MINFLUX microscopy resolves subunits of the cardiac ryanodine receptor and its 3D orientation in cells

Figures

• 

  Fig. 1. 3D MINFLUX images of RyR2 sdAB labeling in HEK293 cells. A. x-y overview of anti-GFP sdAB labeled RyR2s in a HEK293 cell stably expressing RyR2D4365-GFP. B. The corresponding x-z view illustrates the complex 3D distribution of the localizations. C. A small region (dashed rectangle in A), shows that localizations that appear adjacent in an x-y view extend over considerable distances along the z-axis (D). Color indicates z-elevation. Scale bars A, B: 500 nm; C, D: 100 nm.

• 

  Fig. 2. Determination of effective labeling efficiency of anti-GFP sdABs in co-cultures of cells expressing RyR2D4365-GFP or U2OS-Nup96-mEGFP. A MINFLUX 3D image of two adjacent ROIs (green boxes) that were imaged time multiplexed in one imaging run. The left ROI was selected in an U2OS-Nup96-mEGFP cell while the right one was located in a HEK293 cell stably expressing RyR2D4365-GFP. Color indicates z-elevation. B 3D view of MINFLUX localizations in the U2OS-Nup96-mEGFP containing ROI with overlaid templates that identify locations and 3D orientation of NPCs. C Quantitative analysis of NPC data shown in B by overlaying the cumulative histograms of labeled NPC segments (0–16) with predicted curves for various values of pLE and line of best-fit (dashed) at pLE = 60.1 ± 0.2 %. D Across experiments on average an effective labeling efficiency of 51.2% ± 2.3% was obtained (N = 10 MINFLUX data sets from n = 3 biological replicates). The inset shows the average number of site visits per U2OS-Nup96-mEGFP site, which is very close to 1 (1.18 ± 0.09, N = 10 MINFLUX data sets from n = 3 biological replicates). Box plots show the median (center line), the 25th and 75th percentiles (bounds of the…

• 

  Fig. 3. Marker systems for RyR2 MINFLUX imaging and observation of tetramers. A Schematic of relative target and marker sizes showing a primary/secondary AB labeling approach versus a GFP-sdAB or TagRFP-sdAB approach. A microtubule in cross-section illustrates target protein size (diameter ~25 nm) which is similar to the size of RyR2 (panel inspired by ref. 20). B Schematic of RyR2D4365-GFP with approximate locations of anti-GFP sdAB binding sites according to structural data from ref. 14. C Overview of a MINFLUX 3D data set of RyR2D4365-GFP labeling in a HEK293 cell indicating a smaller region of interest (ROI, white box) enlarged in (D). D 3 ROIs (i-iii) are indicated where the localization pattern is compatible with several subunits belonging to an RyR2 tetramer. E Enlarged view of ROIs in (D), rotated into the plane of the candidate RyR2 tetramer and also overlaid with the likely RyR2 orientation in (F). Color indicates z-elevation. Scale bars C: 1 µm, D: 200 nm, E, F: 20 nm.

• 

  Fig. 4. Clustering of RyR2 subunit localizations in HEK293 cells. A Overview of RyR2 subunit clustering in a HEK293 cell expressing RyR2D4365-GFP. Clustering was determined with DBSCAN (ε = 100 nm) and using the 3D coordinates of subunit locations. Coloring indicates cluster size. B x-z view of the large yellow colored cluster in (A). C, D. x-z and y-z views of the loop-region indicated in B. Coloring in B-D indicates z elevation. E Boxplots of mean cluster sizes measured in SU/cluster as counted from data (Measured) and after correction with an indicative effective labeling efficiency (Corrected). N = 10 MINFLUX data sets from n = 3 biological replicates. F Fractions of RyRs present in the three largest cluster indicating that a sizable fraction and frequently the majority of RyRs are within a few giant clusters. N = 10 MINFLUX data sets from n = 3 biological replicates. G Mean cluster sizes across all data sets obtained by using labeling efficiency corrected data from E and division by 4 to obtain RyR2 numbers from SU numbers. N = 10 MINFLUX data sets from n = 3 biological replicates. Box plots show the median (center line), the 25th and 75th percentiles (bounds of the box), and the…

• 

  Fig. 5. MINFLUX imaging of RyR2s in myocytes from PA-RFP RyR2 knock-in mice. A RyR2 distribution in an isolated myocyte from a PA-RFP RyR2 mouse, stained with an anti-tagFP sdAB-Alexa647 and imaged in the confocal mode of the MINFLUX microscope. N > 5 independent repeats with similar results. B Enlargement of a region at the surface sarcolemma (green rectangle in A). C A MINFLUX DNA-PAINT 3D image, localizations are colored by the size of clusters as number of clustered subunits (SU) from DBSCAN cluster analysis. D Statistical analysis yields a mean cluster size of 9.0 ± 0.5 SUs or 16.5 ± 0.9 SUs when only looking at clusters with more than one SU – labeled “CS > 1 SU” (N = 22 MINFLUX data sets from n = 5 cell isolations). E A detail view containing 2 large clusters and a number of isolated SUs (arrows). F The majority of RyR2s reside in large clusters containing ≥50 SUs with on average 72.4 % ± 2.3 % of all RyR2s in these large clusters (N = 22 MINFLUX data sets from n = 5 cell isolations). G A large RyR2 cluster with a number of SUs in tretrad arrangements as indicated by RyR2-shaped overlays. H (i) Statistical analysis across datasets gives a mean RyR2 cluster size of 2.3 ± 0.1 RyR2s…

• 

  Fig. 6. Comparison of MINFLUX DNA-PAINT 3D RyR2 data with cluster morphologies from EM tomography data. A RyR2 positions determined by EM tomography as traced from ref. 4, Fig. 1 therein, overlayed with the locations where sdABs can bind to TagRFP residues (see also Supplementary Fig. 7). From all possible binding positions labeled positions were selected randomly (shown as blue dots) with a probability pSU = 0.5. B 3D localization errors (σ = 3 nm) were added to traced EM data from ref. 26, Fig. 2 therein (i) and4, Fig. 1 therein (ii) to generate simulated subunit clusters which appear broadly similar to actual MINFLUX 3D subunit clusters (iii,iv). Scale bars A: 100 nm; B: 250 nm.