Quantitative Detection of 125IdU-Induced DNA Double-Strand Breaks ...

RADIATION RESEARCH

158, 486–492 (2002)

0033-7587/02 $5.00 q 2002 by Radiation Research Society. All rights of reproduction in any form reserved.

Quantitative Detection of

IdU-Induced DNA Double-Strand Breaks with g-H2AX Antibody

125

Olga A. Sedelnikova,a,1 Emmy P. Rogakou,a,2 Igor G. Panyutinb and William M. Bonnera a

Laboratory of Molecular Pharmacology, Center for Cancer Research, National Cancer Institute, and b Department of Nuclear Medicine, Clinical Center, National Institutes of Health, Bethesda, Maryland 20892

DNA integrity play major roles in tumorigenesis (1, 2). However, DNA DSBs do arise, accidentally from metabolic errors and environmental insults and intentionally during several important cellular functions such as V(D)J (3) and meiotic recombination (4). To maintain chromosomal integrity, a multiplicity of cellular responses are involved in rejoining a DNA double-stranded end with either the original partner or a new compatible partner (5). One of the first responses when DSBs are introduced into the DNA of eukaryotic cells is the immediate and massive phosphorylation of the serine residue in the conserved motif SQ (E/D)(I/L/Y) at the COOH terminus of histone H2AX homologues, yielding a specific modified form named gH2AX in mammals (6). g-H2AX formation is rapid; its half-maximal amounts are reached by 1–3 min postirradiation, and maximal amounts are reached by 10–30 min. At the maximum, approximately 1% of the H2AX becomes gphosphorylated per gray of ionizing radiation. A polyclonal antibody, raised in rabbits against a synthetic phosphorylated peptide containing the mammalian g-H2AX COOH terminal sequence (7), reveals that g-H2AX molecules appear in discrete nuclear foci within 1 min after exposure of cells to ionizing radiation. These observations can be accommodated in a model in which each g-H2AX focus represents one DNA DSB. It was the purpose of this study to test this prediction. To make a rigorous determination of the quantitative relationship between numbers of foci and numbers of DNA DSBs, we used 125I, an Auger-electron emitter with a very shortrange radiation effect. 125I can be incorporated into cellular DNA by using iododeoxyuridine (IdU), a 125I-labeled precursor of DNA synthesis, which is efficiently incorporated into DNA upon addition to cultured cells (8–10). It has long been established that when 125I is incorporated into DNA, it introduces close to one DSB per decay (11, 12). Detailed studies with synthetic DNA oligonucleotides demonstrated that the majority of DSBs are produced within 10 base pairs from the decay site with an efficiency of 0.8 DSB per decay (13). This is in contrast to non-Auger-electron ionizing radiation in which DNA DSBs are formed with low efficiency and damage may be more widespread and heterogeneous. The amount of 125I incorporated and hence the number of DNA DSBs can be determined from measurement of the

Sedelnikova, O. A., Rogakou, E. P., Panyutin, I. G. and Bonner, W. M. Quantitative Detection of 125IdU-Induced DNA Double-Strand Breaks with g-H2AX Antibody. Radiat. Res. 158, 486–492 (2002). When mammalian cells are exposed to ionizing radiation and other agents that introduce DSBs into DNA, histone H2AX molecules in megabase chromatin regions adjacent to the breaks become phosphorylated within minutes on a specific serine residue. An antibody to this phosphoserine motif of human H2AX (g-H2AX) demonstrates that g-H2AX molecules appear in discrete nuclear foci. To establish the quantitative relationship between the number of these foci and the number of DSBs, we took advantage of the ability of 125I, when incorporated into DNA, to generate one DNA DSB per radioactive disintegration. SF-268 and HT-1080 cell cultures were grown in the presence of 125IdU and processed immunocytochemically to determine the number of g-H2AX foci. The numbers of 125IdU disintegrations per cell were measured by exposing the same immunocytochemically processed samples to a radiation-sensitive screen with known standards. Under appropriate conditions, the data yielded a direct correlation between the number of 125I decays and the number of foci per cell, consistent with the assumptions that each 125I decay yields a DNA DSB and each DNA DSB yields a visible g-H2AX focus. Based on these findings, we conclude that g-H2AX antibody may form the basis of a sensitive quantitative method for the detection of DNA DSBs in eukaryotic cells. q 2002 by Radiation Research Society

