Quantitative analysis methods for single-molecule localization microscopy: molecular orientation, nanoparticles, and counting

Single-molecule localization microscopy (SMLM) is a transformative imaging technique that enhances the effective resolution by an order of magnitude when compared to conventional widefield microscopy. This improved resolution empowers researchers to investigate molecular complexes at the nanometer scale, offering profound insights into biological structures. SMLM achieves this breakthrough through the strategic labeling of molecules of interest with fluorophores that undergo stochastic transitions between active and inactive states. The resulting blinking fluorophores are typically imaged over thousands of frames, where in each frame only a sparse subset of fluorophores emits light. Consequently, each fluorophore creates an isolated diffraction-limited spot in the image, permitting precise nanometer-level localization through the fitting of a point-spread function (PSF) model.
This thesis delves into various facets of quantitative parameter estimation within the realm of SMLM. To begin with, I demonstrate the capability of in-focus PSF fitting to discern the molecular orientation of fluorophores that are fixed in orientation. This approach is applied to investigate the orientation of fluorophores incorporated into DNA-origami constructs with deterministic position and orientation.
In the subsequent part of this thesis, I introduce a novel method, extending SMLM to the localization of fluorophores situated on spherical nanoparticles (NPs). NPs have gained prominence in diverse applications, including biosensing, drug delivery, and photo-thermal therapy. Their functionality critically depends on the distribution and number of functional groups on their surface. SMLM is, in principle, ideal for examining the surface functionalization of individual NPs in an aqueous environment. However, the presence of the NP distorts the fluorophore’s PSF, leading to a systematic bias in conventional image analyses that employ Gaussian PSFs. To address this challenge, a fully analytical PSF model for emitters in proximity to spherical NPs is developed. The analytical model offers a substantial computational advantage of four orders of magnitude over numerical approaches and is thus feasible to use directly in image analysis. This model is subsequently applied to 2D experimental images obtained through DNA-PAINT on DNA-coated gold NPs, demonstrating the extraction of 3D positions of functional groups with remarkable 5 nm precision. These findings reveal heterogeneous surface coverage of DNA on the NPs.
Orthogonal to counting the functional groups on an NP by resolving them in a super-resolution microscopy image, one can quantify them based on the blinking kinetics. In the last part of this thesis, I present an innovative method for quantifying the number of molecules within an assembly imaged via DNA-PAINT. This approach relies on a combination of statistical measures and time-series analysis, enabling the determination of molecule counts solely from a single intensity time-trace recorded during localization microscopy, without any a priori knowledge or calibration.
Taken together, here I present quantitative approaches for characterizing the orientation, position, and number of fluorophores, which will be widely applicable in the field of nano- and biotechnology.