Differential regulation of the cellular response to DNA double-strand breaks in G1

Abstract

Double-strand breaks (DSBs) are potentially lethal DNA lesions that can be repaired by either homologous recombination (HR) or nonhomologous end-joining (NHEJ). We show that DSBs induced by ionizing radiation (IR) are efficiently processed for HR and bound by Rfa1 during G1, while endonuclease-induced breaks are recognized by Rfa1 only after the cell enters S phase. This difference is dependent on the DNA end-binding Yku70/Yku80 complex. Cell-cycle regulation is also observed in the DNA damage checkpoint response. Specifically, the 9-1-1 complex is required in G1 cells to recruit the Ddc2 checkpoint protein to damaged DNA, while, upon entry into S phase, the cyclin-dependent kinase Cdc28 and the 9-1-1 complex both serve to recruit Ddc2 to foci. Together, these results demonstrate that the DNA repair machinery distinguishes between different types of damage in G1, which translates into different modes of checkpoint activation in G1 and S/G2 cells.

Figure 1. Cell cycle regulation of checkpoint and repair foci

(a) IR-induction of foci in G1-arrested cells. Cells expressing Rfa1-YFP (W3775-12C), Mre11-YFP (W3483-10A), Ddc2-YFP (W3792-4B), Ddc1-YFP (W3923-12B) or Rad52-YFP (W3749-14C) were grown to OD600 ≈ 0.1 in the absence of exogenous DNA damage in SC medium supplemented with 2 µg/ml α-factor for 90 minutes and the cultures subsequently exposed to IR. Images were acquired 60 minutes after exposure to IR. The G1-arrest was confirmed by FACS analysis. Scale bar, 3 µm. Arrowheads mark foci in representative cells. (b) Reduced efficiency of Rad52 focus formation in response to IR in G1-arrested cells. After 120 minutes, approximately 10% of cells form foci consistent with the percentage of the population escaping arrest as measured by FACS analysis. Points and error bars represent the mean and SEM, unless otherwise noted. (c) Rad52 focus formation upon entry into S phase. Cells expressing Rad52-YFP (W4965-8B) were arrested with 7.5 µg/ml α-factor, exposed to IR, and then incubated for another 30 minutes before they were filtered and placed in fresh medium. (d) IR exposure induces BrdU foci during G1. After exposure to IR, subnuclear BrdU foci are visible in G1-arrested cells (W8127-21B) and approximately 60% (18/29 at 30 minutes after exposure, time point shown) colocalize with Rfa1 foci.

Figure 2. Differential recruitment of checkpoint and repair proteins to IR- and endonuclease-induced DSBs

(a) Focus formation in response to an I-SceI-induced DSB. Expression of the I-SceI endonuclease was induced in cells harboring a fluorescently-marked I-SceI cut-site (I-SceIcs) and either Mre11-YFP (W4363-4B), Rfa1-YFP (W4362-1C), Ddc1-YFP (W4688-11D), Ddc2-YFP (W4364-9B) or Rad52-YFP (W4365-5B). The I-SceIcs is adjacent to an array of 336 copies of the Tet operator sequence in cells expressing the tet repressor fused to monomeric RFP, marking the locus as a discrete RFP dot. Focus formation and the percentage of colocalization with the I-SceI cut-site for each protein is plotted as a function of time after I-SceI was induced by addition of galactose. YFP, RFP, DIC and YFP/RFP merged images of representative cells are shown and selected foci indicated by arrowheads. Mre11 is recruited equally well to the I-SceI cut-site in G1 and S/G2/M. In contrast, Rfa1, Ddc1, and Ddc2 form very few foci in G1 cells in response to an I-SceI DSB. (b) Proteins involved in DSBR do not respond to IR by forming foci at doses resulting only in SSBs. Dose dependency curves for Rfa1, Ddc1, Ddc2 and Rad52 foci, respectively, shown in the colored lines. The solid line indicates the predicted percentage of cells in the population receiving 1 or more SSBs, the dotted line indicates the predicted percentage of cells receiving 1 or more DSBs in response to IR, and the dashed line represents the predicted percentage of cells receiving 2 or more DSBs, assuming that 17 SSBs are generated for every DSB and that the number of breaks induced per cell follows a Poisson distribution (Friedland et al., 1999). None of the proteins form foci in response to low IR doses predominantly resulting in SSBs 60 minutes after exposure.

