A protocol to evaluate changes in DNA damage levels and DNA repair capacity that may be induced by chronic in vivo low dose irradiation in mouse spleen lymphocytes, by measuring phosphorylated histone H2AX, a marker of DNA double-strand breaks, using flow cytometry is presented.
Low dose radiation exposure may produce a variety of biological effects that are different in quantity and quality from the effects produced by high radiation doses. Addressing questions related to environmental, occupational and public health safety in a proper and scientifically justified manner heavily relies on the ability to accurately measure the biological effects of low dose pollutants, such as ionizing radiation and chemical substances. DNA damage and repair are the most important early indicators of health risks due to their potential long term consequences, such as cancer. Here we describe a protocol to study the effect of chronic in vivo exposure to low doses of γ- and β-radiation on DNA damage and repair in mouse spleen cells. Using a commonly accepted marker of DNA double-strand breaks, phosphorylated histone H2AX called γH2AX, we demonstrate how it can be used to evaluate not only the levels of DNA damage, but also changes in the DNA repair capacity potentially produced by low dose in vivo exposures. Flow cytometry allows fast, accurate and reliable measurement of immunofluorescently labeled γH2AX in a large number of samples. DNA double-strand break repair can be evaluated by exposing extracted splenocytes to a challenging dose of 2 Gy to produce a sufficient number of DNA breaks to trigger repair and by measuring the induced (1 hr post-irradiation) and residual DNA damage (24 hrs post-irradiation). Residual DNA damage would be indicative of incomplete repair and the risk of long-term genomic instability and cancer. Combined with other assays and end-points that can easily be measured in such in vivo studies (e.g., chromosomal aberrations, micronuclei frequencies in bone marrow reticulocytes, gene expression, etc.), this approach allows an accurate and contextual evaluation of the biological effects of low level stressors.
Significant controversy over the potential harmful effects of either low or very low doses of ionizing radiation and public’s fear of radiation, driven by images of the Hiroshima and Nagasaki atomic bombings and of rare nuclear plant accidents (exacerbated by mass media), has led to very strict radiation protection regulation and standards that are potentially not scientifically justified. In the last three decades, numerous reports have documented both the lack of harmful and the presence of potentially beneficial biological effects induced by low dose radiation1-4. The major radiation health risk factor is the probability of cancer, estimated based on epidemiological studies of atomic bomb survivors receiving high or medium doses of radiation. Linear extrapolation of these data (so called linear-no-threshold or LNT model) is used to estimate risks of cancer at low doses. However, this approach has not received world-wide scientific acceptance and is heavily debated5.
It is evident that more studies are required to clarify this issue and possibly improve radiation protection standards. Such studies should involve chronic treatments (best approximation of environmental and occupational exposures), in vivo animal models (best for extrapolating effects to human) and such end-points as DNA damage rates, DNA repair and mutagenesis. It is known that DNA is the primary target for damaging radiation effects and incomplete or mis-repair may lead to mutagenesis and cancer development6.
DNA double-strand breaks (DSB) are one of the most deleterious types of DNA lesions and may lead to cell death and tumorigenesis7. It is, therefore, important to be able to reliably and accurately measure the level of DSB after exposure to low dose radiation and/or other stressors, such as chemical pollutants. One of the most sensitive and specific markers of DNA DSB is phosphorylated histone H2AX, called γH2AX8, yet other markers and methods have been suggested9,10. It is estimated that thousands of H2AX molecules, in the vicinity of an induced DSB, are involved in the formation of γH2AX enabling the detection of individual DSB by immunofluorescent labeling with an anti-γH2AX antibody and fluorescence microscopy11. The response is very quick, reaching its maximum between 30 and 60 min. Evidence exists that γH2AX facilitates repair of DNA DSB by attracting other repair factors to the sites of breaks and by modifying chromatin structure to anchor broken DNA ends and provide access for other repair proteins (reviewed in 12). Upon completion of repair of DNA DSB, γH2AX gets de-phosphorylated and/or undergoes degradation, and newly synthesized H2AX molecules replace γH2AX in the affected areas of chromatin10. Monitoring formation and loss of γH2AX can, therefore, provide an accurate estimate of DNA DSB repair kinetics. This approach has been used to study repair of DSB in various human tumor cell lines irradiated with high doses of radiation and its rate and residual DSB levels have been shown to correlate with radiosensitivity13-15.
