HIV-associated neurocognitive disorders (HAND) affect up to 70% of HIV positive individuals and are the leading cause of dementia in patients under 40 years. Despite this, the molecular mechanisms involved in the onset of HAND are not well understood. Among a number of plausible etiological agents of HAND, HIV-Tat has been shown to be neurotoxic in vitro and in vivo, but the basis of its induced neuronal dysregulation remains relatively poorly characterised, giving rise to various competing theories. This thesis describes differential, quantitative proteomic analyses of HIV-Tat-treated neuronal cells in vitro, the goal being to gain deeper insight into the underlying molecular basis of this HIV-Tat-mediated dysregulation, as well as to potentially inform better patient treatments in the future. To achieve this goal, deep, quantitative proteomic analysis of HIV-Tat treated SILAC-labelled SH-SY5Y neuroblastoma cells was carried out, alongside transcriptomic analysis of the same system in which 3077 proteins were identified and quantified with 407 proteins and 1074 genes being differentially expressed. Subsequently, label-free proteomics analysis was used to study the ability of lithium - a proposed new treatment for HAND - to suppress the HIV-Tat induced dysregulated molecular phenotype in SH-SY5Y cells in which 3757 were identified and quantified with 360 and 531 being significantly differentially expressed in HIV-Tat and HIV-Tat + lithium treated cells, respectively.

Airborne infectious diseases, such as tuberculosis (TB), are spread by airborne infectious particles (viable particles with potential for TB infection) in exhaled air from infectious individuals in enclosed spaces. Exhaled air is the carrier of airborne infectious particles and carbon dioxide is used as a surrogate of this exhaled air. Using carbon dioxide as a surrogate for exhaled air, we modified the Wells-Riley model and the prior modified versions of the model, and obtained a exible but sensitive mathematical model that predicts the risks of airborne infectious diseases, such as TB under steady- state and non-steady-state conditions, without assumptions of well mixed airspace and equilibrium conditions. Applying experimental data from in vivo studies to the mathematical model developed in this study, we explored the probability of exposed guinea pigs acquiring infection in these in vivo stud- ies and quantified the number of surviving airborne infectious particles (infective organisms) required to reach the alveolar to establish infection. Our study shows that the number of infective organisms reported in the in vivo studies might have been markedly underestimated. In this study, we investi- gated TB transmission in congregate settings, such as schools, households, public transport, prisons and health care settings and suggested preventive measures. TB transmission in these locations is attributable to numerous factors, including overpopulation and air pollution, which acts as a carrier of airborne infectious particles. We explored the impact of effective contact rate on TB epidemiology using a mathematical model we developed that consists of five states of susceptible, primary infection, reinfected, active TB and treated individuals. An infectious individual with varying effective contact rate (ranging from 5 to 30 per year) was introduced among 100; 000 fully susceptible individuals and we observed the number of primary infection and reinfected individuals at stability points of a TB epidemic. We found that the number of primary infection individuals decreases at the stability point, while that of reinfected individuals increases with increasing effective contact rate. This implies that a large number of active TB cases might be reinfected individuals. Using an age-structured mathemat- ical model developed in this study that incorporates vaccination, we explored TB disease progression in different age groups (from 0 to ≥ 75 years). We found that TB disease progression is age dependent. High TB notification rate was detected for the age groups [0 - 5); [15 - 25); [45 - 55) and [55 - 65) years, and the lowest TB notification rate was detected in the age group [5 - 15) years. Furthermore, we noted that vaccination decreases active disease progression for the age groups [0 - 5) to [15 - 25) years, while TB notification remains high for the age groups [25-35) to ≥ 75) years. The findings in this study suggest that active disease progression depends on age and average duration of the waning of the vaccine effect.