Figure 3. Regulation and kinetics of Rfa1 focus formation

(a–c) Quantitation of Rfa1 foci in response to IR. Cells were arrested by either in G1 by α-factor (a and c) or in G2 by nocodazole (b) and then exposed to IR. Median values for the quantitation of individual Rfa1 focus intensity are plotted as dashed. (a) Rfa1 focus formation in G1. Approximately 80% (0.6 / 0.75 = 0.80) of the cells receiving a DSB in G1 (75%) form an Rfa1 focus (60%) when cells are held in G1, however these foci are fainter than in budded cells. (b) Rfa1 focus formation in G2. Rfa1 foci in G2-arrested cells are 4- to 6-fold brighter, and the focus intensity increases at a rate 3- to 6-fold faster than foci in G1. (c) Cell cycle regulation of Rfa1 foci. The intensity of individual IR-induced Rfa1 foci was measured and the median plotted for cells arrested by α-factor in G1 and released at time = 100 min after exposure to IR. Upon entry into S phase, Rfa1 focus intensity increases at a rate that is 3- to 6-fold higher than in G1 cells, indicating that a faster rate of resection is taking place.

Figure 4. G1 regulation of DSB resection

(a) Rfa1 focus formation in yku70Δ cells. Wild-type (W5713-18A) and yku70Δ (W5713-16D) cells were arrested with α-factor for 210 minutes after which an I-SceI-mediated DSB was induced by addition of galactose. Over the course of three hours, only 20% of wild-type cells accrue Rfa1 foci at the marked DSB. In contrast, 45% of yku70Δ cells have foci at the DSB by three hours after I-SceI induction. (b) Rad53 phosphorylation is not activated in response to a single DSB, while there is constitutive checkpoint activation in yku70Δ cells, as indicated by the Rad53 smear at time 0. (c) Rfa1 focus formation and Sml1 degradation in response to an endonuclease-mediated DSB in G1. As in (a), wild-type (W7542-11D) cells were arrested in G1, and followed by induction of I-SceI endonuclease. The cells forming Rfa1 foci in response to DSB induction also degrade Sml1, indicating downstream checkpoint induction though Rad53 phosphorylation is not visible by western blot. (d) Rfa1 focus formation in sae2Δ cells. Cells were arrested in G1 then exposed to IR. The abrogation of SAE2 (W5071-5D), leads to a slower rate of Rfa1 focus formation in G1 cells, as sae2Δ cells have fewer Rfa1 foci at 30 and 60 minutes after exposure to IR compared to WT cells.

Figure 5. Independent recruitment of checkpoint machinery 9-1-1 and Ddc2-Mec1 requiresCdc28 activity

(a) Ddc2 focus formation in mec3Δ G1 cells. mec3Δ cells were arrested in 2 µg/ml α-factor for 90 minutes and the cultures subsequently exposed to IR, and images were acquired at the stated timepoints. Ddc2 focus formation does not occur in mec3Δ cells (W5358-9A) arrested in G1. (b) Rad53 phosphorylation in WT and mec3Δ G1 cells. Rad53 is phosphorylated in WT G1 cells after exposure to 40 Gy IR, while mec3Δ cells do not. (c) Rfa1 focus formation in mec3Δ G1 cells. Deletion of MEC3 does not prevent Rfa1 focus formation during G1 (W5793-10B). (d) Ddc2 focus formation in S/G2 mec3Δ cells with and without Cdc28 kinase activity. Ddc2 forms foci in budded cells, regardless of the presence of an active DNA damage clamp. When Cdc28 kinase activity is abrogated by the addition of the inhibitor 1-NM-PP1 to a strain containing the cdc28-as1 allele (W7832-2A), no Ddc2 focus formation is observed after 40 Gy IR, indicating that Cdc28 kinase activity is required for recruitment of Ddc2-Mec1 to the sites of DNA damage when Mec3 is absent (W7832-1A). (e) Rad53 phosphorylation in WT and mec3Δ cells. Inhibition of cdc28-as1 cells with 1-NM-PP1 does not alter Rad53 phosphorylation in WT or mec3Δ cells.

