Cell culture and treatment
MC3T3-E1 cell line was obtained from ATCC (Manassas, VA, catalog #CRL-2593). MC3T3-E1 cells were cultured with α-MEM (α-minimal essential medium) containing 10% (v/v) foetal bovine serum (FBS), and 1% (v/v) penicillin/streptomycin. MC3T3-E1 cells were cultured by seeding on a cover glass at 2X104 cell/cm2 in six well plates. The next day, when the confluence of the cells was approximately 70–80%, 50 µM H2O2 was treated with culture media for 24 h (St. Louis, MO, USA, catalog #H1009). After 24 h, cells were fixed with 4% paraformaldehyde. The fixed cells were subjected to IF.
Immunofluorescence staining of γH2AX and TIF (telomere dysfunction-induced foci)
For IF staining of γH2AX, MC3T3-E1 cells were cultured on coverslips, fixed for 15 min with 4% paraformaldehyde, and permeabilised for 15 min with 0.1% Triton X-100 in phosphate-buffered saline containing Tween 20 (PBST, 0.1% Tween 20). Cells were blocked with 3% Bovine Serum Albumin (BSA) and incubated with Anti-γH2AX (Ser139) at the final concentration of 1:200 (cell signalling technology, catalogue #9718) and the secondary antibody reaction was performed using Alexa Fluor 488 antibody (Invitrogen, USA) in the dark. Subsequently, the nucleus was stained with DAPI and mounted.
Following the procedure mentioned earlier, damaged cells underwent a secondary antibody assay to detect TIF. After the second antibody reaction, an additional fixation was performed with 4% paraformaldehyde and hybridized using a Cy3-labeled CCCTAA telomeric Peptide nucleic acids (PNA) probe (Panagene, Daejeon, Republic of Korea). The 4′,6-diamidino-2-phenylindole (DAPI) staining (ImmunoBioScience) was used to identify the nuclei. γH2AX foci were quantified and visualised using a confocal microscope (LSM 800; Carl Zeiss, Germany).
Quantification of foci
After IF, the number of foci in the nuclei of at least 100 cells was determined. When capturing an image under a confocal microscope, Foci-Xpress was used for comparison with ImageJ and manual counting.
1) Classical methods using ImageJ and manual counting
Two methods were used to count foci. First, manual counting of the image files was performed. In the manual method, damaged cells were manually defined, as in the case of more than 10 foci in the nucleus.
ImageJ software was used for counting. In this method, the nucleus was selected by designating the ROI as DAPI. The nucleus selected by ROI is measured to the count of total cells and the foci are captured by command of ‘Find maxima…’ in an ImageJ tool. Image J-based quantification defined cells with 10 or more foci with an intensity of 10000 or higher as damaged cells using ‘Find-maxima...’.
2) Foci-Xpress
In the Fiji-based counting tool, the size of the nuclei was measured, except for the pixel size within 100 µm² based on the Image J pixels. For foci counting, cells with 10 or more foci and a degree of brightness of 10000 or more were defined as damaged cells (Fig. 1B). The parameters were defined for all confocal image files. After the parameters were set, the Foci-Xpress separated the channels of the image (Fig. 1A; step of the split channel). Subsequently, the nuclei stained with DAPI in each channel were defined as ROIs, as set in the parameters (Fig. 1A; step of Detect Nucleus Area). The number of foci of the marker (γH2AX) in the designated ROI was quantified (Fig. 1A; steps of detecting marker and acquiring marker data) with a preset parameter (intensity ≥ 10000, 10 ≥ γH2AX foci; Fig. 1, B). The quantified data are presented as a single CSV file. All the parameters were applied to the confocal images for each independent experiment.
Development of a novel foci quantification tool
This script is based on ImageJ/Fiji software (National Institutes of Health)34. To read the image files, the BFImport function was used to import data from many life science file formats. In the image data, two or three channels represent the marker and background. In almost all cases, the blue channel indicates the background and the red or green channel is the marker. To split and detect the channels, the getLut function in the ImageJ/Fiji script was used. The Threshold of the Huang2 method35, which is an alternative to the Huang method with changes for 16 bits was used as the threshold to obtain temporary background binary images. Subsequently, the filling of holes and a Gaussian Blur filter with a parameter of three sigma are applied to reduce salt-and-pepper and random noises. The threshold of the Huang2 method was applied once more to obtain a background binary image with noise reduction. The watershed algorithm was applied to a noise-reduced background image to split the overlapping background36. Finally, the background is detected by analysing particle modules in ImageJ with foci sizes predefined by users to count and measure objects from pre-processed binary or thresholded images.
Marker images based on the ROIs of the detected background were saved to each image for subsequent processing. The marker Images for each ROI were loaded separately. To process the marker images, Bernsen’s auto local threshold with a 40-pixel radius was applied to the marker image. The markers were detected by particle analysis modules in ImageJ to count and measure objects from predefined binary or threshold images. Because the marker particles are smaller, there is no size limitation. If both red and green marker images were available, the minimum and maximum values of the two marker intensities were obtained.
Each experimental image is processed separately. After the separation process, each experiment was summarised using the marker property data. Mark brightness and counts were used for summarisation and selection. The markers were selected based on criteria for marker brightness. Finally, the program makes the recapitulation comprising the average intensity, number of selected nuclei, number of total nuclei, and ratio of the selected nuclei.