Yeast Stn1 promotes MCM to circumvent Rad53 control of the S phase checkpoint

Treating yeast cells with the replication inhibitor hydroxyurea activates the S phase checkpoint kinase Rad53, eliciting responses that block DNA replication origin firing, stabilize replication forks, and prevent premature extension of the mitotic spindle. We previously found overproduction of Stn1, a subunit of the telomere-binding Cdc13–Stn1–Ten1 complex, circumvents Rad53 checkpoint functions in hydroxyurea, inducing late origin firing and premature spindle extension even though Rad53 is activated normally. Here, we show Stn1 overproduction acts through remarkably similar pathways compared to loss of RAD53, converging on the MCM complex that initiates origin firing and forms the catalytic core of the replicative DNA helicase. First, mutations affecting Mcm2 and Mcm5 block the ability of Stn1 overproduction to disrupt the S phase checkpoint. Second, loss of function stn1 mutations compensate rad53 S phase checkpoint defects. Third Stn1 overproduction suppresses a mutation in Mcm7. Fourth, stn1 mutants accumulate single-stranded DNA at non-telomeric genome locations, imposing a requirement for post-replication DNA repair. We discuss these interactions in terms of a model in which Stn1 acts as an accessory replication factor that facilitates MCM activation at ORIs and potentially also maintains MCM activity at replication forks advancing through challenging templates. Supplementary Information The online version contains supplementary material available at 10.1007/s00294-022-01228-0.


Introduction
Cells must tolerate various forms of DNA replication stress, ranging from extrinsic mutagens to endogenous physiological perturbations. A particularly severe form of replication stress arises when cellular dNTPs are reduced; this stress can be experimentally induced using the ribonucleotide reductase (RNR) inhibitor hydroxyurea (HU). Reducing dNTPs slows advance of replication forks, greatly increasing the likelihood of replication fork collapse (Poli et al. 2012). The S phase checkpoint is a stress response pathway that is activated to safeguard against such catastrophes. In budding yeast, the S phase checkpoint consists of a core signaling axis of three protein kinases: Mec1, Rad53 and Dun1 (Giannattasio and Branzei 2017;Pardo et al. 2017). Mec1 is recruited to single-stranded DNA (ssDNA) that accumulates at stressed replication forks, after which Mec1 phosphorylates and activates Rad53. Rad53 then phosphorylates Dun1 to complete the signaling cascade. Dun1 plays a multi-faceted role in upregulating RNR, leading to expansion of dNTP pools (Zhou and Elledge 1993;Huang et al. 1998;Zhao and Rothstein 2002;Lee et al. 2008;Wu and Huang 2008). Rad53 controls other aspects of the checkpoint, including delaying activation of DNA replication origins (ORI) that normally fire later in the S phase program (referred to here as Rad53-checked ORIs; Santocanale and Diffley 1998;Shirahige et al. 1998;Feng et al. 2006), stabilizing replication forks (Lopes et al. 2001;Sogo et al. 2002;Cotta-Ramusino et al. 2005;Bermejo et al. 2011;Rossi et al. 2015;Colosio et al. 2016;Gan et al. 2017;Chappidi et al. 2019;Devbhandari and Remus 2020;Cabello-Lobato et al. 2021), and preventing premature extension of the bipolar Communicated by Michael Polymenis. mitotic spindle that assembles in HU-arrested yeast cells (Krishnan et al. 2004;Bachant et al. 2005;Julius et al. 2019). These responses synergize to allow DNA synthesis to proceed at a slow but steady rate in HU (Alvino et al. 2007;Poli et al. 2012;Zhong et al. 2013), to circumvent dNTP depletion (Morafraile et al. 2015), and to retain the capacity for accurate chromosome segregation once the extended S phase has been completed (Feng et al. 2009).
