Cell Cycle-specific Measurement of γH2AX and Apoptosis After Genotoxic Stress by Flow Cytometry

Ramon Lopez Perez1,2, Franziska Münz1,2, Jonas Kroschke1,2, Jannek Brauer1,2, Nils H. Nicolay1,2,3, Peter E. Huber1,2 1CCU Molecular and Radiation Oncology, German Cancer Research Center

Keywords: This Month in JoVE, Issue 151, DNA double-strand breaks, cell cycle distribution, apoptosis, DNA damage response, ionizing radiation, carbon ion radiation, genotoxic stress, flow cytometry

Abstract
The presented method or slightly modified versions have been devised to study specific treatment responses and side effects of various anti- cancer treatments as used in clinical oncology. It enables a quantitative and longitudinal analysis of the DNA damage response after genotoxic stress, as induced by radiotherapy and a multitude of anti-cancer drugs. The method covers all stages of the DNA damage response, providing endpoints for induction and repair of DNA double-strand breaks (DSBs), cell cycle arrest and cell death by apoptosis in case of repair failure. Combining these measurements provides information about cell cycle-dependent treatment effects and thus allows an in-depth study of the interplay between cellular proliferation and coping mechanisms against DNA damage. As the effect of many cancer therapeutics including chemotherapeutic agents and ionizing radiation is limited to or strongly varies according to specific cell cycle phases, correlative analyses rely on a robust and feasible method to assess the treatment effects on the DNA in a cell cycle-specific manner. This is not possible with single-endpoint assays and an important advantage of the presented method. The method is not restricted to any particular cell line and has been thoroughly tested in a multitude of tumor and normal tissue cell lines. It can be widely applied as a comprehensive genotoxicity assay in many fields of oncology besides radio-oncology, including environmental risk factor assessment, drug screening and evaluation of genetic instability in tumor cells.

Introduction

The goal of oncology is to kill or to inactivate cancer cells without harming normal cells. Many therapies either directly or indirectly induce genotoxic stress in cancer cells, but also to some extend in normal cells. Chemotherapy or targeted drugs are often combined with radiotherapy to enhance the radiosensitivity of the irradiated tumor1,2,3,4,5, which allows for a reduction of the radiation dose to minimize normal tissue damage.
Ionizing radiation and other genotoxic agents induce different kinds of DNA damage, including base modifications, strand crosslinks and single- or double-strand breaks. DNA double-strand breaks (DSBs) are the most serious DNA lesions and their induction is key to the cell killing effect of ionizing radiation and various cytostatic drugs in radiochemotherapy. DSBs do not only harm the integrity of the genome, but also promote the formation of mutations6,7. Therefore, different DSB repair pathways, and mechanisms to eliminate irreparably damaged cells like apoptosis have developed during evolution. The entire DNA damage response (DDR) is regulated by a complex network of signaling pathways that reach from DNA damage recognition and cell cycle arrest to allow for DNA repair, to programmed cell death or inactivation in case of repair failure8. The presented flow cytometric method has been developed to investigate the DDR after genotoxic stress in one comprehensive assay that covers DSB induction and repair, as well as consequences of repair failure. It combines the measurement of the widely applied DSB marker γH2AX with analysis of the cell cycle and induction of apoptosis, using classical subG1 analysis and more specific evaluation of caspase-3 activation.

The combination of these endpoints in one assay not only reduces time, labor and cost expenses, but also enables cell cycle-specific measurement of DSB induction and repair, as well as caspase-3 activation. Such analyses would not be possible with independently conducted assays, but they are highly relevant for a comprehensive understanding of the DNA damage response after genotoxic stress. Many anti-cancer drugs, such as cytostatic compounds, are directed against dividing cells and their efficiency is strongly dependent on the cell cycle stage. The availability of different DSB repair processes is also dependent on the cell cycle stage and pathway choice which is critical for the repair accuracy, and in turn determines the fate of the cell9,10,11,12. In addition, cell cycle-specific measurement of DSB levels is more accurate than pooled analysis, because DSB levels are not only dependent on the dose of a genotoxic compound or radiation, but also on the DNA content of the cell. The method has been used to compare the efficacy of different radiotherapies to overcome resistance mechanisms in glioblastoma13 and to dissect the interplay between ionizing radiation and targeted drugs in osteosarcoma14,15 and atypical teratoid rhabdoid cancers16. Additionally, the described method has been widely used to analyze side effects of radio- and chemotherapy on mesenchymal stem cells17,18,19,20,21,22,23,24, which are essential for the repair of treatment-induced normal tissue damage and have a potential application in regenerative medicine.

