Professional Documents
Culture Documents
Dist inct domains in Nbs1 regulat e irradiat ion-induced checkpoint s and apopt osis
Pet er McKinnon
Recognit ion, signaling, and repair of DNA double-st rand breaks produced by ionizing radiat ion in mam…
Larry T hompson
T he FHA and BRCT domains recognize ADP-ribosylat ion during DNA damage response
Chao-yie Yang
brief communications
Activation of the ataxia telangiectasia mutated (ATM) However, by counting the number of B lymphocytes in mitosis
kinase triggers diverse cellular responses to ionizing radi- under the microscope before and after irradiation, a milder G2
ation (IR), including the initiation of cell cycle check- arrest in H2AX-deficient cells was apparent at the lowest dose
points1. Histone H2AX, p53 binding-protein 1 (53BP1) and examined (0.75 Gy)13. To determine whether H2AX is involved in
Chk2 are targets of ATM-mediated phosphorylation2–5, but the G2–M checkpoint, we exposed activated H2AX+/+, H2AX−/− and
little is known about their roles in signalling the presence ATM−/− splenic B cells to various doses of IR and measured the pro-
of DNA damage. Here, we show that mice lacking either portion of mitotic cells by flow cytometry using anti-phospho-his-
H2AX or 53BP1, but not Chk2, manifest a G2–M check- tone H3 antibodies15. In the absence of IR, the frequency of mitot-
point defect close to that observed in ATM−/− cells after ic cells was comparable in H2AX+/+ and H2AX−/− B cells (Fig. 1a).
exposure to low, but not high, doses of IR. Moreover, However, 1 h after exposure to 0.5 Gy, 60–80% fewer H2AX+/+ cells
H2AX regulates the ability of 53BP1 to efficiently accu- entered mitosis, whereas almost no change was observed in the fre-
mulate into IR-induced foci. We propose that at threshold quency of H2AX−/− mitotic cells (Fig. 1a, b). H2AX−/− B cells had a
levels of DNA damage, H2AX-mediated concentration of partially impaired G2–M checkpoint at doses up to 5 Gy (Fig. 1c),
53BP1 at double-strand breaks is essential for the ampli- beyond which a normal number of H2AX−/− cells arrested in G2. In
fication of signals that might otherwise be insufficient to contrast, ATM−/− cells had a deficient G2–M checkpoint at both low
prevent entry of damaged cells into mitosis. and high doses of IR16. The degree of the checkpoint defect seemed
to be less severe in H2AX−/− MEFs, compared with H2AX−/− B cells
T
he ‘mean lethal’ dose of IR in humans is estimated to be 4 Gy, (Fig. 1c, d). However, an impaired IR-induced arrest in G2 was con-
which is defined as the dose that causes 50% mortality sistently observed at the lowest IR doses in the absence of H2AX.
(http://www.nrc.gov/). Among the cellular responses to IR, cell The tumour suppressor 53BP1 is a putative mammalian homo-
cycle checkpoints function to protect cells from permanent damage logue of scRad9, an important checkpoint regulator in yeast that
and transformation by delaying progression of the cell cycle until forms irradiation-induced foci which colocalize with γ-H2AX
the DNA is repaired. However, most of the current research proto- foci4,5,17. To determine whether 53BP1 is involved in γ-H2AX for-
cols used for the analysis of IR-induced cellular checkpoints are per- mation or the G2–M checkpoint, we analysed the response of
formed at doses above 4 Gy. The large-scale DNA damage produced 53BP1−/− cells (see Methods) to IR. In contrast to H2AX−/− cells,
by high-dose irradiation may be sufficient to activate redundant which had impaired 53BP1 foci formation (see below)13, H2AX was
checkpoint signalling cascades, and therefore may confound inter- phosphorylated and γ-H2AX foci formed normally in 53BP1−/− B
pretations concerning the role of a number of proteins in regulating cells (Fig. 2a). However, the G2–M checkpoint analysis showed that
the response to DNA damage. For example, conflicting conclusions similar to H2AX-deficient cells, 53BP1−/− cells were less responsive
have been drawn as to the nature of damage-induced cell cycle arrest to IR, exhibiting a fivefold greater threshold than normal before a
in cells from patients with Nijmegen breakage syndrome (NBS)6,7. significant number of cells arrested in G2 (Fig. 2b). These results,
These data are not necessarily inconsistent, as several groups have together with the finding that 53BP1−/− mice have an almost identi-
shown that NBS cells exhibit G1–S and G2–M checkpoint defects at cal phenotype as H2AX−/− mice (including growth retardation, radi-
low, but not at high doses, of IR8–10. Similarly, Chk2−/− mouse embry- ation sensitivity and genomic instability (ref. 13; I.W. and J.C.,
onic fibroblasts (MEFs) were shown to have a dose-dependent G1–S unpublished observations), place 53BP1 downstream of H2AX in
checkpoint defect11,12. Together, these results suggest that deficien- the response to DNA damage.
