MSc., Faculty of Science, School of Statistics and Actuarial Science, University of the Witwatersrand, 2011 / The Human Immunodeficiency Virus (HIV) epidemic is currently one of the greatest challenges and most important health issues in the world. South Africa has one of the fastest growing epidemics in the world therefore reliable estimates of prevalence and incidence are required for understanding the magnitude of the epidemic and improving the methods of prevention. This study examines the estimation of HIV incidence from a cross-section of people, using one of the laboratory methods that discover recent HIV infection in blood samples. The incidence estimate is obtained at a single point in time, thereby saving time and cost expended in following a cohort over a period of time. It also examines incidence from pooled blood samples, and evaluates the assumptions of the different methods of estimating HIV incidence, comparing each of them; and checking the sensitivity of the estimates to the assumptions. Results from the simulation study shows that accurate estimates of incidence can be obtained by pooling blood samples; and these estimates are obtained at a fraction of the cost of individual testing.

Highly active antiretroviral therapy has successfully limited HIV-1 disease progression to AIDS, but is consistently compromised by the emergence and transmission of HIV-1 drug resistant strains. As a result, a continued search for novel anti-HIV-1 agents with improved pharmacological profiles has become fundamental. Chrysotherapy has been in use since the early 1920s in treatment of rheumatoid arthritis, and has since been investigated for various ailments including HIV/AIDS. This study evaluated 45 synthetic gold compounds for drug like properties using theoretical and experimental techniques with the aim of generating sufficient data to considerably aid the rational design of new anti-HIV agents. Theoretical techniques applied included the Osiris Property Explorer and the Lipinski’s Rule of Five which assessed drug-likeness and bioavailability respectively. In vitro studies included aqueous solubility assays, cytotoxicity (PM1 cell lines and PBMCs) assays, antiviral assays in PBMCs, direct enzyme (RT and IN) inhibition, and the effect of serum protein binding and biological stability on antiviral efficacy. An overall low druglikeness score and an intermediate bioavailability were predicted by the Osiris Molecular Property Explorer. Low drug-likeness was suggested to be due to a high frequency of foreign fragments in the synthetic gold compounds, while their high molecular weight reduced bioavailability. In general gold compounds exhibited cytotoxicity properties and moderate aqueous solubility in vitro. Overall, the 45 synthetic gold compounds did not show activity against HIV-1 replication in vitro. Seven compounds (AB05-AB11) exhibited direct HIV-1 RT inhibition, and compounds AB39 and AB04 demonstrated moderate direct HIV-1 IN inhibition, but this activity was abrogated in PBMC inhibition assays. Serum binding, compound stability and cytotoxicity were all implicated in the lack of HIV-1 inhibition in PBMCs. To this end, data obtained was sufficient to aid in the future rational design of second generation HIV RT and IN inhibitors with acceptable pharmacological properties.

Ph. D, Faculty of Science, University of Witwatersrand, 2011 / The evolution of drug resistance in human immunodeficiency virus (HIV) infection has been a focus of research in many fields, as it continues to pose a problem to disease prevention and HIV patient management. In addition to techniques of molecular biology, studies in mathematical modelling have contributed to the knowledge here, but many questions remain unanswered. This thesis explores the application of a number of hybrid stochastic/deterministic models of viral replication to scenarios where viral evolution may be clinically or epidemiologically important. The choice of appropriate measures of viral evolution/diversity is non-trivial, and this impacts on the choice of mathematical techniques deployed. The use of probability generating functions to describe mutations occurring during early infection scenarios suggest that very early interventions such as pre-exposure prophylaxis (PrEP) or vaccines may substantially reduce viral diversity in cases of breakthrough infection. A modified survival analysis coupled to a deterministic model of viral replication during transient and chronic treatment helps identify clinically measurable indicators of the time it takes for deleterious rare mutations to appear. Lastly, persistence of problematic mutations is studied through the use of deterministic models with stochastic averaging over initial conditions.