In previous work, we identified Stn1 as an additional protein connected to the budding yeast S phase checkpoint ). Stn1 was initially identified as a component of the conserved Cdc13-Stn1-Ten1 (CST) complex (Grandin et al. 1997;Rice and Skordalakes 2016). In yeast, CST binds telomere DNA repeats and protects chromosome ends from exonuclease digestion during S phase (Garvik et al. 1995;Nugent et al. 1996;Lin and Zakian 1996;Maringele and Lydall 2002;Jia et al. 2004;Bertuch and Lundblad 2004;Zubko and Lydall 2006;Vodenicharov and Wellinger 2006;Xu et al. 2009;Dewar and Lydall 2012;Langston et al. 2020). Yeast Stn1 also binds the Pol12 subunit of the DNA polymerase α/DNA primase complex (Polα; Grossi et al. 2004;Petreaca et al. 2006). The Stn1-Polα interaction plays a conserved role in chromosome end replication by stimulating Polα priming and fill-in synthesis of telomerase-generated ssDNA overhangs (Qi and Zakian 2000;Grossi et al. 2004;Petreaca et al. 2007;Puglisi et al. 2008;Chen and Lingner 2013). Importantly, however, other observations implicate CST in genome-wide aspects of DNA replication, particularly under conditions of DNA replication stress (Stewart et al. 2018). In a previous publication, we showed STN1 overproduction (STN1 OP) causes yeast cells to become extremely sensitive to HU and other replication stressors ). Remarkably, STN1 OP also phenocopies rad53 S phase checkpoint defects in HU, including activation of later-firing, Rad53-checked ORIs and premature spindle extension. Rad53 is activated normally in STN1 OP cells, indicating upstream events in S phase checkpoint signaling are not perturbed by excess Stn1. While this suggests STN1 OP acts downstream or in parallel to Rad53 to antagonize checkpoint effector responses, the underlying mechanisms remain to be defined.
Rad53 checks the firing of late ORIs through phosphorylation and inhibition of two proteins, Dbf4 and Sld3, required for activation and assembly of the Cdc45-GINS-MCM (CMG) replicative helicase (Lopez-Mosqueda et al. 2010;Zegerman and Diffley 2010;Duch et al. 2011). Dbf4 is a cyclin-like activator for Cdc7, the yeast Dbf4-dependent protein kinase (DDK; Jackson et al. 1993;Bousset and Diffley 1998). The DDK plays an essential role in ORI firing by phosphorylating paired Mcm2-7 hexamers (MCM) at licensed ORIs (Labib 2010). One consequence of MCM phosphorylation is to recruit Sld3 (Fang et al. 2016;Deegan et al. 2016). Sld3, in parallel, is phosphorylated by S phase forms of Cdk1 (S-CDK), leading to phospho-adapted interactions that recruit Cdc45 and GINS (Tanaka et al. 2007;Zegerman and Diffley 2007;Muramatsu et al. 2010). MCM activation corresponds with conformational changes that melt ORI DNA, with each strand entering the interior core of one MCM hexamer in the necessary configuration for bidirectional DNA unwinding (Georgescu et al. 2017;Douglas et al. 2018;Meagher et al. 2019). The DDK and S-CDK, thus, control parallel pathways activating MCM. Rad53 antagonizes both pathways in response to replication stress, imposing a robust check on further ORI firing.
An additional Rad53 checkpoint function circumvented by STN1 OP is to prevent defective extension of the mitotic spindle during an extended S phase. Restraint of spindle extension has generally been considered to be controlled through a separate Rad53 cell cycle arrest pathway, unrelated to regulation of ORI firing and replication fork stabilization. In a recent study, however, we presented evidence that spindle extension in HU-treated rad53 mutants is actually a consequence of a primary defect in DNA replication control (Julius et al. 2019). First, double mutant combinations predicted (rad53 mcm2-1, rad53 mcm5-1) or demonstrated (rad53 dbf4-zn) to reduce ORI firing in HU, or to suppress exonuclease processing of reversed replication forks (rad53 exo1-∆), suppressed the rad53 spindle extension defect. Second, the dbf4-zn allele was preferentially defective for initiating ORI firing adjacent to centromeres (CENs). Third, unregulated Exo1 activity in rad53 mutants generated CEN ssDNA and perturbed kinetochore assembly. Based on these findings, we proposed that the critical role for Rad53 in restraining spindle extension in HU is to stabilize replication forks in proximity to CENs. In the absence of this protective function, exonucleolytic degradation of CEN DNA disrupts kinetochore integrity and S phase spindle force balancing mechanisms. Here we have utilized this revised conception of the spindle extension defect in HU as a convenient genetic readout to assess pathways through which STN1 OP antagonizes the S phase checkpoint. Our results indicate Stn1 is likely to act in concert with both the DDK and the MCM complex to efficiently activate ORI firing-a function revealed most prominently when ORIs fire in an unscheduled manner in the absence of the S phase checkpoint.

Yeast culture
Relevant S. cerevisiae strains and plasmids are listed in figure legends. Cells were cultured in standard formulations of yeast extract/peptone/dextrose (YPD) and synthetic complete minimal (SC) media, with 2% glucose or 2% galactose as a carbon source. Cultures for microscopy were 1 3 supplemented with 50 μg/mL adenine to quench autofluorescence associated with the ade2 mutation in our strain backgrounds. For G 1 synchronization/release, cells were treated with 10 μg/mL α-factor (Bio-Synthesis Corp.), typically for 1.5 h, washed in water, and released into desired culture media. Yeast transformation, strain construction, and other genetic manipulations were performed according to standard techniques (Guthrie and Fink 1991). HU was purchased from either Sigma-Aldrich or Fisher Scientific. Yeast twohybrid analysis was performed using strain PJ69-4a (James et al. 1996).

