published_or_final_version / abstract / Paediatrics and Adolescent Medicine / Master / Master of Philosophy

published_or_final_version / Surgery / Master / Master of Philosophy

published_or_final_version / Microbiology / Master / Master of Philosophy

published_or_final_version / Medicine / Master / Master of Philosophy

Liver transplantation (LT) is the most effective treatment for hepatitis B virus (HBV) related liver failure and hepatocellular carcinoma (HCC). Nevertheless, HBV and HCC recurrence rate remains high after LT. Previous studies have shown that HBV reactivation is associated with HCC recurrence and poor prognosis after LT. The main objectives of this study are to investigate the risk factors for HBV and HCC recurrence after LT, the efficacy of antiviral drugs to prevent HBV reactivation and the underlying mechanisms contributing to HBV reactivation. Firstly, we investigate the risk factor for HBV and HCC recurrence in 551 HBsAg seropositive LT patients, of whom374 had no tumor and 177 had HCC. All patients received indefinite antiviral treatment after LT. The study showed that pre-LT HBV DNA levels and HCC recurrence were significantly associated with HBV reactivation after LT. Younger age, lower Child-Pugh score, beyond UCSF criteria, higher AST level, salvage LT, older donor, HBsAg seropositive at the last follow-up and HBV reactivation after LT were independent risk factors for HCC recurrence. HCC recurrence alone accounts for poor overall survival. The sequence analysis identified drug-resistant mutants as the main contributors to HBV reactivation. In addition, wild-type (antiviral drug-sensitive) HBV reactivation was identified in patients with HCC recurrence. Secondly, we investigate the efficacy of antiviral drugs monotherapy (Lamivudine or Entecavir) in preventing HBV reactivation. This study showed that patients receiving lamivudine (LAM) experienced significantly greater HBV reactivation and HCC recurrence than those receiving entecavir (ETV). In patients with no tumors, HBV reactivation was found in the LAM groups but not in the ETV groups, due to the appearance of a LAM drug-resistant mutant. In patients with HCC recurrence, HBV reactivation was found in both treatment groups. Wild-type HBV reactivation was identified in 17% (5/29) and 100% (1/1) of HCC patients receiving LAM and ETV respectively. This suggests that, although ETV had higher genetic barriers to HBV drug resistance; it still cannot prevent wild-type HBV reactivation in HCC-recurrent patients. Thirdly, we investigate the expression of HBV markers in HCC and adjacent non-tumor tissues. Origin of circulating HBV was identified using genetic distance analysis of HBV isolated from different compartments (i.e. HCC and adjacent non-tumor tissues). The study showed that, in some HCC cases, the expressions of HBsAg and HBV replicative efficiency are higher in HCC tissues than in adjacent non-tumor tissues. Moreover, through genetic distance analysis, we demonstrated that HBV reactivation could originate from recurrent HCC. These data suggest that HCC supports HBV replication and that HBV is secreted from recurrent HCC. Finally, we demonstrate that the up-regulation of drug-specific ABC-transporters is significantly associated with patients with HCC recurrence. In vitro studies also showed that the up-regulation of ABCG2 contributes to antiviral drug-resistant. Finally, we demonstrate that the up-regulation of drug-specific ABC-transporters is significantly associated with patients with HCC recurrence. In vitro studies also showed that the up-regulation of ABCG2 contributes to antiviral drug-resistant. / published_or_final_version / Surgery / Doctoral / Doctor of Philosophy

No description available.

