Standard media for the culture of yeast, SC (synthetic complete, dextrose), SC-HIS (SC lacking histidine), SC-TRP (SC lacking tryptophan), SC-URA (SC lacking uracil), YP (yeast extract, peptone) and YPD (YP, dextrose), are described by Burke et al. (¹⁹). YP(A)D contains YPD with 80 mg/l adenine. CAN plates contain SC medium and 60 mg/l of canavanine.

Relevant yeast strains are listed in Table 1. The mec1-21 strain YA197 (Y620), and the YA184 and YA185 strains used to PCR amplify sml1::URA3 and mec1-Δ::TRP1, respectively, are derived from W303; all other strains are of the S288c background. Strains used to measure SCE contain two overlapping his3 fragments, positioned in tandem at trp1, and were derived from YB163 (²⁰,²¹). Diploid strains were used to measure translocations that were derived from a cross of one haploid (YB109) that contains the his3 fragments on one copy of chromosomes II and IV, and another which did not contain the his3 fragments (YA102) (²²). To measure heteroallelic recombination, we replaced the ade2-101 alleles in YB109 and YA102 with ade2-n (YB318) and ade2-a (YB315), respectively, by two-step gene replacement using the plasmid pKH9 (²³). Heteroallelic recombination was measured by selecting for Ade⁺ recombinants.

We used the mec1-21 missense mutant to measure spontaneous recombination. The original MATamec1-21 strain (Y620) (⁶) was backcrossed 10 times with strains in the S288c background [YB163, YA166 (²¹) and YB315] to generate meiotic segregants that either do (YB312) or do not (YB316, YB314) contain his3 recombinational substrates to measure SCE. We introduced the sml1::KanMX and sml1::URA3 allele in yeast strains by PCR-mediated gene replacement; the primers used for amplifying sml1::KanMX or sml1::URA3 knockout fragments were the same: 5′-CATATCGTTACTGTTTTGGAACATCGC-3′ and 5′-TAAAGGGAAAGGAAAATGCACG-3′. The construction of mec-Δ1 sml1::KanMX (YB327) was described earlier (¹⁷).

To measure translocations and heteroallelic recombination in mec1 strains, mutations were introduced into two haploids by either genetic crosses or by one-step gene replacement (²⁴); one haploid contains the his3 recombination substrates and ade2-n (YB318), while another (YB315) contained ade2-a but no recombination substrates. YB318 was crossed with YB313 to generate the MATa-inc ade2-n mec1-21 meiotic segregant (YB319) that contains the GAL1::his3-Δ5′ and trp1::his3-Δ3′. YB325 is a homozygous mec1-21 diploid that was then used to measure translocations and heteroallelic recombination.

Additional checkpoint mutants were made by either one-step gene disruption (²⁴) or by genetic crosses and screening the phenotype of the appropriate meiotic segregant. The primer pairs used to amplify dun1::KanMX fragments were, 5′AGAAGCCCCTGAATACCATAAATA3′ and 5′CGATGTCAGAGATTTAGAGGAAAAA3′, respectively. We made the mec1-21 dun1::KanMX sml1::URA3 haploid (YB380) by introducing sml1::URA3 into mec1-21 dun1::KanMX (YB369) by one-step gene disruption (²⁴). All gene disruptions were confirmed by PCR.

The rates (events per cell division) of spontaneous SCE, heteroallelic recombination, translocations and mutations in CAN1 were determined by the method of the median (²⁵), as previously performed (²¹). Rates of spontaneous heteroallelic recombination were determined on cells inoculated on YP(A)D, on which there is no growth advantage for Ade⁺ recombinants. Rates of mutations in CAN1 were determined by selecting for resistance to canavanine. We determined the statistical significance by the Mann–Whitney U-test (²⁶).

