Analysis of Phase Heterogeneity in Lipid Membranes Using Single-Molecule Tracking in Live Cells

Abstract

In live cells, the plasma membrane is composed of lipid domains separated by hundreds of nanometers in dynamic equilibrium. Lipid phase separation regulates the trafficking and spatiotemporal organization of membrane molecules that promote signal transduction. However, visualizing domains with adequate spatiotemporal accuracy remains challenging because of their subdiffraction limit size and highly dynamic properties. Here, we present a single lipid-molecular motion analysis pipeline (lipid-MAP) for analyzing the phase heterogeneity of lipid membranes by detecting the instantaneous velocity change of a single lipid molecule using the excellent optical properties of nanoparticles, high spatial localization accuracy of single-molecule localization microscopy, and separation capability of the diffusion state of the hidden Markov model algorithm. Using lipid-MAP, individual lipid molecules were found to be in dynamic equilibrium between two statistically distinguishable phases, leading to the formation of small (∼170 nm), viscous (2.5× more viscous than surrounding areas), and transient domains in live cells. Moreover, our findings provide an understanding of how membrane compositional changes, i.e., cholesterol and phospholipids, affect domain formation. This imaging method can contribute to an improved understanding of spatiotemporal-controlled membrane dynamics at the molecular level.

From Homogeneity to Turing Pattern: Kinetically Controlled Self-Organization of Transmembrane Protein

“…In addition to the previously introduced RD, phase separation is another mechanistic model used to explain the spatial patterns of biomolecules. Biocondensates created through phase separation have dynamic properties such as size variability due to fission–fusion and irregular spacing between condensates. , In contrast to phase separation, our research revealed a remarkable consistency in the spacing and size of fenestrae (Figure d–f and Figure S4). On this basis, we proposed that Turing’s RD model, which has inherent regularity due to wavelike solutions (Supplemental Note), might be more appropriate for explaining the kinetically controlled spatial organization of fenestrae.…”