Protocol
1. Preparation

1. Prepare ≥1 x 105 cells/sample in any type of culture vessel as starting material.
1. For example, conduct a time-course experiment after exposure of U87 glioblastoma cells to ionizing radiation: Irradiate sub-confluent U87 cells in T25 flasks in triplicates for each time point. Choose early time-points (15 min up to 8 h after irradiation) to follow the kinetics of DSB repair (γH2AX level) and late time points (24 h up to 96 h) to assess residual DSB levels, cell cycle effects and apoptosis.
NOTE: The protocol is not restricted to irradiation experiments or any specific cell line. It has been tested with numerous cell lines of all types, from different species and for various treatment conditions.
2. Prepare the following solutions including 10% excess volume.
1. Prepare 2 mL per sample of fixation solution composed of 4.5% paraformaldehyde (PFA) in phosphate-buffered saline (PBS). Prepare the solution fresh. Dilute PFA in PBS by heating to 80 °C with slow stirring under the fume hood. Cover the flask with aluminum foil to prevent heat loss. Let the solution cool to room temperature and adjust the final volume. Pass the solution through a folded cellulose filter, grade 3hw (see the Table of Materials).
CAUTION: PFA fumes are toxic. Perform this step under a fume hood and dispose PFA waste appropriately.
2. Prepare 3 mL per sample of permeabilization solution composed of 70% ethanol in ice-cold H2O. Store at -20 °C.
3. Prepare 7 mL per sample of washing solution composed of 0.5% bovine serum albumin (BSA) in PBS.
4. Prepare 100 µL per sample of 3% BSA in PBS as antibody diluent.
5. Prepare 100-250 µL per sample of DNA staining solution composed of 1 μg/mL 4′,6-diamidin-2-phenylindol (DAPI) in PBS.

3. Set the centrifuge for 15 mL tubes to 5 min at 200 x g and 7 °C. Let the centrifuge cool down and use these settings for all centrifugation steps.

2. Sample Collection

1. If processing adherent cells (e.g., U87 glioblastoma cells grown in T25 flasks with 5 mL of Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum at 37 °C and 5% carbon dioxide atmosphere), continue with steps 2.1.1. and 2.1.2. For suspension cells proceed directly to step 2.1.2.
1. Collect the medium in a centrifugation tube. Detach the cells using a routine cell culture method, which may include the use of trypsin, ethylenediaminetetraacetic acid (EDTA) or other cell detachment agents.
1. For U87 cells, prewarm PBS and trypsin/EDTA (see the Table of Materials) to 37 °C, wash the cell layer with 1 mL of PBS, incubate the cells for 1-2 min with 1 mL of trypsin/EDTA and support cell detachment by tapping at the flask. Collect all washing solution and the cell suspension in the tube with the medium.

2. Centrifuge the cells, discard the medium and resuspend the cells in 1 mL of PBS.

2. Pipet the cells up and down several times to ensure a single cell suspension and transfer the cell suspension into a tube with 2 mL of fixation solution (4.5% PFA/PBS, 3% final concentration).
CAUTION: PFA fumes are toxic. Perform this step under a fume hood and dispose PFA waste appropriately.
3. Incubate the cells for 10 min at room temperature.
4. Centrifuge the cells and discard the supernatant by decantation.
5. Loosen the cell pellet by tapping onto the tube and resuspend the cells in 3 mL of 70% ethanol. Proceed directly with the next step or store the samples at 4 °C for up to several weeks.
3. Washing and Staining

1. Centrifuge the cells and discard the supernatant by decantation.
2. Loosen the cell pellet by tapping onto the tube. Resuspend the cells in 3 mL of washing solution (0.5% BSA/PBS), centrifuge and discard the supernatant.
3. Repeat the washing step 1x with 3 mL, and then 1x with 1 mL washing solution. In the last step, discard the supernatant carefully by pipetting. Take care not to aspirate the pellet.
4. Dilute the antibodies against γH2AX, phospho-histone H3 (Ser10) and caspase-3 (see the Table of Materials) in 100 µL/sample with antibody diluent (3% BSA/PBS).
5. Loosen the cell pellet by tapping onto the tube and resuspend the cells in 100 µL of the antibody solution prepared in step 3.4. Keep the samples in the dark from this step onward.
6. Incubate the samples for 1 h at room temperature.
7. Centrifuge the cells and discard the supernatant carefully by pipetting. Take care not to aspirate the pellet.
8. Loosen the cell pellet by tapping onto the tube and resuspend the cells in 100-250 µL of DNA staining solution (1 µg/mL DAPI/PBS).
1. Use 100 µL if 1-2 x105 cells are present and increase the volume for higher cell numbers (250 µL for ≥1 x 106 cells). Proceed directly with the next step or store the samples in the dark at 4 °C for up to 2 weeks.
NOTE: For some cell types optimizing the DAPI concentration can help to improve the separation of the cell cycle phases.
9. Pipet the samples through the cell strainer cap of a sample tube with a mesh pore size of 35 µm.