INTRODUCTION

The integrity of chromosomal DNA is essential to its information transfer functions such as replication and transcription as well as to the mechanical segregation of chromosomes during mitosis and meiosis. Thus a single unrejoined double-strand break (DSB) in the DNA may be an oncogenic and even lethal lesion, since both informational redundancy and linear continuity are compromised; evidence is accumulating that defects in the maintenance of Address for correspondence: NIH, NCI, CCR, Bld. 37/5A5050, 9000 Rockville Pike, Bethesda, MD 20892; e-mail: [email protected]. 2 Present address: Erasmus University, 3000 DR Rotterdam, The Netherlands. 1

486

QUANTITATIVE DETECTION OF DNA DSBs WITH g-H2AX ANTIBODY

487

FIG. 1. Method of quantitative detection of 125IdU disintegrations in 125IdU-treated SF-268 cells. Panel A: Calibration samples with 125IdU in the range of the amount given to the asynchronous cells for 20 h (3.7 3 102211 3 104 Bq) were exposed to a radiation-sensitive screen together with the same fixed cell samples used for immunocytochemistry (panel B) and were analyzed with Bio-Imaging Analyzer. Panel C: Calibration curve. The integrated pixel values (intensity of darkness) of the calibration samples (panel A) plotted as a function of their dpm values measured with a g-ray counter. Panel D: The final numbers of 125IdU disintegrations per cell plotted as the amount of 125IdU given to the cells. The dpm values of the cell samples (panel B) were obtained using the calibration curve. To find the number of 125IdU disintegrations per cell, the dpm were multiplied by the number of minutes the cells were exposed to 125IdU and then divided by the number of cells in the sample.

488

SEDELNIKOVA ET AL.

FIG. 2. g-H2AX foci formation in synchronized human cells from 125I decay. Panel A: Cells synchronized at the G1/S-phase boundary were released into S phase and incubated for 30 min with 125IdU. Panel B: 91% of the cells in the culture incorporated 125I into DNA, compared to 47% of the cells in asynchronous culture. Panel C: After fixation and processing, maximum projections (showing all recorded foci) of cells incubated with different amounts of 125IdU were analyzed. Panel D: Values for the average numbers of foci per cell were plotted as a function of the


I radioactivity in the samples; this value is then compared to the number of g-H2AX foci appearing in cells. This general strategy was also used by Jakob et al. (14) to answer a different question; they demonstrated a 1:1 correlation between the number of CDKN1A (p21) foci per cell and the number of accelerated heavy-ion traversals through nuclei, based on ion flux measurements. Here, using cells with 125I-labeled DNA, we show that there is a direct correlation between the number of g-H2AX foci per cell and the number of 125I decays per cell. Given that each 125I decay yields a DNA DSB, each DNA DSB is shown to yield a visible g-H2AX focus. 125

MATERIALS AND METHODS Cell Culture and Labeling Conditions Cells of the cell lines SF-268 (from a human astrocytoma) and HT1080 (from a human fibrosarcoma) were obtained from the ATCC (Manassas, VA) and maintained according to ATCC recommendations at 378C in a humidified atmosphere with 95% air/5% CO2. SF-268 cells were cultured in RPMI 1640 medium containing 50 U/ml penicillin, 50 mg/ml streptomycin, and 10% fetal bovine serum (FBS); HT-1080 cells were cultured in DMEM containing 50 U/ml penicillin, 50 mg/ml streptomycin, and 10% heat-inactivated FBS. All cell culture reagents were purchased from Life Technologies (Gaithersburg, MD). For labeling with 125 I, cells were seeded on Labtek II four-well glass slides (Nalge Nunc International, Naperville, IL) at a density of 2 3 105 cells per well and grown in the presence of indicated amounts of 125IdU (ICN Biomedical, Costa Mesa, CA). After 20 h, the radioactive medium was removed and the cells were rinsed twice with PBS (phosphate-buffered saline), replenished with complete medium containing 10–4 M thymidine, and incubated for 1 h. After the cultures were washed with PBS, they were fixed in 2% paraformaldehyde for 30 min. SF-268 cell cultures were synchronized in two steps, first with serum deprivation and second with aphidicolin. Cultures were maintained in serum- and isoleucine-deprived medium for 48 h (15, 16), then in complete medium containing 10 mg/ml aphidicolin (Sigma, St. Louis, MO) for 15 h, then in complete medium lacking aphidicolin. The efficiency of synchronization was evaluated using a FACSCalibur flow cytometer and Cell Quest 3.3 software (Becton Dickinson, San Jose, CA). Cultures were incubated with the indicated amounts of 125IdU for 30 min, 4 h after removal of aphidicolin, then incubated in medium containing 10–4 M thymidine for 1 h. After washing with PBS, the cultures were fixed in 2% paraformaldehyde for 30 min. The homogeneity of 125IdU incorporation was evaluated by autoradiography as described previously (17–19). Laser Scanning Confocal Microscopy After cell fixation, immunocytochemical procedures were performed as described previously (7). Briefly, the cell preparations were permeabilized in 70% ethanol at 2208C for 5 min, blocked with 8% BSA for 1 h, incubated with the g-H2AX primary antibody at 800-fold dilution of the immune serum for 2 h, incubated with FITC-conjugated goat anti-rabbit secondary antibody (Oncogene Research Products, Cambridge, MA) at 200-fold dilution (stock solution concentration was 100 mg/ml) for 1 h, stained with propidium iodine, and mounted under cover slips. The mi-

