Telomeres are specialized nucleoprotein structures that protect chromosome ends from DNA damage response (DDR) and DNA rearrangements. resolved individual fluorophores, forming clusters of localizations corresponding to telomeres. To ensure that every cluster corresponded to a telomere, the localizations were overlaid with wide-field images, and the data were filtered to reject groups of signals that did not correspond to an image of a telomere, had Rabbit polyclonal to PAI-3 a very low number of localizations (<50), or were not properly clustered. Many telomeres buy ZM 323881 hydrochloride adopted roughly an ovoid structure, but the heterogeneity of shapes suggested a buy ZM 323881 hydrochloride considerable plasticity of telomeres (Fig. 1C; Supplemental Fig. S2). Figure 1. Human telomeres are heterogeneous in length, size, and shape. (= 412) than in HeLa S (mean = 299), although their ratio was not in proportion with their average lengths, which differed by a factor of three. Finally, fluorescence intensity in fluorescence in situ hybridization (FISH) experiments has been correlated with telomere length in numerous studies (Poon and buy ZM 323881 hydrochloride Lansdorp 2001). Altogether, these analyses suggest that the small subset of DDR-positive telomeres in TRF2-depleted cells that had a larger = 1.4 + 0.6/?0.5. STORM image acquisition STORM imaging was performed on a custom-built STORM microscope with a 100 100-m2 FOV as described previously (Douglass et al. 2016). The large FOV of this microscope allowed for the simultaneous imaging of between 10 and 30 nuclei; a flat illumination pattern ensured uniform fluorophore photoswitching across the FOV. For each condition and replicate, three to five FOVs were acquired, depending on the density of the cells. For the present work, an additional laser (Coherent Sapphire, 488-nm peak emission wavelength, 50 mW) was introduced into the setup to image Alexa 488 IF. A dichroic filter (Chroma, Z488bcm) was used for beam combining, and fluorescence emission in the Alexa 488 channel was filtered with a GFP emission filter (Chroma, ET525/50m). Individual coverslips containing fixed and labeled HeLa cells were placed in a custom-built sample holder containing 1000 L of imaging buffer (see below) supplemented with an oxygen-scavenging system. Before each STORM acquisition, a wide-field image of the FOV was acquired: one for the Alexa 647 channel (50-msec exposure time at 1.4 mW in the objective back focal plane [BFP]) and one for the Alexa 488 channel (500-msec exposure at 0.1 mW in the objective BFP). For STORM acquisitions, 20,000 frames per FOV at 10-msec exposure time and zero interframe delay were acquired with 590 mW of 647-nm laser power in the objective BFP; only the Alexa 647 channel was acquired in STORM. A 405-nm laser light was applied at frame number 10,000 and steadily ramped upward between 0 and 4. 0 mW in the objective BFP through the end of the acquisition. The 405-nm laser light was applied to return Alexa 647 fluorophores to the emitting state and achieve more complete spatial sampling. The STORM imaging buffer with oxygen-scavenging system was described previously (Olivier et al. 2013) and uses millimolar concentrations of polyunsaturated hydrocarbon cyclooctatetraene to boost photon yields during STORM imaging. All reagents were purchased from Sigma-Aldrich. The images shown in Figure 1A were taken on an inverted Nikon N-STORM microscope with a 100/1.49 N.A. apo TIRF objective (Nikon) and an EMCCD camera (Andor, iXon3 897). A 500-mW 640-nm laser (Coherent Sapphire) and a 100-mW 402-nm laser (Coherent Sapphire) were used to induce fluorophore photoswitching and control the switching rate, respectively. Molecule localization and drift correction (using cross-correlation) for data in Figure 1A only were performed in the Nikon NIS-Elements software version 4.30.01. Before the STORM acquisition, wide-field images of the DAPI and Cy5 channels were acquired. The probe used in this experiment was the Cy5-(CCCTAA)3 PNA probe (Eurogentec, PN-TC055-005), and the DNA was labeled with DAPI. The oxygen-scavenging system used for STORM imaging was glucose oxidase/catalase-based and prepared as described previously (Olivier et al. 2013). Filtering and cluster analysis of STORM data The filtering and analysis pipeline used in this work consists of seven discrete steps that were applied to each FOV individually (Supplemental Fig. S1). Unless otherwise stated, analyses were performed in a custom-written Python analysis library (B-Store, versions 0.1.1 and 0.2.0; https://github.com/kmdouglass/bstore) for Python 3.5. Computing localizations from raw image stacks Input data for the analysis pipeline originated from STORM acquisitions and consisted of stacks of images of single.