4. Measurement

1. Place the samples on ice, start the flow cytometer (see the Table of Materials) configured with an optical setup according to Table 1 and press the Prime button. If required, switch on the ultraviolet laser (355 nm wavelength) separately and set the power to 20 mW using the appropriate software.
2. Open the acquisition software (see Table of Materials), log in and create a new experiment by clicking the New Experiment button on the
Browser toolbar.
3. Use the Inspector window to customize the name of the experiment and choose ’5 Log Decades’ for plot display.
4. Click the New Specimen button in the Browser toolbar and expand the new entry by clicking the ‘+’ symbol at its left side to show the first Tube. Select the respective icon and type to rename the Specimen (e.g., cell type) and the Tube (sample identifier). Click the Tube Pointer of the first Tube (arrow-like symbol at its left) to turn it green (active).
5. Open the Parameters tab in the Cytometer window and choose the parameters according to Table 1. Delete all unnecessary parameters. NOTE: the parameter names may vary depending on the custom presets (e.g., Cy3 instead of Alexa555). Make sure that the selection matches the optical filters and detectors in Table 1. The light paths of all fluorophores are fully independent in this setup and compensation of spectral overlap is not required; however, it might be necessary if another optical setup is used.
6. Open the Worksheet window and create plots according to Figure 1. Draw 2 dot plots and 4 histograms using the corresponding toolbar buttons and click the axis labels to choose the appropriate parameters (front scatter (FSC-A) versus side scatter (SSC-A), DAPI-W versus DAPI-A and a histogram for DAPI-A and each antibody-coupled fluorophore.
7. Attach a control sample to the cytometer and press the Run button on the instrument. Select the first tube in the Browser window of the software and click the Acquire Data button in the Acquisition Dashboard. Adjust the sample injection volume using the Low, Mid or High buttons and the fine tuning wheel on the instrument. Preferably work at Low setting, but try to acquire at least 100 events/second (see Acquisition Dashboard).
8. Adjust the detector voltages for FSC, SSC and DAPI in the Parameters tab of the Inspector window using the dot plots in Figure 1 as a guideline. Switch to logarithmic scale for the FSC and SSC parameters if the cell population appears too dispersed on a linear scale.
9. Press the Standby button on the cytometer and continue with the worksheet setup in the software.
1. Use the Polygon Gate tool to define the Cells population in the FSC-A versus SSC-A plot and the Rectangle Gate tool to define the SingleCells population in the DAPI-W versus DAPI-A plot. Press Ctrl + G keys to show the Population Hierarchy and click on the default gate names to rename them.

2. Subsequently right-click on all histograms and choose Show Populations | SingleCells from the context menu.

10. Optimize the detector voltages for the antibody-coupled fluorophores to cover the full dynamic range by subsequently acquiring control and treated samples. Maximize the signal-to-noise ratio and avoid detector saturation. Make sure that the Alexa488 peak in the SingleCells population is neither truncated in the control nor the treated sample.
11. Press the Standby button on the cytometer and optionally perform steps 4.11.1-4.11.3 in the Worksheet window of the software to get a rough estimate of treatment effects during sample acquisition.
1. Select SingleCells in the Population Hierarchy and use the Rectangle Gate tool to define the G1 population in the DAPI-W versus DAPI-A plot. Right-click at the Alexa488 histogram and choose Show Populations | G1 from the context menu.
2. Right-click at the Alexa488 histogram and choose Create Statistics View from the context menu. Right-click at the Statistics View and choose Edit Statistics View. Go to the Statistics tab and activate the checkbox for the median of the Alexa488 signal in the G1 population (deactivate all other options).
3. Select SingleCells in the Population Hierarchy and use the Interval Gate tool to define the subG1, M and Casp3+ populations in the DAPI-A, Alexa555-A (Cy3-A in Figure 1) and Alexa647-A histograms.
12. Press the Run button on the cytometer and measure the samples using the Acquisition Dashboard in the software. Set the stopping gate to
All events or the Cells gate (if numerous small particles are present) and the number of events to record to 10,000.
13. Click Next Tube to create a new sample, rename it in the Browser window, click Acquire Data to start the acquisition and Record Data to start recording.
14. Select File | Export | FCS files from the menu bar to export the data. Optionally select File | Export | Experiments to save the experiment as an additional zip file to enable reimporting the experiment at a later time point.
NOTE: The setup of existing experiments can be easily reused with Edit | Duplicate without data.

