Novel approach to measure the size of plasma-membrane nanodomains in single molecule localization microscopy – University of Copenhagen

Forward this page to a friend Resize Print Bookmark and Share

DynaMo Center of Excellence > DynaMo News > Publications > Novel approach to meas...

24 June 2015

Novel approach to measure the size of plasma-membrane nanodomains in single molecule localization microscopy

Ziomkiewicz I, Sporring J, Pomorski TG, Schulz A (2015) Novel approach to measure the size of plasma-membrane nanodomains in single molecule localization microscopyCytometry A 87: 868–877.

Abstract

Many membrane proteins are not evenly distributed over the plasma membrane, but gathered in domains assumed to have a particular lipid composition. Using single molecule localization microscopy (SMLM) we have immunolocalized a glycosylphosphatidylinositol (GPI)-anchor protein that labels nanodomains in a specialized plant cell type, and compared the suitability of three methods to estimate their size. As conventional methods full width at half maximum (FWHM) and the full diameter (FWMin) of domains were used. A boundary detection method of the domain area (DA) was performed in order to take irregular shapes into account. In order to compare the influence of the chosen measurement methods, we have developed a MatLab program that allows for automated analysis of domain sizes from multiple SMLM images and provides the statistics of three key features of domains: FWHM and FWMin along their long and short axes as well as the DA, derived from the molecular density. Domains formed by the GPI-anchor protein are approximating elliptical shapes. Direct and indirect immunolabeling resulted in a statistically significant difference in apparent domain size, reflecting the fact that the secondary antibody molecules extend the uncertainty along the nanodomain border. FWMin values along the long and short axis give good estimates of regular, geometrically centred domain shapes, while the DA value matches regular as well as irregular shapes best, as derived from computer-generated, irregular point clusters.

Link to article