Factors determining DNA double strand break repair pathway choice in G2 phase.

Transcription

1 Supplementary Data Factors determining DN double strand break repair pathway choice in G2 phase. tsushi Shibata 1, Sandro Conrad 2, Julie irraux 1, Verena Geuting 2, Olivia arton 2, mani Ismail 1, ndreas Kakarougkas 1, Katheryn Meek 3 Gisela Taucher-Scholz 4, Markus Löbrich 2* and enny. Jeggo 1* Figure S1. nalysis of IR induced foci (IRIF) in G1 and G2 cells. and ) phidicolin (H) was added immediately after 2 Gy X-rays. H treatment blocks the replicative polymerases and hence progression from S to G2 phase. H induced replication stress induces an obvious pan-nuclear H2X/R signal in S phase (eucher et al, 29). Cells are stained with H2X or R and CEN-F (a G2 marker). G1 cells show very low CEN-F signal and clear IRIF. Early S cells show very low CEN-F signal but strong pan-nuclear H2X/R signal. Late S phase cells have a mild CEN-F signal + pan-nuclear H2X/R signal. G2 cells have a greater CEN-F signal and clear H2X/R IRIF. M phase cells are distinguished from other cell cycle phases by morphology and cytoplasmic CEN-F signal. H does not affect DS repair including H2X, R, Rad51 foci formation or SCE formation in cells derived from G2 phase cells and blocks cell cycle progression from S to G2 even in checkpoint defective cells. Full controls for the use of H have been included in (eucher et al, 29; Shibata et al, 21). Figure S2. IR induced R/Rad51 foci co-localise with DSs in G2 phase. ) Cells were irradiated with 2 Gy X-rays and H added immediately. Visible R foci were not detected until ~2 hr after X-rays. However IR induced R foci at >2 hr strongly co-localize with the DS marker, 531. ) IR induced H2X and 531 co-localize in G2 cells. C) IR induced Rad51 foci at >2 hr co-localize with H2X foci. The number of R foci at 2 h post X-rays is similar to the number of unrepaired DSs in HR defective cells (e.g. RC2 cells) (which is also identical to the number of unrepaired DSs in TM defective cells). dditionally, NHEJ defective cells are impaired in the fast component of DS repair in G2 phase. Thus we enumerate R/Rad51 foci at >2 hr after IR as a monitor of resection. We relate this to the number of DSs induced as assessed by H2X numbers 15 min post treatment in Figure 1. Figure S3. Substantial persistent R foci and a DS repair defect in HR defective cells after carbon-ion irradiation. ) Cells were irradiated with 2 Gy carbon ions, and stained with R and CEN-F. 5% of R foci were repaired at 24 hr post carbon irradiation in wild type cells (Figure 1D). In contrast, >9% of R foci remained up to 24 hr post carbon irradiation in RC2 deficient cells. Importantly, the number of R foci at 24 hr in RC2 deficient cells is not drastically changed compared to 2 hr, indicating that the number of R foci at 2 hr