Spindle length analysis
Spindle length distributions in fixed cell samples was performed as previously described (Bachant 2005). Cells harboring SPC42-GFP were released from G 1 arrest into fresh media containing 200 mM HU. After 2.5 h, culture aliquots were briefly (1-5 min) fixed either using 1% formaldehyde diluted in phosphate buffered saline (PBS). Samples were washed into PBS and stored at 4 °C. DNA staining was performed using 4′6-diamidino-2-phenylindole (DAPI; Vecta-Shield, Vector Laboratories). Cells were visualized on either Nikon E-800 or Nikon Eclipse 80i microscopes equipped with florescence optics and 100 × (Plan Apo, 1.40 NA) objectives. The distance between Spc42-GFP spindle pole foci and bud circumference measurements were performed using the MetaMorph (Molecular Devices) suite of software tools.

Chromosome spreads
To detect Stn1-HA on chromatin, 5 mL triplicate cell cultures were grown to logarithmic phase in appropriate selective media. Cells were collected by brief centrifugation and pellets were resuspended in 1 mL ZK buffer (25 mM Tris pH 7.5, 0.8 M KCl) supplemented with 40 μL of 1 M DTT, incubated for 2 min at room temperature. Samples were spheroplasted by addition of 5 μL of zymolyase solution (20 mg/mL zymolyase 100 T, 2% glucose, 50 mM Tris pH 7.5) and 2 μL of BME, and incubated for 15 min at 30 °C. The spheroplasted cells were washed with ice cold MES solution (1 M Sorbitol, 0.1 M MES pH 6.5, 1 mM EDTA, 0.5 mM MgCl 2 ), then resuspended in 300 μL of MES solution. 20 μL of the cells were spotted onto a pre-cleaned glass slide, followed by addition of 40 μL of PFA solution (3% paraformaldehyde, 3.4% sucrose) and 80 μL of 1% lipsol. After 2 min of lysis, an additional 80 μL of PFA solution was added, and lysates were spread across the glass slides with a clean glass Pasteur pipette. Slides were dried at room temperature overnight. Prior to immunostaining, slides were washed with 0.2% Photoflo (Kodak) for 30 s and PBS for 5 min, then blocked with 350 μL of TBS containing 10 mg/ mL BSA for 15 min at 4 °C. Excess blocking solution was drained, and 80 μL of primary antibody (mouse anti-HA, 12CA5 from Roche) was added at a 1:200 dilution in TBS containing 10 mg/mL BSA. Cover slips were applied, and slides were incubated at 4 °C in a wet chamber overnight. Slides were washed twice with TBS, drained, and 80 μL of secondary antibody (FITC-conjugated goat anti-rat from Sigma) was added at a 1:500 dilution in TBS containing 10 mg/mL BSA. Cover slips were added, and slides were stored in the dark from this point on. After a 2 h incubation at 4 °C, slides were washed with TBS twice, and dried at room temperature for 4 h. DNA was stained with DAPI as described above, cover slips were applied, and samples were visualized by fluorescence microscopy.

Cell viability assays
5 mL cell cultures were grown to logarithmic phase. Cell concentration was determined by hemocytometery, after which cells were diluted into fresh media containing 200 mM HU. An aliquot of the culture was immediately removed and diluted as calculated so that 100 μL of the dilution yielded ~ 500 colony forming units when plated on solid media, providing an initial time point. At desired times, additional culture aliquots were removed, diluted in a similar 1 3 fashion and plated to determine colony forming units. Colonies were counted after incubation for 3-5 days, and the fraction of surviving cells relative to the initial timepoint was determined.