cies in some ATM kinase substrates may result in intermediate ATM phosphorylates numerous targets that mediate cell cycle
checkpoint defects. checkpoints and DNA repair, including Chk2, Nbs1, 53BP1 and
Phosphorylation of histone H2AX (γ-H2AX) is among the ear- H2AX3−5,18. In NBS cells, the G2–M checkpoint defect observed after
liest ATM-dependent responses to DSBs2,3. Recently, it has been low-dose IR correlated with impaired ATM-dependent phosphoryla-
demonstrated that loss of H2AX in mice results in genomic insta- tion of Chk2 (ref. 9). Moreover, a recent study reported a modest
bility and defects in concentrating various DNA repair factors into defect in the phosphorylation of Chk2 and Brca1 in human cells in
IR-induced nuclear foci13,14. No severe abnormalities in IR-induced which 53BP1 expression was blocked with inhibitory RNA
cell cycle checkpoints were observed in H2AX deficient cells13,14. (siRNA)19. To determine whether H2AX and 53BP1 affect the ability
a
Control IR (0.5 Gy)
104 104
b
102 102
H2AX+/+ 110
100 100
100 101 102 103 104 100 101 102 103 104
DNA content
c d
100 100
Percentage of P-H3 cells
90 H2AX+/+ 90 H2AX+/+
80 H2AX–/– 80 H2AX–/–
70 ATM–/– 70
60 60
50 50
40 40
30 30
20 20
10 10
0 0
0 2 4 6 8 10 0 2 4 6 8 10
Dose (Gy) Dose (Gy)
Figure 1 H2AX is required for low-dose IR-induced G2 arrest. a, Flow cyto- H2AX+/+ and H2AX−/− B cells after exposure to 0.5 Gy IR. c, The percentage of
metric profiles of the cell cycle distribution of H2AX+/+ and H2AX−/− stimulated B mitotic cells before and after exposure to IR (0, 0.5, 1, 2.5, 5 and 10 Gy) was
cells before and 1 h after treatment with 0.5 Gy IR. Propidium iodide staining for analysed in H2AX+/+, H2AX−/− and ATM−/− stimulated B cells. d, The induction of the
DNA content (x axis) and for histone H3 phosphorylation (y axis) is used to distin- G2–M checkpoint in H2AX+/+ and H2AX−/− MEFs was analysed 1 h after exposure to
guish mitotic cells from G2 cells. b, The percentage of mitotic cells positive for different doses of IR, as described in c. Data are representative of at least three
phoshorylated histone H3 (P-H3) (irradiated/unirradiated) (as a function of time in independent experiments.