Hepatitis B virus (HBV) has been found to be highly endemic in Africa and south east Asia. In southern Africa, subgenotype A1 and genotype D prevail while in south east Asia genotype B and C predominate. Infection with HBV can lead to a wide spectrum of clinical presentations ranging from an asymptomatic carrier state to self-limited acute or fulminant hepatitis to chronic hepatitis with progression to cirrhosis and hepatocellular carcinoma (HCC). It has been shown that viral factors as well as a number of host and environmental factors can influence the course of HBV infection. Development and progression of various liver diseases are associated with either an increase or decrease in hepatocyte apoptosis. Dysregulated apoptosis itself may be a fundamental feature of most acute and chronic human liver diseases. The purpose of this study was to characterise the subgenotype A1 and genotype D HBV infection, prevailing in South Africa. To control for the influence of host factors on HBV infection as well as to avoid the use of in vitro cell lines, such as Huh-7, that have defective apoptotic pathways, the in vivo urokinase plasminogen activator severe combined immunodeficient (uPA-SCID) transgenic mouse model was utilised. The HBV infection of the transgenic mice infected with HBV positive sera containing either subgenotype A1 wildtype, subgenotype A1 with the G1862T mutation, subgenotype A2 or genotype D, was compared. For the first time, we were able to demonstrate the successful infection of the uPA-SCID transgenic mouse model with subgenotype A1 of HBV. The successful establishment of the in vivo HBV infection with different genotypes or subgenotypes in the uPA-SCID transgenic mice was demonstrated by the increase of HBV DNA levels, the presence of cccDNA and HBV transcripts as well as the detection of the core and/or surface HBV antigens in the liver tissue of the chimeric mice. Differences between the HBV infections with the various genotype/subgenotypes were observed. Subgenotype A1 with the G1862T mutation showed the earliest detection and therefore highest levels of cccDNA as well as the highest HBV DNA levels when compared to the other strains. The highest HBV DNA levels were recorded for the subgenotype A1 G1862T infected transgenic mouse followed by genotype D, subgenotype A2 and the lowest levels observed in the subgenotype A1 wild-type infected transgenic mouse. HBsAg was also only detected in the livers of mice infected with subgenotype A1 with the G1862T mutation. HBcAg staining in the chimeric liver was positive when the mice were infected with genotype D, which concurs with previous observations that genotype D is characterised by high HBcAg expression. Subgenotype A1 with the 1862 mutant showed the highest levels of apoptosis as a result of the abnormal precore precursor protein accumulation shown to be associated with this 1862 missense mutation. Thus different genotypes and subgenotypes as well as variations within genotypes can influence HBV infection. Moreover, the results of these experiments in the immunocompromised chimeric mice, grafted with liver cells from a single donor, suggests that even when host and environmental risk factors are controlled for, the subgenotype or genotype can influence the course of infection. The limitations of the uPA-SCID transgenic mouse model include the lack of an immune system and the short life-span of the animal; therefore a population based study was carried out to investigate the influence of host factors on HBV infection in various disease groups. The study cohort comprised 635 serum samples from South Africa, China and Japan. Of these samples, 564 were HBsAg-positive and the remaining 71 HBsAg-negative, HBV DNA negative controls. The study cohort included asymptomatic carriers; chronically infected HBV patients as well as patients with HBV associated HCC. Possible associations were determined between HBV genotype, HBV viral load, apoptosis levels, disease group and the age and gender of the patient where available. Apoptosis levels were quantified by the measurement of cleaved cytokeratin 18 (M30) in serum. Patients infected with genotype C or subgenotype A1 were shown to possess a higher odds ratios of developing HCC compared to subgenotype B2 or genotype D, respectively. Significantly higher HBV viral loads were observed in genotype C compared to subgenotype B2. Among the Asian cohort, it was also shown that the male gender was positively associated with high viral loads in HCC patients. Moreover, a positive association between higher HBV viral load levels and HCC in the South African cohort was observed. Male gender, older age, HBV viral load, subgenotype A1 and the presence of the G1862T mutation were shown to be positively and significantly associated with higher levels of apoptosis. In this study it was discovered that the levels of cleaved cytokeratin 18 could potentially be used as a biomarker for the severity of HBV infection because a significant difference was observed with the apoptosis levels between the asymptomatic and HCC patient disease groups. We conclude that even when the influence of host and environmental factors is controlled for, as is the case in the chimeric mouse model, the HBV genotype can affect the progression of infection. Moreover, it was shown in the population based study that the effect of HBV genotype on the outcome of HBV infection can be influenced by host factors. The subgenotype A1 G1862T mutation was shown in both studies to affect both HBV infection and apoptosis. This suggests that HBV variants should be investigated to ascertain their potential impact on the course of HBV infection as it may differ from the wild-type. Apoptosis was shown to be associated with HBV infection in both studies and could possibly be an ideal marker of the progression of HBV infection. These findings are important in helping us to understand factors influencing the course of HBV infection. We have therefore shown in both the studies that differences do exist between the South African subgenotype A1 and genotype D, and that these differences should be taken into consideration for the future evaluation of HBV infection and treatment of South African HBV infected patients. Moreover, cleaved cytokeratin 18 may provide an ideal surrogate marker for HBV disease progression and monitoring.