489

croscopy was done with a Nikon PCM 2000 laser scanning confocal microscope (Nikon Inc, Augusta, GA) using SimplePCI software (Compix Inc., Cranberry Township, PA) for image processing. Optical sections (0.5 mm) through the thickness of the cells were imaged and combined in a maximum projection so that all the visible foci were recorded. The projection was saved as a BMP file and brought into Paint Shop Pro 7 (Jasc Software, Minneapolis, MN) for presentation. Foci were counted by eye. Quantification of the Number of

IdU Disintegrations

125

The numbers of IdU disintegrations accumulated per Lab-Tek II well during the period of cell incubation with 125IdU were determined on the same fixed cell samples used for quantification of foci. The sample slides were placed in contact with a radiation-sensitive screen that was scanned in the BAS 1500 Bio-Imaging Analyzer (Fuji Medical Systems, Stanford, CT). The image densities (Integrated Pixels) were compared to those obtained with known amounts of 125IdU to determine the 125I dpm (disintegration per minute) values. The numbers of cells per well were determined by visually counting cells in 20 fields from an upper left corner of the well to its lower right corner and normalizing those values to the well area. The number of disintegrations per well was multiplied by the duration of exposure of cells to 125IdU and divided by the number of cells in the well to derive a value for average disintegrations per cell. The final numbers of 125IdU disintegrations per cell were compared to the numbers of g-H2AX foci per cell. Figure 1 illustrates this method of quantitative detection of 125IdU disintegrations in 125IdU-treated SF-268 cells. 125

RESULTS

g-H2AX Foci in Synchronized Cells Pulsed with

125

IdU

From previous work (7), it is known that g-H2AX foci form within minutes of irradiation, increase in intensity for 30 min after irradiation, and then decrease slowly over a period of hours. Thus a short incubation and chase period would ensure that the maximum number of g-H2AX foci formed due to 125I disintegration were present at cell fixation. To ensure that all cells had incorporated 125IdU during the 30-min incubation, SF-268 cultures were synchronized (see the Materials and Methods) and 125IdU was added in early S phase. Under these experimental conditions, all cells would have incorporated similar amounts of 125IdU, and most if not all of the 125I disintegrations would be expected to result in the g-H2AX foci still being present when the cells were fixed. This protocol resulted in 97% synchronized cells (Fig. 2A) and similar amounts of 125I being incorporated in each cell (Fig. 2B). The average number of foci per cell increased when the cultures were incubated with increasing amounts of 125IdU (Fig. 2C, 3.7 3 102 to 7.4 3 104 Bq). As is typical of many tumor lines, some SF-268 cells contained one to three foci in the absence of 125IdU treatment (Fig. 2C, 0 Bq). The same cultures were analyzed for the amount of incorporated 125IdU per cell, and the average

← values for the average numbers of decays per cell. The average numbers of foci per cell in 20 cells: control, 2.6 6 2.5; 3.7 3 102 Bq, 3.2 6 2.3; 1.8 3 103 Bq, 4.4 6 2.4; 1.8 3 104 Bq, 4.8 6 4.0; 3.7 3 104 Bq, 8.5 6 4.2; 7.4 3 104 Bq, 18 6 7.1. Values for the average number of decays per cell were measured as shown in Fig. 1. Panel E: Distributions of numbers of foci in the cells of different populations were recorded. Red, control; purple, 1.8 3 103 Bq; beige, 3.7 3 103 Bq; yellow, 1.8 3 104 Bq; green, 3.7 3 104 Bq; blue, 7.4 3 104 Bq.