of ATM to be activated by DNA damage, we used western blotting scRad53 is essential for IR-induced arrest in Saccharomyces cerevisi-
to examine hyperphosphorylation of Chk2 in response to IR. After ae23. A recent study showed that decreased levels of 53BP1 in human
treating B cells with IR, we observed a dose-dependent decrease in cells results in an impaired G2 and intra-S phase checkpoint arrest
the mobility of Chk2 (Fig. 3a, b), which reflects its phosphorylation that correlates with a reduction in the Thr 68-phosphorylated form
state20. Comparison of the IR-induced mobility shift of Chk2 (or of Chk2 (Chk2-T68-P) and a failure of Chk2-T68-P to concentrate
Nbs1, data not shown)13 identified no large differences between in IR-induced foci19. To evaluate the role of Chk2 in the G2–M
H2AX+/+ versus H2AX−/− or 53BP1−/− versus 53BP1+/+ cells, even at checkpoint, we analysed the IR response of Chk2-deficient B cells
low IR doses. In marked contrast, no Chk2 or Nbs1 (data not and MEFs12 (Fig. 3d, e). In marked contrast to H2AX and 53BP1,
shown) migration delay was observed in ATM−/− B cells at doses up Chk2 seemed to be dispensable for a proper induction of the G2–M
to 5 Gy (Fig. 3a), demonstrating the ATM-dependence of these checkpoint at both low and high doses of radiation (Fig. 3d, e).
phosphorylation events at low doses of IR18. At 10 Gy, however, we To determine what aspects of 53BP1 function could be affected
did detect phosphorylation of Chk2 and H2AX in ATM−/− B cells by H2AX deficiency, we performed a detailed analysis of the
(Fig. 3a, c), consistent with previous studies suggesting the exis- responses of 53BP1 to IR in H2AX+/+ and H2AX−/− cells. As previ-
tence of redundant pathways for activation of these proteins at high ously reported for human cells4,5,17, 53BP1 formed nuclear foci very
doses20–22. However, ATM was required for phosphorylation of rapidly after IR in mouse cells, which colocalized with γ-H2AX foci
H2AX at low doses (Fig. 3c)3. Although we cannot rule out the pos- (Fig. 4a, b). 53BP1 foci were already detectable 5 min after irradia-
sibility that there are partial defects in phosphorylation of check- tion, but the average number of foci per cell increased markedly by
point kinase substrates in H2AX−/− and 53BP1−/− cells19 (see below), 1 h post-irradiation (Fig. 4c)17, which paralleled the onset of G2–M
our data indicate that a major impairment of ATM function is checkpoint activation (Fig. 1b). In contrast, 53BP1 foci formation
unlikely to be the primary cause of the G2–M checkpoint defect. was severely compromised in H2AX−/− cells (Fig. 4b)13. Although
scRad9-dependent phosphorylation of the Chk2 homologue H2AX did not affect the overall levels of 53BP1, phosphorylation of
Dose (Gy)
a
0 0.5 1 2.5 5 10
53BP1–/– γ-H2AX
b
Control IR
100
Dose (Gy)
γ-H2AX foci
Figure 2 53BP1 regulates the G2–M checkpoint. a, γ-H2AX formation was analysed by western blotting in 53BP1−/− stimulated B cells (top). The distribution of γ-H2AX
foci (green) was analysed in 53BP1+/+ and 53BP1−/− stimulated B cells before and after exposure to 1 Gy (bottom). DNA was counterstained with the TOPRO-3 dye (red).
Similar results were obtained after treatment with 3 Gy (data not shown). b, The percentage of mitotic cells positive for phoshorylated histone H3 (P-H3) before and after
exposure to IR was analysed in 53BP1+/+ and 53BP1−/− stimulated B cells, as described in Fig. 1. Data are representative of at least three independent experiments.