The purpose of this dissertation was to provide an overview of acute intermittent porphyria, focussing on the structure and function of the enzyme, porphobilinogen deaminase (PBGD), as well as experimentally demonstrating the use of kinetic, structural and thermodynamic approaches to shed light on the enzyme reaction. The key focus was to investigate the effect of three mutations of the active site lysine 98 residue (K98) on the enzyme’s stability and mechanism. Two clinically relevant PBGD mutants, the K98E and K98R were expressed. Both of these mutants have previously been described in patients. We engineered and expressed an additional mutant, K98A, in order to investigate the effect of charge at this residue. The K98E, K98R and K98A recombinant proteins were successfully engineered, expressed and purified. These mutations were kinetically characterised, and the low enzyme activity supports the fact that the K98E and the K98R are known-disease causing mutations. The negligible activity of the K98A and K98R mutants was predicted as a result of a loss of DPM co-factor binding, which was analysed and proved with a co-factor spectral shift assay. Further attempts to examine the interaction of co-factor binding involved removal of the bound cofactor from wild type enzyme, in order to investigate the possible interaction of the ‘apo’- enzyme with the DPM co-factor. However, no results were obtained to elucidate this interaction, largely due to the highly unstable nature of the generated ‘apo’-enzyme. Native polyacrylamide gel electrophoresis (PAGE) was performed in order to observe changes in enzyme-substrate complexes between the wild type and the different mutant proteins. The enzyme-substrate complexes for the wild type were clearly shown, however we could not do so in our mutant proteins. The secondary structure estimations as well as the conformational stability of the mutants were tested with the use of circular dichroism. Far- and near-UV analysis provided insight into the effect of each mutation on the enzyme’s secondary and tertiary structure respectively. Results indicate that the different mutations cause marginal alterations in secondary structure, and resulted in changes of aromatic ring conformations in the near-UV analysis. Finally, modelling of each mutation to known crystal structures of the human enzyme was done in order to provide a rationalisation of kinetic and conformational data. Although this provided only a static image and estimation of the structural effect of each mutation, it did allow for some speculation in order to rationalise the kinetic and conformational data obtained. Overall, this work illustrates how the characterisation of expressed, purified, AIP-associated mutant enzymes aids our understanding of the complex structure and mechanism of the PBGD enzyme.

Kaposi's sarcoma-associated herpesvirus (KSHV), a gamma-herpesvirus with a particularly high seroprevalence in Sub-Saharan Africa (SSA), is the etiological agent of the endothelial tumour Kaposi's sarcoma (KS), the most common acquired immunodeficiency syndrome (AIDS)-related malignancy worldwide and particularly in SSA. It also causes primary effusion lymphoma (PEL), multicentric Castleman disease (MCD) and KSHV inflammatory cytokine syndrome (KICS). AIDS-related deaths have declined, due to global scale-up of antiretroviral therapy (ART). However, the vast majority of these occurred in SSA, where tuberculosis (TB) is the leading cause of mortality among human immunodeficiency virus (HIV)-infected individuals, accounting for a third of all AIDS-related deaths. The exceptionally high burden of suspected TB in SSA causes misdiagnosis or delayed diagnosis of diseases mimicking TB, such as several pathologies associated with KSHV. KSHV infection is essential but insufficient for the development of KS and other KSHV-associated pathologies; precipitating factors, such as HIV-related immune suppression and potentially genetic predisposition, are required. The erythropoietin-producing hepatocellular carcinoma (Eph) receptor A2 protein (EPHA2) tyrosine kinase receptor is a promising candidate for studies on genetic variants as it potentially acts on two levels: susceptibility to KSHV infection (being one of the key receptors utilised by KSHV for cell entry and intracellular trafficking) and susceptibility to KS development (being implicated in oncogenesis). Despite the high seroprevalence in SSA, the contribution of dysregulated KSHV lytic replication or host KSHV receptor variations to disease outcome in HIV-infected patients is unknown. We hypothesised that KSHV lytic reactivation plays yet unrecognised roles for morbidity and mortality in high HIV settings and to this end, we conducted a cohort study of 682 HIV-positive critically ill patients admitted to Khayelitsha Day Hospital, South Africa, investigated for TB, and followed for 12-weeks to ascertain vital status. We demonstrated that elevated blood KSHV viral load (VL) was a strong predictor of death in hospitalised HIV-infected patients without microbiologically proven TB. Further, we identified and validated variants in the EPHA2 protein tyrosine kinase and sterile alpha motif domains that were significantly associated with susceptibility to infection, KS development and/or KSHV VL in 300 South African HIV-infected patients, by aggregate by-gene analysis. In order to elucidate the functional significance of the identified EPHA2 missense mutations, we knocked out endogenous EPHA2 by CRISPR/Cas9 in the human endothelial cell line, HuARLT2, and reintroduced the wild type and mutant EPHA2 open reading frames by lentiviral transduction. These engineered cells were assessed for baseline EPHA2 phosphorylation levels and susceptibility to KSHV infection utilising recombinant KSHV in binding, internalisation and infection assays. We found that the EPHA2 mutant c.2254T>C (p.Leu700Pro) in the tyrosine kinase domain, associated with KS in our patient cohort, was deficient in tyrosine phosphorylation and less permissive to rKSHV infection when introduced as a single mutation or as a double mutant together with c.2257A>C (p.Asp701Ala) which was found to be in linkage disequilibrium with it. Another tyrosine kinase domain variant, c.2688G>S (p.Ala845Pro), found to be overrepresented among KS patients, had enhanced baseline tyrosine phosphorylation levels. These findings validated the patient-derived data on the molecular level by assigning functional consequences to some mutants which might have implications for the development of future biomarkers predicting KS susceptibility in high-risk populations. In summary, this novel research contributes to the understanding of KSHV-associated pathology and disease outcome. It identified KSHV VL as a potential biomarker to predict KSHV-associated diseases and mortality and assessed the contribution of KSHV entry receptor EPHA2 variations to KSHV-associated pathologies, with potential clinical implications, by facilitating the development of novel diagnostic and surveillance tools.