We modified this experimental approach and applied it to an in vivo mouse study to examine the effects of low doses of chronic γ- and β-irradiation on DNA DSB levels and repair (Figure 1). Firstly, we demonstrate a method to perform a long-term chronic exposure of mice to β-radiation either emitted by tritium (hydrogen-3) in the form of tritiated water (HTO) or as organically bound tritium (OBT) dissolved in drinking water. The two forms are expected to accumulate and/or distribute differently in the body and therefore, produce different biological effects. Both forms are potential hazards in nuclear industry. This treatment is paralleled by chronic exposure to γ-radiation at an equivalent dose rate to allow a correct comparison of the two radiation types, which is crucial for the evaluation of their relative biological effectiveness. Beta-radiation is composed of electrons, making it very different from the γ-radiation, high energy photons. Due to this difference, Β-radiation represents mostly internal health hazard and may produce different biological effects compared to γ-radiation. This complication resulted in significant controversies over the regulation of exposure to β-radiation emitted by HTO. Thus, regulatory levels of HTO in drinking water for the public vary from 100 Bq/L in Europe to 75,000 Bq/L in Australia. It is, therefore, important to compare biological effects of HTO to equivalent doses of γ-radiation. Secondly, the rate of DNA DSB is measured in isolated splenocytes upon the completion of chronic exposures using immunofluorescently labeled γH2AX detected by flow cytometry. This allows the evaluation of the extent of DNA damage inflicted by the in vivo exposures. However, it is sensible to expect that such low level exposures may not produce any detectable rates of DNA DSB; instead, some hidden changes/responses may be expected that would affect the cells’ ability to repair DNA damage. These changes, if found, may be either stimulatory (producing a beneficial effect) or inhibitory (producing a deleterious effect). The proposed protocol allows revealing such changes by challenging the extracted splenocytes with a high dose of radiation that produces a significant amount of damage (e.g., 2 Gy producing approximately 50 DNA DSB per cell or a 3 – 5 fold increase in total γH2AX level). Subsequently, the formation and loss of γH2AX, reflecting the initiation and completion of DSB repair, is monitored by flow cytometry. In this way, not only basal and treatment-induced levels of DNA DSB can be measured, but also its potential effect on the cells’ ability to respond and repair DNA damage induced by much higher stressor levels.
All mouse handling and treatment procedures should adhere to rules set forth by the legislature and/or animal care programs and approved by a local animal care committee. All methods described in this protocol were performed in accordance with the guidelines of the Canadian Council on Animal Care with the approval of the local animal care committee.
CAUTION: All work with radioactivity (including, but not limited to handling of tritium, external γ-irradiation, handling radioactive animal tissues, bedding waste) should adhere to rules set forth by the legislature and/or radiation protection authorities and be performed by authorized personnel in a certified laboratory and/or facility.
1. Animal Work
2. Sacrifices and Sampling
3. Preparation of Splenocyte Cultures
4. Challenging Irradiation and Fixing
5. Immunofluorescent Labeling with Anti-γH2AX Antibody
Note: Typically, 15 – 30 samples per day can be immunofluorescently labeled and assayed by flow cytometry in a single run. To ensure an accurate comparison between treatment groups, include sample(s) from each group. See also Reference 16 for further details.
Figure 2 shows examples of flow cytometry graphs expected for splenocytes prepared using the method described here. Cells are first gated based on the <electronic volume-side scatter> scatter plot (Figure 2A and 2D; electronic volume is equivalent to forward scatter). FL3/propidium iodine histograms (Figure 2B and 2E) confirm normal cell cycle distribution. Mean γH2AX signal calculated using the FL1 channel confirms a > 2-fold increase in DNA DSB in 2-Gy irradiated cells (Figure 2F) compared to the untreated control (Figure 2C). Next, we examined the sensitivity of the described method of measuring immunofluorescently labeled γH2AX by flow cytometry in mouse splenocytes. Figure 3A shows γH2AX levels, indicative of DNA DSB, in mouse splenocytes 1 hr after the ex vivo exposure to either low (0.1 Gy) or high (2 Gy) doses of γ-radiation relative to the untreated control. It can be seen that 0.1 Gy induced a slight increase in the DSB level and that increase was statistically significant. The strong 3-fold induction was expected and detected after the 2 Gy exposure. To validate the specificity of γH2AX immunofluorescent labeling as described in the protocol, labeled cells were examined with a fluorescence microscope. The marked foci-like pattern observed indicates that the fluorescence signal originates from the individual DNA DSB within the chromatin (Figure 3B).