490

SEDELNIKOVA ET AL.

FIG. 3. Formation of g-H2AX foci in asynchronous human cells from 125I decay. Exponentially growing cultures of SF-268 (panel A) and HT-1080 cells were incubated for 20 h with 125IdU. Panel B: Most of the SF-268 cells (96%) incorporated 125I, but the amounts were heterogeneous. After fixation and processing, maximum projections (showing all recorded foci) of HT-1080 (panel C) and SF-268 (panel D) cells incubated with different amounts of 125IdU were analyzed. Panel E: Values for the average numbers of foci per cell were plotted as a function of the values for the average numbers of decays per cell for HT-1080 (circles) and SF-268 (diamonds). The average numbers of foci per cell (SF-268) for 20 cells: control, 2.1 6 2.3; 3.7 3 102 Bq, 2.5 6 2.5; 1.8 3 103 Bq, 4.3 6 3.7; 3.7 3 103 Bq, 4.8 6 4.6; 1.8 3 104 Bq, 21.1 6 16.3; 3.7 3 104 Bq, 30.9 6 20.7; 7.4 3 104 Bq, 58.8 6 24.3. Values for the average number of decays per cell were measured as shown in Fig. 1.

numbers of decays per cell during the experiment were calculated. When the average numbers of foci per cell were plotted as a function of the average number of 125I decays per cell (Fig. 2D), the graph revealed a direct correlation between the number of 125I decays and the number of foci per cell. Theoretically, the numbers of radioactive decays

per cell in a population are expected to follow a Poisson distribution. The distribution of numbers of foci in these cell populations was examined and was found to be consistent with the Poisson distribution (Fig. 2E). The slope of the line (Fig. 2D) is 0.8 foci per decay. This value is not sensitive to the length of the labeling and chase


period; when the chase was lengthened to 2 h, the ratio of foci to decays remained unchanged. When coupled with the in vitro efficiency of DNA DSB formation from 125I decay of 80% (13), the value of 0.8 foci per decay indicated that each DNA DSB yields a visible g-H2AX focus.

g-H2AX Foci in Asynchronous Cells The results presented above demonstrate a one-to-one relationship between g-H2AX foci and DNA DSBs in SF268 cells. Since cell synchronization can be laborious and many cell lines cannot be efficiently synchronized, we decided to compare the results obtained with synchronized cultures with those for cultures grown using more conventional protocols. Thus cultures of SF-268 and of HT-1080 cells were grown exponentially overnight in the presence of 125IdU. HT-1080 is a human fibrosarcoma line that we were not able to synchronize efficiently. Both HT-1080 and SF-268 cells responded to increasing 125IdU incorporation with the formation of increasing numbers of discrete gH2AX foci throughout the nuclei (Fig. 3C and D). Again, some cells in both lines contained one to three foci in the absence of 125IdU (0 Bq in Fig. 3C and D). Application of more than 11 3 104 Bq 125IdU to HT-1080 cells or 8 3 104 Bq to SF-268 cells (data not shown) inhibits formation of foci. We counted the foci in 15–20 cells and plotted the number of foci per cell as a function of the final number of 125I decays per cell (Fig. 3E). The graph shows a direct correlation between the number of 125I decays and the number of foci per cell, about 0.3 foci per decay for HT-1080 cells and 0.24 foci per decay for SF-268 cells. As expected, these values are significantly lower than those values obtained in synchronized cell cultures because the length of the incubation necessary to label all cells in the culture permitted significant numbers of foci formed from 125I decay to disappear (7). However, the important finding is that under these conditions different cell lines have very similar responses and yield similar numbers of g-H2AX foci per 125I decay. DISCUSSION

Core histones are stable constituents of nucleosomes. As a core histone and when specifically modified in response to a DNA DSB, H2AX would anchor a signal to the nucleoprotein fiber marking the DNA break site. H2AX is phosphorylated rapidly and extensively to form g-H2AX in response to agents that cause DNA DSBs, with approximately 0.03% of the H2AX becoming phosphorylated per DNA DSB. In a typical mammalian cell with H2AX as 10% of the H2A complement, 0.03% corresponds to about 2000 H2AX molecules (6). Immunocytochemical studies show that these molecules are massed in foci situated at or very near the sites of the breaks (7). Here we demonstrate