The integrated psychotherapy model (IPM) is an insight-oriented, integrative therapeutic approach that weaves psychoanalytic, cognitive-behavioral, and humanistic approaches into a treatment methodology. This model is new and untested; therefore, its therapeutic effectiveness is unknown. The purpose of this study was to measure the treatment effectiveness of IPM using Bell's Object Relations and Reality Testing Inventory, the Constructive Thinking Inventory, and the Working Alliance Inventory. Participants in the study included 19 undergraduate psychology students volunteering for extra credit and 11 clients of counseling psychology graduate students. This quasi-experimental, pretest-posttest, nonequivalent group study involved 9 sessions of IPM for the treatment group and 9 classes in a general psychology course for the comparison group. An analysis of covariance using the pre-post testing of object relations and reality testing, productive and unproductive thinking, and working alliance measured changes in these constructs and determined the therapeutic effectiveness of IPM. Results revealed that there were no differences between the experimental and comparison groups. Although no significant differences were demonstrated when comparing pre and post testing, this study demonstrated that 9 sessions of IPM did not harm those who underwent the treatment; this finding is positive given the need for further research to potentially validate the IPM as a new and effective integrative model for psychotherapy. It is recommended that a similar study be repeated with more seasoned IPM therapists, a longer treatment period, and the focus of change on client symptoms.

Parasitic manipulation of host behavior lies at the intersection of disease, animal behavior, and coevolutionary processes. In many of these interactions, the underpinning biology is brought into sharp focus as they are obligate relationships, under strong selection to bring about specific changes in host behavior that determine if the parasite will transmit or die. However, experimental and molecular techniques to understand these interactions are still developing and identification of the mechanisms of manipulation is a primary goal in the field. As such, we investigated host-parasite interactions between Camponotus floridanus (Florida carpenter ant) and Ophiocordyceps camponoti-floridani (Florida zombie ant fungus) from multiple molecular perspectives. By combining genome, gene expression, protein-interaction, and metabolite data from multiple experiments, we analyzed parasitic manipulation in a multiomic framework. We considered the most robust hypotheses of how parasitic manipulation occurs to be those supported by multiomic data. Two major avenues of parasitic influence on host behavior appear to be direct interference with neurotransmission and dysregulation of core cellular pathways that affect behaviors. For example, heightened expression of host dopamine synthesis enzyme genes, predicted binding of secreted parasite proteins to dopamine receptors, and reduced dopamine precursor abundance during displays of manipulated behavior all correlate the dysregulation of dopaminergic processes to manipulation phenotypes. We discuss numerous possible hypotheses, many with multiomic support, some without. We predict that modification of host behavior is a complex and multi-layered process that integrates multiple mechanisms we propose here.