NIH3T3, No IR
NIH3T3, 6 Gy
Dose (Gy)
a b Dose (Gy)
0 0.5 1 2.5 5 10
0 0.25 0.5 1 2.5 5 10
Chk2PP
H2AX+/+ Chk2 Chk2PP
53BP1+/+
Chk2
Chk2PP
H2AX–/– Chk2
53BP1–/– Chk2PP
Chk2
Chk2PP
ATM–/– Chk2
d
Percentage of P-H3 cells
100
90 Chk2+/+
80 Chk2–/–
70
c Dose (Gy) 60
50
0 0.5 1 2.5 5 10 40
30
H2AX+/+ γ-H2AX 20
20
0
H2AX–/– 0 2 4 6 8 10
Dose (Gy)
e
ATM–/–
Percentage of P-H3 cells
γ-H2AX 100
90 Chk2+/+
80 Chk2–/–
70
60
50
40
30
20
20
0
0 2 4 6 8 10
Dose (Gy)
Figure 3 Chk2 phosphorylation in ATM−/−, H2AX−/− and 53BP1−/− cells and H2AX−/− and ATM−/− stimulated B cells exposed to several doses of IR, as described in
G2–M checkpoint in Chk2−/−− cells. a, b, Lysates from wild-type, H2AX−/− and ATM− a. d, e, The induction of the G2–M checkpoint in Chk2+/+ and Chk2−/− stimulated B
/−
(a) or 53BP1+/+ and 53BP1−/− (b) stimulated B cells exposed to varying doses of IR cells (d) or MEFs (e) was analysed 1 h after exposure to different doses of radiation,
were used to analyse the phosphorylation status of Chk2 by western blotting. The IR- as described in Fig. 1. In both d and e, the percentage of mitotic cells are scored as
induced migratory shift is shown for 3T3 cells. Chk2pp indicates the position of phos- being positive for phosphorylated histone H3 (P-H3).
phorylated Chk2. c, γ-H2AX formation was analysed by western blotting in wild-type,
a d H2AX+/+ H2AX–/–
0 1 3 0 1 3 (Gy)
53BP1S25P
53BP1
Normalized 53BP1S25P
levels (arbitrary units)
0.6
0.5 H2AX–/–
b Control IR H2AX+/+
0.4
0.3
0.2
H2AX+/+ 0.1
0
0 1 3
Dose (Gy)
H2AX–/–
e Low number of DSBs
ATM
53BP1
H2AX γ-H2AX
c
γ-H2AX
45 Modified chromatin
Number of foci per cell
40
35
30
25
53BP1
20
15 Foci, signal amplification
10
5
0
0 5 10 60 16 hr
G2 Mitosis
Time (min)
Figure 4 H2AX regulates IR-induced 53BP1 foci formation. a, Colocalization recognizes 53BP1 phosphorylated at Ser 25 (53BP1S25P)27,28. The immunoprecipi-
of γ-H2AX with 53BP1 foci. Wild-type MEFs were exposed to 1 Gy IR and then tates were analysed by Western blotting with 53BP1-specific antibodies. 15 µg of
incubated at 37 °C for 1 h before staining for γ-H2AX (red) and 53BP1 (green). these lysates were used for western blot analysis of 53BP1 levels. The ratio of
Images were merged to determine colocalization (yellow). b, Distribution of 53BP1 phosphorylation on Ser 25 to the total amount of 53BP is plotted below
53BP1 (green) was analysed in H2AX+/+ and H2AX−/− MEFs before and 1 h after e, A model for the ATM/H2AX/53BP1-mediated regulation of the G2–M check-
exposure to 1 Gy IR. Similar results were obtained after exposure to 3 Gy (data point. In response to low-dose IR, activated ATM phosphorylates a wide range of
not shown). DNA was counterstained with DAPI (blue). c, The number of foci per targets, including H2AX and 53BP1. H2AX is essential for the accumulation of
cell (H2AX+/+ MEFs) averaged over 25 cells per data point. d, H2AX+/+ and H2AX− 53BP1 and other factors into the chromatin microenvironment, within megabases
/−
stimulated B cells were untreated or exposed to IR and nuclear extracts were from the actual DSB. This localized concentration of signal transduction complex-
prepared 1 h after treatment. 200 µg of nuclear extracts were immunoprecipitat- es would generate sufficient amplification of the initial signal to abort the entry of
ed with a phospho-specific antibody (see Supplementary Information, Fig. S1) that G2 cells into mitosis.