No description available.

Dans le royaume végétal, les plantes doivent faire face à de nombreux facteurs de stress environnementaux abiotiques et biotiques. La diversité des agents pathogènes conduit encore aujourd’hui à des pertes économiques importantes en impactant le rendement et la qualité des fruits des plantes cultivées. Si pour certains agents pathogènes les traitements chimiques sont adaptés, ils restent inopérants contre les virus menant à l’utilisation de moyen empirique fastidieux à mettre en place et souvent onéreux. La tomate est une plante d’intérêt agricole majeur dont l’un des challenges les plus importants consiste à lutter contre les maladies et ravageurs pouvant provoquer de graves dégâts industriels. Parmi ces fléaux, le PepMV fait partie d’un groupe de virus émergeants dont les connaissances générales sont encore très limitées et méritent d’être approfondies compte tenu de son impact agronomique important. Puisqu’aucune résistance naturelle n’a encore été identifiée chez la tomate vis-à-vis du PepMV, il devient urgent de développer des moyens de lutte efficace afin d’améliorer la production maraichère. Les études développées lors de ce doctorat se sont donc concentrées sur les nouvelles stratégies d’amélioration des plantes en identifiant dans un premier temps des mécanismes d’interaction du pathosystème PepMV/tomate par la mise en place d’une stratégie sans a priori en levure. Cette étape a permis de déterminer des gènes candidats de tomate dont certains potentiellement indispensables au cycle viral du PepMV. Des cribles par la méthode TILLING ont été réalisés avec les facteurs hôtes de sensibilité afin d’acquérir, si possible, des plantes possédant une résistance durable, à large spectre vis-à-vis de nombreux virus ou agents pathogènes et commercialisables en Europe car non OGM. D’autres candidats ne pouvant pas être exploités par une perte de fonction, qui conduirait à une sensibilité accrue des plantes suite à l’infection du PepMV, ont alors fait l’objet d’une étude scientifique plus fondamentale. Ces études ont permis de souligner le mécanisme et le rôle de résistance de plusieurs candidats dans le pathosystème PepMV/tomate. / Within the plant kingdom, crops struggle with numerous abiotic and biotic stresses. The fabulous diversity of pathogens still leads to dramatic economic losses by impairing yield and fruit quality. Chemical treatments are available and effective against some pathogens but unfortunately drive to critic environmental damage and remain useless to face viruses. Nowadays, best ways to eradicate or at least to restrain viruses are empirical and expensive. Tomato is one of the most cultivated plants among vegetable crops in agriculture and the major challenge consists in a intense battle against serious damages on industrial benefits caused by pathogens diseases. The emergent virus PepMV is a worldwide scourge and the limited knowledge deserves to be explored. Unfortunately, natural resistance in tomato has not been identified yet and consequently, it appears as an emergency to develop effective strategies in order to improve tomato production. At first, this study has been based on new crops improvement strategies and focused on identifying mechanisms involved in the PepMV/tomato pathosystem by a yeast two hybrid strategy. This approach leads to the determination of tomato candidate genes potentially essential for the viral cycle of PepMV. TILLING screening method has been used for host susceptible factors to possibly acquire marketable non OGM and sustainable resistant plants to a wide range of pathogens. To go one step further with other candidates for which a loss of function strateg might lead to amplified plant susceptibility following by PepMV infection, fundamental investigation has been pursued. Mechanisms and implications of a few candidates in the PepMV/tomato pathosystem have been unraveled.

Vita. / Digitized by Kansas Correctional Industries

No description available.

No description available.

No description available.