Slack, Sarah Louise; (2022) A high-throughput approach to vaccine bioprocess development. Doctoral thesis (Eng.D), UCL (University College London).

Preview

Text SarahLouiseSlackThesisFinalSubmission.pdf - Accepted Version Download (6MB) | Preview

Abstract

Vaccination is currently the predominant tool in the prevention of infectious disease. Each year, an estimated 2-3 million lives are saved worldwide and infant mortality has been significantly reduced. Despite substantial recent advances, vaccine manufacturing can still be laborious owing to difficulties in development, lengthy clinical trials, and stringent regulations. In light of the SARS-CoV-2 (Covid-19) pandemic, the need for a development platform which can rapidly screen potential candidates and/or a vaccine scaffold capable of adaptability to new disease targets has never been more apparent. To meet this need, the breadth of vaccine types under exploration has rapidly expanded. DNA and RNA vaccines offer the opportunity for rapid manufacture but can be poorly immunogenic, whilst subunit vaccines can require complex processing. Virus-like particles (VLPs) have the potential to address these two factors. Tandem Core VLPs, expressed in the methylotrophic yeast Pichia pastoris, are an exciting alternative to current manufacturing methods. They have excellent potential, both as standalone vaccines for the virus from which they are derived, or as scaffolds for the display of foreign antigens. The hepatitis B core antigen (HBC) can spontaneously self-assemble, forming icosahedral particles that are inherently immunogenic. Tandem Core HBC VLPs have been genetically modified in the major insertion region (MIR) enabling surface display of up two epitopes of interest when assembled. For HBC VLPs to be considered a viable vaccine candidate, their bioprocessing must be optimized. Currently, there are various issues to address including problems with formation, solubility and immunogenicity, which are often clone dependent. In this work, Tandem Core VLPs, consisting of genetically linked HBC monomers carrying different epitopes in the MIR will be used to develop a high-throughput platform and explore the impact of different inserts on VLP production and processing. Influenza will be used as a model pathogen owing to its persistence as a public health threat. The aim of this work is to develop a vaccine platform, defined by Adalja et al., (2019) as “a technology in which the underlying, nearly identical mechanism, device, delivery vector or cell line was employed for [design of] multiple target vaccines”. The equipment and methods developed in this work were considered to enable: (1) thorough investigation of three HBC VLP candidates in an attempt to identify a universal bioprocess, irrespective of surface displayed epitopes; (2) formation of a small-scale high throughput platform which could be implemented for rapid screening of new disease targets or to allow fine-tuning of processes for epitope-dependent optimisation. After initial studies using an ‘empty’ HBC VLP (HBC-K1,K1), the ambr®250 modular was used to investigate upstream bioprocessing of three influenza specific candidates (HBC -HA2,3M2E, -LAH3,K1 and -3M2E,K1), exploring different fermentation induction strategies and to identify epitope related differences. Following this, the most readily soluble candidate (HBC-LAH3,K1) was selected for further upstream optimisation combining ambr® 250 experimentation with statistical Design of Experiments (DoE). An improved process was identified enabling an increase in VLP titre, a 34% increase in biomass compared to the initial condition, and a 6% decrease in process time compared to methanol induction. This process was then applied to the production of the alternative VLP constructs. The improved feeding regime resulted in higher biomass and soluble HBC yield for all three VLPs. Subsequent downstream process studies on the primary recovery of VLP candidates was then necessary to account for the reduced volumes associated with miniaturised fermentation studies, and to bridge the gap between upstream processing and purification. Building on previous work, a high-throughput, small scale cell disruption method was investigated using Adaptive focused acoustics®. A 96-well plate workflow was demonstrated, enabling suitable VLP release and recovery with a ~99.7% reduction in sample volume, in comparison to high pressure homogenisation (HPH). Finally, chromatography screening was undertaken using high-throughput PreDictor® plates to rapidly identify separation conditions for the various vaccine candidates. Studies were conducted to investigate suitable resins and binding/elution conditions and to determine the influence of the physicochemical properties of the displayed epitopes on separation performance. Multiple resins were identified as being suitable for VLP purification, and results were useful to manipulate chromatographic separation (5mL column scale) conditions for the VLPs to achieve improved product yield and purity profiles. Overall, this research suggests that a high-throughput vaccine development platform can be realised through the integration of numerous small-scale single-use equipment, techniques and methodologies. Namely, the use of the ambr®250 bioreactors, AFA® cell disruption in 96-well plates and 96-well PreDictor™ resin plates. Combined with statistical DoE, this platform can be used to rapidly optimise production and purification conditions for novel vaccine technologies such as HBC Tandem Core VLPs. The improved bioprocessing of these constructs paves the way for future vaccine candidates which exploit HBC as a vaccine scaffold. These findings have implications for reducing the time taken to develop vaccine manufacturing processes and prepare for disease outbreaks based on ‘Pathogen X’.

Download activity - last month

Export as CSVJSONXML

Download activity - last 12 months

Export as XMLJSONCSV

Downloads by country - last 12 months

Export as CSVJSONXML

Archive Staff Only

View Item