There are an estimated 350 million people chronically infected with Hepatitis B Virus (HBV), of which approximately 600 000 die each year from HBV complications including cirrhosis and liver cancer. The X protein from HBV (HBx) has been implicated in the progression of chronic HBV to liver cancer and has been reported to manipulate several critical cellular pathways. These include the cell cycle, the tumour suppressor protein p53, protein degradation and signal transduction pathways. The role of these interactions in HBV replication and the viral lifecycle is currently unknown. The lack of animal models and infectable cell lines together with solubility and stability issues related to the HBx protein have made progress difficult. The reliance on approximate cellular and animal models has yielded many discordant studies that have confounded our interpretations of the role of HBx. There have been no novel approaches attempting to express HBx at a quantity and quality sufficient for high resolution X-ray and nuclear magnetic resonance structural determination. Additionally no bioinformatic analyses have been applied to HBx, and thus distinctive features of HBx that may be responsible for these challenges have not been reported. This thesis describes the detailed experimentation to express and purify HBx in a functional, soluble and stable form. The study focussed on Saccharomyces cerevisiae and Semliki Forest Virus (SFV) expression systems, together with the use of a solubility-enhancing Maltose Binding Protein protein tag (MBP). The S. cerevisiae-based pYES2 and YEp and mammalian expression vectors showed production of HBx protein. However HBx that had been expressed using S. cerevisiae and human cells could not be reliably detected in Western blots using antibodies raised against E. coliexpressed HBx. This result was despite the positive visualisation of HBx using the same antibodies and immunofluorescence microscopy. This validated previous reports describing the variable antigenicity of HBx. Furthermore these findings supported the decision to develop eukaryotic-based HBx expression vectors as results suggested structural differences between eukaryote and prokaryote expressed protein. HBx was subsequently detected and purified in a soluble and active form using an MBP tag as well as a SFV expression vector. All of these options provide an excellent point from which further work at optimising HBx expression and structural elucidation can occur. Bioinformatic analysis of HBx suggested the presence of protein disorder and protease sensitive sites within the negative regulatory domain of HBx. Literature descriptions of the molecular promiscuity that protein disorder allows, offers an explanation for the presence of the discordant findings on HBx interactions and functions. It is generally accepted that proteins containing disorder are tightly regulated and thus experimental systems employing overexpression methodologies may encourage cellular toxicity and non-specific interactions through the use of short linear motifs. Evolutionary analysis of HBx sequences revealed that the eight HBV genotypes (A-H) showed concordance regarding synonymous and non-synonymous substitutions at the overlapping and non-overlapping domains of hbx. Substitutions in hbx were most common at positions where a synonymous substitution occurred in the overlapping partner gene. The presence of sites under positive, neutral and negative selection were identified across the length of HBx. The different genotypes showed positive selection indicating selective pressures unique to each, thus offering a contributing explanation for the variable disease severity observed between the subtypes. Overall, this thesis has provided novel methods to express and purify HBx in S. cerevisiae and mammalian cells. These methods, together with an increased understanding of the nature of HBx sequences through bioinformatic analysis, pave the way to conduct both structural studies and biological assays to elucidate the genuine roles of HBx in the HBV lifecycle and its contribution to the progression to liver cancer.

Hepatitis B virus (HBV) is a major global public health burden, with over 350 million people chronically infected. This results in approximately 600,000 liver cancer-related deaths annually. Chronic HBV infections are normally managed with long-term anti-HBV therapeutics, such as reverse transcription inhibitors, which target post-transcriptional viral processes without affecting the cccDNA. Treatment failure however is largely as a result of the stability of this episomal viral DNA. The cccDNA minichromosome serves as a reservoir of HBV DNA and is capable of re-establishing viral replication following withdrawal of treatment. Designer nucleases, like transcription activator-like effector nucleases (TALENs), have recently been used to create double stranded breaks (DSBs) at target-specific endogenous DNA loci. These nucleases are designed as pairs, which upon dimerisation cleave double-stranded DNA. Subsequent activation of the cellular non-homologous end-joining (NHEJ) pathway often results in targeted mutagenesis at the DSB site. As TALENs may be designed to bind to any DNA sequence, they are commonly used as genetic engineering agents. Inactivation of HBV cccDNA, using these engineered TALENs, presents a unique approach to disabling viral replication permanently. To investigate this, a panel of TALENs targeting the core (C), surface (S) and two different polymerase (P1 and P2) regions of HBV cccDNA were generated using a Golden gate modular assembly approach. TALENs were initially tested in two liver-derived cell lines. Firstly as transient co-transfections in Huh7 cells using a HBV replication competent plasmid, followed by long term investigations in HepG2.2.15 cells which model HBV replication in vitro. Inactivation of HBV was determined by measuring markers of viral replication, whilst TALEN-mediated targeted disruption was verified by T7 endonuclease 1 (T7E1) or CELI endonuclease assays and sequencing. In vitro, the S TALEN inhibited HBsAg secretion by 80% in Huh7 cells and 60% in HepG2.2.15 cells. Furthermore, S TALEN-mediated targeted disruption occurred in 35-47% of cccDNA copies, whilst the C TALEN resulted in 11% targeted disruption of cccDNA in without inhibition of HBsAg expression. The P2 TALEN showed no anti-HBV efficacy, however the P1 TALEN inhibited HBsAg expression by up to 60% without any evidence of site-directed cleavage. As this TALEN binding site spans the HBV Enhancer I sequence, knock-down of HBsAg expression is most likely to occur as a result of transient transcriptional repression. To confirm whether permanent repression of HBV transcription could be achieved, a KRAB-based transcription activator-like repressor (rTALE) targeting the HBV pre-S2 promoter was generated. Using an in vitro reporter gene assay, the pre-S2 rTALE inhibited luciferase expression by up to 90%. However this was only achieved using high molar concentrations of the repressor, suggesting multiple rTALEs may improve HBV transcriptional repression. As the S and C TALENs displayed significant anti-HBV efficiency in vitro, they were tested in a murine hydrodynamic injection model of HBV replication. In vivo, the S TALEN inhibited HBsAg secretion by 95% and induced disruption in 77–87% of HBV DNA targets. In addition the C TALEN inhibited HBcAg expression and induced disruption in 78-93% of HBV DNA targets. Additionally, serological analysis showed a reduction in circulating virions and no apparent liver toxicity, as determined by real-time PCR (qPCR) and aspartate transaminase (AST)/ alanine aminotransferase (ALT) liver function tests respectively. Deep sequencing at the S and C TALEN binding sites showed targeted mutagenesis of HBV DNA in samples extracted from murine hepatocytes transfected with TALENs, however wild-type sequences were exclusively detected in mice that had not been treated with anti-HBV TALENs. Furthermore, frameshift deletions were predominantly detected indicating major disruptions in the HBV surface and core sequences. These results indicate that TALENs designed to disable and silence HBV cccDNA are effective both in vitro and in vivo and as such provide a promising therapeutic approach to treat this serious infection.

