Advancing super-resolution imaging: innovations for enhanced single-molecule localisation microscopy

Thesis

Single-molecule localisation microscopy is a super-resolution microscopy method capable of attaining resolutions beyond the diffraction limit of light. Its application to the life sciences has led to unique insights into the functions and organisation of biological molecules in cells.

Despite its benefits, its routine usage faces numerous challenges. This thesis examines several important aspects of these, namely: the detrimental effects of sample drift, insufficient resolution to resolve many different types of protein complexes, the perturbation of native cellular states through fluorescent labelling, and the low signal-to-noise ratio that obfuscates single-molecule detection.

To address sample drift, a novel microscope design is presented and its sample drift is quantified, demonstrating negligible mechanical drift and obviating the need for drift correction methods. This reduces experimental complexity and improves the accessibility and reliability of single-molecule localisation microscopy.

Advances in enhancing resolution are presented through implementing mirror-enhanced fluorescence in combination with a denoising algorithm. Using both technologies conferred an approximately 15 nm enhancement in resolution, allowing a wider range of biological molecules to be visualised.

Different methods to label proteins with nanobodies are compared, thereby outlining the detrimental effects of popular fluorescent labelling methods, as well as highlighting that different experimental outcomes can be achieved by altering the location of the amino acid used to conjugate nanobodies to dyes.

Finally, an evaluation and characterisation of a novel deep learning denoising algorithm is presented, demonstrating that it improves signal-to-noise ratio and resolution for certain types of noise. This evaluation aids other researchers aiming to use denoising algorithms, and underlines the strengths and limits of a state-of-the-art denoising methodology.

Overall, advances to the instrumentation, fluorescent labelling, and software that underpin single-molecule localisation microscopy are presented in this thesis. These advances are expected to aid further methods development, as well as increase the reproducibility and accessibility of applying single-molecule localisation microscopy to investigate biological systems.

Authors: Tang, M

• DOI: 10.5287/ora-8njbm1mqr
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: microscopy; single-molecule localisation microscopy; super-resolution microscopy; biological imaging; optics
• Subjects: Life Sciences; Image analysis; Biophysics; Optics