Figure 6. RPA checkpoint function is required to maintain Ddc1 foci

(a) Ddc1 focus formation in rfa1-t11 cells. Ddc1 forms foci in rfa1-t11 cells (W5872-3C), however these foci do not persist, and fully disappear by 120 min after exposure to IR, unlike in WT cells where the foci persist for over 3 hours. (b) Ddc2 focus formation in rfa1-t11 cells. Ddc2 focus formation in rfa1-t11 cells (W5873-9B) is reduced from 80% to 60% compared to WT, but is not completely abrogated. (c) Ddc2 focus formation in rfa1-t11 mec3Δ cells in the presence or absence of Cdc28 activity. We find that RPA and Cdc28 act independently to recruit Ddc2. Although rfa1-t11 cdc28-as1 cells (W7848-9B) are partially compromised for Ddc2 focus formation, inhibition of Cdc28 kinase activity further reduces Ddc2 focus formation in rfa1-t11 cells (~30% compared to ~70%), while Ddc2 foci still form in 1/3 of the population. On the other hand, rfa1-t11 mec3Δ cells (W7848-7A) are compromised for Ddc2 focus formation since rfa1-t11 mec3Δ cells, even in the presence of Cdc28 kinase activity, only form Ddc2 foci in 10% of budded cells in response to IR, compared to 70% in rfa1-t11 cells, and 40‐50% in rfa1-t11 cells where Cdc28 kinase activity is inhibited. (d) Inhibition of Cdc28 activity reduces Rad53 phosphorylation in rfa1-t11 cells. After exposure to 40 Gy IR, Rad53 is phosphorylated in rfa1-t11 cells (left). Inhibition of Cdc28-as1 by addition of 1-NM-PP1 reduces Rad53 phosphorylation.

Figure 7. Models for cell cycle regulation of DSB repair and checkpoint activation

(a) Cell cycle regulation of DSB repair. When a DSB is generated (1), ends may be either ‘ragged’ as those produced following exposure to IR (left) or ‘clean’ such as those resulting from an endonuclease-mediated cleavage (right). The MRX complex recognizes both kinds of ends (2) while during G1, clean ends are preferentially bound by Yku70/Yku80 (3). Once Yku70/Yku80 is bound, the ends are not released until ligation is complete or until the cell has entered into S phase (4, 5). Upon entering S phase, Yku70/Yku80 dissociates from the ends, freeing them for digestion by nucleases resulting in ssDNA. RPA binds the ssDNA, independently recruiting the 9-1-1 (Ddc1-Mec3-Rad17) and Ddc2-Mec1 checkpoint complexes (5, right). ‘Ragged’ DNA ends, on the other hand, are not bound by Yku70/Yku80 during G1. Instead, these free ends are processed by one or more nucleases into 3’ ssDNA tails, recognized and bound by RPA. RPA then recruits Ddc1-Mec3-Rad17, which in turn is required for Ddc2-Mec1 focus formation in G1. (4, left). The homologous recombination machinery, here shown as Rad52 and Rad51, cannot be recruited however, until the cells have entered into S phase (6). (b) Ddc2-Mec1 checkpoint activation. In G1 cells, 9-1-1(Ddc1-Mec3-Rad17) is absolutely required for recruitment of Ddc2 into foci and subsequent Rad53 phosphorylation. In S/G2, Cdc28 activity and RPA modification on Lysine 45 (orange star) act in conjunction with the 9-1-1 complex to promote Ddc2 focus formation and downstream checkpoint events. See discussion for details.