5. Data Evaluation

1. Drag and drop the ‘.fcs’ files into the sample browser of the flow cytometric analysis software (see Table of Materials). Apply the gating strategy shown in Figure 2. Make sure that the gates fit to the corresponding population in all of the samples before proceeding with the next daughter gate.
1. To apply changes to all samples, select the changed gate in the sample browser, copy it by pressing Ctrl + C, select the parent gate, press Ctrl + Shift + E to select the equivalent nodes in all samples and press Ctrl + V to paste or overwrite the gate. Do not use group gates.
2. Double-click on the first sample in the browser to open the SSC-A vs. FSC-A plot. Use the Polygon tool in the toolbar to define the
Cells population (Figure 2, plot 1), excluding debris from the analysis. Make sure that the gate is wide enough towards the upper right
corner to accommodate treatment-related shifts, but restrict the border facing to the lower left corner of the plot to reliably exclude the cell.
3. Double-click on the Cells gate to open a new plot window and change the axes to DAPI-W (vertical) vs. DAPI-A (horizontal) by clicking on the axis labels. Use the Rectangle tool to define the SingleCells population (Figure 2, plot 2), excluding cell doublets or clumps from the analysis (doublets of G1 cells have the same DAPI-A intensity as G2 and M cells, but a considerably higher DAPI-W value). NOTE: Varying cell counts in different samples will cause shifts in the overall DAPI-A signal strength due to equilibrium binding of DAPI to DNA. This will not affect the cell cycle analysis, but it may be necessary to adjust the right border of the single cell gate sample by sample to account for these shifts.
4. Double-click on the SingleCells gate to open a new plot window and change the axes to show the DAPI-A histogram. Use the Bisector tool to distinguish single cells with normal DNA content (CellCycle population) from apoptotic cells with degraded DNA (subG1 population). Subsequently select the new gates in the browser and press Ctrl + R to rename them accordingly (Figure 2, plot 3).
5. Select the CellCycle gate in the browser and choose Tool | Biology | Cell Cycle… from the menu bar to open the cell cycle modelling tool. Choose Dean-Jett-Fox25,26 in the Model section to estimate the frequency of cells in G1, S and G2/M phase (Figure 2, plot 4). Use constraints only in case of poor modelling performance (minimize the Root Mean Square deviation between model and data).
6. Create gates for G1 (Ellipse tool), S (Polygon tool) and G2 + M (Ellipse tool) phase in the DAPI-W vs. DAPI-A plot of the CellCycle population to enable cell cycle-specific γH2AX measurement (Figure 2, plot 5). Do not use the modelling tool for automatic cell cycle gating based on the DAPI-A histogram, as this can be inaccurate.
NOTE: It may be necessary to move the gates along the DAPI-A axis sample by sample to account for the aforementioned shifts in overall DAPI signal strength. Only move the three gates as a group and do not change the shape of individual gates to avoid bias.
7. Use the Bisector tool to distinguish phospho-histone H3-positive (M) and -negative (M-) cells in the Alexa555-A histogram of the CellCycle population (Figure 2, plot 6). Hold Ctrl and select the gates G2 + M and M-. Press Ctrl + Shift + A to create the G2 (G2+M & M-) gate.
8. Use the Bisector tool to distinguish caspase-3-positive (Casp3+) and -negative (Casp3-) cells in the Alexa647-A histogram of the SingleCells population (Figure 2, plot 7). Set the threshold such that the average of the Casp3+ population in the untreated controls amounts to ~0.8% to assure high sensitivity and minimize assay-to-assay variations.
9. Press Ctrl + T to open the Table Editor and configure it according to Figure 3. Drag and drop the different populations from the sample browser into the Table Editor and double-click on the rows to change the statistic, parameter and name settings. Remove the unnecessary rows that are automatically added after the drag and drop of the cell cycle modelling icon by selecting the rows and pressing Del.
10. Choose To File, the format and destination in the Output section of the menu ribbon and click Create Table to export the data as ‘.xlsx’ file.
2. Use table calculation software (see Table of Materials) for further data analysis according to Supplementary File 1 (FACS_Analysis_Template_(1).xlsx).
1. Correct the frequencies of cells in the different cell cycle phases such that their sum amounts to 100% by applying the formula X’ = X * 100 / Σ(all cell cycle phases), with X’: corrected value, X: raw value, to each cell cycle phase.
NOTE: Deviations from a sum of 100% occur due to inaccuracies in the cell cycle modelling, but are usually small (<5%).