491

a 1:1 numerical correspondence between DNA DSBs and g-H2AX foci. One question that may arise is why there are not two gH2AX foci per DNA DSB since a DSB generates two DNA ends. Rogakou et al. (7) showed that in most cases, the two DNA ends from a DSB may stay very close and thus may not be resolvable. However, in those uncommon instances where broken chromatids were found, both broken chromatid ends were capped with g-H2AX foci. However, being uncommon, these instances would not be expected to increase the ratio of g-H2AX foci to DNA DSBs significantly. Also it has been suggested recently that in addition to the direct DSB caused by 125I decay, there may be other DSBs in neighboring helices, since the mean diffusion distance of an •OH is 6 nm, or about the radius of a nucleosome particle. This could place a second, indirect DSB 80 bp away in the adjacent helix of the same nucleosome or up to 1200 bp away in an adjacent nucleosome (20). Since g-H2AX foci cover megabase equivalents of chromatin, such indirect DSBs would be included in the same focus. g-H2AX foci may play an essential role in the efficient recruitment of proteins involved in the repair of the DNA DSBs (21). This role may be to mark the site of the damage very much the way a marker buoy marks the sites of underwater items. It is also possible that the H2AX g-phosphorylation alters chromatin structure to facilitate repair or to stabilize the break region so that the DNA ends remain in proximity. There is evidence that g-H2AX is a recruitment signal. RAD51 and BRCA1 foci were not formed if formation of g-H2AX was prevented by incubation of the cell cultures with the PI-3 kinase inhibitor wortmannin before and during exposure to ionizing radiation, but these foci were formed if incubation with wortmannin was delayed until 5 min postirradiation, after g-H2AX foci had formed but before RAD51 and BRCA1 foci had formed. Thus this result indicates that g-H2AX foci may be essential for these other proteins to form foci at sites of DNA DSBs even several hours later (21). g-H2AX has been shown to appear in all situations examined that result in DNA DSBs: environmentally produced DNA DSBs (7), metabolically produced breaks during replicational stress (22, 23), and programmed breaks as during meiotic recombination (24) and V(D)J recombination (25). V(D)J recombination is notable because no more than two g-H2AX foci were visualized per nucleus, in agreement with the number of genomic sites of V(D)J recombination. In contrast, the number of DNA DSBs from incorporated 125IdU may be varied, and under these conditions, we have established that close to one g-H2AX focus forms per DNA DSB. Thus the technique has potential applications for quantifying numbers of DNA DSBs from a variety of causes in a variety of eukaryotic cells. Received: January 11, 2002; accepted: May 3, 2002

492

SEDELNIKOVA ET AL.

REFERENCES 1. D. O. Ferguson and F. W. Alt, DNA double strand break repair and chromosomal translocation: Lessons from animal models. Oncogene 20, 5572–5579 (2001). 2. A. J. Pierce, J. M. Stark, F. D. Araujo, M. E. Moynahan, M. Berwick and M. Jasin, Double-strand breaks and tumorigenesis. Trends Cell Biol. 11, S52–S59 (2001). 3. S. D. Fugmann, A. I. Lee, P. E. Shockett, I. J. Villey and D. G. Schatz, The RAG proteins and V(D)J recombination: Complexes, ends and transposition. Annu. Rev. Immunol. 18, 495–527 (2000). 4. S. Keeney, C. N. Giroux and N. Kleckner, Meiosis-specific DNA double-strand breaks are catalyzed by Spo11, a member of a widely conserved protein family. Cell 88, 375–384 (1997). 5. J. E. Haber, Partners and pathways repairing a double-strand break. Trends Genet. 16, 259–64 (2000). 6. E. P. Rogakou, D. R. Pilch, A. H. Orr, V. S. Ivanova and W. M. Bonner, DNA double-stranded breaks induce histone H2AX phosphorylation on serine 139. J. Biol. Chem. 273, 5858–5868 (1998). 7. E. P. Rogakou, C. Boon, C. Redon and W. M. Bonner, Megabase chromatin domains involved in DNA double-strand breaks in vivo. J. Cell Biol. 146, 905–915 (1999). 8. L. E. Feinendegen, V. P. Bond and W. L. Hughes, 125I-DU (5-iodo2-deoxyuridine) in autoradiographic studies of cell proliferation. Exp. Cell Res. 43, 107–119 (1966). 9. A. I. Kassis, P. Y. Wen, A. D. Van Den Abbeele, J. BaranowskaKortylewicz, G. M. Makrigiorgos, K. R. Metz, K. Z. Matalka, C. U. Cook, S. K. Sahu and S. J. Adelstein, 5-[125I]-29deoxyuridine in the radiotherapy of brain tumors in rats. J. Nucl. Med. 39, 1148–1154 (1998). 10. N. Miyazaki and Y. Fijiwara, Mutagenic and lethal effects of [5125 I]iodo-29-deoxyuridine incorporated into DNA of mammalian cells, and their RBEs. Radiat. Res. 8, 456–465 (1981). 11. R. E. Krisch and R. D. Ley, Induction of lethality and DNA breakage by the decay of iodine-125 in bacteriophage T4. Int. J. Radiat. Biol. 25, 21–30 (1974). 12. R. E. Krisch, F. Krasin and C. J. Sauri, DNA breakage, repair and lethality after 125I decay in rec1 and recA strains of Escherichia coli. Int. J. Radiat. Biol. 29, 37–50 (1976). 13. P. N. Lobachevsky and R. F. Martin, Iodine-125 decay in a synthetic oligodeoxynucleotide. I. Fragment size distribution and evaluation of breakage probability. Radiat. Res. 153, 263–270 (2000).

