In the foregoing project, an investigation was made into the relative KIR gene frequencies of three South African cohorts. Playing an important part in innate immunity, KIR fill a vital gap between viral onsets and cell mediated humeral immunity. Being able to sense when cells are abnormal, NK cells possess the ability to destroy cells which show altered HLA molecules during KIR/HLA interaction. Ethnic cohorts that were investigated included African black, mixed ancestry and the Caucasian populations. From these individuals DNA material was extracted using a âsalting outâ method before SSP-PCR genotyping. Seventeen primer pairs were used in the identification of individual KIR genes. PCR products were electrophoresed against a molecular weight marker in order to verify the correct fragment size. Products were viewed on a UV light where observations were noted, and indicated as present or absent. Data was recorded onto a spreadsheet indicating the absence or presence of each particular gene. Tabulated results were used in the construction of graphs as well as Ï2 calculations. These graphs were used in the critical analysis of linkage disequilibrium as well as comparative analysis between the ethnic cohorts. Findings indicate that all framework genes are present in all cohorts. The Black African and mixed ancestry cohorts have not been genotyped for the KIR genes before. Investigation within non-framework genes revealed the identification of several new haplotypes, with the majority observed within the mixed ancestry cohort. Positive linkage disequilibrium was detected between 2DL2-2DS2 and 2DL5B-2DS5 for both the black African and Caucasian cohorts while 2DL1-2DL2 and 2DL5B- 2DS5 linkages were found in the mixed ancestry population. No negative linkages were observed for any of the three cohorts.

Chronic myelogenous leukaemia (CML) is a malignant clonal disorder that results in the uncontrolled production of white blood cells. This disease is a result of a reciprocal translocation of the long arms of chromosome 9 and 22 resulting in a shortened chromosome 22, harbouring the BCR-ABL fusion gene, known as the Philadelphia chromosome. The BCR-ABL oncogene encodes for a constitutively activated tyrosine kinase that interferes with normal cell differentiation and apoptosis. CML can be effectively treated with tyrosine kinase inhibitors, such as imatinib mesylate (GleevecÃ). However, some CML patients experience adverse drugs reactions (ADRs) to imatinib and cannot be treated at the recommended dose. There is a concern that lowering the dose of imatinib to reduce the side effects can result in the development of resistant cancer cells, and thus a cessation in treatment is rather recommended. Imatinib is metabolized by the cytochrome P450 enzyme, CYP3A4. However, if the ADRs were a result of decreased metabolic effect of CYP3A4, it would be possible to reduce the dose of imatinib without effecting efficacy. It is hypothesised that single nucleotide polymorphisms (SNPs) can alter the catalytic activity of the CYP3A4 enzyme. Thus a decrease in metabolic rate can result in ADRs due an increased exposure to the drug. Therefore, the aim of this study was to determine whether SNPs in the CYP3A4 gene are associated with ADRs from imatinib treatment. In this study, the DNA sequence of the CYP3A4 gene from 25 CML patients treated with imatinib were compared to a reference DNA sequence obtained from Genbank. The SNPs identified during this study was statistically analysed, and their association with the presence of ADRs was determined using the online statistics package, SNPator. A total of six SNPs were detected, I193I, T15871G, CYP3A4*1G, C23187T, I369V and G73239A. Of these, I369V and G73239A are novel and not described previously in literature. It was found that I369V resulted in an amino acid change, involving a substitution of isoleucine with valine. The remaining SNPs identified in this study were located in intron regions, with the exception of I193I which is a synonymous SNP. There is little information available on the frequency of SNPs located in introns, since these SNPs are generally regarded to have no impact on the expression or activity of a protein. However, in this study an SNP located in intron 10 was significantly associated with the presence ADRs. Current hypothesises suggest that intron SNPs could affect the expression levels of a protein by influencing the splicing efficiency of mRNA and subsequently translation efficacy. Future research needs to elaborate on the role of CYP3A4*1G on CYP3A4 expression as well as on the prevalence of other alleles identified in this study in South African populations.