2 represents the peak of R foci formation. This result also suggests that carbon ion induced DSs that become associated with R are repaired by RC2 dependent HR. ) >~9% of H2X foci persist in Rad54 -/- MEFs up to 24 h post carbon-ion irradiation. The number of H2X foci was enumerated after 2 Gy carbon ions. These results indicate that DS repair post carbon ions is repaired by RC2/Rad54 dependent HR. G2 cells were identified as speckled p-histoneh3 Ser1 staining in MEFs. C) The kinetics of DS repair was examined by enumerating H2X foci in HSF1 primary cells in G1 (left) or G2 (right) cells. % of repair kinetics is shown in Figure 1 and. Cells were exposed to 2 Gy carbon ions, 2 Gy X-rays and 2 (for G1) or 5 M (for G2) Etp for 15 min and stained with H2X and CENF. *The number of H2X foci post carbon ions was examined at the indicated time points except for 6 hr. D) R or Rad51 foci numbers in G2 phase were enumerated following exposure to 2 Gy carbon ions, 2 Gy X-rays and 5 M Etp in HSF1 primary cells. % of R or Rad51/ H2X foci at 15 min is shown in Figure 1C. The doses chosen generate similar numbers of H2X foci at 2 h after treatment, since R/Rad51 foci numbers reach the maximal levels at 2 hr post treatment as discussed in Supplementary Figure S2. The induction of DS formation after carbon ion irradiation might be underestimated if they arise in close proximity. Nevertheless, importantly, given that we observe fewer not more H2X foci, it is unlikely that this can explain the magnitude of the increase in R/Rad51 foci that we observe. *The number of R post carbon ions was examined at the indicated time points except for 6 hr. Figure S4. Etp can induce HC-DSs, but they are 2-3 fold lower than X-ray-induced HC-DSs. ) Not all persistent H2X foci co-associate with pk-1 foci. Cells lacking XLF, a core NHEJ protein, were exploited to examine a situation where EC and HC-DSs might persist. t 8 h post 3 M Etp. for 3 min, the % of pk-1 positive H2X foci is approximately 2% in XLF cells, which corresponds to the percentage of HC-DN. Since Etp induces a lower percentage of HC-DSs, 1R (WT) htert cells were treated with 15 M Etp for 3 min, which results in ~8 H2X foci remaining at 8 hr similar to the number remaining in XLF cells treated as described above. In these 1R htert cells nearly all (84 %) of H2X foci co-localise with pk-1 foci. Thus, pk-1 does not form at all persisting H2X foci. Rather, those persisting in control cells represent the slowly repaired HC-DSs. ) Etp can induce DSs at HC regions in NIH3T3 cells. To induce a similar number of DSs (~3 H2X foci), cells were treated with 2 M Etp 3 min or irradiated with 1 Gy X-rays. fter fixation, cells were stained for H2X (green) and DI (blue). H2X (red) in right panel represents regions of H2X and dense-di chromocentre overlap as determined by softworx Suite software. Etp induced DSs can localise to densely staining DI regions but the percentage is less than for X-ray induced DSs. Figure S5. nalysis of R foci in Ku8+RC2 sirn cells. DN-K expression in CHO V3 cell lines.

3 ) Ku8 sirn significantly increases R foci numbers post IR. Quantification is shown in Figure 3. Knockdown efficiency was assessed by immunoblot as shown in right panel. -tubulin was used as a loading control. ) DN-Kcs expression levels in CHO V3 cell lines. DN-Kcs null hamster V3 cells express human DN-Kcs with or without mutations in the CDE phosphorylation sites. Ku8 and -tubulin were used as a loading control. The level of DN-Kcs expression between V3+ WT and particularly between the V3+CDE S>D mutant and V3-CDE S> mutant is similar but the magnitude of resection at 1 h post treatment is very different. The signal intensity of DN-Kcs was quantified with ImageJ. Normalized signal intensities with Talin (cytoskeletal protein) are shown on the bottom of the blots. Figure S6. CtI sirn leads to reduced IR induced Rad51 foci formation in G2 but does not affect activation of TM signalling. ) Cells were irradiated with 3 Gy IR, and stained with Rad51 and CEN-F. Cells containing >5 Rad51 foci were scored as positive. Consistent with ablated R foci formation following CtI sirn treated cells, CtI sirn also significantly reduced Rad51 foci formation (right panel). ) CtI sirn reduced Chk1 S317 phosphorylation, but did not affect TM dependent pk-1 S824 and pchk2 T68. Further, IR induced ptm S1981 foci form normally in CtI knockdown G1 and G2 cells. Figure S7. model for the regulation of DS repair pathway usage at two-ended DSs in G2. We propose that both DN damage complexity and chromatin complexity can influence the speed of DS repair and that DSs, which are not efficiently rejoined by NHEJ, preferentially undergo DS end-resection. 1) DSs that arise in EC-DN without high DN complexity undergo efficient repair with faster kinetics by NHEJ regardless of cell cycle phase. 2) when a highly complex DS arises in EC-DN, it is repaired with slow kinetics in both G1 and G2. In G2, NHEJ stalls and the DSs undergo resection and subsequent repair by HR. 3) HC is also a barrier slowing repair by NHEJ. In G1 phase, an HC-DS is repaired slowly by NHEJ. In G2 phase, NHEJ stalls, allowing resection and subsequent repair by HR. Repair pathway usage at an HC-DS can be switched from HR to NHEJ when CtI dependent DS end resection does not occur. Whereas CtIdependent resection commits to HR. TM has direct and indirect roles in regulating DS end resection by promoting K-1 dependent heterochromatin relaxation and by activating CtI by phosphorylation. The precise magnitude of DN damage complexity required to stall or slow NHEJ is currently unclear but the association of base damage at DSs does not appear to restrict NHEJ usage since most X or -ray induced DSs fall into this class. dditionally, it is unclear whether there is an impact of lesion complexity on HC-DSs e.g. whether DSs without any additional lesions can be repaired rapidly (and without undergoing resection) in G2 phase. Our findings do not eliminate the possibility that some Etpinduced DSs arising in HC-DN can be repaired with fast kinetics without undergoing resection.