HIV-1 subtype C is currently the most prevalent in the epidemiology of HIV/AIDS cases in subSaharan Africa. Most clinical protease inhibitors (PIs) were designed against subtype B and are reported to have reduced activity against subtype C proteases. Our initial hypothesis was to create an Escherichia coli based life-or-death selection system for the screening of potential PIs against HIV-1 subtype C protease (PR). This system was engineered by inserting an HIV PR cleavage sequence between the export signal peptide of the commonly used TEM-1 β-lactamase, which upon co-expression of the HIV PR in vivo, would cleave the modified β-lactamase thus preventing its translocation to the periplasmic space. This would result in the host cells' sensitivity to β-lactam antibiotics supplemented in the growth media. The presence of an inhibitor would restore resistance and therefore ‘life'. Despite validation of the E. coliscreening system using the Tobacco Etch Virus (TEV) protease, co-expression of HIV protease subtype C did not inhibit cell growth. Further investigations revealed PR C activity was inhibited by an endogenous E. coli protein. The inhibitor was isolated from E. coli crude cell lysates using ammonium sulphate precipitation, gel filtration and anion exchange chromatography fractionation. It was identified using peptide fingerprinting mass spectrometry (PMF), as alkyl hydroperoxide reductase C22 subunit (AhpC22). Mass-Assisted Laser Desorption-Ionisation-time of flight (MALDI-TOF) analysis of the precursor pre-incubated with AhpC22 revealed reduced autocatalytic cleavage occurring at the N-terminus of PR C. Inhibition kinetics using a recombinant source of AhpC22 characterized the enzyme as a non-competitive inhibitor of PR C activity with an inhibition constant (Ki ) of 0.88 µM. We also describe a protocol to express, purify and refold the HIV-1C protease which is well known for aggregation into inclusion bodies.

Sea turtles are long-lived, globally distributed animals with a complex life-history. Individuals from different populations often share the same foraging areas (mixed stock aggregations). Understanding patterns of dispersal and connectivity between reproductive populations and mixed stock aggregations is fundamental for the development of effective conservation plans. Recently, green sea turtle (Chelonia mydas) populations in several reproductive areas have increased, providing an opportunity to evaluate how demographic changes in reproductive areas impact dispersal to, and the composition of, mixed stock aggregations. In this dissertation, I evaluated how dispersal from reproductive populations in the Greater Caribbean to mixed stock aggregations may have changed over time (Chapter 2). I analyzed mitochondrial DNA haplotypes from samples collected from nesting females captured at Melbourne Beach, Florida, USA, and in-water juveniles from two mixed stock aggregations in central Florida (Indian River Lagoon and Trident Basin) over two time periods. Over a 15-year period there were small variations in the composition of the mixed stocks, without a clear relationship to the recent growth in reproductive populations. I developed a modification to the established "many-to-many" mixed stock model to use the distance between source populations and mixed stock aggregations to weight model estimates. In Chapter 3 I created a simulation to understand how sample size and the level of similarity in relation to haplotype frequency between source populations can impact mixed stock model estimates. I determined that a minimum of 150 samples from each mixed stock aggregation is required to accurately estimate contributions from source populations to mixed stock aggregations for most cases using data currently available in the literature. Improving the resolution of the genetic marker used (i.e., increasing the distinction of haplotype frequencies between source populations) can produce similar results using a smaller number of samples. Finally, in Chapter 4 I evaluated genetic structure of green turtle populations in the Greater Caribbean using a next-generation sequencing approach. I used the same sampling scheme as Chapter 2, with samples from a nesting beach (Melbourne Beach, FL) and two mixed stock aggregations (Indian River Lagoon and Trident Basin, Florida). I identified 4 distinct populations within the samples, and similar to the mtDNA assessment in Chapter 2, the genomic approach also showed small variations in the composition of mixed stock aggregations over a 15-year period. I used a coalescent model to evaluate how these populations diverged from one another, and found strong support for current gene flow among all 4 populations. Results from my analyses reiterate the complexity of sea turtle's dispersal dynamics, and the level of connectivity among populations in the Greater Caribbean. Future studies using mixed stock analysis should consider sample size with more than 150 samples per mixed stock aggregation and the use of more refined genetic markers. Also, genomic assessments of across multiple reproductive aggregations are required for a deeper understanding of other aspects of their ecology.

