Separable functions of Tof1/Timeless in intra-S-checkpoint signalling, replisome stability and DNA topological stress

Abstract The highly conserved Tof1/Timeless proteins minimise replication stress and promote normal DNA replication. They are required to mediate the DNA replication checkpoint (DRC), the stable pausing of forks at protein fork blocks, the coupling of DNA helicase and polymerase functions during replication stress (RS) and the preferential resolution of DNA topological stress ahead of the fork. Here we demonstrate that the roles of the Saccharomyces cerevisiae Timeless protein Tof1 in DRC signalling and resolution of DNA topological stress require distinct N and C terminal regions of the protein, whereas the other functions of Tof1 are closely linked to the stable interaction between Tof1 and its constitutive binding partner Csm3/Tipin. By separating the role of Tof1 in DRC from fork stabilisation and coupling, we show that Tof1 has distinct activities in checkpoint activation and replisome stability to ensure the viable completion of DNA replication following replication stress.


Introduction
The faithful replication of the genome by the DNA replication machinery is hindered by a range of exogenous and endogenous factors that are capable of disrupting replication forks.
Such cellular events are a prominent feature of the early stages of cancer and have been collectively referred to as replication stress (RS) (Halazonetis et al, 2008). RS occurs when either the helicase or polymerase activities of the replisome are impeded. Potential impediments include chemical changes to the DNA, stable DNA binding protein complexes, nucleotide deficiency and DNA topological stress (Zeman & Cimprich, 2014;Keszthelyi et al, 2016).
On encountering RS, the replisome is thought to be stabilised at the replication fork until the impeding context is removed or bypassed and replication restarted (Marians, 2018). A failure to stabilise the replisome is associated with erroneous processing of the replication fork, leading to either toxic recombination intermediates or disruption of replication restart.
The activation of the checkpoint pathways is essential for replication fork stabilisation and restart (Saldivar et al, 2017). Following RS, ATR type kinases (Mec1 in S. cerevisiae) are activated by the accumulation of excessive RPA-coated single stranded DNA (Costanzo et al, 2003;Zou & Elledge, 2003). ATR/Mec1 then acts with mediator proteins to activate effector kinases that promote fork stability, regulate nucleotide metabolism, and minimise the firing of further replication origins (Pardo et al, 2017). The mediator proteins required for effector kinase activation can act either at the fork, commonly referred to as the DNA replication checkpoint (DRC), or independently of the fork, referred to as the DNA damage checkpoint (DDC). Mediator proteins in the DRC include the replisome proteins Mrc1/Claspin (Sc Mrc1) (Alcasabas et al, 2001) and Tof1/Swi1/Timeless (Sc Tof1) (Foss, 2001). Tof1 appears to act as a nexus for various processes relating to both stabilising the replication fork during RS and ensuring faithful chromosome inheritance. In addition to its role in activating effector kinases, Tof1, along with Mrc1 and the Tof1 interacting protein Csm3/Swi3/Tipin (Sc Csm3) are required to stimulate DNA replication in vivo and in vitro (Tourriere et al, 2005;Yeeles et al, 2017;Aria et al, 2013;Cho et al, 2013), and for coupling of helicase and polymerase activities (Katou et al, 2003;Bando et al, 2009;Errico et al, 2007;Smith et al, 2009;Lou et al, 2008). Independently of Mrc1, the Tof1-Csm3 complex is also required for pausing of replication forks at stable protein-DNA barriers (Tourriere et al, 2005;Calzada et al, 2005;Hodgson et al, 2007) and to focus the action of topoisomerases ahead of the fork to prevent excessive fork rotation during DNA replication (Schalbetter et al, 2015). This latter activity has been hypothetically linked to the observation that the C terminus of Tof1 and the type IB topoisomerase Top1 interact (Park & Sternglanz, 1999), potentially ensuring a relative enrichment of Top1 at the replication fork (Schalbetter et al, 2015;Keszthelyi et al, 2016). Tof1 is also required to recruit Top1 to replicating regions in vivo (Shyian et al, 2019). In addition, both tof1 and csm3 (but not mrc1 ) cells are acutely sensitive to the chemotherapeutic drug camptothecin (CPT) that stabilises the covalently DNA-bound intermediate of Top1 (Redon et al, 2006;Rapp et al, 2010;Hosono et al, 2014). The potential linkage between Tof1-Csm3-dependent activities of fork pausing, resolving DNA topological stress and resistance to CPT treatment is unclear.
To attempt to define the molecular relationships between the range of processes that involve Tof1, we have generated a series of truncations across the C terminus of Tof1 and assessed whether these mutations are sufficient for the functions of Tof1 in activating the DRC, replication fork pausing, helicase-polymerase coupling and preventing excessive fork rotation in response to DNA topological stress.