2. Normalize the median γH2AX intensities to the DNA content in the different cell cycle phases by dividing the values in S phase by 1.5 and in G2 and M phase by 2.0.
3. To calculate the combined normalized γH2AX level in the whole cell population, use the formula IA = IG1 * G1 + IS * S + IG2 * G2 + IM * M, where IA, IG1, IS, IG2, IM are normalized median γH2AX intensities of all, G1, S, G2, M cells respectively and G1, S, G2, M are corrected frequency of cells in the respective cell cycle phase.
4. For the normalized γH2AX levels and the frequency of subG1 and caspase-3-positive cells, subtract the average value of the untreated controls from each sample.
5. Calculate the mean and standard deviation of each parameter from all replicate samples and plot the results into diagrams.

Representative Results

Human U87 or LN229 glioblastoma cells were irradiated with 4 Gy of photon or carbon ion radiation. Cell cycle-specific γH2AX levels and apoptosis were measured at different time points up to 48 h after irradiation using the flow cytometric method presented here (Figure 3). In both cell lines, carbon ions induced higher γH2AX peak levels that declined slower and remained significantly elevated at 24 to 48 h compared to photon radiation at the same physical dose (Figure 4A). This indicated that carbon ions induced higher peak levels of DNA double-strand breaks (DSBs) than photons that were repaired less efficiently. For both radiation types, the γH2AX levels were highest in G1 cells, probably because DSB repair is limited to the pathway of non-homologous end-joining at this stage of the cell cycle. In line with the higher DSB induction rate and slower repair kinetics, carbon ions induced a stronger and longer-lasting cell cycle arrest in G2 phase (Figure 3B) and a higher rate of apoptosis than photons (Figure 4C). Carbon ion radiation may therefore help to overcome radioresistance mechanisms observed in glioblastoma upon classical radiotherapy with photons (see Lopez Perez, et al.1 for detailed discussion).

The results shown in Figure 4 are examples of an optimal outcome of this method, showing clear differences between untreated and irradiated cells and statistically significant differential effects between different treatments. In cases where induction of DSBs and apoptosis is less clear, it is important to include positive controls. As shown here, cells fixed 1 h after photon irradiation are good positive controls for γH2AX induction. Photon radiation is also known to efficiently induce apoptosis in lymphocytes, which makes them useful as positive controls for apoptosis 2-4 days after irradiation. Another possibility is to use drugs for apoptosis induction, such as the proteasome inhibitor MG132 or a death receptor ligand (Figure 5).
Another important aspect for an optimal outcome of the assay is the quality of the cell cycle profiles. In Figure 4B the G1 and G2 peaks were narrow and clearly separated, making it easy to apply cell cycle modelling and to define the gates for cell-cycle-specific measurement of γH2AX. Depending on the cell type and the strength of the treatment effect (Figure 6A,B), the resolution of the different cell cycle phases can be significantly worse. In some cases, the DAPI concentration has a big impact on the quality of the cell cycle profile and needs to be adjusted (Figure 6C). In cases where no clear G1 peak is detectable due to the treatment, the cell cycle modelling tool cannot be applied accurately. It is advisable then to only use manual gating in the DAPI-A versus DAPI-W plot based on controls with a well-defined cell cycle profile, as described in the protocol.