In situ Klenow primer extension on chromosomal ssDNA and Southern blotting
To detect ssDNA regions of chromosomes, we modified a previously described in situ ssDNA labeling method (Feng et al. 2011). Logarithmic phase cultures were washed and resuspended in 50 mM EDTA, and the cell concentration of each sample was determined by hemocytometery. For each sample, ~ 10 9 cells were placed in a final volume of 500 μL. Cell suspensions were warmed to 55 °C and mixed with 500 μL of low melt agarose (Invitrogen) that had been dissolved in a 1:100 dilution of 1× TBE. The agarose/cell mixture was pipetted into plug molds and allowed to solidify at room temperature for 15 min. Solidified plugs were treated with 5 mL of spheroplasting solution (1 M sorbitol, 20 mM EDTA, 10 mM Tris pH 7.5, 14 mM BME, 0.5 mg/ mL Zymolyase 20 T) for 4 h at 37 °C. Plugs were washed with SDS solution (1% SDS, 100 mM EDTA, 10 mM Tris pH8) twice for 15 min each, and incubated in SDS solution at 37 °C overnight with gentle rocking. The following day, plugs were washed with NDS solution (1% sarkosyl, 10 mM Tris base, 0.5 M EDTA pH 9.5) 3 times for 30 min each, followed by 5 washes with TE for 30 min each, then stored in 4 °C. Multiple plugs were prepared for each sample.
To perform the Klenow reactions, two plugs per sample were each pre-equilibrated in 5 mL of TMB (50 mM Tris pH 6.8, 5 mM MgCl 2 , 10 mM BME) for 30 min at room temperature. One plug was mixed with 400 μL of TMB buffer, 10 μL of dNTPs (10 μM each dNTP), 10 μL of random hexamer primers at 10 μM (Thermo Scientific), 100 units of Exo − Klenow polymerase (New England BioLabs), and 50 μL of 10 × Klenow buffer. The second plug was treated identically, but no Klenow was added. The samples were incubated at 37 °C for 2 h and then washed with TE. Plugs were then pre-equilibrated with 1 × β-agarase buffer for 30 min on ice, heated to 65 °C to melt agarose, and treated with 5 units of β-agarase (New England BioLabs) for 1 h at 42 °C. The salt concentration was adjusted to 0.5 M NaCl, 0.8 M LiCl, 0.3 M NaO-Ac, samples were cooled on ice for 15 min, and DNA was precipitated with isopropanol. The DNA was then washed with cold 70% ethanol, dried, and resuspended in 40 μL of TE.
Prior to electrophoresis, reaction products were denatured by addition of 10 μL of 1 M NaOH and 1 μL of 0.5 M EDTA, followed by boiling at 95 °C for 5 min. Samples were cooled on ice for 5 min then electrophoresed on large format 1% agarose gel overnight at 50 V. DNAs were transferred to a nylon membrane (Hybond-XL, Amersham) overnight and crosslinked with 120 mJ of UV (Stratagene). The membrane was blocked with Church's buffer (1% BSA, 1 mM EDTA, 0.5 M phosphate buffer, 7% SDS) overnight at 55 °C with gentle rocking. 25 μL of P 32 radiolabeled TG 1-3 probe was added, and the blot was incubated overnight at 55 °C with gentle rocking. The following day, the blot was washed three times with 1 L of washing solution (4 × SSC, 0.1% SDS), and exposed to X-ray film for 5 days at −80 °C. After development, the membrane was stripped by boiling in 0.1% SDS three times for 15 min each, then blocked with Church's buffer as before. 25 μL of P 32 radiolabeled rDNA probe was added, and the blot was processed for autoradiography as before.

STN1 overproduction displays genetic interactions with DUN1, RNR2 and MCM
To identify genetic pathways affected by STN1 OP, we chose an approach based on our previous finding that STN1 OP checkpoint defects could be suppressed in a pol12-40 mutant ). The Pol12-40 mutant protein is partially defective for binding Stn1 (Petreaca et al. 2006; see schematic of Stn1 domains and interactions, Fig. 1A), and retention of excess Stn1 on chromatin spreads is greatly reduced in pol12-40 mutants ), implying STN1 OP acts through POL12. Extending this logic, we over-expressed STN1 under control of the galactose-inducible GAL promoter in a collection of mutants defective for DNA replication control or tolerating DNA replication stress. Transformants were evaluated on galactose media for enhancement/suppression of mutant growth defects or enhancement/suppression of STN1 OP HU sensitivity. One interaction we identified was that STN1 OP was synthetically lethal with a dun1-∆ mutant, even in the absence of HU (Fig. 1B). Synthetic lethality with dun1-∆ was also observed following OP of a Stn1 fragment (Stn1 288−494 ; amino acids 288-494) comprising two C-terminal winged helix domains (Fig. 1A). stn1 288−494 OP was assessed because this is the minimal STN1 region required for OP checkpoint phenotypes ). We additionally examined the effect of STN1 and STN1 288−494 OP in a recessive loss of function rnr2-1 temperature sensitive mutant (Zhou and Elledge 1992). OP of STN1 and STN1 288−494 strongly inhibited rnr2-1 growth on galactose media, even at a permissive temperature of 23 °C (Fig. 1C). These observations suggest STN1 OP imposes an essential requirement for Dun1-mediated upregulation of RNR.