Results
In order to dissect which domains of the 1238 amino acid (aa) long Tof1 protein are required for each of its characterized functions we generated a series of premature stop codons in a TOF1 ORF sequence optimised for budding yeast expression (Yeeles et al, 2017). Premature stop codons were introduced at aa 627, aa 762, aa 830 aa 997 and aa 1182 (figure 1A). The truncated proteins were expressed from the endogenous TOF1 locus. Truncation of Tof1 at each of these positions generated proteins of the predicted size, which were expressed to levels comparable both with exogenous codon-optimised wt TOF1 and with endogenous Tof1 protein (figure 1B).
We first set out to establish which of the truncated Tof1 proteins could suppress the excessive fork rotation phenotype of tof1 deleted cells, as visualised by Southern blotting of replication products (Schalbetter et al, 2015). In cells with wildtype function of TOF1, episomal plasmids (figure 2A) accumulate only modest levels of DNA catenanes during S phase following Top2 inactivation. This demonstrates that fork rotation is relatively infrequent and therefore that DNA topological stress is primarily resolved by Top1 ahead of the replication fork in this context (Schalbetter et al, 2015). In contrast, tof1 cells accumulate hypercatenated plasmids during S phase indicating that, in this context, DNA topological stress is resolved far more frequently by fork rotation and action of topoisomerases behind the fork (figure 2B). To assess whether the truncated proteins were capable of suppressing hypercatenation and thus restoring the primacy of Top1 action ahead of the fork, we replaced the TOF1 gene with each of the tof1 truncation alleles encoding the truncated proteins into tof1 top2-4 cells containing the plasmid pRS316. Following synchronisation in G1, we cultured the cells at the restrictive temperature to ablate Top2 activity, and released the cells into the cell cycle for one passage through S phase. We then harvested the cells prior to mitosis, preventing further cell cycle progression with the microtubule depolymerising drug nocodazole. We extracted DNA and assessed the frequency of DNA catenation introduced into the plasmid using two-dimensional gel electrophoresis and Southern blotting. As expected, expression of TOF1 wt completely suppressed the excessive fork rotation and DNA catenation of tof1 top2-4 cells ( Figure 2C). Expression of tof1 1182 which lacks only the final 57 amino acids of the C terminus of Tof1 also fully rescued the excessive fork rotation ( Figure 2H). However, expression of Tof1 proteins truncated at 627, 762, 830 and 997 did not rescue this effect ( Figure 2D-G). This indicates that the region of Tof1 between aa 997 and aa 1182 is required to suppress excessive fork rotation. This matches the region of Tof1 (997-1226) that is capable of supporting a two-hybrid interaction with Top1 (Park & Sternglanz, 1999) and recruiting Top1 to the replication fork (Shyian et al, 2019). This is consistent with the model that preferential recruitment of Top1 to the replisome stimulates its action ahead of the fork (Schalbetter et al, 2015;Keszthelyi et al, 2016).
To further examine the functionality of the Tof1 truncated proteins we next set out to determine which of the mutants were able to support replication fork pausing at a replication block. We cloned the S. cerevisiae rDNA region that contains the replication fork barrier (RFB) (corresponding to Chromosome XII sequence : 459799 -460920) into the multicopy yeast episomal plasmid pRS426 (figure 3A). Transplanting the RFB sequence into an artificial location pauses replication forks in a Fob1 dependent manner, but does not fully arrest ongoing replication (Calzada et al, 2005). It therefore appears to act in a similar manner to other endogenous protein pausing sites (Hodgson et al, 2007)(cartoon example shown in figure 3B). Cells without Tof1 did not pause at the RFB site on the plasmid (figure 3C), whereas cells expressing TOF1 wt fully supported pausing at this location (figure 3D).
However, expression of tof1 627 and tof1 762 did not rescue pausing at the RFB site (figures 3E and 3F ). Together this data indicates that the last 408 aa of Tof1 are not required for replication fork pausing at protein barriers to DNA replication and therefore that the role of Tof1 in replication fork pausing is independent of its role in inhibiting fork rotation.
Deletion of TOF1 results in hypersensitivity to CPT, an agent which stabilises the Topoisomerase 1 covalent complex (Top1cc) to DNA, forming a DNA protein cross-link (DPC). Deletion of CSM3 results in similar sensitivity to CPT treatment while mrc1 cells display only mild sensitivity to this agent (Redon et al, 2006). Therefore, CPT sensitivity appears to be a marker for functions of the Tof1-Csm3 heterodimer that are independent of their association with Mrc1. The cause of the acute toxicity of tof1 or csm3 cells to CPT is currently unclear. On the leading strand a DPC and an adjacent single strand DNA break would be predicted to either inhibit replication fork progression or generate a double strand break (Strumberg et al, 2000;Sparks et al, 2019). Alternatively, CPT treatment has also been found to cause increased DNA topological stress on cellular DNA (Koster et al, 2007), potentially generating a situation where efficient recruitment of topoisomerases to the fork by Tof1 is required for ongoing elongation. Our data has established that tof1 truncations that cannot support relaxation of DNA topological stress through Top1 recruitment can still support fork pausing at a RFB complex. We therefore examined how expression of each of the truncation mutations suppressed the sensitivity of tof1 cells to CPT. Cells expressing tof1 627 or tof1 762 were highly sensitive to CPT while expression of tof1 830, tof1 997 and 1182 provided wt levels of resistance to CPT (figure 3J). Therefore, truncation mutants that do not support Tof1's role in promoting Top1 action ahead of the fork (tof1 830 and tof1 997) are normally resistant to CPT. Conversely, tof1 truncations that do not support replication fork pausing (tof1 627 and tof1 762) cannot suppress CPT sensitivity. We conclude that CPT sensitivity is closely associated with the ability of the Tof1-Csm3 complex to support pausing of the fork at a protein block to replication.
Replisome pausing induced by nucleotide depletion with Hydroxyurea (HU) also requires Tof1 function. Following acute treatment of cells with HU both Tof1 and Mrc1 are required to maintain the coupling of helicase and polymerase activities at the fork (Katou et al, 2003).