The methods of measuring dNTPs in yeast are as described in Chabes et al., (¹³). At a density between 0.5 × 10⁷ and 1.5 × 10⁷ cells/ml, ∼1 × 10⁸ cells were harvested by filtration through 25 mm White AAWP nitrocellulose filters (0.8 µm, Millipore AB, Solna, Sweden). The filters were immersed in 500 µl of ice-cold extraction solution (12% w/v trichloroacetic acid, 15 mM MgCl₂) in Eppendorf tubes. The following steps were carried out at 4°C. The tubes were vortexed for 30 s, incubated for 15 min and vortexed again for 30 s. The filters were removed and the supernatants were collected after centrifugation at 20 000g for 1 min and added to 800 µl of ice-cold Freon-trioctylamine mixture [10 ml of Freon (1,1,2-trichlorotrifluoroethane, Aldrich, Sigma-Aldrich Sweden AB, Stockholm, Sweden, 99%) and 2.8 ml of trioctylamine (Fluka, Sigma-Aldrich Sweden AB, Stockholm, Sweden, >99%)]. The samples were vortexed and centrifuged for 1 min at 20 000g. The aqueous phase was collected and added to 800 µl of ice-cold Freon-trioctylamine mixture. The mixture was vortexed and centrifuged as described earlier. A 475 µl aliquot of the aqueous phase was pH-adjusted with 25 µl of 1M NH₄HCO₃, pH 8.9, the deoxyribonucleotides were separated from the ribonucleotides by boronate affinity chromatography (Affi-Gel 601, Bio-Rad) and quantified by HPLC. Another 47.5 µl aliquot of the aqueous phase was mixed with 152.5 µl of water and used for the HPLC quantification of NTP pools. Separation of nucleotides was done on a Partisphere 5 SAX column (PolyLC Inc., Columbia, MD, USA) using a UV-2075 Plus detector (Jasco, Mölndal, Sweden). Nucleotides were isocratically eluted with 2.5% acetonitrile, 0.36 M ammonium phosphate, pH 3.4 buffer.

To detect Rad53 phosphorylation after HU and MMS exposure, protein samples were obtained from log phase cells (A₆₀₀ = 0.5–1) exposed to 200 mM HU or 0.01% MMS for 2 h. Proteins (10 µg) were separated on 10% acrylamide/0.266% bis-acrylamide gels, transferred to nitrocellulose membrane and exposed to a Rad53 antibody (yC-19, Santa Cruz). The secondary antibody was peroxidase conjugated goat anti-mouse IgG (Jackson ImmunoResearch Laboratory, Inc.). The signals were developed using ECL western lightning kit (NEL102, PerkinElmer Life Science, Inc.).

We reasoned that if lower dNTP levels correlate with higher recombination rates in mec1 mutants, then SML1 mutations would decrease homologous recombination in mec1-21 mutants. We previously observed a 6-fold increase in the rate of spontaneous SCE in the mec1-21 mutant (YB312) compared to wild-type (¹⁸). We observed no difference in the rates of spontaneous recombination between sml1 (YB326) and wild-type (YB163); however the 6-fold increase in rates of SCE was reduced to wild-type levels in mec1-21 sml1 (YB329, Table 2).

Rates of heteroallelic recombination and translocations in mec1-21 were 10-fold higher and 23-fold higher than wild-type, respectively (Table 3). We determined whether mec1-21 hyper-recombination between homologs and non-homologs also required SML1. Heteroallelic recombination between ade2 alleles can be visualized using a colony pigment assay, where Ade⁺ recombinants appear as white colony sectors (Figure 2). Heavily white sectored colonies indicated that more heteroallelic recombination occurred in mec1-21 (YB325) than in wild-type (YB348), while the mec1-21 sml1 mutant (YB375) exhibited less visible sectoring than mec1-21 cells (Figure 2).

We then measured rates of translocations and heteroallelic recombination in cells grown on medium containing excess adenine (YPAD) so that there is no growth advantage for Ade⁺ cells. While the rates of heteroallelic recombination and translocations in the sml1 (YB323) and wild-type (YB348) were similar (Table 3), mec1-21 sml1 diploid mutant (YB375) exhibited 8- and 4-fold lower rates of heteroallelic recombination and translocations, respectively, compared to mec1-21. However the rates of recombination in mec1-21 sml1 were still between 2- and 5-fold higher than those observed in wild-type (Table 3). Thus, SML1 deletion in mec1-21 diploid partially suppresses spontaneous heteroallelic recombination and translocations.

