PM-NANOMAPS · Dynamic spatial organization of plasma membrane proteins at the nanoscale

Dynamic spatial organization of plasma membrane proteins at the nanoscale

The overall objective of the project was to obtain quantitative structural information of the plasma membrane (PM) organization at the nanoscale. Two main problems have hindered the direct visualization of membrane components at the molecular level with fluorescent microscopy in previous experiments. First, the specific labeling of membrane components such as proteins, glycans, or lipids without perturbing their native distribution. Second, the diffraction-limit of fluorescence microscopy restricting the spatial resolution to approximately half of the wavelength of light in the image plane. To overcome these problems, we applied super-resolution (SR) fluorescence microscopy techniques, such as direct Stochastic Optical Reconstruction Microscopy (dSTORM), combined with a bioorthogonal labeling strategy based on “click-chemistry”. While the initial specific objective of the project was to study the dynamics of PM components, ongoing research lines at the host group showed that the aforementioned experimental approach was not well suited. The main problem was that the laser intensities required to produce artifact-free super-resolved dSTORM images impeded SR live-cell imaging due to cell damage. Therefore, the final results are focused on the static organization of PM proteins and glycans after fluorescence staining and fixation as described below: 1- Fluorescence staining of PM proteins using bioorthogonal “click chemisty” reactions: Bioorthogonal fluorescence labeling strategy based on “click chemistry” allowed the visualization of large populations of PM proteins and glycans with minimal perturbation of their native distribution. We first metabolically labeled newly synthesized proteins with clickable amino acids containing an azide group, e.g. L-azidohomoalanine (AHA), incorporated co-translationally into nascent proteins by replacing endogenous methionine. Alternatively, we used azido sugars, e.g. peracetylated N-azidoacetylgalactosamine (Ac4GalNAz), incorporated post-translationally into membrane glycans including glycoproteins. Once these proteins were delivered to the PM, they were stained with fluorophores via copper-catalyzaed (CuAAC) or copper-free azide-alkyne cycloadditions (SPAAC). Fluorophore titrations revealed optimal labeling efficiency conditions leading to localization densities from 400 to 2000 localizations per square micrometer. Protein synthesis inhibitors confirmed the specificity of this labeling strategy. 2- dSTORM imaging (artifacts and 2D Projections of 3D cell membrane structures): SR localization microscopy based on the photoswitching of standard organic fluorophores, i.e. directSTORM, allowed the detection of PM proteins with single-molecule sensitivity. The images recorded for the four labeling schemes described above showed highly dense homogenous organization of PM components. Artifacts due to improper fluorophore photoswitching and two-dimensional (2D) projections of three-dimensional (3D) cell structures were identified. For example, overlapping membranes leaded to overestimation of protein content, and vesicle-like structures located in the proximity of the PM appeared as protein clusters. 3- Quantitative estimation of densities and spatial distribution of PM proteins and glycans: Due to the intrinsic photoswitching of fluorophores a single labeled protein can be detected multiple times. Therefore, localization densities measured with dSTORM imaging do not directly reflect the molecular content of PM proteins. To overcome this problem, reference samples were prepared to evaluate the number of repeated localizations per fluorophore. Our results lead to estimated densities form 50 to 350 molecules per square micrometer depending on the labelling scheme used. Besides estimation of molecular content, the spatial distribution of membrane proteins was quantitatively analyzed using different methods, e.g. density maps, pair correlation function, and K Ripley’s function analysis. Comparison of the spatial distribution obtained experimentally with simulated data allowed to determine the degree of clustering of PM proteins at different length scales. Main deviation from homogeneity appeared at distances related with the experimental precision (20 to 30 nm) and multiple detections of single fluorophores. However, a certain but low degree of clustering was also observed at larger scales (from 30 to 800 nm) as revealed by K Ripley’s function analysis. CONCLUSIONS: The plasma membrane (PM) is involved in several cell functions such as immune response, cell adhesion, and tissue formation. Whereas its composition has been extensively investigated from the biochemical point of view, direct visualization of PM components at the nanoscale has been hindered especially when using standard fluorescence microscopy. Recently, super-resolution localization microscopy has surpassed the diffraction-limit achieving a spatial resolution down to approximately 20 nm. During the last decade, such emergent techniques (e.g. STORM, dSTORM, PALM and fPALM) have revealed the existence of protein nanoclusters of synaptic proteins, and membrane receptors involved in cell proliferation, cell growth, or immune response. Hence, strengthening the hypothesis of a universal hierarchical spatial organization of PM where membrane proteins are mainly confined forming nanodmains with sizes ranging from few tens to few hundred nanometers. However, to inspect the global distribution of PM components a more general experimental approach devoted to stain and visualize simultaneously large population of proteins is required. This project showed how metabolic labelling and “click chemistry” combined with dSTORM imaging can be used to visualize the spatial distributions of PM components such as membrane proteins and glycans. Our data demonstrate the validity and efficiency of the methodology. Besides the quantitative estimation of PM protein contents, our data revealed how 2D-projections of 3D-inherent structures can produce artifacts. In particular, vesicle-like structures and fluorophore photoswitching induced apparent clusters with a size between a few tens to few hundred nanometers. Finally, statistical methods revealed a certain degree of lateral aggregation in areas without the artifacts previously mentioned in a length scale between 30 and 800 nm. Whereas these results might indicate some spatial organization of PM proteins at the nanoscale, further experimental data with 3D super-resolution microscopy can help in the future to elucidate the organization of PM membrane component completely unbiased by 2D projections of cell structures. IMPACT OF THE TRAINING ACTIVITIES IN THE CAREER DEVELOPMENT OF THE FELLOW: While fulfilling this project the fellow has gone to a complete set of training activities oriented to gain research expertise in the field of super-resolution microscopy, as well as transferable skills to become an independent researcher. The fellow carried out all the stages of the project including: sample preparation, image acquirement, and quantitative analysis of the experimental data. Furthermore, the fellow gained experience in supervising MSc and PhD students, he strengthened his scientific communication skills attending and presenting his results in international conferences, as well as his scientific writing skills preparing peer review articles and grant proposals. All in all, the fellowship enlarged the perspectives of the fellow in terms of career development accomplishing wide experience in super-resolution microscopy what will allow him to continue investigating in the field in the next future.

Текст от CORDIS, на английски · Данни: CORDIS, © Европейски съюз