Representative results of the assessment of DNA DSB rate and their repair in splenocytes of mice exposed for one month to very low concentrations of tritiated water (10 kBq/L) are presented in Figure 4. It can be seen (Figure 4A) that although the treatment resulted in a 10% increase of the basal γH2AX level, the change was not statistically significant (N = 5, one-sample Student’s t test). Furthermore, the measured formation and loss of γH2AX after the challenging 2 Gy irradiation, representing the kinetics of DNA DSB repair, was not different between the cells from the control and HTO-treated mice (Figure 4B).
Figure 1. Experimental diagram to evaluate the effect of chronic in vivo irradiation on DNA damage and DNA repair. After in vivo treatment of mice with chronic irradiation for a desired period of time, mice are sacrificed and splenocyte cultures are established and exposed to a challenging irradiation. Splenocyte aliquots are collected and fixed at various times after the challenging dose. Levels of γH2AX are then measured using immunofluorescent labeling and detection by flow cytometry. The plot on the bottom of the Figure is a schematic of expected data. Gamma-H2AX level drastically increases early after the challenging irradiation followed by a decrease back to the control value, if DNA repair is complete. DNA damage Δ represents the damage produced due to the experimental treatment of mice and DNA repair Δ represents the effect of the treatment of mice on DNA DSB repair capacity.
Please click here to view a larger version of this figure.
Figure 2. Representative flow cytometry raw data. (A, D) Scatter plots and gating regions for splenocytes and cell debris and erythrocytes. (B, E) DNA histograms representing cell cycle distributions. (C, F) γ-H2AX fluorescence histogram and mean fluorescence values. Data in sets (A-C) were obtained from a representative untreated control spleen cells, whereas data in sets (D-F) were obtained from cells irradiated with 2 Gy γ-radiation and harvested 1 hr after the irradiation.
Please click here to view a larger version of this figure.
Figure 3. Detection of γH2AX by flow cytometry is sensitive and specific. (A) Splenocytes freshly isolated from male CBA mice were irradiated with either 0.1 or 2 Gy of γ-radiation, or sham-treated and allowed to develop γH2AX response for 1 hr. Cells were then fixed, immunofluorescently labeled with anti-γH2AX antibody and analyzed by flow cytometry. Normalized mean values ± SD are shown (N = 12). * and *** signify statistical difference with p < 0.05 and p < 0.001, respectfully (one-sample Student’s t-test). (B) Representative microphotographs of untreated control (UT) or 2 Gy irradiated splenocytes immunofluorescently labeled with anti-γH2AX antibody. The foci-like pattern of green fluorescence confirms the specificity of the immunofluorescent labeling for γH2AX.
Please click here to view a larger version of this figure.
Figure 4. Representative results showing the lack of effect produced by 1-month exposure of mice to tritiated water (HTO) on DNA DSB level and repair. (A) Splenocytes were isolated from mice treated with 10 kBq/L of HTO in drinking water and control animals. The level of γH2AX was measured using immunofluorescent labeling and flow cytometry as described in the protocol. Mean γH2AX fluorescence levels relative to those of control ± SD are shown (N = 5). These data were generated from the t = 0 time-point that was used in Figure 3B for the evaluation of DNA DSB repair. (B) Isolated splenocytes were exposed to a challenging 2 Gy dose of γ-radiation and aliquots of cells were fixed at times t = 0, 1 and 24 hrs after the challenging irradiation. Mean fluorescence of γH2AX relative to its own t = 0 control is shown ± SD (N = 5).
Please click here to view a larger version of this figure.
The protocol presented in this paper is useful for conducting large scale mouse in vivo studies examining the genotoxic effects of low levels of various chemical and physical agents, including ionizing radiation. Our unique specific pathogen-free animal facility equipped with a GammaBeam 150 irradiator and a 30 meter long irradiation hall allows conducting life-long studies involving low or very low dose rate irradiations on hundreds or even thousands of mice. Smaller irradiation facilities for chronic low dose irradiation can be set up in a university or hospital environment where radiobiology studies are conducted routinely and radiation protection guidelines/regulations are implemented. However, stricter regulations normally apply to facilities designed for internal animal irradiation where handling of radionuclides, such as tritium, is carried out. Although rare, laboratories certified for such internal radiation exposure studies exist in various research institutions world-wide and their unique capabilities can be used for studies described in this protocol.