We also identified genetic interactions between STN1 OP and mcm2-1, mcm5-1 and mcm7-1, recessive, temperature sensitive alleles that compromise MCM activity and DNA synthesis (Tye 1999). As MCM is a multimeric complex, it is notable that both mcm mutations and altered MCM expression produce a complex assortment of genetic interactions, including co-suppression and dosage enhancement (Yan et al. 1991). Additionally, MCM7 has a second function as a cell cycle regulated transcriptional repressor, with the mcm7-1 mutation increasing the expression of other MCM genes (Fitch et al. 2003). Against this backdrop, we observed that STN1 OP in mcm2-1 and mcm5-1 partially alleviated the extreme HU sensitivity associated with STN1 OP, allowing growth at up to 25 mM HU at a mcm semi-permissive temperature of 30 °C ( Fig. 2A). Moreover, mcm2-1 and mcm5-1 STN1 OP transformants showed a marked improvement in their ability to recover from transient exposure to 200 mM HU at 30 °C (Fig. 2B). Immuno-blotting revealed similar amounts of OP Stn1 accumulated in WT, mcm2-1 and mcm5-1 transformants (Fig. 2C). The ability of mcm2-1 and mcm5-1 to allow STN1 OP cells to recover from transient HU exposure is notable, as our prior work indicates that a dramatic loss of survival following acute HU exposure is typically a consequence of simultaneously deregulating ORI firing and destabilizing replication forks (Desany et al. 1998;Alcasabas et al. 2001;Julius et al. 2019).
As a further connection between STN1 and MCM, we found STN1 OP acted as a dosage suppressor of mcm7-1 (Fig. 2D); a similar suppression of mcm2-1 or mcm5-1 was not observed (not shown). mcm7-1 suppression did not occur following OP of either N-or C-terminal Stn1 regions, indicating the effect required full length Stn1 (Fig. 2D). In sum, these results indicate STN1 OP exhibits complex interactions with MCM. On the one hand, reduced MCM function alleviates STN1 OP HU toxicity. On the other, STN1 OP partially restores viability to at least one MCM loss of function mutant strain.

STN1 OP acts through MCM to induce S phase checkpoint defects
As described in the Introduction, we previously characterized mcm2-1 and mcm5-1 as mutations that suppressed the spindle extension phenotype of HU-treated rad53 mutants (Julius et al. 2019). To see if a similar relationship was observed with STN1 OP, WT, mcm2-1 and mcm5-1 STN1 OP strains were released from a G 1 block into media containing 200 mM HU. Spindle length distributions were evaluated in fixed cells after 2.5 h using a GFP tagged spindle pole body protein (Spc42-GFP; representative micrographs of WT, rad53 and STN1 OP spindle morphologies in HU can be found in (Bachant et al. 2005;Gasparyan et al. 2009;Julius et al. 2019). As expected, WT cells transformed with a vector control displayed the short (1-2 μm) spindles characteristic of HU-arrested cells (% spindles ≥ 3 μm = 6), while 53% of WT/pSTN1 transformants exhibited a heterogenous range of spindle lengths ≥ 3 μm ( Fig. 3A; 3 μm is our threshold for an extended spindle). In comparison, mcm2-1/pSTN1 and mcm5-1/pSTN1 transformants exhibited 1% and 5% spindle extension, respectively (Fig. 3A), a similar extent of suppression to that of HU-treated mcm2-1 rad53 and mcm5-1  figure). Transformants were grown to saturation in selective media. Tenfold serial dilutions (black triangles) were stamped onto selective glucose or galactose solid media containing the indicated concentrations of HU. Duplicate sets of plates were cultured at 23°, 30° and 36 °C. C rnr2-1 (Y221) cells were transformed and analyzed as in (B) rad53 strains (Julius et al. 2019). Unlike the situation with pol12-40 suppression of STN1 OP defects ), mcm2-1 and mcm5-1 did not reduce Stn1 binding to spread chromatin preparations (Fig. 3B).