PhD, Faculty of Health Sciences, University of the Witwatersrand, 2009 / Chronic infection with the hepatitis B virus (HBV) is a major risk factor for cirrhosis and hepatocellular carcinoma, which is the sixth most common cancer worldwide. Available treatment for chronic HBV infection has limited efficacy in preventing associated complications. The compact and multifunctional nature of the viral genome limits its mutability making HBV an ideal candidate for therapy based on nucleic acid hybridisation. The potent and specific gene silencing that can be achieved with RNA interference (RNAi) has fueled interest in exploiting this pathway as a therapeutic modality. Synthetic and expressed RNA sequences have been used to activate RNAi. These engineered sequences mimic natural substrates of the RNAi pathway, which allows them to enter and reprogramme the pathway to effect silencing of intended targets. Tradionally expressed RNAi activators have been transcribed as short hairpin RNA (shRNA) sequences from RNA polymerase III (Pol III) promoters. These shRNA mimic precursor microRNA (pre-miRNA) and consequently enter the RNAi pathway at a relatively late stage. Overexpression of shRNA sequences from Pol III promoters, specifically the U6 promoter, has been associated with toxic side effects and has raised concerns about the use of expressed RNAi activators. Another concern of developing therapeutic RNAi expression cassettes is the emergence of HBV mutants that are resistant to silencing by a single expressed RNAi effecter. These points have highlighted the need for the development expressed RNAi activators that are effective at low concentrations and capable of combinatorial silencing. To address these issues the aim of this study was to assess the feasibility of anti HBV effecter sequences that mimic an early substrate (viz. primary miRNA or pri-miRNA) of the RNAi pathway. Pri-miRNA expression is typically under the transcriptional control of Pol II promoters. Consequently RNAi activators that Abstract - xi - mimic pri-miRNA, so-called pri-miR shuttles, may be expressed from Pol II promoters. Initially a panel of shRNA expression cassettes driven by a Pol III promoter was constructed and silencing of HBV replication assessed. Pri-miR shuttles were then designed by incorporating guide sequences of the most effective anti HBV U6 shRNA into naturally occurring pri-miR-122 and pri-miR-31. Potent inhibition of viral replication was observed with both Pol III and Pol II-driven pri-miR shuttle expression cassettes in vitro and in vivo. Subsequently liver-specific pri-miR-122 and multimeric pri-miR-31 shuttle expression cassettes were created. Pri-miR-122 shuttle sequences expressed from the alpha-1 antitrypsin promoter and HBV basic core promoter exhibited the best liver-specific silencing. Polycistronic pri-miR-31 shuttle sequences were shown to produce multiple RNAi activators capable of silencing multiple target sequences. Silencing by the pri-miR shuttle sequences was independent of toxic effects that arise from induction of the interferon response or saturation of the endogenous miRNA pathway. Pri-miR shuttles clearly represent an improved option for the use of expressed shRNA and brings therapeutic RNAi technology a step closer to clinical application.

MSc (Med), Faculty of Health Sciences, University of the Witwatersrand / The hepatitis B virus remains a major public health problem even after decades of its discovery. Horizontal transmission during early childhood is the predominant mode of transmission in highly endemic regions such as sub-Saharan Africa. Infection exhibits a wide spectrum of clinical manifestations, from an asymptomatic stage to severe liver disease which may result in hepatocellular carcinoma (HCC). The HBV X protein (HBx) has been implicated in carcinogenesis, which often has a poor prognosis, consequently the use of highly specific monoclonal antibodies (mAbs) directed against HBx in an enzyme-linked immunosorbent assay (ELISA) could lead to early identification of HBV carriers at risk of developing liver cancer. A variety of mixed hybridoma cell cultures secreting anti-HBx antibodies were cloned and sub-cloned by “limiting dilution”. Clonal supernatants were assessed for anti-HBx antibody production by Indirect ELISA and Western/Immunoblotting. Monoclonal antibodies were then characterized according to their relative binding affinity (Indirect ELISA) and relative epitope specificity (Competitive ELISA). One of our monoclonal antibodies was found to bind to the same epitope on HBx as the commercial anti-HBx antibody and with the same high affinity. In the developed Sandwich ELISA, our monoclonal antibody proved effective as the „detecting‟ antibody when the commercial anti-HBx antibody was deployed as the „capture‟ antibody. This Sandwich ELISA will be further developed in our laboratory with the object of applying it to patient sera.