In the absence of Tof1 or Mrc1 the helicase advances beyond the point of nascent DNA incorporation, indicating uncoupling of helicase and polymerase activities in the replisome (Katou et al, 2003). This uncoupling is predicted to lead to increased binding of the single stranded DNA binding protein, RPA1, to the single stranded regions generated by uncoupled helicase action. We have experimentally confirmed this prediction using RPA1 ChIP-SEQ in HU treated wildtype, tof1 and mrc1 cells. Release of cells arrested in G1 with alpha factor into media containing 200 mM HU led to strongly increased RPA1 chromatin binding around replication origins in tof1 and mrc1 cells compared to wildtype (figure 4A). Expression of TOF1 wt and tof1 830 both suppressed the accumulation of RPA1 around early firing origins (figure 4B). Thus, expression of the tof1 830 truncation mutant was sufficient to ensure helicase and polymerase coupling. However, expression of either tof1 762 or tof1 627 still led to increased RPA1 around early origins, indicating helicase-polymerase uncoupling. Notably the elevated level of RPA1 around origins in either tof1 762 or tof1 627 was consistently less than the level of RPA1 observed in tof1 cells (figure 4B), indicating that they retained some function in coupling of helicase and polymerase activities, potentially through their mediation activity in the DRC (see below).
We next set out to establish whether expression of the truncated proteins was sufficient to support the mediator function of Tof1 in activation of the DRC. In S. cerevisiae, sustained activation of the effector kinase Rad53 in S phase following HU treatment can be through the DRC mediated by Mrc1 and Tof1, or through the Rad9-mediated DDC (Pardo et al, 2017). Therefore, loss of detectable activation of Rad53 in response to HU only occurs in cells lacking both tof1 and rad9 function. Conversely, HU-dependent Rad53 activation can be rescued by the activity of either Tof1 or Rad9 protein in tof1 rad9 cells. We assayed whether expression of any of the tof1 truncation mutants could rescue Rad53 activation in response to HU treatment in tof1 rad9 cells, using the Rad53 phospho-mobility assay (Pellicioli et al, 1999). In these experiments, treatment of exponentially growing wt cells with 200 mM HU led to a robust auto-phosphorylation-linked mobility shift of Rad53 protein as visualised by western blot. This shift was partially attenuated in tof1 and completely lost in Previous studies have shown that hypo-morphic alleles of some checkpoint components delay but do not prevent the full activation of Rad53 following treatment with HU (Hustedt et al, 2015). To assess whether this was the case for the tof1 627 allele we assayed protein extracts from cells taken at sequential time points following release into 200 mM HU for Rad53 activation (figure 5B). We did not observe a delay in activation of Rad53 in the tof1 627 allele compared to wildtype Tof1, consistent with this allele activating the intra-S-checkpoint with wildtype kinetics.
The separation of function observed between the roles of Tof1 in checkpoint signalling and in replisome coupling/pausing provided us with an opportunity to differentiate between the contribution of these two Tof1 functions to genome stability following RS induced replisome stalling. Following treatment of budding yeast with 200 mM HU, replication forks stall shortly after origin firing, activating the DRC. DRC activation is essential to maintain cellular viability during incubation in HU (Lopes et al, 2001;Tercero et al, 2003).
Checkpoint-dependent viability is thought to be due to regulation of several pathways including inhibition of late origin firing (Paciotti et al, 2001), inhibition of nucleases that could process the arrested fork (Segurado & Diffley, 2008;Cotta-Ramusino et al, 2005) and stabilisation of replisome components at the arrested fork (Lopes et al, 2001;Tercero et al, 2003). The role of Tof1 in maintaining genome stability in response to RS could be through checkpoint signalling, through a general role in maintaining replisome integrity, or a combination of the two. To investigate the contribution of these two roles to genome stability we compared the response to arrest and subsequent wash off of 200 mM HU in tof1 rad9 cells, where the checkpoint activation of Rad53 is ablated, to tof1 rad9 cells complemented with either TOF1 wt, tof1 627 or tof1 830, all of which generate Tof1 protein capable of activating the DRC in tof1 rad9 cells (figure 5C and 5D). Using FACS analysis of DNA content, we observed that uncomplemented tof1 rad9 cells were unable to complete DNA replication following removal of 200 mM HU, consistent with previous analysis of checkpoint-defective cells (Desany et al, 1998;Lopes et al, 2001) (figure 5C).
Complementation of these cells with wildtype Tof1 resulted in rapid completion of DNA replication ( figure 5C). In addition, complementation of tof1 rad9 cells with either of the checkpoint competent truncation mutants tof1 627 or tof1 830 also resulted in rapid apparent completion of DNA replication (figure 5C), consistent with the checkpoint function of Tof1 alone being sufficient for resumption of DNA replication following HU arrest. In addition to testing DNA content we also re-plated these cells following release from the acute HU treatment onto YPD plates to assess whether they could survive the treatment. As expected, tof1 rad9 cells failed to form colonies following acute HU treatment whereas TOF1 wt rad9 cells displayed efficient colony-forming capability (figure 5D). Despite their apparent resumption in DNA replication following acute HU treatment, tof1 627 cells generally failed to recover from the HU-induced arrest, although they were more viable than tof1 rad9 cells ( figure 5D). This indicates that despite DRC activation and apparent efficient resumption of DNA replication, tof1 627 expressing cells were still subject to frequent unrecoverable lesions following fork stalling. Consistent with this interpretation both tof1 627 and tof1 762 cells were acutely sensitive to chronic treatment with HU compared to the other tof1 truncation mutants ( figure 5E). indicating that stalled forks in these cells were able to restart replication following release from the HU block (figure 6B right). As predicted, tof1 rad9 cells had a much smaller increase in copy number around origins consistent with poor levels of recovery or restart from HU-arrested forks in this genetic background ( Figure 6B right). In contrast to these two extremes, tof1 627 cells exhibited a similar pattern of increased copy number around origins fired in HU to wt and tof1 830, but often failed to increase copy number to the same level as seen in late replicating regions (figure 6B bottom right).
These findings indicate that Tof1 proteins that can mediate DRC activation can promote efficient restart around HU-stalled forks. However, replication forks associated with Tof1 proteins that cannot support replisome coupling or pausing then go on to exhibit elongation defects that often prevent the completion of DNA replication. This leads to toxic lesions if the Rad9-dependent repair pathways are absent.
In summary, we have found that the function of Tof1 in checkpoint activation only requires . This is consistent with the tof1 627 and tof1 762 truncations compromising the stability of the interaction between Tof1 and Csm3, leading to partial destabilisation of the Csm3 protein.