14. B. Jakob, M. Scholz and G. Taucher-Scholz, Immediate localized CDKN1A (p21) radiation response after damage produced by heavyion tracks. Radiat. Res. 154, 398–405 (2000). 15. H. Lee, J. M. Larner and J. L. Hamlin, A p53-independent damagesensing mechanism that functions as a checkpoint at the G1/S transition in Chinese hamster ovary cells. Proc. Natl. Acad. Sci. USA 94, 526–531 (1997). 16. P. M. O’Connor and J. Jackman, Synchronization of mammalian cells. In Cell Cycle—Materials and Methods (M. Pagano, Ed.), pp. 63–74. Springer, Berlin, Heidelberg and New York, 1995. 17. O. A. Sedelnikova, I. G. Panyutin, A. R. Thierry and R. D. Neumann, Radiotoxicity of iodine-125-labeled oligodeoxyribonucleotides in mammalian cells. J. Nucl. Med. 39, 1412–1418 (1998). 18. R. Baserga and D. Malamud, Autoradiography: Techniques and Applications. Harper and Row, New York, 1969. 19. W. E. Stumpt and H. F. Solomon, Eds. Autoradiograpy and Correlative Imaging. Academic Press, San Diego, 1995. 20. M. A. Walicka, S. J. Adelstein and A. I. Kassis, Indirect mechanisms contribute to biological effects produced by decay of DNA-incorporated iodine-125 in mammalian cells in vitro: Double-strand breaks. Radiat. Res. 149, 134–141 (1998). 21. T. T. Paull, E. P. Rogakou, V. Yamazaki, C. U. Kirchgessner, M. Gellert and W. M. Bonner, A critical role for histone H2AX in sequential recruitment of repair factors to nuclear foci after DNA damage. Curr. Biol. 10, 886–895 (2000). 22. I. M. Ward and J. Chen, Histone H2AX is phosphorylated in an ATRdependent manner in response to replicational stress. J. Biol. Chem. 276, 47759–47762 (2001). 23. C. L. Limoli, E. Giedzinski, W. M. Bonner and J. E. Cleaver, UVinduced replication arrest in the xeroderma pigmentosum variant leads to DNA double-strand breaks, g-H2AX formation, and Mre11 relocalization. Proc. Natl. Acad. Sci. USA 99, 233–238 (2002). 24. S. K. Mahadevaian, J. M. A. Turner, E. P. Rogakou, F. Baudat, J. Blanco-Rodrı´guez, M. Jasin, W. M. Bonner and P. S. Burgoyne, Initiation of homologous recombination precedes synapis during mammalian meiosis. Nat. Genet. 27, 271–276 (2001). 25. H. T. Chen, A. Bhandoola, M. J. Difilippantonio, J. Zhu, M. J. Brown, X. Tia, E. P. Rogakou, T. Brotz, W. M. Bonner and A. Nussenzweig, Response to RAG-mediated V(D)J cleavage by NBS1 and g-H2AX. Science 290, 1962–1964 (2000).