Globally, South Africa is the eighth largest producer of GM crops and also imports GM food. In addition to the promise of increased agricultural production, the introduction of GM crops is also having an impact on society in terms of consumer acceptance and trade. As a result, most countries manage GMOs in terms of development, use and application as well as require mandatory GM labelling for consumer preference. With an increase in GM developments, monitoring the food chain in terms of GM labelling and unapproved GM events will continue to pose a regulatory challenge. The aims of this thesis were the following: 1. To determine the uptake of GM food into the food chain; 2. To study the application of voluntary GM labelling; 3. To investigate the impact of mandatory GM labelling; and 4. To establish a monitoring system to detect illegal GMOs in South Africa. Until 2005 it was assumed that there were only low levels of GM crop in the food chain, based on production volumes. However, results from this thesis have shown that 76% of food products tested positive for the presence of GM in 2005. There was also no consideration of mandatory GM labelling as it was thought that voluntary GM labelling was successfully being applied in South Africa. Despite this, 31% of products labelled to indicate an absence of GM, such as âGMO freeâ, ânon-GMâ and âorganicâ, contained genetic modification above 1%, and 20% of these contained more than 5% genetic modification. These results demonstrated the extent of GM in the food chain in South Africa and highlighted the fact that voluntary GM labelling does not protect consumers against misleading claims. In 2008, the Consumer Protection Act mandated the labelling of GM in food products and ingredients. However, there was a lot of uncertainty as to how this would impact the food industry. The subsequent research on the impact of mandatory GM labelling in South Africa determined that 67% of maize and 54% of soybean products will have to be labelled for GM content. In addition to this, GM was also detected in 50% of products labelled to indicate an absence of GM. Furthermore, results indicated that the use of either a 1% or 5% threshold does not make a considerable difference in terms of the number of products implicated. The use of the term âmay contain genetic modificationâ as suggested by draft regulations to the Consumer Protection Act may provide a cost effective manner in which GM labelling can be applied in a developing country similar to South Africa, as it would reduce costs in terms of GM detection. The draft regulations for the Consumer Protection Act also make provision to indicate the absence of GM below a threshold that does not included terminology such as âGMO freeâ or ânon-GMâ. Furthermore, the draft regulations do not require third party verification and compliance will mainly be self-regulating. The implication of this is that consumers or consumer groups will become responsible for policing the application of GM labelling in South Africa. Finally, this thesis presents a GM monitoring scheme for unapproved GMOs, that have not been proven safe for human health and/or the environment. The scheme has the advantage of being cost effective and can be applied to the regulatory situation in any country, taking approved GM events into consideration. The scheme was applied to off-the-shelf food products in South Africa to determine the presence of illegal GMOs. Even though no unapproved GM events were detected, a potential illegal import of GM soybean event A2704-12 was found. It was also found that an approved GM soybean event was comingled with rice and wheat products, although not indicated in the ingredients. The research emanating from this thesis has contributed to inform discussions that have resulted in the inclusion of mandatory GM labelling in the Consumer Protection Act 68 of 2008. It is hoped that the research on the application of mandatory GM labelling and the monitoring for unapproved GM events in the food chain will have a similar impact on the regulatory system in South Africa.

