DNA polymerase stabilization at stalled replication forks requires Mec1 and the RecQ helicase Sgs1

Abstract

To ensure proper replication and segregation of the genome, eukaryotic cells have evolved surveillance systems that monitor and react to impaired replication fork progression. In budding yeast, the intra-S phase checkpoint responds to stalled replication forks by downregulating late-firing origins, preventing spindle elongation and allowing efficient resumption of DNA synthesis after recovery from stress. Mutations in this pathway lead to high levels of genomic instability, particularly in the presence of DNA damage. Here we demonstrate by chromatin immunoprecipitation that when yeast replication forks stall due to hydroxyurea (HU) treatment, DNA polymerases alpha and epsilon are stabilized for 40-60 min. This requires the activities of Sgs1, a member of the RecQ family of DNA helicases, and the ATM-related kinase Mec1, but not Rad53 activation. A model is proposed whereby Sgs1 helicase resolves aberrantly paired structures at stalled forks to maintain single-stranded DNA that allows RP-A and Mec1 to promote DNA polymerase association.

Fig. 1. Sgs1 helps stabilize DNA pol ε at replication forks on HU. (A) Primers amplify genome regions corresponding to two early-firing origins: ARS607 (m, black bars), and non-origin sites at +4 kb (n, light blue) and +14 kb (o, yellow), and ARS305 (x, grey), and non-origin sites +3.5 kb (y, green) and +10 kb (z, dark blue). Probes for the late-firing origin ARS501 (l, white) and the late-replicating Tel VII-L subtelomeric region (t, red) monitor background signals and late origin activation. Active and inactive origin positions (ARS) are described in Raghuraman et al. (2001) and Reynolds et al. (1989). ChIP was performed at 30°C on Myc-tagged DNA pol ε as described in Materials and methods for synchronized wild-type (GA-1296) or isogenic sgs1Δ (GA-2206) cells released into medium containing 0.2 M HU for the times indicated. The height of the bars represents the real-time PCR signal as fold increase of IP over beads alone control, for primers near ARS607 (B and C) and ARS305 (D and E). (F) FACS analysis shows that S phase is completed by 40 min at 30°C after release from pheromone for sgs1 (GA-2209) and isogenic wild-type cells (GA-893).

Fig. 2. Sgs1p is associated with replication forks. (A) Time course ChIP was performed on Myc-tagged DNA pol ε for wild-type (GA-1296) and (B) isogenic sgs1Δ (GA-2206) cells after synchronization and release into YPAD without HU at 16°C. Primer sites are described in Figure 1. Quantitation by real-time PCR is described in Materials and methods and Figure 1. (C) FACS analysis shows similar progression through S phase for wild-type and sgs1Δ strains in the absence of HU at 25°C (for more extensive analyses of such profiles see Versini et al., 2003). (D) ChIP was performed at the indicated time points at 16°C for Myc-tagged Sgs1 from GA-1699 cells, as in (A). Amplified ARS305 fragments, quantitation, primers and colour coding are described in Figure 1. (E) As (D), except that cell growth was at 30°C and cells were released from G₁ into 0.2 M HU. (F) Isolated S phase nuclei from GA-1699, carrying Myc-tagged Sgs1, were incubated in an S phase nuclear extract supplemented with nucleotides and digoxigenin-derivatized dUTP (Pasero et al., 1999). After 30 min incorporation at 25°C, fixed nuclei were probed with mAb 9E10 (Sgs1, in red) and rabbit anti-digoxigenin to detect newly synthesized DNA (green). On average, 48% of the Sgs1 signal co-localizes with Dig-dUTP (see arrows).

Fig. 3. DNA pol ε stability requires Sgs1 helicase activity. (A) Map of the Sgs1 helicase domain (grey), Walker A motif (black) and point mutation K₇₀₆A (Lu et al., 1996). (B) Wild-type (GA-1296) cells bearing pRS415 vector (WT, black), or sgs1Δ cells (GA-2206) bearing either pRS415 vector (Δ, white), pRS415 with full-length SGS1 (FL, dark grey) or pRS415-sgs1-hd, expressing helicase-dead protein (hd, light grey), were grown into SC-Leu, synchronized and released into either 0.2 M HU for the indicated times (upper panel) or increasing amounts of MMS for 1 h for viability assays (see Materials and methods). (C) Time course ChIP for DNA pol ε was performed with strains described in (B), as described in Figure 1. Upper and lower graphs show data for ARS305 (probe x) and Tel VIIL (probe t), respectively.

Fig. 4. DNA pol ε is lost from stalled forks in a mec1 mutant. Time course ChIP for DNA pol ε was performed on isogenic wild-type (A, GA-1296), mec1-1 (B, GA-1306) and rad53Δ (C, GA-2208) cells as described in Figure 1. Both mutants carry sml1Δ, which has no effect on S phase checkpoints (Zhao et al., 1998). Primer sites and colour coding for fragments are described in Figure 1A. (D) Spindle length and nuclear separation were monitored with TAT1 antibody and DAPI as described (Shimada et al., 2002) for isogenic wild-type (GA-1296) and mec1-1 (GA-1306) cells. Following G₁ arrest and release into HU, 200 cells were measured at each indicated time point. (E) FACS analysis at 30°C on wild-type and mec1-1 cells after α-factor block and release into 0.2 M HU.

Fig. 5. Loss of sgs1 impairs DNA pol α association at a stalled replication fork. Time course ChIP for DNA pol α was performed as described in Figure 1 for (A) wild-type (GA-1296) and (B) sgs1Δ (GA-2206) cells. In (C), ChIP for Orc2 was performed in wild-type (GA-893) and sgs1Δ (GA-2209) isogenic strains. Primer sites and colour coding for fragments are described in Figure 1A and, in (C), wild-type data are introduced for the ARS305 region (labelled x^WT) as stippled open bars. (D) Cultures of GA-1759 (expressing Myc-tagged Sgs1 and HA-tagged Rpa1) and GA-1699 (expressing Myc-tagged Sgs1) were arrested in G₁ with α-factor and released into S phase in either the presence or absence of 0.2 M HU. Detection of the indicated proteins in WCEs is shown in the top panel. WCEs were used for IP by incubating with Dynabeads coupled to anti-pol α, anti-HA or anti-Myc (see Materials and methods). Stringent washing was followed by elution and western blot analysis with the monoclonal antibodies 24D9 (pol α), 9E10 (Myc for Myc-Sgs1) or 12CA5 (HA for HA-Rpa1). Probing the Sgs1 IP for Rfc3 was also negative (data not shown).

Fig. 6. Proposed mechanisms for stabilizing polymerases at stalled replication forks. When replication stalls, the integrity of the replisome must be maintained to ensure the resumption of fork elongation. Potential inappropriate strand pairing between nascent strands or within a repeat-containing nascent strand might be reversed by Sgs1p helicase. This could suppress recombination and provide ssDNA for Mec1 kinase binding. We do not exclude the participation of other factors, but note that genomic instability increases 1000-fold upon mutation of both sgs1 and mec1 (Myung and Kolodner, 2002).