The analysis of a cell cycle profile without a clear G1 peak can be further complicated if the treatment leads to low cell numbers that result in large shifts in the overall DAPI signal strength. In this situation, it can be difficult to judge, if a peak is the G1 or the G2 peak. To avoid this
problem, the cells can be counted with a Neubauer chamber or by other means after the last washing step (step 3.3) and the cell numbers can be adjusted. The peak positions in the treated cells will then match with the controls. Drug-mediated apoptosis-induction in U87 glioblastoma cells. U87 cells were treated with 5 ng/mL of a modified TNF-related apoptosis-inducing ligand (see Table of Materials) plus 2.5 µM MG132 for 20 h before fixation to validate the apoptosis measurement by active caspase-3 and subG1 analysis. (A) Histogram of the caspase-3 signal of treated versus untreated cells. (B) DAPI histogram of treated cells with a subG1 population, indicating DNA degradation. The histogram of caspase-3-positive events is overlaid in red, demonstrating that most of the apoptotic cells had not yet degraded their DNA at this time point. Please click here to view a larger version of this figure. Factors influencing the quality of cell cycle profiles. Too harsh treatments complicate cell cycle analysis. (A) No G1 peak is detectable 48 h after treatment of LLC cells (Lewis lung carcinoma) with 20 Gy of photon radiation or (B) 96 h after treatment of HS68 fibroblasts with 100 mJ of UV radiation. (C) DAPI concentration in different cell lines: In HS68 fibroblasts (upper panel) the G2/M peak (see arrows) was equally well-separated from the G1 peak for DAPI concentrations between 0.01 and 1.0 µg/mL. In contrast, DAPI concentrations below 0.75 µg/ mL resulted in poor separation and shape definition of the G2/M peak (see arrows) in MRC-5 fibroblasts (lower panel). Please click here to view a larger version of this figure.

Discussion

The featured method is easy to use and offers a fast, accurate and reproducible measurement of the DNA damage response including double- strand break (DSB) induction and repair, cell cycle effects and apoptotic cell death. The combination of these endpoints provides a more complete picture of their interrelations than individual assays. The method can be widely applied as a comprehensive genotoxicity assay in the fields of radiation biology, therapy and protection, and more generally in oncology (e.g., for environmental risk factor assessment, drug screening and evaluation of genetic instability in tumor cells). The most critical step in this protocol is the fixation of the cells. To obtain a clean sample with a clearly defined cell population and optimal resolution of different cell cycle phases, it is important to give adherent cells enough time to detach from the culture vessel and form a completely round shape. The cells should be transferred into the fixation solution as a single cell suspension without cell clumps. To separate them, the cells must be pipetted up and down or passed through a fine needle several times in difficult cases. The fixation time should be kept constant between all samples and should not exceed 20 min including the time for centrifugation before resuspension of the cells in 70% ethanol. Prolonged fixation will lead to loss of accessible epitopes for antibody binding and will decrease the signal strength and sensitivity of the method.

The main limitation of the technique is that it does not provide information on the quality of the γH2AX foci other than the intensity in the whole nucleus. Particularly large γH2AX foci as well as the occurrence of a pan-nuclear γH2AX staining have been linked to clustered DSBs1,27,28,29, which are highly complex lesions that are very difficult to repair30. Such clustered lesions seem to be characteristic for particle radiation such as clinically applied carbon ion radiation and may be the reason for the superior biological effectiveness of heavy ion radiation compared to photons31,32,33. Analysis of the γH2AX foci size may therefore be of interest as an indicator of the damage complexity. Although it is possible to use the present protocol with an image stream cytometer to visualize the γH2AX foci, the achievable resolution cannot compete with a classical microscopic approach. However, we have successfully prepared microscopic slides from the remaining samples after flow cytometric measurement (Supplementary Figure 1) and evaluated several features including γH2AX foci count, size and pan-nuclear intensity using a semi-automated imaging and analysis system. To prepare the samples for microscopy, 30-50 µL of the stained cells were spread over the surface of a 24 mm x 24 mm cover glass, air dried overnight at room temperature in the dark and then embedded with mounting medium on a glass slide.

In comparison with microscopic evaluation of DSB levels by γH2AX foci counting, the flow cytometric measurement is superior in throughput, sampling rate, dynamic range and accuracy. The dynamic range of microscopic foci counting is limited by the optical resolution, leading to the inability to distinguish overlapping foci29,34. In practice, we have observed a linear relationship between radiation dose and γH2AX foci numbers below 2 Gy, and a strong saturation effect above 2 Gy in U87 glioblastoma cells1. In contrast, the intensity of the γH2AX signal measured by flow cytometry increased linearly with dose for the whole dose range tested (0-8 Gy). The accuracy of microscopic foci counting is limited by the inability to distinguish between different cell cycle phases without additional stainings35. The number of DSBs induced in a cell is not only dependent on the radiation dose, but also on the DNA content, and hence the cell cycle stage. G2 cells for example have twice as much DNA as G1 cells and the average number of DSBs induced at a certain radiation dose is twice as high for G2 cells compared to G1. This is clearly reflected in the γH2AX signal intensity of G2 versus G1 cells at early time points after radiation measured by flow cytometry. Therefore, it is important to evaluate DSB levels in a cell cycle-specific manner to obtain accurate results. A pooled analysis of cells in different cell cycle phases, like usual in microscopic approaches, can be biased by shifts in the cell cycle distribution particularly due to radiation-induced G2 arrest. To account for this effect and to compare DSB repair characteristics between different cell cycle stages, the γH2AX levels can easily be normalized to the DNA content in the flow cytometry method as indicated in the protocol.

