Comprehensive mutational analysis of the checkpoint signaling function of Rpa1/Ssb1 in fission yeast

Replication protein A (RPA) is a heterotrimeric complex and the major single-strand DNA (ssDNA) binding protein in eukaryotes. It plays important roles in DNA replication, repair, recombination, telomere maintenance, and checkpoint signaling. Because RPA is essential for cell survival, understanding its checkpoint signaling function in cells has been challenging. Several RPA mutants have been reported previously in fission yeast. None of them, however, has a defined checkpoint defect. A separation-of-function mutant of RPA, if identified, would provide significant insights into the checkpoint initiation mechanisms. We have explored this possibility and carried out an extensive genetic screening for Rpa1/Ssb1, the large subunit of RPA in fission yeast, looking for mutants with defects in checkpoint signaling. This screen has identified twenty-five primary mutants that are sensitive to genotoxins. Among these mutants, two have been confirmed partially defective in checkpoint signaling primarily at the replication fork, not the DNA damage site. The remaining mutants are likely defective in other functions such as DNA repair or telomere maintenance. Our screened mutants, therefore, provide a valuable tool for future dissection of the multiple functions of RPA in fission yeast.


INTRODUCTION
The integrity of the genome is crucial for the survival of all living organisms. To maintain 56 genome integrity, several mechanisms have evolved in eukaryotes: the accurate DNA 57 replication machinery, repair pathways that deal with replication errors and various types of 58 DNA damage, and the mechanisms that control telomere homeostasis. Overseeing these 59 cellular processes is the checkpoint system that coordinates their activities with the cell cycle 60 progression [see review (1)]. All these DNA metabolic processes involve ssDNA as a transient 61 intermediate that has to be recognized and properly protected from nuclease attack. Due to its 62 abundance and high affinity for ssDNA, RPA is the major ssDNA binding protein in 63 eukaryotes [see reviews (2, 3)]. RPA is also known as replication factor A (RFA) and ssDNA 64 binding protein (SSB). It was originally purified as a protein required for replication of simian 5 The recruited TopBP1 can stimulate ATR kinase via its ATR activation domain and thus 90 enhance the checkpoint signaling (13). Like TopBP1, ETAA1 also activates mammalian ATR 91 kinase both in vitro and in vivo (14) similar to the budding yeast Ddc2, Dna2, and Dpb11 that 92 activate Mec1 kinase, the ATR ortholog (15,16,17). Proteomics analyses have identified 93 hundreds of phosphorylation targets of ATR in both mammalian and yeast cells (18,19,20), 94 including RPA. Phosphorylation of Rpa2 by ATR at the phosphorylation domain is believed to 95 regulate the functions of RPA (21). 96 Most of the yeast genetic studies of RPA are carried out with the budding yeast S. 97 cerevisiae. An early study by gene disruptions showed that all three RPA subunits are essential 98 for cell survival in budding yeast (22). Although deletion of ssb3 gene encoding the small 99 subunit in fission yeast is not lethal, the null mutant is sensitive to the genotoxins that disrupt 100 DNA replication (23). Due to the essentiality of RPA, studying its functions in vivo has been 101 challenging, and relying on the mutants that allow cell survival and, in the meantime, are 102 defective in checkpoint, or other functions. One such separation-of-function mutant is the 103 budding yeast rfa1-t11, which was first reported 25 years ago (24). This mutant carries a single 104 mutation that converts Lys 45 residue to glutamic acid in the N-terminal DBD-F domain. The 105 rfa1-t11 mutant is defective in homologous recombination (25, 26) and Mec1-mediated 106 checkpoint signaling (7). Another rfa1 allele identified by an earlier study also showed a 107 checkpoint defect at the G1/S and the S phase checkpoint pathways (27). Since the rfa1-t11 108 mutation is in the DBD-F, it is generally believed that the mutation interrupts the interaction 109 between RPA and Ddc2, the ATRIP homolog in budding yeast, leading to the defect in Mec1 110 kinase signaling. However, not all previous studies are consistent with the checkpoint sensor 111 function of RPA (28,29,30). A more recent structural study showed that while the N-terminus 112 