Loss of STN1 function suppresses rad53 S phase checkpoint defects
The results presented so far are consistent with the idea that STN1 OP activates the MCM complex to induce firing of checked ORIs in HU. In this regard, it is interesting that OP of DBF4 is similar to OP of STN1 in being sufficient to circumvent Rad53 control of both checked ORI firing and spindle extension in HU (Mantiero et al. 2011;Tanaka et al. 2011;Julius et al. 2019). Furthermore, in searching for potential physical interactions between Stn1 and DNA replication factors we identified a two-hybrid interaction between Stn1 and Dbf4 (Supplemental Fig. 1). From this, one possibility is that excess Stn1 circumvents the Rad53 check on ORI firing in HU through DDK activation of MCM, prompting us to examine the consequences of eliminating DBF4 in STN1 OP cells. Such a genetic test is possible using a gain of function mcm5 allele, mcm5-bob1, that bypasses the requirement for the DDK in ORI firing, allowing cells to proliferate in the absence of either Dbf4 or Cdc7 (Hardy et al. 1997;Hoang et al. 2007;Miller et al. 2014). Importantly, however, the Rad53 check on late ORI firing remains largely intact in mcm5-bob1 mutants due to the necessity of also circumventing Rad53 inhibition of Sld3 (Zegerman and Diffley 2010), and ~ 90% of mcm5-bob1 cells arrest with short spindles in HU (Fig. 5C). Thus, STN1 OP in dbf4-∆ mcm5-bob1 permits an epistatic test of whether STN1 OP defects are directed exclusively through DBF4.
stn1 mutants accumulate ssDNA at telomeres, arising from defects in telomere replication and chromosome end protection (Grandin et al. 1997). If Stn1 functions more globally in DNA replication, we hypothesized Stn1 might also prevent accumulation of ssDNA at interior chromosomal regions. To test this, we modified a previously described procedure for in situ labeling of chromosomal ssDNA (Feng et al. 2011). stn1 1−281 , WT and mec1-21 cells were embedded in agarose, permeabilized, and DNAs complementary to ssDNA regions were synthesized using random oligonucleotide primers and Klenow DNA polymerase. After a denaturation step and nucleic acid recovery, short primer extension products were separated from the larger mass of chromosomal DNA by electrophoresis and analyzed by Southern blotting. As expected, stn1 1−281 , but not WT or mec1 controls, displayed a ssDNA signal when the blots were probed with a telomeric DNA repeat sequence (TG 1-3 panel, Fig. 7A). Re-probing the blot with a repetitive sequence within the rDNA locus revealed stn1 1−281 also accumulated ssDNA at this internal chromosomal region Fig. 6 stn1 1-186 is a suppressor of rad53 spindle extension in HU. A stn1 1−186 , rad53-21 (hc2804) and rad53-21 stn1 1−186 (hc2806) strains harboring SPC42-GFP were released from a G 1 arrest into 200 mM HU media. The distance between Spc42-GFP spindle pole foci was evaluated at 2.5-h post-release. Numbers on histograms indicate the percentage of cells with spindles ≥ 3 μm. B The same strains were processed as in (A), except in this experiment both bud circum-ference (as a metric for elapsed time in S phase) and spindle length were quantified. Color coding on graphs: cells with spindles ≤ 3 μm, green; cells with spindles ≥ 3 μm and bud circumferences ≤ 15 μm (small-to medium-budded cells), red; cells with spindles ≥ 3 μm and buds ≥ 15 μm (medium-to large-budded cells), orange. The percentage of total cells with extended spindles is shown on the right-hand side of each graph 1 3 (rDNA panel, Fig. 7A). ssDNA at the rDNA locus was also observed in mec1-21, which is known to accumulate replicative DNA damage (Feng et al. 2011). Treating stn1 1−281 cells with HU showed that blocking DNA synthesis considerably reduced both telomeric and rDNA signals. In contrast, mec1−21 cells, which fail to check ORI firing and experience replication catastrophes in HU, did not display such a reduction. HU-treated mec1−21 samples also showed a fainter rDNA signal in the absence of Klenow (Fig. 7B). To explain this, chromosome fragmentation and nuclease assault in HU-treated mec1 cells (Cha and Kleckner 2002;Feng et al. 2009Feng et al. , 2011 may generate rDNA fragments that are small enough to enter the gel and be visualized by our method. stn1 1−281 was used for the experiment shown in Fig. 7A because this allele causes a more severe growth defect than stn1 1−186 . However, stn1 1−186 was subsequently found to also accumulate ssDNA at the rDNA locus (not shown).

Discussion
In this study, we used premature spindle extension in HU to assess the genetic basis for S phase checkpoint defects in STN1 OP cells. This approach was predicated on our recent observations that mutations that suppress rad53 checkpoint defects in DNA replication control (mcm2-1, mcm5-1, dbf4zn, exo1-∆) co-suppress defective spindle extension, suggesting a mechanistic coupling between these phenotypes (Julius et al. 2019). Based on this previous study, we proposed spindle extension is an indirect consequence of replication fork catastrophes in the vicinity of centromeres. These catastrophes occur due to simultaneously deregulating two key Rad53 effector responses: loss of the check on ORI firing, which exacerbates nucleotide depletion, and loss of replication fork stability, which generates ssDNA. A key finding of the work presented here is that spindle extension induced by STN1 OP in HU is suppressed by the same set of DNA replication and nuclease mutations that suppress rad53, indicating gain of STN1 and loss of RAD53 deregulate similar processes in DNA replication control. As discussed below, our data cumulatively suggest DNA replication functions for Stn1 likely converge on the MCM complex.