Extensions of the present protocol are possible with relatively little extra effort. Additional antibodies with compatible fluorophores labels can be included in step 3.4 without the need for any extra steps during sample preparation. If a more detailed analysis of apoptosis is desired, caspase-7, -9 and -8 antibodies could be included. As another extension, we have recently included a live and dead cell dye, a troponin T antibody and counting beads in the protocol to specifically analyze cardiomyocytes co-cultured with fibroblasts after irradiation. The addition of the live/dead cell stain and the counting beads expands the capability of the method to include fibroblast proliferation and non-apoptotic cell
death of the cardiomyocytes as further endpoints that provide useful additional information on the cell fate after genotoxic stress. The extended protocol is available upon request.

Disclosures
The authors have nothing to disclose.

Acknowledgments
We thank the Flow Cytometry Facility team at the German Cancer Research Center (DKFZ) for their support.

References

1. Jensen, A., et al. Treatment of non-small cell lung cancer with intensity-modulated radiation therapy in combination with cetuximab: the NEAR protocol (NCT00115518). BMC Cancer. 6, 122, (2006).
2. Oertel, S., et al. Human Glioblastoma and Carcinoma Xenograft Tumors Treated by Combined Radiation and Imatinib (Gleevec).
Strahlentherapie und Onkologie. 182 (7), 400-407, (2006).
3. Timke, C., et al. Combination of Vascular Endothelial Growth Factor Receptor/Platelet-Derived Growth Factor Receptor Inhibition Markedly Improves Radiation Tumor Therapy. Clinical Cancer Research. 14 (7), 2210-2219, (2008).
4. Zhang, M., et al. Trimodal glioblastoma treatment consisting of concurrent radiotherapy, temozolomide, and the novel TGF-β receptor I kinase inhibitor LY2109761. Neoplasia. 13 (6), 537-49 (2011).
5. Blattmann, C., et al. Suberoylanilide hydroxamic acid affects γH2AX expression in osteosarcoma, atypical teratoid rhabdoid tumor and normal tissue cell lines after irradiation. Strahlentherapie und Onkologie. 188 (2), 168-176, (2012).
6. Hoeijmakers, J.H. DNA Damage, Aging, and Cancer. New England Journal of Medicine. 361 (15), 1475-1485, (2015).
7. Rodgers, K., McVey, M. Error-Prone Repair of DNA Double-Strand Breaks. Journal of Cellular Physiology. 231 (1), 15-24, (2016).
8. Ciccia, A., Elledge, S.J. The DNA Damage Response: Making It Safe to Play with Knives. Molecular Cell. 40 (2), 179-204, (2010).
9. Rothkamm, K., Krüger, I., Thompson, L.H., Löbrich, M. Pathways of DNA double-strand break repair during the mammalian cell cycle. Molecular and Cellular Biology. 23 (16), 5706-15, (2003).
10. Escribano-Díaz, C., et al. A Cell Cycle-Dependent Regulatory Circuit Composed of 53BP1-RIF1 and BRCA1-CtIP Controls DNA Repair Pathway Choice. Molecular Cell. 49 (5), 872-883, (2013).
11. Bakr, A., et al. Functional crosstalk between DNA damage response proteins 53BP1 and BRCA1 regulates double strand break repair choice. Radiotherapy and Oncology. 119 (2), 276-81, (2015).
12. Mladenov, E., Magin, S., Soni, A., Iliakis, G. DNA double-strand-break repair in higher eukaryotes and its role in genomic instability and cancer: Cell cycle and proliferation-dependent regulation. Seminars in Cancer Biology. 37-38, 51-64, (2016).
13. Lopez Perez, R., et al. DNA damage response of clinical carbon ion versus photon radiation in human glioblastoma cells. Radiotherapy and Oncology. 133, 77-86, (2019).
14. Oertel, S., et al. Combination of suberoylanilide hydroxamic acid with heavy ion therapy shows promising effects in infantile sarcoma cell lines. Radiation oncology. 6 (1), 119, (2011).