Discussion
Defining how the Timeless protein functions to maintain genome stability, both generally and following RS has been complicated by its pleiotropic functions. Here we have shown that distinct regions of Tof1 are required for its roles during DNA replication.
We find that the far C terminal region (>997aa) of Tof1 is required to supress excessive fork rotation in cells. This region is also sufficient to mediate a two-hybrid interaction with the type 1B topoisomerase Top1 (Park & Sternglanz, 1999). During DNA replication Top1 can only act ahead of the replication fork to facilitate unwinding of the parental DNA duplex (Schalbetter et al, 2015;Keszthelyi et al, 2016). Therefore, the recruitment of Top1 to the replisome via the C terminal region of Tof1 will bias topoisomerase action to the region ahead of the fork, providing a ready explanation for why this region is required to suppress the alternate pathway of duplex DNA unwinding, specifically fork rotation and the action of Type II topoisomerase behind the fork. During preparation of this study others have also shown that a similar C terminal region of Tof1 is required to recruit Top1 to active replication forks, consistent with this model (Shyian et al, 2019). Interestingly, the SV40 T antigen helicase also directly recruits Top1 to the SV40 replisome to facilitate DNA unwinding (Simmons et al, 1996) suggesting that direct replisome recruitment of Top1 could be a general feature of eukaryotic replisomes. It remains to be seen if topoisomerase recruitment to the eukaryotic replisome is normally achieved via Timeless family proteins or if this function could be swapped between different replisome components in different organisms.
Distinct from the far C terminus, the middle region of Tof1 between amino acids 627 and 830 are required for fork pausing at the RFB protein barrier and also to maintain coupling of helicase and polymerase activities during HU induced RS. Expression of tof1 627 and tof1 762 truncations ablates the interaction between Tof1 and Csm3. This also appears to be linked to a general partial loss of Csm3 protein levels in cells, suggesting that at least part of the function of this region is to stabilise the Tof1-Csm3 interaction. We speculate that the Tof1-Csm3 heterodimer plays a central role in replisome structure that facilitates coordination of CMG and replicative polymerases consistent with observations across several systems (Katou et al, 2003;Cho et al, 2013;Errico et al, 2007). Current data argues that functional consequences of the Tof1-Csm3 interaction extend beyond the mapped 627-830 aa region of Tof1. For example, we have previously shown that csm3 cells do not preferentially target Top1 activity to the region ahead of the fork, despite full length Tof1 still being expressed (Schalbetter et al, 2015). This argues that the Tof1-Csm3 interaction is required to appropriately configure the Tof1-Top1 interaction within the replisome. This region of Tof1 that stabilises its interaction with Csm3 is also required to suppress the sensitivity of tof1 cells to the Top1cc complexes generated by CPT. Since the C-terminal region required for the Tof1-Top1 interaction is dispensable for suppression of CPT sensitivity in tof1 cells, it appears that the coordination of Top1 action with the replisome by Tof1 does not influence CPT sensitivity. Rather our data argue that CPT sensitivity is closely linked with stable fork pausing and helicase-polymerase coupling. We speculate that Tof1 activity allows the fork to pause when encountering a Top1cc complex on the leading strand, preventing polymerase runoff (Strumberg et al, 2000) and the generation of a potentially lethal single ended DNA DSB.
Our data argue that the roles of Tof1 in responding to DNA topological stress are distinct from its roles in pausing at replication fork blocks. In contrast to our data, others have recently shown a partial loss of pausing at the endogenous RFB through expression of a Tof1 protein truncated at 981 amino acids, which cannot recruit Top1 to the fork. At present the reason for the variation in fork pausing efficiency of our truncations at 830 and 997 relative to a truncation at 981 is unclear. Further additional loss of pausing at the endogenous RFB was observed in the absence of Top2 and the presence of the truncated Tof1 protein (Shyian et al, 2019). This suggests that, at least at the endogenous RFB, the action of either Top1 or Top2 ahead of the fork promotes fork pausing. Whether this model reflects a general model for fork pausing at protein blocks to replication or is specific to the endogenous RFB remains to be determined.
Finally, we find that the N-terminal half (627 aa) of Tof1 is sufficient for efficient intra-Scheckpoint signalling in response to HU. Both Mrc1 and Tof1 have been identified as required for efficient checkpoint signalling in HU (Alcasabas et al, 2001;Foss, 2001). Mrc1 alone appears capable of facilitating the sensor/effector kinase interaction (Berens & Toczyski, 2012), whereas evidence that Tof1/Timeless acts as a mediator of the checkpoint in the absence of Mrc1/Claspin is lacking. Rather it appears more likely that Tof1 is required to promote association of Mrc1 with the replisome (Bando et al, 2009;Yeeles et al, 2017), thus indirectly promoting the role of Mrc1 in the DRC. We speculate therefore that the N-terminal half of Tof1 is sufficient for Mrc1 to be associated with the replisome in a manner compatible with mediating DRC signalling. DRC activity following fork stalling ensures that the replisome and fork structure is maintained in a form capable of efficiently restarting. Tof1's role as a mediator protein provides a straightforward explanation of why its absence results in problems in fork restart (Tourriere et al, 2005). However, the replisome instability caused by loss of Tof1 also suggests that replisome structure without Tof1 is not proficient in restart following HU arrest irrespective of the whether or not the checkpoint is activated. The separation of function shown by the tof1 627 mutant, which is proficient for checkpoint signalling but not proficient for replisome coupling or fork pausing, has allowed us to show that Tof1 has important roles in completing DNA replication following RS beyond its role in DRC activation activity and DRC dependent restart. Our data argues that Tof1 is required to stabilise restarted forks for subsequent elongation of the replisome. The reason for fork failure is unclear but we speculate that tof1 mutants deficient in coupling helicase and polymerase activity in HU are also defective in efficient recoupling during recovery from HU treatment, leading to frequent fork failure distal from the restart site. Notably, the sensitivity of tof1 mutants to HU is highly dependent on the activity of Rad9, consistent with the Rad9 dependent repair pathways being required for tof1 linked DNA lesions. Although our data show that Tof1 has roles in recovering from RS beyond mediation of the DRC, we cannot rule out a role for Tof1 downstream of DRC activation. The phosphorylation state of Tof1 is partially dependent on Mec1/Tel1 activity (de Oliveira et al, 2015) and it is known that Tof1 phosphorylation can influence fork pausing (Bastia et al, 2016). Therefore, checkpoint dependent phosphorylation of the C terminal half of Tof1 could be important for recovery of the fork following RS.
In summary, our data argues that the N-terminal half of Tof1 is linked to Mrc1 replisome functions, and that the far C terminus regulates Top1 association with the replisome. We find that an internal region of Tof1, between aa 627 and aa 830 that is required for stable binding to Csm3, is crucial for fork pausing and replisome coupling of helicase and polymerase activities. These data indicate that Tof1 has roles both in co-ordinating general replisome architecture independently through both Mrc1 and Csm3 interactions, as well as specifically recruiting other activities such as Top1 to the replisome.
See Table 2 for all plasmids used in this study.
Generation of yeast strains expressing truncated forms of Tof1 was carried out in two steps.