CML is a cancer of the white blood cells and it effects on average one individual in every 100,000. Since it was first described in 1845 by John Hughes Bennett and the subsequent discovery of the Philadelphia chromosome by Nowell and Hungerford in 1960, this hematopoietic malignancy has received much attention in terms of scientific study. Elucidating the pathogenic pathway has lead to the development of targeted therapy. In 2001 imatinib mesylate was introduced as first line therapy for CML. The success of imatinb was illustrated during the IRIS trial by Real-time quantification of BCR-ABL mRNA. BCR-ABL expression levels are correlated to disease stage and progression. BCR-ABL mRNA quantification is therefore the most accurate and sensitive prognostic marker to monitor CML patients. Hence, Real-time PCR for BCRABL has been introduced in many international laboratories to allow for accurate and reliable monitoring to improve and manage patient treatment. Standardization became problematic due to the ease of method development and robustness for Real-time quantification of BCR-ABL mRNA by different laboratories. As a result a plethora of methods for Real-time quantification of BCR-ABL mRNA have been published. This is especially problematic for laboratories with limited means undertaking to develop and implement such a method. Since there are no standardized guidelines, in-house development is required. Furthermore, availability of commercial copy number standards for control and target genes makes it difficult to implement any one method from the literature especially since there is criticism for the genes where standards are commercially available. From a thorough analysis of the literature, problem areas considering RNA extraction, the choice of priming for cDNA synthesis, primers and probes for Real-time PCR as well as a specific control gene together with copy number standards and reference material were clearly defined. Based on this information, best laboratory practice regarding common methodology from literature was established. Only recently through an initiative known as Europe Against Cancer (EAC) has there been a concerted effort to facilitate regional standardization of Real-time quantification of BCR-ABL mRNA. During this study a modified EAC method for Real-time quantification of BCRABL mRNA was developed and validated with the emphasis to improve reproducibility. Instead of ABL or BCR, GUS was used as control gene based on recommendations from literature. Based on statistical analysis it was concluded that the modifications did not bias the percentage BCR-ABL result. It cannot be emphasised enough that standardization for Real-time monitoring of BCR-ABL is most crucial as it will ultimately facilitate molecular laboratories to develop this diagnostic with much greater ease. In order for standardization to be realized, copy number standards as well as reference material for quality control purposes needs to become more readily available. In addition to that, specific guidelines for assay criteria such as appropriate Ct values and analysis of data must also be developed. By streamlining Real-time quantification of BCR-ABL the treatment and monitoring of CML patients can be improved on a global scale.

Tissue factor is a transmembrane glycoprotein that functions as the primary initiator of coagulation in response to mechanical or chemical damage. Due to its key position within the coagulation cascade it also plays an important role in the pathology of thrombosis and thrombotic complications associated with cardiovascular disease as well as in non-thrombotic disorders and diseases such as obesity, diabetes, cancer and HIV-AIDS. Recognising the potential in tissue factor inhibition as a novel approach to antithrombotic therapy, our laboratory utilized phage display technology in a previous study, in order to identify a 26 kDa single chain antibody fragment which functionally inhibits human tissue factor. In the current study, the tissue factor inhibitory single chain variable fragment (TFIscFv) was expressed by means of the pIT2 plasmid vector by Escherichia coli HB2151. This expression system was utilised in an up-scale setting in an attempt to improve the TFI-scFv yield. Although functional TFI-scFv was successfully purified from the culture media by means of Protein A affinity chromatography, the process was hampered by large sample volumes, low levels of expression as well as the high cost involved in Protein A purification. Due to an initial focus on improving TFI-scFv yield through the processing of larger sample volumes rather than the improvement of the expression system, immobilised nickel affinity chromatography was investigated as a more cost effective alternative to Protein A affinity chromatography. It was found that the original expression system was incompatible with immobilised nickel affinity chromatography as the protein was expressed into the culture media. The culture media contained nickel chelating elements that stripped the nickel from the column and consequently prevented TFIscFv purification. Subsequently the TFI-scFv gene was isolated, cloned into an over-expression system and modified to redirect the expression to the bacterial cytoplasm. Although TFI-scFv was successfully redirected to the bacterial cytoplasm and purified by means of nickel affinity chromatography, it was found that expression was hampered due to the presence of rare codons. The detrimental effect of rare codons on TFI scFv yield was addressed through the modification of the expression host by the coexpression of the pRARE plasmid as well as by the rare codon optimization of the TFI-scFv gene sequence for expression in E. coli. Although the co-expression of the pRARE plasmid only slightly improved TFI-scFv yield, a sufficient amount of TFIscFv was generated for functional testing. The modified TFI-scFv displayed a similar inhibition effect with reference to the original construct. The rare codon optimisation resulted in a substantial increase in TFI-scFv yield but consequently resulted in the loss of solubility and the production of inclusion bodies. Although the loss of TFIscFv solubility is unwanted, the high level of expression achieved provides an ideal platform for the further development and characterization TFI-scFv in animal thrombosis models.

