Development of a Low-Cost Biosensor for Tuberculosis Diagnosis

Abstract

Tuberculosis is a deadly disease that is a major public health issue, especially in low-resource communities. However, there are shortcomings associated with current diagnostic tools that are hindering the eradication of this disease. The overarching goal of this research was to create a diagnostic device for tuberculosis that is inexpensive and easy to use while still exhibiting excellent diagnostic sensitivity.

This thesis will describe the multifaceted approach that was taken to develop such a device. To begin, the rationale for choosing tuberculosis as a biosensing target will be provided, and the shortcomings of current tuberculosis diagnostic tools will be summarized. This information is included to provide context for the experimental work that was conducted, much of which focused on the construction and testing of a paper-based lateral flow assay for the detection of the tuberculosis antigen lipoarabinomannan. The assay was constructed using an anti-lipoarabinomannan DNA aptamer sequence that was previously identified in the literature, and various signal generation methods, including aptamer-labelled gold nanoparticles and the catalysis of chromogenic reactions by aptamerlabelled enzymes, were employed. Aptamers were chosen over antibodies in this research to increase the stability and reduce the cost of the lateral flow assay. Although the assay constructed in this research was ultimately unable to successfully detect lipoarabinomannan, this thesis will describe the many insights that were gained regarding the challenges of developing such a sensor and suggest potential solutions to these challenges.

Additional experimental work described in this thesis focused on the testing of horse spleen ferritin loaded with synthetic ferrihydrite cores as a potential replacement for peroxidase enzymes that are commonly used in biosensing. To facilitate the implementation of these catalysts into biosensors such as lateral flow assays, the ferritin was also modified with biotin groups, and work was undertaken to formulate a stable aqueous formulation of a peroxidase substrate that was compatible with the ferritin. Catalysis studies showed the ferritin could successfully catalyze the same reactions as peroxidase enzymes. With some additional optimization of the substrate formulation, horse spleen ferritin holds great promise as a low-cost, highly stable alternative to these ubiquitous enzymes.

In parallel with the experimental work that took place in the laboratory, in silico experiments were also conducted to analyze the aptamer that was utilized in this research. Docking and molecular dynamics studies involving the aptamer and a fragment of lipoarabinomannan revealed a potential binding site on the aptamer. However, inconsistencies between the results of these simulations and experimental work reported in the literature highlighted the shortcomings of the computational models of the aptamer and antigen that were generated in this research. Further work is required to produce more realistic simulations. Finally, a novel, easily multiplexable sensor architecture is proposed in this thesis, and the computational modelling that was conducted to construct this sensor in silico is described. The computational modelling allowed for optimization of the sensor design, and it is hoped that further computational studies will enable the eventual in vitro implementation of this sensor to create low-cost, highly sensitive diagnostic tools for diseases such as tuberculosis.