Within the protocol, we use the immunofluorescent labeling of γH2AX to accurately evaluate DNA DSB rates produced by experimental exposures. More importantly, the introduction of a challenging irradiation delivered ex vivo to extracted splenocytes makes it possible to evaluate the intactness of the DNA DSB response (γH2AX level at 1 hr after the challenge) and of DNA DSB repair (γH2AX level at 24 hr compared to 0 hr). It should be noted that quantification of γH2AX foci per cell by fluorescence microscopy may provide higher sensitivity compared to flow cytometry. For example, doses as low as 0.001 Gy were reported to produce significant increases in the number of γH2AX foci per cell. However, the use of flow cytometry-based detection of γH2AX provides advantages over fluorescence microscopy analysis because it requires less time for measurements, it does not require highly specialized/trained personnel and it avoids subjective decision-making related to γH2AX foci recognition and counting. These advantages are critical for large scale animal studies. A typical animal experiment with 10 treatment groups and 10 mice per group will result in 100 x 3 = 300 spleen samples, if the method suggested (Figure 1) is followed. To generate data from this number of samples, approximately 100 hrs will be required using flow cytometry. Alternatively, if done by manual microscopic analyses, the same task would take at least 300 hrs, using a conservative estimate of 40 min of highly demanding microscope time per sample. Although attempts to automate analysis and scoring of γH2AX foci have been made17-19, they all have a number of limitations. In particular, the round shape and small size of mouse splenocytes and lymphocytes makes it impractical to resolve all individual γH2AX foci microscopically. While adherent cells have been successfully scored by automated microscopy systems for γH2AX foci, no success has been demonstrated for either mouse splenocytes or lymphocytes.
The best way to eliminate day-to-day variability in the measured values of γH2AX signal by flow cytometry is to include experimental group and control samples in the same run and to calculate relative γH2AX changes within every run. Additional recommendations include: a) a historical control consisting of untreated and 2 Gy irradiated sample is used in every run; b) one person performs all the antibody labeling procedures; c) a single stock of buffers and a single batch of antibody, both primary and secondary, is used for the entire set of samples within one experiment. It is also vital to perform a flow cytometer quality control test recommended by the instrument manufacturer routinely and prior to each run of samples. This is required to verify the instrument’s optical alignment and fluidics system and also helps minimizing day-to-day variation in measured γH2AX fluorescence between samples to be compared or pooled in one group.
It should be pointed out that the use of a larger number of time-points (e.g., 1, 2, 4, 6 hrs) post challenging irradiation may help analyze and measure short term DNA DSB repair kinetics in greater detail16. However, overall repair potential and the likeliness of delayed health effects would be represented by the residual damage measured only at 24 hrs and compared to the one observed at the initial time point (t = 0)13-15 (Figure 4).
Since basal γH2AX levels are known to increase in old animals or senescent cells 20, for more accurate interpretation of the results obtained in long-term exposure mouse studies (> 6 months) using the described method, the use of a young untreated group, in addition to an age-matched control, would be advisable. Although some reports indicate that γH2AX may increase in proliferating cells, which would not be related to DNA DSB, this can easily be controlled by measuring S-phase and G2/M-phase cell fractions using DNA histograms (Figure 2A, D). If certain experimental conditions still prevent a researcher from using γH2AX as a marker of DNA damage (due to the reasons outlined above or their combination), another protein forming DNA damage foci, such as 53BP1, can be used 21.
Although we have not attempted using this method for non-lymphoid tissues, such as muscle, heart, brain and others, it would seem difficult to process them for flow cytometry. Tissue sections and microscopy for γH2AX foci counting would be the method of choice, if analysis of such tissues is required. However, other lymphoid cells, such as peripheral blood lymphocytes, bone marrow lymphocytes or thymocytes, can be used for measuring γH2AX levels by the described method. The advantage of using spleen lymphocytes is the relative simplicity of cell suspension preparation and the large number of cells that can be isolated from a spleen. This becomes even more advantageous if additional end-points are included in the study. In our experience, a typical yield of splenocytes per mouse allows for the collection of at least 10 aliquots, each suitable for any of the following assays: flow cytometry, RT-qPCR for gene or miRNA expression, western blot, genomic DNA for CpG methylation assay. Thus, additional splenocyte aliquots collected in parallel to those for the γH2AX end-point may be used to measure the expression of genes involved in DNA damage signaling and repair responses. Such genes as GADD45A and CDKN1A, that are normally transcriptionally up-regulated in response to cytotoxic treatments22, may provide a clue as to whether cells are properly responding to the challenging stress influence. Similarly, protein levels and post-translational modifications involved in stress responses may be assessed. Finally, it is not expected that any of the steps of the extraction procedure, if performed as described here, can induce γH2AX in splenocytes.
