Restriction lift date: 2021-05-05
Bacterial inhabitants of tumours: methods for exploration and exploitation
Walker, Sidney P.
University College Cork
The presence of bacteria in patient tumours of various types has been reported by numerous groups since 2014, but the findings of these and many similar studies remain contentious. Tumour samples provide many obstacles to carrying out robust and reliable microbial surveys, primarily the anticipated low biomass of these samples, which leaves them vulnerable to environmental contamination. While the debate over the presence or absence of bacterial communities in these tumours continues, it impedes any research into how such bacteria might be utilised in medicine. Larger sample numbers are required, from diverse tumour tissues within the human body, and these must be analysed in a reproducible and accurate manner to allow for the drawing of definitive conclusions in this debate. To accommodate this requirement, the primary methodological aspects of this thesis were: i) The assembly and validation of a contamination control pipeline using recent advances is bioinformatic contamination control detection. ii) The development and validation of a bacterial DNA extraction protocol for formalin fixed, paraffin embedded (FFPE) samples, with accompanying FFPE biological standards for use as controls. A key aim of this thesis was to increase the accuracy and reproducibility of research into clinical tissue biopsies by eliminating the role of contamination, and to expand the applicability of FFPE tissues which represent an invaluable resource of samples for analysis. In this thesis, ecological surveys of a variety of related environments were conducted with the common goal of characterising a detectable bacterial community and identify potential bacterial biomarkers unique to these host environments. Regardless of whether or not a consistently present and detectable tumour microbiome exists, tumours possess several phenotypes making them hospitable environments for bacteria to colonise. Where the unique physiology of tumours is seen as an obstacle for traditional cancer treatments, they represent an opportunity for bacterial-mediated solutions. Therefore, findings from sequencing-based research of host environments have potential to be translated into the use of administered bacteria as delivery vehicles to locally produce biomolecules. There are two considerations in this context, requiring two very different applications of bioinformatics. i) The first is to identify which bacteria colonize the desired niche in body; this can be a ‘foreign’ body such as a tumour (Chapter 3 and 4), or parasite (Chapter 5), or a distal niche such as the gut. ii) The second, often under-considered parameter, relates to what these bacteria produce. Synthetic biology presents enormous scope for sophisticated medical therapy mediated by novel synthetic proteins. However, the task of getting a bacterial cell to successfully express and secrete a stable protein that it does not produce naturally is far from trivial, and is becoming a key aspect of the synthetic biology field. To facilitate this synthetic protein aspect, a novel strategy for the performance prediction of designed protein constructs was developed. This tool was able to predict the overall performance of a protein construct in vitro using only in silico derived data. This thesis aimed to develop novel strategies for the analysis of bacterial communities within tumours by i) increasing the sample sizes available to future projects by enabling the use of FFPE samples and ii) improving the accuracy of analysis by designing bioinformatics analysis pipelines appropriate for these samples. This enabled further research concerned with finding differentially present taxa between the tumour and surrounding environment, which have the potential for use as therapeutic vectors. As the key aim is to establish the presence of potential bacterial therapeutic vectors rather than to establish the role these bacteria play in tumorigenesis, this approach is easily translatable to other foreign bodies, and could therefore be validated in a parasitic nematode model.
Microbiome , Big data , Tumour
Walker, S. P. 2020. Bacterial inhabitants of tumours: methods for exploration and exploitation. PhD Thesis, University College Cork.