Media and Cell-Cycle Synchronisation
For plasmid catenation experiments in top2-4 pRS316-containing strains, cells were grown to mid-log phase in synthetic complete media without uracil +2% glucose, before being resuspended in in YP 2% Glucose (YPD). Cells were arrested in G1 by addition of 10 µg/ml alpha factor peptide (Genscript) for 1.5 hrs, after which a second dose of alpha factor (5µg/ml) was added. When >90% of cells were unbudded, cultures were shifted to the restrictive temperature for top2-4 (37⁰C), for one hour before release into S-phase by washing three times with YPD at 37⁰C. Time 0 indicates the time from addition of the first wash. 50 µg/ml nocodazole was added to cultures at 45 minutes to prevent mitotic entry, and at 80 minutes from release 10 ml samples for 2D gel and Southern blotting analysis were collected by centrifugation and snap-freezing the pellets in liquid nitrogen.
For fork pausing experiments cells containing pRS426-RFB were grown to mid-log phase in synthetic complete media without uracil +2% glucose. 10ml cultures were collected by centrifugation and snap-freezing the pellets in liquid nitrogen.
For experiments involving treatment with HU, cells were grown in YPD to mid-log phase before either HU being added to 200 mM for 120 minutes (for Rad53 activation experiments For RPA ChIP experiments cells were grown in YP +2% raffinose to mid-log phase at 25 °C, before being arrested with 10 µg/ml alpha factor peptide. After 1hr 45 min 2% galactose and an additional 5 µg/ml alpha factor was added. After 2hr, when cells were >90% unbudded, 25 µg/ml doxycycline was added. 15 minutes after doxycycline addition temperature was switched to 37°C and incubated for 1hr. Cells were then released by washing 3 times with pre-heated YP 2% raffinose 2% galactose with 25 µg/ml doxycycline and resuspended in the same media supplemented with 200 mM HU. Time 0 was taken as the time from the first wash. Samples were then incubated for 1 hr at 37°C before being fixed by resuspending in YP + 1% formaldehyde (Sigma) for 45 min at 25 °C. 125 mM glycine was then added followed by a 5 min incubation at 25 °C. Cells were washed with PBS before being pelleted and snap-frozen in liquid nitrogen.

