Recent applications of superresolution microscopy have led to the believe that virtually all membrane proteins are organized in clusters. Single molecule-based superresolution microscopy exploits the possibility to localize single dye molecules with an accuracy far below the optical diffraction limit. The key idea is to stochastically switch the fluorescent probes between a dark and a bright state. Typically, many thousand images are recorded, and high-resolution localization maps are reconstructed from the determined molecular positions. Recently, however, concerns about the existence of nanoclusters have been fueled by the notion that virtually all fluorescent probes show complex blinking behavior including long-lived dark states, which leads to overcounting due to repeated observation of single molecules.
In this project I will develop a novel experimental procedure, combining single molecule localization microscopy and microfluidics to enable the reliable identification of nanoclusters. The concept of this project is based on our recently published method to detect molecular clustering by varying the density of fluorescent labels. To this end, comparative analysis of images recorded at different label-densities allows for discriminating clustered versus non-clustered protein distributions. However, limitations in sensitivity introduced particularly by cell-cell heterogeneities still impair the applicability of the method.
To enable label titration microscopy on single cells, I will design a microfluidic biochip for application on a single molecule microscopy system. This will greatly enhance the sensitivity of the method with respect to the detection of clusters at large backgrounds of non-clustered molecules, and cluster scenarios which are currently difficult to identify. Additional computational improvements of the analysis software will allow for quantitative analysis of protein nanoclustering.
I will validate the method based on three pillars: i) thorough in silico testing with Monte Carlo simulations, based on the characterization of the blinking behavior of commonly used fluorescent labels; ii) as an in vitro test system I will use fluorescently labelled DNA-origami to artificially create nanoclustered fluorophore distributions; iii) for characterization and validation of the new method in a real live scenario, I will analyze proteins with known distributions in the plasma membrane. I will use clathrin-coated pits and LFA-1 as positive controls for molecular clustering at different length scales, and GPI-anchored mGFP as a negative control for uniformly distributed membrane proteins.
Ultimately, I will apply the novel method for the characterization of key signaling proteins at the plasma membrane of Jurkat T-cells. Planned target molecules are the T-cell receptor complex, the co-receptor protein CD4, the signaling kinase Lck and the adaptor protein LAT, which have been frequently reported to show nanoclustering. Given the concerns raised above it seems crucial to revisit the topic with an alternative quantitative approach in order to clarify, which model of the plasma membrane is correct. On the long term, the new method shall become an easy to use technique for the community to address protein clustering in various cellular or tissue environments.