15. Blattmann, C., et al. Suberoylanilide hydroxamic acid affects γH2AX expression in osteosarcoma, atypical teratoid rhabdoid tumor and normal tissue cell lines after irradiation. Strahlentherapie und Onkologie. 188 (2), 168-176, (2012).
16. Thiemann, M., et al. In vivo efficacy of the histone deacetylase inhibitor suberoylanilide hydroxamic acid in combination with radiotherapy in a malignant rhabdoid tumor mouse model. Radiation Oncology. 7 (1), 52, (2012).
17. Nicolay, N.H., et al. Mesenchymal stem cells retain their defining stem cell characteristics after exposure to ionizing radiation. International Journal of Radiation Oncology Biology Physics. 87 (5), 1171-1178, (2013).
18. Nicolay, N.H., et al. Mesenchymal stem cells are resistant to carbon ion radiotherapy. Oncotarget. 6 (4), 2076-87 (2015).
19. Nicolay, N.H., et al. Mesenchymal stem cells exhibit resistance to topoisomerase inhibition. Cancer letters. 374 (1), 75-84, (2016).
20. Nicolay, N.H., et al. Mesenchymal stem cells maintain their defining stem cell characteristics after treatment with cisplatin. Scientific Reports.
6 (January), 20035, (2016).
21. Nicolay, N.H., et al. Mesenchymal stem cells are sensitive to bleomycin treatment. Scientific Reports. 6 (May), 26645, (2016).
22. Rühle, A., et al. Cisplatin radiosensitizes radioresistant human mesenchymal stem cells. Oncotarget. 8 (50), 87809-87820 (2017).
23. Münz, F., et al. Human mesenchymal stem cells lose their functional properties after paclitaxel treatment. Scientific Reports. 8 (1), 312, (2018).
24. Rühle, A., et al. The Radiation Resistance of Human Multipotent Mesenchymal Stromal Cells Is Independent of Their Tissue of Origin.
International Journal of Radiation Oncology Biology Physics. 100 (5), 1259-1269, (2018).
25. Dean, P.N., Jett, J.H. Mathematical analysis of DNA distributions derived from flow microfluorometry. Journal of Cell Biology. 60 (2), 523-527 (1974).
26. Fox, M.H. A model for the computer analysis of synchronous DNA distributions obtained by flow cytometry. Cytometry. 1 (1), 71-77, (1980).
27. Costes, S. V, Boissière, A., Ravani, S., Romano, R., Parvin, B., Barcellos-Hoff, M.H. Imaging features that discriminate between foci induced by high- and low-LET radiation in human fibroblasts. Radiation Research. 165 (5), 505-515, (2006).
28. Meyer, B., Voss, K.-O., Tobias, F., Jakob, B., Durante, M., Taucher-Scholz, G. Clustered DNA damage induces pan-nuclear H2AX phosphorylation mediated by ATM and DNA-PK. Nucleic Acids Research. 41 (12), 6109-6118, (2013).
29. Lopez Perez, R., et al. Superresolution light microscopy shows nanostructure of carbon ion radiation-induced DNA double-strand break repair foci. FASEB. 30 (8), 2767-2776, (2016).
30. Schipler, A., Iliakis, G. DNA double-strand-break complexity levels and their possible contributions to the probability for error-prone processing and repair pathway choice. Nucleic Acids Research. 41 (16), 7589-7605, (2013).
31. Stenerlow, B., Hoglund, E., Carlsson, J. DNA fragmentation by charged particle tracks. Advances in Space Research. 30 (4), 859-863 (2002).
32. Friedland, W., et al. Comprehensive track-structure based evaluation of DNA damage by light ions from radiotherapy-relevant energies down to stopping. Scientific Reports. 7, 45161, (2017).
33. Pang, D., Chasovskikh, S., Rodgers, J.E., Dritschilo, A. Short DNA Fragments Are a Hallmark of Heavy Charged-Particle Irradiation and May Underlie Their Greater Therapeutic Efficacy. Frontiers in Oncology. 6, 130, (2016).
34. Böcker, W., Iliakis, G. Computational Methods for analysis of foci: validation for radiation-induced gamma-H2AX foci in human cells. Radiation Research. 165 (1), 113-124 (2006).
35. Löbrich, M., et al. γH2AX foci analysis for monitoring DNA BOS172722 double-strand break repair: Strengths, limitations and optimization. Cell Cycle. 9
(4), 662-669, (2010).