TCA Extraction
10 ml of mid-log yeast cultures were centrifuged at 3500 rpm for 5 min and the resulting cell pellets were snap-frozen in liquid nitrogen before storing at -80°C. All further steps were carried out on ice, and centrifugation and homogenisation steps at 4°C. 200 µl of 20% TCA was added to thawed cell pellets and the cell suspension was transferred to screw-cap tubes containing 500 µl of 0.5mm zirconia/silica beads (BioSpec Products). Cells were homogenised using a FastPrep-24 (MP Biomedicals) on max speed (6.5 m/s) for 4 x 1 min pulses, with 1 minute on ice in between pulses. Beads were separated from the lysate by piercing the tubes and centrifugation of the mixtures into fresh tubes at 3000 rpm for 2 min.
Beads were washed once with 600 µl 5% TCA and centrifuged again into tubes containing the cell lysate. Cell extract/TCA mixtures were then centrifuged at 13000 rpm for 5 min before removing all TCA from the resulting pellets. This step was repeated once to ensure complete removal of all TCA. To the pellets 200 µl of 1 X sample buffer was added before boiling samples for 5 min. Samples were spun at 13000 rpm for 5 min and the resulting supernatants were collected and stored at -20°C for SDS-PAGE and western blotting analysis.

SDS-PAGE and western blotting
Protein extracts prepared by TCA extraction were run on either 8%, 10% or 12% SDS-PAGE gels before being wet-transferred to nitrocellulose membranes at 50V for 90 min, 4°C.
Proteins were visualised by staining in Ponceau for 30 seconds before membranes were blocked in 5% milk (Marvel) PBS 0.2% Tween-20 (PBS-T). All primary antibodies were diluted in 5% milk PBS-T and incubated overnight at 4°C. In between primary and secondary antibody incubations membranes were washed 3 times with PBS-T for 15 min. Secondary antibodies were diluted in 5% milk PBS-T and incubated with membranes for 1 hr at room temperature. Proteins were detected using Western Lightning Plus-ECL (Perkin-Elmer, NEL104001EA) and images were acquired on an ImageQuant LAS4000 system (GE Healthcare). Densitometry analysis was carried out using ImageQuant TL software. For plasmid DNA catenation analysis of pRS316, purified DNA was nicked with Nb.Bsm1 (New England Biolabs, R0706S) according to manufacturer's instructions.
For fork pausing analysis of pRS416-RFB, purified DNA was digested with BamHI-HF (New England Biolabs, R3136S) and SnaBI (New England Biolabs, R0130L) according to manufacturer's instructions.
After nicking/digestion the DNA was precipitated, washed and solubilised as above with the addition of 300 mM Sodium Acetate pH 5.2 at the first ethanol addition.

2-D Gel Electrophoresis of replication intermediates or catenated plasmid replication products
Purified DNA was separated in the first dimension by electrophoresis in 0.4%
For DNA catenation analysis of plasmid pRS316 from top2-4 cells, 1st dimension gels were run at room temperature for 16-18 hrs at 30V. A portion of the gel containing small amount of each DNA sample was excised and stained in 0.5 µg/ml Ethidium Bromide 1XTBE to reveal extent of genomic DNA mobility. The remaining non-stained gel slices containing the plasmid were excised and embedded in 1.2% MegaSieve/MegaBase Agarose 1X TBE and run in 1XTBE at 4°C for 16-17 hrs, 120V.
For analysis of paused replication intermediates from plasmid pRS426-RFB, for 1st dimension gels were run at room temperature for 15.5 hrs at 30V. 1st dimension gels were stained in 0.5 µg/ml Ethidium Bromide in 1XTBE and gel slices containing the replication intermediates were excised and embedded in 1% MegaSieve/MegaBase Agarose 1X TBE 0.3 µg/ml Ethidium bromide gels. 2nd dimension gels were run in 1XTBE at 4°C for 8 hrs, 120V with re-circulation of the running buffer.