It would be beneficial and highly recommended to sample additional organs and tissues during sacrifices to supplement DNA damage and repair data obtained for splenocytes as per presented protocol with other end-points. For example, we routinely collect bone marrow cells for the in vivo bone marrow micronucleus assay, various tissues for senescence-associated β-galactosidase and DNA repair gene expression measurements. The ability to collect other tissues would depend on the available technical support during sacrifices; however, it can significantly help in the proper interpretation of results and the construction of a much more biologically relevant systemic picture of the response to the low level treatment being investigated.
The authors have nothing to disclose.
The authors would like to acknowledge the contribution of our colleagues Sandrine Roch-Lefevre and Eric Gregoire of the Institute of Radioprotection and Nuclear Safety (Paris, France). This work was supported by the Government of Canada Science and Technology program at Canadian Nuclear Laboratories (Chalk River, Ontario, Canada), the CANDU Owners Group (Toronto, Ontario, Canada), the Canadian Nuclear Safety Commission and by the Institute of Radioprotection and Nuclear Safety (Paris, France).
HTO: tritiated water, [3H] | locally obtained from a nuclear reactor | stock activity 3.7 GBq/mL; can be substituted with HTO from Perkin Elmer | |
OBT: Alanine, L-[3-3H]: organically bound tritium (OBT) | Perkin Elmer | NET348005MC | 1mCi/mL (185 MBq) |
OBT: Glycine, [2-3H]: organically bound tritium | Perkin Elmer | NET004005MC | 1mCi/mL (185 MBq) |
OBT: Proline, L-[2,3-3H]: organically bound tritium (OBT) | Perkin Elmer | NET323005MC | 1mCi/mL (185 MBq) |
tritiated water, [3H] (HTO) | Perkin Elmer | NET001B005MC | substitute for HTO of local origin |
GammaBeam 150 irradiator | Atomic Energy of Canada Limited | locally manufactured | can be substituted with another g-radiation source of sufficiently low activity |
tween-20 | Sigma Aldrich | P1379-500ML | |
RPMI | Fisher Scientific | SH3025501 | Hyclone RPMI 1640 with L-Glutamine and HEPES 500mL |
fetal bovine serum | Sigma Aldrich | F1051-100ML | |
anti-gH2AX antibody, clone JBW301 | Millipore | 05-636 | |
Alexa fluor-488 goat anti-mouse antibody | Life Technologies (formerly Invitrogen) | A21121 | |
propidium iodine | Sigma Aldrich | P4864-10ML | 1mg/mL |
ethanol | Commercial Alcohols | P006-EAAN | 500ml bottles Absolute Ethanol |
1.5 mL tubes | Fisher Scientific | 2682550 | microcentrifuge tubes |
15 mL tubes | Fisher Scientific | 05-539-5 | sterile polypropylene centrifuge tubes |
liquid nitrogen | Linde | P110403 | |
12 x 75 mm mL uncapped glass tubes | Fisher Scientific | K60B1496126 | disposable borosilicate glass tubes with plain end |
T25 flasks | VWR | CA15708-120 | nunc tisue culture 25ml flask (supplier no. 156340) |
scissors | Fine Science Tools | 14068-12 | Wagner scissors 12cm sharp/sharp |
forceps, straight | Fine Science Tools | 11008-13 | Semken forceps 13cm, straight |
forceps, curved | Fine Science Tools | 11003-12 | Narrow Pattern forceps 12 cm, curved |
60 mm petri dishes | VWR | CA25382-100 | BD Falcon tissue culture dish 60x15mm |
cell strainers, 70 mm | Fisher Scientific | 08-771-2 | Falcon cell strainers 50/case |
PBS recipe: 1 tablet dissolved in 200mL of deionized water, adjust pH to 7.4 if needed. | Sigma Aldrich | P4417-100TAB | Phosphate Buffered Saline Tablets |
TBS recipe: to make 10X Stock, | |||
30g TRIS HCl | Sigma Aldrich | T3253-1KG | Trizma hydrochloride |
88g NaCl | Fisher Scientific | S271-500 | Sodium Chloride |
2g KCl | Fisher Scientific | P217-500 | Potassium Chloride |
Dissolve in 1L of deionized water, adjust pH to 7.4 |