53BP1 was reduced by 40% in H2AX−/− cells, relative to controls 1 h a large increase in the local concentration of 53BP1 in the chro-
after treatment with 3 Gy (Fig. 4d; also see Supplementary matin region adjacent to the DSB (Fig. 4e). The accumulation of
Information, Fig. S1). Although it is unclear whether 53BP1 is 53BP1 and other factors in close proximity to a DSB would func-
phosphorylated only when bound to damaged DNA, these results tion as a platform for interactions between multi-component sig-
suggest that H2AX may not be essential for the initial recruitment nal transduction complexes, facilitating the activation of the check-
of 53BP1 to sites of DSBs. Consistently, we have found that point. This model predicts that H2AX and 53BP1 deficiency should
although H2AX is a central regulator of IR-induced focus forma- result in impaired phosphorylation of downstream effectors.
tion, the initial migration of 53BP1 (and other signalling proteins, Indeed, we have observed a partial defect in phosphorylation of
such as Nbs1) to DSBs is independent of H2AX (A. C., O. F-C. and 53BP1 in H2AX−/− cells and a recent study reported that 53BP1 reg-
A.N., unpublished observations). ulates Brca1 phosphorylation19, which is essential for the G2–M
On the basis of these experiments, we conclude that abrogation checkpoint15. H2AX-mediated concentration of signalling com-
of the G2–M checkpoint in H2AX−/− cells in response to low-dose plexes at DSBs would be critical under conditions in which there is
IR is independent of Chk2, but is linked to the inefficient accumu- a low level of DNA damage. Above a threshold, there would be suf-
lation 53BP1 at sites of damage. We propose the following model in ficient activation of checkpoint effectors to propagate signalling
which ATM-mediated phosphorylation of histone H2AX results in cascades along pathways independent of H2AX.
RECEIVED 14 OCTOBER 2002; REVISED 30 OCTOBER 2002; ACCEPTED 31 OCTOBER 2002; ACKNOWLEDGEMENTS
PUBLISHED 25 NOVEMBER 2002.
These studies were in part motivated by discussions with T. Halazonetis, who suggested examining the
1. Shiloh, Y. Biochem. Soc. Trans. 29, 661–666 (2001).
effects of low-dose IR, and we thank T. Halazonetis for sharing unpublished results. We also thank M.
2. Rogakou, E. P., Pilch, D. R., Orr, A. H., Ivanova, V. S. & Bonner, W. M. J. Biol. Chem. 273, Lichten, J. Chung, A. Lee, S. Petersen and A. Singer for critical comments on the manuscript, and M.
5858–5868 (1998). Kruhlack for assistance with microscopy. P.B.C was supported by a grant from The Robert Welch
3. Burma, S., Chen, B. P., Murphy, M., Kurimasa, A. & Chen, D. J. J. Biol. Chem. 276, 42462–42467 Foundation.
(2001). Correspondence and requests for material should be addressed to A.N.
4. Anderson, L., Henderson, C. & Adachi, Y. Mol. Cell. Biol. 21, 1719–1729 (2001). Supplementary Information accompanies the paper on www.nature.com/naturecellbiology.
5. Rappold, I., Iwabuchi, K., Date, T. & Chen, J. J Cell Biol 153, 613–620 (2001).
6. Jongmans, W. et al. Mol. Cell. Biol. 17, 5016–5022 (1997).
7. Yamazaki, V., Wegner, R. D. & Kirchgessner, C. U. Cancer Res. 58, 2316–2322 (1998). COMPETING FINANCIAL INTERESTS
8. Girard, P. M., Riballo, E., Begg, A. C., Waugh, A. & Jeggo, P. A. Oncogene 21, 4191–4199 (2002). The authors declare that they have no competing financial interests.