Southern Blotting
Following 2-D electrophoresis, gels were washed sequentially in Depurination buffer Catenated pRS316 plasmids or replication intermediates from pRS416-RFB were probed with DNA amplified from pRS316 (probing specifically for the URA3 gene). Labelling and detection used random prime labelling incorporating fluorescein tagged dUTP (Roche).
Following probing, hybridized fluorescein tagged dUTP was detected with alkaline phosphatase tagged anti fluorescein Fab fragments (Roche), revealed with CDP-Star (GE Healthcare) and non-saturating exposures acquired on an ImageQuant LAS4000 system (GE Healthcare). Densitometry analysis was carried out using ImageQuant TL software.

FACS analysis of DNA content
For analysis of cell cycle progression, 0.5 ml of yeast culture was pelleted by centrifugation at 13000 rpm for 15 seconds before removal of all growth media. The pellets were resuspended in 0.5 ml 70% ethanol and stored at 4°C before processing and analysis.
Fixed cells were washed in 50 mM Tris pH 8.0 and 10 mg/ml RNaseA (Sigma-Aldrich) was added. Cells were incubated overnight at 37⁰C, pelleted and re-suspended in freshly made 5 mg/ml pepsin in 5 mM HCl and incubated again at 37⁰C for 30 min. Fixed cells were washed once more in 50 mM Tris pH 8.0 and re-suspended in 0.5 mg/ml propidium iodide in 50mM Tris pH 8.0. Samples were sonicated for 5 seconds each on low power to reduce clumping before analysis using the BD Accuri™ C6 Plus (BD Biosciences).

Drug sensitivity assays
Yeast cells were grown to mid-log phase before being serially diluted 10-fold in YPD. 5 µl of each dilution was spotted onto YPD plates containing the indicated dose of drug or control reagent and incubated for 24-28 hr at 25°C before imaging.

Colony Survival Assays
Yeast cells were grown to mid-log phase before being arrested in G1 by the addition of 10 µg/ml alpha factor peptide. When cells were >90% unbudded they were released into the cell cycle in the presence of 200 mM HU for 1 hr. Following the HU treatment cells were counted, diluted in YPD medium and plated onto YPD plates. Colonies were counted 48 hrs after plating and the viability was calculated as the percentage of plated cells able to form colonies. Statistical significance was calculated using an unpaired Students t-test.
200 μl of 0.5 mm zirconia/silica beads were added to samples and cells were lysed using the FastPrep-24 (MP Biomedicals) on max speed (6.5 m/s), with 5 rounds of 1 min each. Lysate was spun out and IP buffer (0.1% SDS, 1.1% Triton-X-100, 1.2 mM EDTA, 16.7 mM TRIS HCl (pH8), cOmplete Tablets Mini EDTA-free EASYpack (Roche), PhosSTOP (Roche)) was added to a final volume of 1 ml. Samples were sonicated using the Focused-Ultrasonicator (Covaris) (Average incident power -7.5 Watts, Peak Incident Power -75 Watts, Duty Factor -10 %, Cycles/Burst -200, Duration -20 min). The sample was centrifuged for 20 min at 13000 rpm at 4°C. Supernatant was then diluted to 7.5 ml with IP buffer. 75 μl protein A Dynabeads (Invitrogen) and 75 μl protein G Dynabeads (Invitrogen), were washed 3 times in IP buffer before adding to the sample and incubating for 2 h at 4°C. 2 ml of the supernatant was taken to 15 ml falcon tubes, and the rest was kept at -20°C as an input sample. To the 2 ml sample RPA1 antibody (1:10000, Agrisera, AS07214) was added followed by overnight incubation on a rotating wheel at 4 °C.
A mix of Dynabeads, Protein A (30 μl) and Protein G (30 μl), was washed 3 times in IP buffer. This was added to each sample and incubated at 4°C for 4 h. Supernatant was removed and beads were washed at 4°C for 6 min in TSE-150 (1% Triton-X-100, 0.1% SDS, 2 mM EDTA, 20 mM Tris HCl (pH8), 150 mM NaCl), followed by TSE-500 (1% Triton-X-100, 0.1% SDS, 2 mM EDTA, 20 mM Tris HCl (pH8), 500 mM NaCl), followed by LiCl wash (0.25 M LiCl, 1% NP-40, 1% dioxycholate, 1 mM EDTA, 10 mM Tris HCl (pH8)) and finally Tris-EDTA (TE pH8). Elution was carried out in 400 μl elution buffer (1% SDS, 0.1M NaHCO3, for 30 min at room temperature. At the same time 50 μl from the input sample was added to 150 μl of elution buffer. 20 μl of 5 M NaCl and 10 μl of 10 mg/ml proteinase K (Invitrogen) was then added to the input, and 40 μl and 20 μl to the IP samples respectively. These were incubated at 65°C overnight. Then 10 μl of DNase-free RNase (Roche) was added to the input and 20 μl to the IP samples, and they were left at 37°C for 30 min. All DNA was purified with a Qiagen PCR purification kit and eluted in 40 μl H2O. 34 ul from the RPA1 samples and 1 ng DNA in 34 ul water from the input were used for library preparation. 5 µl 10 x NEB2.1 buffer and 5 µl of random primers (8N, 3 mg/ml stock) were added and the samples were boiled at 95°C for 5 min and immediately placed on ice for 5 min. 5 µl 10 x dNTPs with dUTP instead of dTTP (2 mM each) and 1 µl T4 polymerase (NEB) were added and the mixture was incubated at 37°C in a thermal cycler for 20 min, and 5 µl 0.5 M EDTA (pH 8) was immediately added to stop the reaction. The resulting dsDNA was used to create libraries using the Ultra II library kit (NEB) as per the manufacturer's instructions with 13 cycles at the amplification step.
Paired end sequencing was performed using the MySeq (75bp reads from each side) or NextSeq 500 (42 bp reads from each side) systems to result >2 million reads. The data then was sorted into 50 bp bins, normalized to have a mean value of 1, and used for meta data analysis using custom-made R programs.