Relationships between STN1, RAD53 and the DDK in the S phase checkpoint
In STN1 OP cells, Rad53 exhibits the electrophoretic mobility shift characteristic of Rad53 auto-phosphorylation, indicating checkpoint signaling upstream of Rad53 is not disrupted by excess Stn1 ). Since we show Stn1 likely binds Dbf4, it remains possible that STN1 OP interferes with the ability of Rad53 to complex with and phosphorylate this effector substrate. Alternatively, STN1 OP could interfere with the checkpoint indirectly, by circumventing Rad53 regulatory mechanisms. Our observations lead us to favor the latter interpretation, with Stn1 Fig. 7 Accumulation of ssDNA damage in stn1 mutants. A WT (hc160), stn1 1−281 (hc671) and mec1-21 (hc30) strains were grown to logarithmic phase at 30 °C or additionally treated with media containing 200 mM HU for 3 h. For each sample, a fixed number of cells were suspended in agarose plugs, spheroplasted, and either treated or not treated with hexameric primers and Klenow DNA polymerase. Extension products corresponding to chromosomal ssDNA were analyzed by Southern blotting with either telomeric (upper panel, TG 1-3 ) or rDNA (lower panel) probes. B WT (hc160), stn1 1−281 (hc671), rad6-∆ (JBY285), stn1 1−281 rad6-Δ (hc2636) strains all harboring pSTN1-URA3 (pVL1046) were grown to saturation in selective media and tenfold serial dilutions were stamped onto either Ura − or 5'-FOA containing media. Plates were incubated for 4 days at 30 °C functioning as an accessory replication factor rather than a Rad53 anti-checkpoint. First, STN1 OP is lethal to dun1 and rnr2 mutant strains, suggesting an increased demand on dNTP pools even in the absence of HU. Second, a nonessential role for Stn1 in DNA replication is supported by the observations that stn1 mutants accumulate ssDNA outside of telomeres and impose a requirement for post-replication DNA repair. Third, STN1 OP acts as a dosage suppressor of mcm7-1. The basis for this suppression remains to be determined. Since mcm7-1 abolishes transcriptional repression of other MCM genes, it is possible overproduced Stn1 may suppress mcm7-1 through processes other than restoring Mcm7 activity within the Mcm2-7 hexamer. Fourth, the epistasis of the stn1 1−186 phenotype in rad53-21 in stn1 1−186 rad53-21 double mutants indicates that Stn1 is necessary to manifest rad53 HU phenotypes. One genetic interpretation is thatlike Dbf4, Sld3, and Exo1-Stn1 is another effector that is negatively regulated by Rad53, with the absence of inhibition leading to deregulated ORI firing and fork destabilization at centromeric regions (Supplemental Fig. 2). Whether Stn1 is a direct target of Rad53 in the S phase checkpoint will be important to assess in future studies.
One possibility we addressed in this study is that STN1 OP might activate ORI firing in HU by promoting DDK activity towards MCM substrates. This is because, first, we detected a likely physical interaction between Dbf4 and Stn1, and, second, to our knowledge, only STN1 OP and DBF4 OP have been shown to be sufficient to override Rad53 inhibition of ORI firing Mantiero et al. 2011;Tanaka et al. 2011;Julius et al. 2019). How increased DDK activity circumvents the parallel Rad53 check on Sld3 is not clear. The minimal essential role for the DDK in activating MCM is to relieve an auto-inhibitory activity with the N-terminus of Mcm4 (Sheu and Stillman 2010). However, deletion of this domain (Mcm4 ∆74-174 ) is not sufficient to bypass the Rad53 check on late ORI firing; it is also necessary to simultaneously bypass Rad53 inhibition of Sld3 (Sheu et al. 2016). The same pattern is observed with mcm5-bob1, which is also proficient for the Rad53 check on ORI firing (Zegerman and Diffley 2010). DBF4 OP enriches Sld3 and other initiation factors at late firing ORIs (Tanaka et al. 2011). Such enrichment could conceivably circumvent the Rad53 block on Dbf4 and Sld3 at the S phase checkpoint.
