Regulation of spatial-temporal dynamics of exocytosis by the exocyst complex

Neuronal migration is essential for proper development of the cerebral cortex. As a first step, a postmitotic cell extends its leading process, presumably by adding new membrane at the growing tip, which would enable directed locomotion. The goal of the first part of the present study was to determine if biosynthetic exocytic pathway is polarized in migrating cells and whether polarized exocytosis promotes directed cell migration. A promising candidate for controlling the spatial sites of vesicle tethering and fusion at the plasma membrane is an octameric protein complex called the exocyst. We found that cell migration in a wound assay, as well as cortical neuronal migration during embryonic development, was impaired when the exocyst complex was perturbed. By combining total internal reflection fluorescent microscopy (TIRF) and a stochastic model of exocytosis, we found that vesicle exocytosis is preferentially distributed close to the leading edge of polarized cells, that the exocytic process is organized into spatial hotspots" and that the polarized delivery of vesicles and clustering in hotspots depend on the intact exocyst complex. The exocyst complex achieves this spatial regulation by determining the sites at the membrane where secretory vesicles tether. Thus, our study supports the notion that polarized membrane traffic regulated by the exocyst is an essential component of cell migration and that its deficit may lead to cortical abnormalities involving cortical neuronal malpositioning.;Recent advances in cellular imaging have enabled the monitoring intracellular processes, but designing mathematical tools that can formally describe their dynamics, remains an emerging challenge. In the second part of this work, I implemented a novel approach based on hidden Markov models (HMMs) in order to characterize insulin-controlled membrane trafficking events critical for the homeostasis of blood glucose levels. While a Poisson process-based model provides a static snapshot of a process at one time point, the HMM reveals how a process evolves in real time and is thus particularly suitable for analyzing imaging data. Markov models describe stochastic processes whose state changes over time. Hidden Markov models describe stochastic processes whose states we cannot observe directly; we can only observe the outcomes of the states. In live imaging, we measure variables related to the events of interest that we observe. Based on the sequence of outcome of these measurements, a HMM can reveal a sequence of hidden states that generated the outcomes. It thus enables us to analyze precisely the spatial-temporal dynamics of a process. We propose that HMMs represent a promising tool in the field of biomedical imaging, where we are inherently interested in describing the dynamics of a given process in real time.