Local image correlation methods for the characterization of subcellular structure and dynamics by confocal and super-resolution microscopy

This thesis work aspires to present a new concept for the application of correlation techniques to the study of the cellular environment. By exploiting local analysis in combination to a fast fit-free technique (the phasor approach) we provide an exhaustive high-resolution analysis of structural and dynamic properties while maintaining a reasonable computation time. The dissertation will be articulated as follows: In CHAPTER 1 we aim to provide the reader with a description of the techniques that will be exploited during the rest of the dissertation together with the open questions and problematics that our techniques will try to answer to. In CHAPTER 2 we present the local analysis concept and its application to a correlation technique capable of measuring size and concentration (ICS). We will show how we coupled ICS to the phasor approach to create a technique (PLICS) for the assessment of size heterogeneity. PLICS will be demonstrated with simulations as well as with cellular samples and will be applied to the study of endocytic vesicles uptake and to the characterization of other organelles. In CHAPTER 3 the concept is extended to two-colors samples for the determination of local inter-structure distance (PLICCS). We will present a pattern analysis method we developed that exploits this information in order to evaluate the relative distribution of the structures imaged in the two channels, comparing it to a random distribution. This method will be validated with simulations and applied to the study of replication-transcription collisions. Successively, we will show that PLICCS can be converted to a localization algorithm for single particle tracking that will be used for tracking membrane receptors in living neurons. CHAPTER 4 will describe the extension of our local analysis to RICS, a correlation technique capable of measuring the diffusion coefficient of a fluorescent probe. The resulting algorithm (L-RICS) provides high resolution diffusion maps that will be used to characterize the diffusion of a fluorescent probe (GFP) within the nucleus and nucleolus of living cells. We will show that the algorithm can be implemented also in non-linear scanning systems. CHAPTER 5 will conclude the dissertation by introducing advanced correlation methods for the analysis of non-Brownian diffusion and their coupling to super-resolution techniques. In particular, we will present a super-resolution correlation technique (SPLIT) recently developed capable of analyzing the cellular environment and a microcamera-based approach (Airyscan comprehensive correlation analysis) we developed for the parallel implementation, in super-resolution, of several complementary correlation techniques.