Section: New Results

Correlation-based method for membrane diffusion estimation during exocytosis in TIRFM

Participants : Ancageorgiana Caranfil, Antoine Basset, Charles Kervrann.

The dynamics of the plasma membrane of the cell is not fully understood yet; one of the crucial aspects to clarify is the diffusion process during exocytosis. Several observation methods exist, including TIRFM (Total Internal Reflection Fluorescence Microscopy), that has successfully been used to determine the successive steps of exocytosis. However, computing characteristic values for plasma membrane dynamics is problematic, as the experimental conditions have a strong influence on the obtained data and a global model cannot be determined. The goal of this study was to build a correlation-like method to estimate local diffusion parameters in TIRFM images. Using a correlation approach similar to TICS (Temporal Image Correlation Spectroscopy) with an adapted local model, we have developed a novel correlation-based method to estimate the diffusion coefficient for every diffusion event in TIRFM images. We turned the non-linear model of the TICS method into a linear one, and made it rely on less parameters than the other estimation methods. Results are excellent for sequences with a good signal-to-noise ratio (see Fig. 8 ); however, time and space dependencies are introduced with the presence of moderate-to-strong image noise. Although only synthetic images have been used so far, studies of real-life TIRFM images are forthcoming, along with refinements to make the method robust to noise.

Collaborators: Perrine Paul-Gilloteaux and Francois Waharte (UMR 144 CNRS-Institut Curie, PICT-IBiSA).

Figure 8. Left: first six instances of a TIRFM image sequence showing a diffusion event. Right: the correlation-based method is applied on the TIRFM sequence; both the computed values of the autocorrelation, for different values of the temporal lag, and the fitting function for these values are represented.