Sync-SEQ
Pellets from 2 ml cultures were resuspended in 500 μl SDS buffer (1% SDS, 10 mM EDTA, 5M Tris HCl, cOmplete Tablets Mini EDTA-free EASYpack (Roche), PhosSTOP (Roche)). μl of 10 mg/ml proteinase K (Invitrogen) was then added followed by an overnight incubation on 65°C. Then 10 μl of DNase-free RNase was added to the sample and they were incubated at 37°C for 30 min. DNA was then purified with a Qiagen PCR purification kit and eluted in 50 μl H2O. 50 ng of DNA in 50 ul water was used for library preparation using the Ultra II library kit (NEB) as per the manufacturer's instructions with 6 cycles at the amplification step.
Paired end sequencing was performed using NextSeq 500 (42 bp reads from each side) systems to result >2 million reads.
The resulting ratios were smoothed by a moving average of 2 and plotted using custom-made R programs.

Acknowledgments
We

Declaration of Interests
The authors declare no competing interests.

Data Availability
All strains are available on request. All sequencing is available on request. endogenous TOF1 and codon-optimised TOF1 wt and tof1 mutants were C-terminally tagged with TAP epitope and lysates were checked for expression using peroxidase anti-peroxidase (PAP) which immuno-reacts with the protein A portion of TAP tags. Ponceau stain of blotted membrane is shown to illustrate protein content of lanes. A) Plasmid pRS426-RFB used for analysis of fork pausing at the RFB. The two unique restriction digest sites used to linearize the plasmid for 2D replication fork gel analysis and the Y structure anticipated to be generated by fork pausing at the RFB pause site are shown. B) Schematic of the replication fork Y arc predicted to be generated by pRS426-RFB digestion with BamHI and SnaBI in wildtype cells and resolved by 2D gel electrophoresis. Arrow indicates the accumulation of replication intermediates generated by pausing at the RFB on pRS426. C-I) 2D gel analysis of pRS426-RFB in exponentially growing cells containing   HU. The indicated cell mutants were arrested in 200 mM HU following release of a G1 synchronised culture. After 60 minutes in HU, the HU containing media was washed off and cells allowed to recover and resume passage through S phase. Shown are FACS profiles of DNA content before and after the HU arrest point. Time points indicate the time in minutes after release from the HU arrest (T0 indicates time from the first wash). D) Viability of cells following cell cycle arrest induced by acute HU treatment. G1arrested cells were released from the block into YPD media containing 200 mM HU for 1 hour, after which the cultures were diluted and plated to YPD plates. Colony number was counted 48 hours after plating and results from 4 repeats were quantified and displayed in histogram (bottom). P values: *=<0.05, **=<0.01, ***=<0.001. E) Viability of cells during chronic treatment with HU. Viability spot assay of TOF1 wt, tof1Δ, tof1 627, tof1 762, tof1 830, tof1 997 and tof1 1182 cells on YPD or YPD plus 10 or 40 mM HU for three days at 25°C.

Figure 6: Replication in tof1 627 cells is reduced after release from HU treatment in rad9Δ background
A) Experimental set up for sync-seq experiments in cells arrested in 200 mM HU and released to follow replication fork restart. B) Overlaid relative copy number of TOF1 wt rad9Δ and tof1Δ rad9Δ compared to tof1 830 rad9Δ (top panels) and tof1 627 (lower panels) after 60 min exposure to 200mM HU (left panels) and 80 min after wash off of HU allowing the cells to recover and resume DNA replication (right panels). Dashed lines represent known origins.