Maize streak virus (MSV; Genus: Mastrevirus; Family: Geminiviridae) causes maize streak disease (MSD); a major biotic threat to maize farming especially in sub-Saharan Africa, and it neighbouring Indian and Atlantic Ocean Islands, where its insect vectors in the genus Cicadulina thrive. Of the eleven known MSV strains (called A through K), only MSV-A is economically significant as it is the only one that causes severe disease in maize. MSV is a single stranded DNA (ssDNA) virus which, like RNA viruses, has high mutation and recombination rates. Given that these processes can sometimes promote viral diversity and result in the rapid evolution of new, fitter MSV variants, continuous genomic surveillance of MSV is therefore important. Based on analyses of full genome sequences, MSV-A has been classified into five subtypes (-A1, -A2, -A3, -A4, and -A6) and more than 20 recombinant lineages. Here, I showed using laboratory-based experiments that maize infecting mastreviruses such as Maize streak Reunion virus (MSRV) and MSV-C which have been found maize plants displaying severe streak symptoms do not in fact cause severe streak symptoms in maize when used to infect maize on their own. Although a mixed infection involving MSRV and MSV-B resulted in slight changes in symptom phenotypes it is unlikely that MSRV and MSV-C are responsible for emerging maize diseases. I carried out model-based phylogenetic and phylogeographic analyses of MSV-A movement dynamics in and out of Madagascar, Ethiopia and Rwanda using newly determined MSV-A genome sequences (Madagascar: n = 56; Ethiopia: n = 84) together with other sequences from GenBank. I showed that most movements of MSV-A into Madagascar have been from East Africa between the early 1990s and 2000s. My inferences show that MSV-A1 variants currently found in Ethiopia likely arrived there from Uganda or Kenya between 1985 and 1988. Similarly the MSV-A1 variants found in Rwanda likely also moved there from Ethiopia, Kenya or Uganda between 2007 and 2011. The time periods over which inferred movements of MSV-A1 into Madagascar, Rwanda and Ethiopia occurred all correspond with the period during which trade between these and other East African nations was being liberalized. Although these temporally-scaled phylogeographic analyses indicated that human activities are likely responsible for some of the long-range movements of MSV-A1 variants (such as movements from East Africa to Madagascar), leafhopper-mediated dissemination of these variants also likely played a major role in long and short distance movements of these variants within both Madagascar and between East African countries. Over 90 years of evolution that yielded MSV-A-ZW-MatA_1994 in the MSV-A1 lineage, produced symptoms that have varied in a less concerted ways, or largely remained unchanged. Major harms (intensity of chlorosis, leaf deformation and stunting) have decreased while the amount of colonized cells (chlorotic areas) that determine onward transmission have increased. These data suggest MSV-A has evolved to optimize the number of cells it infects for effective onward transmission, while reducing excessive harm to its hosts. Altogether, these results suggest (1) synergism potentially plays a role in some instances of severe streak disease and (2) the movement of MSV-A1 within the East African region and Madagascar emphasizes the importance of this MSV-A subtype as a major ongoing threat to maize production within these regions; and (3) over the last 90 years, the MSV-A1 subtype has evolved to produce greater chlorotic areas on the leaves of infected maize plants while at the same time either not increasing or reducing the degrees of chloroplast destruction, stunting and deformation caused by infections: characteristics that may have enhanced the transmissibility of this variant and therefore played an important role in the present rise to dominance of this subtype throughout East Africa and Madagascar.

Parasitoid wasps are ecologically and economically important as biological control agents. However, little is known about the diversity, distribution and biology of most hymenopteran parasitoids due to their small size, morphological conservatism, and complex life styles. The focus of my PhD research was to investigate the evolution and speciation of euphorine braconid wasps, using a combination of multilocus phylogenetics and population genomic techniques combined with traditional taxonomy. The three data chapters of my dissertation are divided into different taxonomic ranks of euphorine braconids, focusing on genera, species, and populations. For chapter 2, I built a multilocus phylogeny of the tribe Euphorini with extensive taxa sampling around the globe. I confirmed the monophyly of Peristenus and Leiophron, two important biological control agents, and provided updated generic concepts and identification resources to aid applied researchers. In Chapters 3 and 4, I focused on cryptic species within the Peristenus pallipes complex in North America. I used an integrative taxonomic approach to resolve the taxonomic confusion within the Nearctic Peristenus pallipes complex (Chapter 3), then I used ddRADSeq to examine their evolutionary relationships with their Lygus hosts (Chapter 4). My dissertation provided a comprehensive analysis of Peristenus at multiple taxonomic ranks using phylogenetics and population genomics, providing insights into their evolutionary history that can be extrapolated into other groups of parasitoid wasps. The results from these studies also advanced our understanding of this group of animals of theoretical, economical, and conservation importance.