STORM imaging and analysis

YJ Yunmin Jung
LW Lai Wen
AA Amnon Altman
KL Klaus Ley
request Request a Protocol
ask Ask a question
Favorite

For STORM imaging, µ-Slide 8-well glass bottom chambered coverslips (Ibidi) were cleaned with 1 M NaOH (Sigma) for 1 h, and coated with PLL (0.1%) for 1 h at RT before use. Fixed cells labeled with AF647- or AF568-conjugated antibodies and suspended in 300 μl PBS were placed in each well incubated for 1 h at RT, and then the buffer was exchanged with blinking buffer [PBS containing 0.5 mg mL−1 glucose oxidase, 40 μg mL1 catalase, 10% (w/v) glucose, 50 mM cysteamine, and 93 mM Tris-HCl (all from Sigma)]. All buffers were filtered with 0.2 μm-pore filters (Corning) and freshly mixed before use. For fiducials, Fluorescent Nanodiamond (FND) (100 nm, Adamas Nano) diluted in water (1:1000) was sonicated for 15 min and then added to the PLL-coated wells for 30 min. Unbound FNDs were washed 5x with distilled water and then blinking buffer was added. STORM imaging was acquired with a UAPON100XOTIRF1.49NA oil objective (Olympus) at epi-illumination mode on Nanoimager (ONI) microscope implemented with a cylindrical lens for localization in z-dimension. The focus was maintained by the autofocus module of the system. Excitation lasers 532 and 640 nm and the two channels were separated with dichroic (640 LP) and emission (584/80 and 685/40) filters. 532 and 640 nm laser lines were operating at a power of 27.5 and 13.8 kW cm−2, respectively. For each channel, 30,000 frames (3000 frames x 10 movies) were recorded with a speed of 10 ms per frame. The pixel size on the detector was 117 nm. The two-channel registration with FND fiducial images in the two-channel and single molecule localization processing was performed using Nanoimager software (version 1.1.6165-012f4ed3; ONI). Data analysis and reconstruction of images were performed with custom-written Matlab (MathWorks, 2018a) scripts. Single molecule localization data were thresholded as: 0–14 nm for “precision in x, y (nm)”;−400 to 400 nm for “z range”; 20,000 for “number of photons”; 100 for “background”; 0–2.5 for “Point spread function σX and σY (pixel)'; 0.6–1.5 for “σX/σY” (pixel)'. Channel correction for individual cell images was done with low-resolution images of the single molecules collected in all frames rendered with the original pixel size by calculating the shifted peak of the 2D cross-correlation image obtained using a 2D Gaussian fit. The tip positions were determined using the 2D rendered image (ignoring z localization) as described in the area segmentation section below. Molecules that are localized within the 9 × 9 pixels (corresponding to 1.053 µm × 1.053 µm) around the tip positions were analyzed for the quantification. 3D-pair-correlation (g(r)) between the l-sel and CD45 was calculated. Briefly, the mean count of all CD45 molecules localized within a distance between r (radii) and r + dr (dr = 10 nm) away from each l-sel molecules was calculated, and then it was divided by the volume of the spherical shell. Finally, this number was divided by the total density of the CD45 molecules.

Do you have any questions about this protocol?

Post your question to gather feedback from the community. We will also invite the authors of this article to respond.

post Post a Question
0 Q&A