All the cells of blood arise from two lineages, the myeloid and the lymphoid lineage. The various cells of blood perform vital functions in the body. These cell counts are closely regulated by regulatory pathways. Mutations within genes that encode for the proteins involved in these pathways can occur. These mutations can cause uncontrolled proliferation of the cells. Myeloproliferative neoplasms are malignancies where there is an uncontrolled increase in the formation of the myeloid cells. The four classical neoplasms are polycythaemia vera, essential thrombocythaemia, primary myelofibrosis and chronic myeloid leukaemia. A mutation (V617F) in the tyrosine kinase, Janus kinase 2, has been found to be the cause of at least three of the classical MPNs. The mutation lies in the domain of the protein that controls its tyrosine kinase activity. The tyrosine kinase thus is constitutively active and causes proliferation of the myeloid cells. The V617F mutation lies in exon 14 and more recently several mutations have been described in the neighbouring exons encoding for the regulatory domain of the gene. Very few studies have been done on the other exons of the JAK 2 gene. In the study it was attempted to screen 15 MPN patients for mutations in the JAK 2 gene. Two different cell populations (lymphocytes and granulocytes) of each patient were screened. It was found that the cell purity was not sufficient in the study and better separation techniques are required for future studies. Only the granulocytes were used for the remainder of the study. High resolution melting curve analysis was used to screen the patients for mutations, however the data did not correlate with the sequencing results and it was decided to proceed with sequencing of all the samples. Seven of the 25 exons of the JAK 2 gene were successfully sequenced. The remaining exons could not be screened due to time constraints and complications such as multiple amplicon formation. Two previously reported single nucleotide polymorphisms were found in exons 11 and 15 in two patients. The clinical significance thereof is uncertain however, the patient whom had the SNP in exon 15 was negative for the V617F mutation and had a MPN. In exon 14 the V617F mutation was identified and the prevalence thereof correlates to that reported in literature. A novel SNP was found in exon 13 of a PV patient negative for the V617F mutation and the significance thereof is also uncertain. Additionally a novel inverse duplication consisting of at least of exon 13 was also identified. No mutations were identified in exons 10, 12, 16 and 17 of the JAK 2 gene. This was, to our knowledge, the first report in South Africa that found the prevalence of the V617F mutation in MPN patients correlating to the prevalence reported in literature. A novel SNP was identified in exon 13 and further studies are needed on the possible effect thereof. The previously reported SNPs found in exons 11 and 15 might be the cause of the formation of a MPN, however further research is needed. A novel duplication variant was also identified and this might be a possible splice variant. The study showed that the region between exons 10 and 15 in the JAK 2 gene is a mutational hotspot and further studies are needed to elucidate the effect thereof.

The focus of this thesis is the inclusion of stochasticity into mathematical models of assembly with particular interest to the in vitro polymerisation of intermediate filaments, one of three components of the cytoskeleton. From the chemical master equation (CME), two additional models (the reaction rate equations or RREs and the two-moment approximation equations or 2MA equations) are derived. As analysis of the CME is generally intractable, we present the stochastic simulation algorithm (SSA) as a means of reproducing the most probable state of the CME at a given time. The results from the SSA are compared to simulations of both the RREs and the 2MA equations and we find that the three models are in good agreement. Further, the numerical results are compared to mean lengths and length distributions of experimental data which all models are shown to mimic. Mathematical analyses of the RREs demonstrate the conservation of mass in the system, and the unique positive equilibrium is proven to be globally asymptotically stable. Further, the 2MA equations are also shown to have conservation of mass and to possess an analogous equilibrium to the one found in the case of the RREs. In general, this study illustrates how randomness can be incorporated in polymerisation models and highlights the advantages and disadvantages of the different approaches.

Human embryonic stem cells (hESCs) replicate in vitro by the process of self-renewal, whilst maintaining their pluripotency. Understanding the pathways involved in the regulation of this process will assist in developing fully-defined conditions for the robust proliferation of hESCs necessary for therapeutic applications.

