True Molecular Scale Visualization of Variable Clustering Properties of Ryanodine Receptors.

DNA-PAINT Achieves Molecular Resolution with Relatively Low Experimental Complexity

Our data confirm that DNA-PAINT can provide very high spatial localization, with comparatively low optical complexity and demands on dye photo-physics, so that we could achieve high precision routinely in relatively complex biological preparations. As shown recently ( Dai et al., 2016 , Jungmann et al., 2014 ), the photon yields for imaging with DNA-PAINT improve the localization precision to 15 RyRs, using the RyR cluster mask for segmentation of both RyR and JPH event data) in exchange-PAINT data. This included per channel calibrations of the RyR and JPH signals, respectively. The ratio r J-R = N R / N J was calculated for each large cluster in this way.

Simulation of Synthetic Data

Centroids of punctate RyR labeling densities were placed iteratively in either (1) a gridded organization at a fixed spacing ( Figure S2 A) or (2) randomly placed and at a variable spacing to the next nearest neighbor as described by a random sample from a normal distribution with a specified σ. The centroids of the puncta were then convolved with a 2D Gaussian model with a σ of 5 nm. The model was then used in the PyME software to generate single-molecule events within the labeled regions and rendered as grayscale images to match the imaging and localization parameters observed in the dSTORM and DNA-PAINT experimental data for simulation of dSTORM and DNA-PAINT imaging, respectively ( Table S2 ).

Imaging Probes, Reagents, and DNA-PAINT Protocol All primary antibodies used here were previously characterized for both immunofluorescence imaging and in vitro analyses of the targets RyR2, JPH2, and CAV3. The RyR2 antibody was a mouse monoclonal IgG (catalog no. MA3-916; Thermo Scientific, DE). See details on antibodies under Supplemental Experimental Procedures . For dSTORM imaging, a secondary antibody conjugated to Alexa Fluor 647 was used and imaged in switching buffer. In DNA-PAINT imaging, the sample was immersed in a buffer (“buffer C” as in Jungmann et al., 2014 ) containing typically 200 pM of either an ATTO 655 or ATTO 550 imager strand, complementary to the docking strand linked to the secondary antibody that we aimed to image. The imager strands reversibly bind to the complementary docking strands. Using TIRF illumination the fluorophores of hybridized imager strands were imaged, while transiently immobilized and appeared in the image as transient fluorescent spots whose shape matches the point spread function (PSF) of the microscope in its focal plane. These events were recorded as a series of image frames ( Figure S1 ).

Statistical Methods

All mean and SDs of measurements presented in the manuscript were calculated using standard statistics routines in NumPy or Excel.

Supplemental Information Document S1. Experimental Procedures, Figures S1–S6, and Tables S1 and S2 Document S2. Article plus Supplemental Information