Electron microscopy combined with spatial analysis: quantitative mapping of the nano-assemblies of plasma membrane-associating proteins and lipids

Abstract

The plasma membrane (PM) is a complex environment consisting of > 700 species of lipids and many different types of membrane-associating proteins. These lipids and membrane proteins are distributed heterogeneously into nanometer-sized domains, called nanoclusters. The lateral spatial segregation in the PM gives rise to different curvature and lipid composition, which determines the efficiency of effector binding and signal transmission. Here, we describe an electron microscopy (EM)-spatial mapping technique to quantify the extent of nanoclusters formation in the PM. The nano-assemblies in the PM are quantified via expressing the GFP-tagged proteins or lipid-binding domains in the cells, which are then immunolabeled with the gold nanoparticles pre-coupled to the anti-GFP antibody. The gold nanoparticles are visualized via the transmission EM at high magnification. The statistical analysis of the Ripley's K-function calculates the spatial distribution of the gold nanoparticles. Important spatial parameters, such as the extent of nanoclustering, the clustered fraction, the number of proteins per cluster, the optimal size of a nanocluster, and the number of proteins localized to the PM, can be calculated. Further detailed aggregation pattern, such as the populations of monomers, dimers, trimers, and higher ordered oligomers, can also be extracted from the spatial analysis. The EM-bivariate analysis quantifies the extent of co-localization between two different components in the PM and provides key information on the protein-protein and the protein-lipid interactions over a long-distance scale from 8 to 240 nm.

Keywords: Electron microscopy; Lipid-anchored proteins; Nanoclustering; Ripley’s K-function; Signal transduction; Spatial distribution.

Fig. 1

Univariate nanoclustering analysis quantifies the lateral spatial distribution of a single population of gold nanoparticles on the intact plasma membrane sheets. A An intact PM sheet of kidney cells ectopically expressing GFP-K-Ras was attached to an EM grid, fixed and immunolabeled with 4.5-nm gold nanoparticles. Gold distribution on the intact PM sheet was imaged using transmission EM at 100,000× magnification. B x and y-coordinates of the gold particles were assigned using ImageJ and were used to calculate the spatial distribution. Blue dots indicate monomers; yellow dots indicate dimers; orange dots indicate trimers; red dots indicate higher ordered oligomers. C The extent of nanoclustering, L(r) − r, was plotted against the length scale, r, in nanometer. L(r) − r values above the 99% confidence interval (99% CI) indicate statistically meaningful clustering

Fig. 2

Bivariate co-clustering analysis characterizes the co-localization between two populations of gold nanoparticles. A An intact PM sheet of kidney cells ectopically co-expressing GFP-K-Ras and RFP-RAF-1 was attached to an EM grid, fixed and immunolabeled with 6-nm gold labeling GFP and 2-nm gold labeling RFP. Gold distribution on the intact PM sheet was imaged using transmission EM at 100,000× magnification. B x and y-coordinates of the gold particles were assigned using ImageJ and were used to calculate the co-localization between the two populations of gold. Smaller black dots indicate 2-nm gold labeling RFP; larger red dots indicate 6-nm gold labeling GFP. C The extent of co-localization, L_biv(r) − r, was plotted against the length scale, r, in nanometer. L_biv(r) − r values above the 95% confidence interval (95% CI) indicate statistically meaningful co-localization