Single Molecule Counting using STORM/PALM

Abstract

Single-molecule counting using STORM/PALMDuring the last decade, several super-resolution microscopy techniques have been developed that enable the circumvention or even the breaking of Abbe's diffraction limit. One approach of achieving super-resolution is to control the emission state of individual fluorophores in time and the concomitant ability to localize active emitters with up to nanometer accuracy. Two of the most commonly implemented super-resolution techniques are stochastic optical reconstruction microscopy (STORM) and photo-activated localization microscopy (PALM). Each technique uses the localization of individual emitters to improve resolution as has been demonstrated with numerous intracellular structures. Beyond that, these techniques open the possibility of counting on a single molecule level. However, the ability to count is fundamentally limited by the fluorophores used for these techniques which blink many times before photobleaching. A typical STORM dye like Cy5 can cycle between tens and hundreds of times between a dark and a fluorescent state before photobleaching, thus leading to a set of localizations scattered around the true position. Whereas this property does not affect the reconstruction of images of extended objects like intracellular filaments or compartments, it severely influences the interpretation and quantification of objects for which the exact stoichiometry can be important like for membrane protein aggregates. To resolve this issue, we have implemented an algorithm that uses the spatial and temporal information of fluorophore localizations from STORM/PALM experiments to obtain a quantitative picture of the underlying molecule distribution. Our algorithm reliably operates on artificial data as well as on experimental data from biological constructs with a well-defined number of attached fluorophores.