Quantitative Analytics for Spectroscopic Single-Molecule Localization Microscopy


Single-molecule localization microscopy (SMLM) has significantly stimulated the development of methods to quantitatively visualize and characterize biomolecules in vitro and in situ. SMLM is a class of super-resolution microscopy (SRM) techniques, which exploits the “on-off” switching of individual fluorescent molecules to estimate their location with nanometer precision and offers spatial resolution up to 10 nm. However, the invaluable spectroscopic information of fluorescent labels has previously been overlooked in conventional SMLM. Recently developed spectroscopic SMLM (sSMLM), integrates a dispersive component into the traditional SMLM system to concurrently capture the spatial and spectral information of each single-molecule emission event. Thus far, the emission spectra have been used to expand the multiplexing capabilities of SMLM and develop functional SRM. While traditional SMLM has widely been adopted by the scientific community, the limited availability of software tools to process sSMLM data has hindered the routine use of sSMLM for SRM studies. Additionally, the benefits of using spectroscopic information to improve quantitative SRM have remained unexplored. This dissertation aims to build an image processing platform for sSMLM and develop three quantitative spectroscopic analysis methods for the characterization of nanostructures. First, we present RainbowSTORM, a freely available ImageJ plug-in, which includes functions to calibrate sSMLM systems, process two-dimensional and three-dimensional sSMLM data, and generate pseudo-colored sSMLM image reconstructions. Second, we develop a regression method to evaluate spectroscopic information and reject signals from fluorescent impurities which can lead to molecular misidentification and degraded spatial resolution in SMLM. Using this method, we quantify immobilized nanorulers and improve sample identification of surface-combed DNA fibers. Third, we develop quantitative spectroscopic analysis for cluster extraction (qSPACE), a post-processing method for the sSMLM variant referred to as spectroscopic point accumulation for imaging in nanoscale topography (sPAINT), which captures the fluorescence induced by transient interactions between Nile Red (NR) dye and polymersomes. We use qSPACE to accurately size and count the polymersomes, while rejecting non-specific interactions between NR and the poly-L-lysine surface. Finally, we develop photon-accumulation enhanced reconstruction (PACER) which numerically increases the photon budget of fluorescent labels and estimates the emitter’s spatial location with improved localization precision. Using PACER, we image quantum dots and Alexa Fluor 647 labeled DNA nanostructures with a minimum spacing of 6 nm to uncover individual molecules that would be unresolvable using conventional SMLM.

Alternate Identifier
Date created
Resource type
Rights statement