pilin phosphorylcholine could be altered by the phase variation of pilin glycosylation on the adjacent pilin monomer. Furthermore, the sites for phosphorylcholine modification are commonly observed in N. meningitidis strains but not in N. gonorrhoeae indicating the importance of phosphorylcholine in pathogenisis of N. meningitidis. In addition, the biosynthesis of phosphorylcholine for pilin post-translational modification still remains a mystery. Bacteria generally obtain choline from the environment. In this thesis, we demonstrated that pilin phosphorylcholine post-translational modification could be endogenously synthesized in N. meningitidis. In summary, this thesis describes the purification method of obtaining pure post-translationally modified pilin from N. meningitidis. The phosphorylcholine post-translation modification sites on pilin have been determined and showed the importance of these sites in antibody binding specificity.

Type 1 Diabetes (T1D) is an autoimmune disease where the immune system destroys the insulin-producing beta cells within the pancreas. Due to the difficulty of obtaining relevant tissue samples from patients at risk of disease, many researchers have utilized the nonobese diabetic (NOD) mouse as a model of T1D due to their natural high susceptibility for this disease which shares many characteristics with human patients. This model has been critical for uncovering many mechanisms involved in the pathogenesis of T1D including the key roles played by autoreactive T cells in the destruction of beta cells. More recently, NOD mice have shown that self-reactive B cells act as important antigen presenting cells for activating and amplifying the T cell response against beta cells. In order to identify faulty self-tolerance mechanisms causing production and activation of B cells recognizing beta cell proteins, we have developed a transgenic mouse model whereby elevated numbers of B cells are made specific for a neo-self antigen whose expression is restricted to beta cells on the T1D-prone NOD genetic background and compared it to that of transgenic mice of the non-autoimmune prone C57BL/6 (B6) genetic background. These studies revealed that NOD and B6 B cells can both be effectively tolerized to the model beta cell-restricted antigen. However, provision of help from activated T cells readily overturned this tolerance on the NOD but not the B6 background. Prior evidence has associated Idd5 (chromosome 1) and Idd9/11 (chromosome 4) diabetes susceptibility loci in NOD mice with the development of self reactive B cells contributing to T1D. The gene encoding CTLA-4 has been identified as the major candidate susceptibility gene within Idd5, thus leading to our studies comparing B cell expression of this molecule in NOD and diabetes-resistant strains. Although almost always associated with down-modulating T cells responses, our studies and that of others confirm expression of CTLA-4 by activated B cells. We encountered B cell expression of CTLA-4 to vary from that of T cells, being expressed earlier and predominantly on the cell surface rather than within intracellular vesicles. Our studies also showed aberrant expression of different splice variants of CTLA-4 by NOD B cells compared to diabetes-resistant mice controlled by genes within and outside the Idd5 genetic locus. Hence, these studies raise the possibility that CTLA-4 may contribute to T1D through its actions on both T and B cells. Given the large nature of the Idd9/11 susceptibility locus in NOD mice and the absence of any strong candidate genes that may influence the diabetogenic capacity of B cells in this strain, we resorted to microarray technology to reveal putative genes within this genomic region with the potential to control the B cell phenotype. We focused our microarray studies on the first transitional (T1) B cell population in the spleen given that it is an important stage of tolerance to peripherally expressed self-antigens which have been found to possess various defects in NOD mice. Comparing gene expression profiles of NOD T1 B cells that expressed susceptibility or resistance alleles at the Idd9/11 locus identified 20 differentially expressed genes with the potential to contribute to development of diabetogenic B cells. Overall, data presented in this thesis provides a greater understanding of the molecular and cellular mechanisms underlying B cell contribution to T1D in NOD mice. These data are hoped to eventually lead to the development of selective strategies for removing or inhibiting only those B cells that contribute to development of T1D while ensuring that humoral immunity to foreign pathogens remains intact in human patients at risk of developing disease.