The MEC1-mediated pathway for regulating dNTP levels involves Rad53 activation, which in turn leads to Dun1 activation (⁷,⁹,²⁹). By western blots, we observed that Rad53 is partially activated to P-Rad53 in the mec1-21 hypomorph, compared to wild-type, after HU and MMS exposure (Supplementary Figure S1). We observed ∼40% of the Rad53 activation (P-Rad53/Rad53) in mec1-21, compared to wild-type, while in mec1-Δ, the level of activation was <30% (n = 2). These data are consistent with observations that mec1-21 exhibits partial checkpoint activation after exposure to agents that cause DNA replication stress (³⁰).

Considering that DUN1 is required for transcriptional induction of the RNR genes (⁷) and is suggested to maintain basal dNTP levels (⁹), we asked whether mec1-21 dun1 mutants would exhibit higher levels of spontaneous recombination, compared to mec1-21. Although there is no difference between spontaneous rates of SCE in dun1 (YB370) and wild-type, we observed a 15-fold increase in the rate of SCE in the mec1-21 dun1 double mutant, compared to wild-type (Table 2). SML1 mutations conferred significantly lower rates of SCE in mec1-21 dun1 (YB369), but were still 3-fold higher than wild-type. These data indicate that DUN1 suppresses recombination in mec1-21, but a SML1 deletion only partially suppresses the hyper-recombination phenotype of the mec1-21 dun1 double mutant.

We expected that the higher and lower rates of homologous recombination exhibited by S-phase checkpoint mutants would inversely correlate with dNTP levels. We measured dNTP levels in wild-type, sml1, mec1-21, dun1, mec1-21 sml1, mec1-21 dun1 and mec1-21 dun1 sml1 strains (Figure 3). In comparison to wild-type, the S288c sml1 strains exhibited increased dNTP levels over 100%, while the S288c dun1 strains exhibited ∼50–70% reduction in dNTP levels. The mec1-21 levels of dNTPs were ∼50–70% of those observed in wild-type strains. Interestingly, the dNTP levels of mec1-21 dun1 strains were similar to mec1-21. These results are consistent with observations that MEC1 is epistatic to DUN1 in controlling dNTP levels (⁶,⁸). We do not know the lowest dNTP level that maintains viability in yeast; thus, it is possible that basal dNTP levels cannot be reduced further than those observed in mec1-21 dun1.

The basal dNTP levels were similar in sml1, mec1-21 sml1 and mecl-21 dun1 sml1, and increased ∼2-fold, relative to wild-type (Figure 3). The basal level of each dNTP was elevated between 2- and 4-fold in both the mec1-21 sml1 and the mec1-21 dun1 sml1 strains, relative to mec1-21 and mec1-21dun1, which exhibit lower dNTP levels than wild-type (Figure 3). These data indicate that deleting SML1 in both checkpoint mutants and wild-type leads to similar dNTP levels. These results are consistent with observations that MEC1 and DUN1 both function in a pathway to degrade Sml1 (⁸,⁹). Thus, higher dNTP levels correlate with suppression of homologous recombination in mec1-21 mutants.

One hypothesis is that higher dNTP levels facilitate DNA replication at stalled replication forks, preventing replication fork regression or the formation of double-strand breaks. Higher dNTP levels can also lead to higher rates of 4NQO-associated mutagenesis (¹²). We therefore measured rates of spontaneous mutation at CAN1 in wild-type, mec1-21, sml1, mec1-21 sml1, mec1-21 dun1, dun1 and mec1-21 dun1 sml1 strains (Table 4). The rate of spontaneous mutation in wild-type (YB163) was 4 × 10⁻⁷, in agreement with Datta et al. (¹¹). In comparison to wild-type, the mutation rate is the same in sml1 and mec1-21 sml1 but lower in mec1-21 dun1 (P < 0.05). The rates of spontaneous mutation are increased ∼2-fold in mec1-21 dun1 sml1 (P < 0.05), in comparison mec1-21 dun1. Thus, higher dNTP levels correlate with the higher rate of spontaneous mutagenesis in mec1-21 dun1 sml1 but not in sml1 strains.