Although Stn1 may function in a positive-acting manner with the DDK, the findings presented here indicate that the ability of STN1 OP to force checked ORI firing in HU cannot be directed exclusively through the DDK. dbf4-∆ mcm5-bob1/STN1 OP cells still exhibit spindle extension in HU which, from a genetic standpoint, argues STN1 OP must have an additional target(s). We therefore propose Stn1 acts in a parallel, potentially reinforcing, pathway with the DDK to activate ORI firing (Supplemental Fig. 2). This is supported by the observation that mcm2-1 and mcm5-1 are the strongest suppressors of HU spindle extension in both rad53 mutants and STN1 OP cells, suggesting STN1 and RAD53 ultimately converge on MCM. In sum, our genetic analysis of gain and loss of STN1 function is consistent with Stn1 participating in a nexus of interactions involving MCM/ CMG, POL12/Polα, and the DDK. During HU challenge, the effect of excess Stn1 within this network is to counteract key aspects of Rad53 DNA replication control. Conversely, when Stn1 fails to act within this network, cell accumulate ssDNA indicative of replicative DNA damage, even in the absence of exogenous replication stress.

Speculative roles for Stn1 in MCM function
While our study does not address the molecular basis for how Stn1 activates MCM, several observations warrant discussion. In a potentially related manner to budding yeast, OP of Stn1 in human cells stimulates firing of dormant ORIs during HU recovery (Wang et al. 2014), while Stn1 depletion decreases ORI activation after replication stress (Wang et al. 2012). Additionally, in a recent study human Stn1 was shown to bind to Mcm4 and Mcm7, as well as to Ctf4/And1 (Wang et al. 2019). Ctf4/And1 functions as an adaptor that links Polα to the replisome and also potentially tethers bidirectional replisomes together (Yuan et al. 2019). In yeast, the Ctf4-Polα linkage is preferentially involved, although not essential, for initiating lagging strand synthesis (Porcella et al. 2020). Knockdown of human Stn1 was found to reduce And1 chromatin association during recovery from replication stress, leading to a model where CST provided a backup mechanism to recruit And1/Polα, thereby stimulating initiation of lagging strand synthesis under challenging firing conditions (Wang et al. 2019). Stn1 has also been shown to stimulate Polα priming/catalysis and replication of ssDNA templates in vitro (Goulian and Heard 1990;Nakaoka et al. 2012), and, from our previous work, disruption of Stn1 binding to the Pol12 subunit of Polα rescued STN1 OP S phase checkpoint defects ). It, therefore, seems likely that there is a conserved role for Stn1 in ORI firing that is closely coupled to Pol12/Polα. If Stn1 plays a conserved role in stimulating Polα activity during stress-related ORI firing, how might a connection with MCM be involved? One possibility is that STN1 facilitates a coupling between CMG activation and Polα recruitment and priming. As revealed in human cells, this may involve bridging interactions between Stn1, MCM and And1/Ctf4 (Wang et al. 2019). Given our finding Stn1 likely also interacts with Dbf4, a related possibility is that Stn1 helps maintain an activated status for MCM. DDK-mediated phosphorylation of MCM is counteracted by the Glc7/PP1 phosphatase, which is recruited to ORIs through the Rap1interacting factor Rif1 (Boos and Ferreira 2019). Thus, Stn1 may be recruited to the replisome not just to facilitate lagging strand synthesis under challenging conditions but also to counteract Rif1 and maintain MCM phospho-activation (Supplemental Fig. 2). Recent evidence suggests DDK activity towards MCM is involved not just in the initial activation of MCM during ORI firing, but also in maintaining CMG activity at challenged replication forks (Cabello-Lobato et al. 2021;Dolson et al. 2021). The role of Rad53 in stabilizing replication forks in HU is also be closely coupled to CMG, with Rad53 blocking CMG advance beyond the site of leading strand synthesis during replication stress (Gan et al. 2017;Devbhandari and Remus 2020). Although the Rad53 mechanism enforcing this coupling is not yet clear, Rad53 docking sites within CMG are likely to be involved (Can et al. 2019). We therefore speculate the similarities between gain of STN1 and loss of RAD53 encompasses both functional populations of MCM. In this view, Stn1 plays an accessory role in activating MCM conversion to CMG during ORI firing and stimulates CMG advance through difficult templates. Disruptions to these functions could lead to the accumulation of ssDNA gaps during S phase and the requirement for post-replication DNA repair we detected in stn1 mutants.
Funding National Institutes of Health, National Science Foundation, Cancer Research Coordinating Committee.

Conflict of interest
The authors declare they have no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.