4 Table. S1 Comparison of DS repair kinetics between Etoposide and -ray induced damage Etoposide -ray Induction of H2X foci ~ No. of pk-1 associated H2X foci at 8 hr % of pk-1 associated DSs at 8 h ~4.6% 12.6% Half-life of faster repair kinetics ~2.2 h 1.5 h* Half-life of slower repair kinetics 12.7 h 1.2 h* *The data of -ray are from Noon et al., NC, 21. References. eucher, irraux J, Tchouandong L, arton O, Shibata, Conrad S, Goodarzi, Krempler, Jeggo, Lobrich M (29) TM and rtemis promote homologous recombination of radiation-induced DN double-strand breaks in G2. EMO J 28(21): Shibata, arton O, Noon T, Dahm K, Deckbar D, Goodarzi, Lobrich M, Jeggo (21) Role of TM and the damage response mediator proteins 531 and MDC1 in the maintenance of G(2)/M checkpoint arrest. Mol Cell iol 3(13):

5 Supplementary Figure S1 G1 Early S Late S G2 M g-h2x CENF Merge + DI 8 hr after 2 Gy IR (+ H) IR G1 Early S Late S G2 M R CENF Merge + DI 2 hr after 2 Gy IR (+ H) IR

6 Supplementary Figure S2 2 Gy IR (hr) G2 phase S phase R 531 Merge+DI R 531 Merge+DI G2 phase S phase g-h2x 531 Merge+DI g-h2x 531 Merge+DI 2 hr C 2 Gy IR (hr) G2 phase S phase g-h2x Rad51 Merge+DI g-h2x Rad51 Merge+DI

7 No. of R or Rad51 foci per cell No. of gh2x foci per cell No. of gh2x foci per cell No. of carbon ion induced gh2x foci per cell Supplementary Figure S3 CENF R DI R DI C WT G1 phase RC2-2 hr 24 hr 2 hr 24 hr Etp X-ray C G2 phase WT Rad54 -/ Time post irradiation (hr) Etp X-ray C12 D 1 35 * Time post damage (hr) G2 phase Time post damage (hr) * Etp (Rad51) X-ray (Rad51) C12 (R) Time post damage (hr)

8 Supplementary Figure S4 1R (WT) htert 8 hr post 15 mm Etp pk-1 gh2x DI + Merge pk-1 associated gh2x / nalysed no. of gh2x foci 151/181 (83.4%) XLF htert 8 hr post 3 mm Etp 38/173 (21.9%) gh2x DI Merge HC foci 2 mm Etp 14.3% 1 Gy X-ray 29.7% *% of HC-gH2X foci

9 siku8 sirc2 siku8 + sirc2 Supplementary Figure S5 4 hr after 1 Gy R CENF DI Merge sicontrol siku8 sirc2 siku8 + sirc2 RC2 Ku8 Ku7 b-tubulin V3 DN-K DN-Kcs Talin (loading control)

10 No. of ptm foci in G2 per cell % of Rad51 foci positive cells in G2 Supplementary Figure S6 Rad51 CENF Merge +DI 1 hr 8 hr Gy 2 hr 2 3 Gy 2 hr 2 Time post irradiation (hr) IR (Gy) CtI b-tubulin ptm CEN-F Merge +DI G1 1 Gy IR pk-1 S824 K-1 pchk1 S317 Chk Gy X-ray, 2 hr G2 3 Gy IR 2 pchk2 T68 1 Chk2

11 Supplementary Figure S7 Euchromatin Heterochromatin (1) (2) (3) DS without complexity Ku/DN-K DS with complexity DS in HC regions HC remodelling K-1 Ku/DN-K TM NHEJ stalling NHEJ [Repair with FST kinetics] TM CtI CDK CtI CtI dependent resection R/Rad51 loading HR [Repair with SLOW kinetics]