Both mec1 and dun1 mutants exhibit lower dNTP levels, compared to wild-type, but only mec1-21 exhibits higher SCE recombination rates. We suggest that the higher recombination rates result from replication forks stalling due to inadequate dNTP levels. MEC1 functions to prevent replication fork collapse, especially at slow zones of replication (³), and when ribonucleotide reductase is inhibited by HU (³⁴). We suggest that the MEC1 function in preventing replication fork collapse reduces spontaneous SCE in dun1 mutants. dun1 mutants do exhibit higher rates of spontaneous heteroallelic gene conversion between homologs (¹⁰), and it is possible that sister chromatid gene conversions are not detected in our assay (³⁵). However, in mec1 hypomorphs, stalled replication forks would lead to more replication fork collapse and recombinogenic lesions. Thus, MEC1 has two functions in suppressing recombinogenic lesions: (i) maintaining dNTP levels and (ii) preventing replication fork collapse.

When SML1 is deleted in either mec1-21 or mec1-21 dun1 strains, we speculate that higher dNTP levels reduce replication fork stalling and thus fewer replication forks collapse. However, we observed that mec1-21 sml1 mutants still exhibited some hyper-recombination between homologs or non-homologous chromosomes when dNTP levels were high. It is likely that we can detect higher levels of translocations in mec1-21 sml1 mutants because there is a lower rate of spontaneous translocations, compared to heteroallelic and SCE, in wild-type. Thus, more recombinogenic lesions still occur in mec1-21, compared to wild-type, even at high dNTP levels.

Although dNTP levels in mec1-21, dun1 and mec1-21 dun1 are similar, the rates of spontaneous recombination are synergistically increased in the double mutant, compared to the single mutants. We suggest that there may be two different reasons for these observations. First, although the overall dNTP levels are the same, the rate of dNTP production may be lower in mec1-21 dun1, compared to dun1, leading to more replication stalling. Second, Schollaert et al. (³⁶) reported that CHK1 is required for hydroxyurea resistance in dun1 or dun1 sml1 mutants, suggesting additional DUN1 functions in promoting replication. This possibility is supported by the observation that SML1 deletion only partially suppresses the hyper-recombination phenotype in mec1-21 dun1. We suggest that MEC1 and DUN1 may have redundant roles in suppressing recombinogenic breaks.

While MEC1 may prevent replication fork collapse, a possible function of DUN1 in preventing recombinogenic lesions is to promote translesion synthesis. DUN1 functions in promoting spontaneous mutagenesis in both wild-type and in pol3 mutants (¹¹). We observed that the rate of spontaneous mutation was similar in dun1 and mec1-21 dun1. Thus, it is possible that failure to promote translesion synthesis in mec1-21 dun1 leads to more DNA replication stalling at sites of spontaneous damage, increasing the number of collapsed replication forks and recombinogenic lesions.

4-NQO-associated mutagenesis increases with higher dNTP levels in strains deleted for genes that encode error-prone polymerases (¹²), suggesting that replicative polymerases function as translesion polymerases on non-bulky lesions when dNTP levels are higher (¹²). However, we did not observe higher levels of spontaneous mutagenesis in sml1, compared to wild-type. Considering that Sml1 is degraded in S-phase, the S-phase dNTP levels are likely similar in wild-type and in sml1 mutants (⁸). We speculate that higher dNTP levels may promote translesion synthesis in cells where replication forks frequently stall, as in S-phase checkpoint mutants when the DNA polymerase encounters slow zones of replication. Here, we speculate that higher dNTP levels in mec1-21 dun1 sml1 promote more translesion synthesis and thus reduce recombinogenic lesions.