When injured, endothelial cells secrete inflammatory cytokines such as interleukin-6 (IL-6), interleukin-8 (IL-8), and tumour necrosis factor-α (TNF-α). These inflammatory cytokines stimulate the endothelial release of ultra large Von Willebrand factor (ULVWF) multimers that bind platelets to form thrombi in small vessels. The interaction between thrombosis and inflammation is not fully elucidated. A disintegrin-like metalloprotease with thrombospondin type I repeats-13 (ADAMTS-13) is freshly released from Weibel-Palade bodies of endothelial cells into the plasma and it cleaves the ultra large and hyperactive VWF multimers into smaller and less active forms. These VWF multimers mediate the initial adhesion of activated platelets, the first step in both inflammation and thrombosis. This process may be affected by the amount of ULVWF released and the processing capacity of ADAMTS-13. Little is known about the initial onset of HIV-associated TTP, a fatal thrombotic disease that is characterised by the absence of ADAMTS-13. The mechanisms underlying the initial onset and/or burst of TTP episodes still remain poorly understood. Interrelated components, such as coagulation factors and inflammatory cytokines, all contribute to the development of TTP, since increased levels of cytokines interleukin-6 and tumour necrosis factor and the coagulation factor, tissue factor is measured in these patients. Therefore, we hypothesised that certain inflammatory cytokines and coagulation factors released during inflammation may stimulate the release of VWF simultaneously while inhibiting the synthesis of ADAMTS-13, which results in an acquired deficiency of plasma ADAMTS-13 and ultimately in a TTP episode. Our aim was to examine the effects of inflammatory cytokines and coagulation factors such as tissue factor and thrombin as well as combinations thereof on the release and cleavage of ULVWF by cultured human umbilical vein endothelial cells (HUVECs). HUVECs were treated with cytokines, IL-6, IL-8, and TNF-α and coagulation factors, tissue factor and thrombin, and their combinations, for 24 hours under static conditions. The cells were then exposed to a shear stress of 2.5 dyne/cm2 to expose the VWF cleaving sites. The VWF, VWF propeptide and ADAMTS-13 secretion were measured by an ELISA technique. ADAMTS-13 content was measured using Western blot technology with densitometry. All treatments and their combinations, excluding IL-6, significantly stimulated the release of VWF and VWF propeptide from HUVECs. The VWF propeptide levels were constantly higher than the major VWF protein levels suggesting that the measurement of the VWF propeptide levels may be a better representation of the amount of VWF secreted from endothelial cells. Tissue factor alone and in combination with inflammatory cytokines increase the amount of VWF release from endothelial cells substantially. This correlates with the situation in thrombotic patients with inflammation where extremely high VWF levels are measured. Densitometric analysis of the Western blots indicated that lower levels of ADAMTS-13 secretion were found with all treatments. These results suggest that inflammatory cytokines such as IL-8 and TNF-α, coagulation factors such as thrombin and tissue factor, as well as combinations thereof, stimulate the release of ULVWF and inhibit the release of ADAMTS-13 in HUVECs, resulting in the accumulation of hyperreactive ULVWF in plasma and on the surface of endothelial cells to induce platelet aggregation and adhesion on the vascular endothelium. Our study may offer a logical explanation of how systemic inflammation and thrombosis might trigger the onset and/or burst of TTP in patients with HIV-associated TTP.

The focus of this thesis is the inclusion of stochasticity into mathematical models of assembly with particular interest to the in vitro polymerisation of intermediate filaments, one of three components of the cytoskeleton. From the chemical master equation (CME), two additional models (the reaction rate equations or RREs and the two-moment approximation equations or 2MA equations) are derived. As analysis of the CME is generally intractable, we present the stochastic simulation algorithm (SSA) as a means of reproducing the most probable state of the CME at a given time. The results from the SSA are compared to simulations of both the RREs and the 2MA equations and we find that the three models are in good agreement. Further, the numerical results are compared to mean lengths and length distributions of experimental data which all models are shown to mimic. Mathematical analyses of the RREs demonstrate the conservation of mass in the system, and the unique positive equilibrium is proven to be globally asymptotically stable. Further, the 2MA equations are also shown to have conservation of mass and to possess an analogous equilibrium to the one found in the case of the RREs. In general, this study illustrates how randomness can be incorporated in polymerisation models and highlights the advantages and disadvantages of the different approaches.