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.06.531248 doi: bioRxiv preprint 6 of Ddc2 binds to the DBD-F of Rfa1, the rfa1-t11 mutation, however, does not significantly 113 affect its binding to Ddc2 (31) although the same mutation affects its binding to Mre11-  Xrs2 complex (32). This suggests that the checkpoint functions of RPA, particularly its 115 checkpoint sensor function in the DRC pathway, remain to be fully understood in vivo. 116 The fission yeast S. pombe is an established model for studying the cellular mechanisms 117 that are conserved in higher eukaryotes. Unlike the budding yeast in which the major 118 checkpoint effector kinase Rad53 (functional homolog of human Chk1) activates both the DRC 119 and the DDC pathways, the DRC and the DDC pathways are mediated by Cds1 (human Chk2) 120 and Chk1 (33) separately, in fission yeast, which promotes unambiguous description of the 121 checkpoint signaling mechanisms. Several RPA mutants have been reported previously in 122 fission yeast that are sensitive to ionizing radiation and genotoxic drugs (23, 34, 35). None of 123 them however appears to have a defined checkpoint defect. We have recently carried out a 124 large-scale genetic screen in fission yeast by random mutation of the genome, looking for 125 mutants that are defective in the DRC signaling pathway. This hydroxyurea (HU)-sensitive or 126 hus screen has identified several previously uncharacterized mutants such as tel2-C307Y in the 127 essential Tel2-Tti1-Tti2 (TTT) complex (36), a series of mutants of rqh1 of a RecQ helicase 128 (37), and two mutants of the essential Smc5/6 complex (38). To our surprise, although this 129 genome-wide screen has identified every single previously known DRC gene multiple times, it 130 did not identify a single RPA mutant, which raises a concern about the checkpoint sensor 131 function of RPA in vivo. To better understand the in vivo checkpoint functions of RPA, we took 132 a targeted forward genetics approach to screening the large subunit Ssb1, aiming to identify a 133 non-lethal mutant that lacks the checkpoint signaling function. Such a mutant, once identified, 134 would provide a much clearer insight into the checkpoint initiation mechanisms at the 135 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made generates replication stress by slowing the polymerase movement at the fork. Since the DRC 160 deals with the replication stress, cells lacking Cds1 are highly sensitive to HU (Fig. 1A). In the 161 presence of MMS, cells lacking Chk1 are highly sensitive, indicating that the DNA damage is 162 mainly dealt with by the DDC. Since the rad3∆ mutant lacks both the DRC and the DDC, it is 163 highly sensitive to both HU and MMS (Fig. 1A). Under similar conditions, the four RPA 164 mutants were found sensitive to both HU and MMS, particularly the ssb1-R339H mutant. 165 We then examined the Rad3-initiated checkpoint signaling in these RPA mutants by 166 Western blotting using the phospho-specific antibodies described in our previous studies (40, was examined and compared with that in wild-type cells. The experiment was repeated three 179 times and the quantitation results are shown in Fig. 1C. We found that in the presence of HU, 180 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made in the presence of HU, we found that Cds1 phosphorylation was unchanged in ssb1-G78E, 183 moderately reduced in ssb3∆, or even higher in ssb1-R339H and ssb1-G78E ( Fig. 1D and E). 184 These results show that although the mutants are sensitive to HU, their mutations do not 185 significantly compromise the Rad3 kinase signaling in the DRC pathway.

186
Since the RPA mutants are sensitive to MMS (Fig. 1A), we next examined the 187 phosphorylation of Chk1 by Rad3 (45, 46), which activates Chk1 in the DDC pathway. Chk1 188 phosphorylation is commonly examined by mobility shift assay (45). Using this assay, we 189 found that after treatment with MMS, Chk1 phosphorylation was significantly increased in (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.06.531248 doi: bioRxiv preprint quantitation methods used and the potential issues with the loading, particularly the IPed Chk1. 204 We conclude that the Rad3 kinase signaling in the DRC and the DDC pathways are minimally 205 compromised or remain functional in the four previously reported RPA mutants, which 206 confirms the earlier results for ssb1-R339H (rad11A) and ssb3∆ mutants (23, 34).

208
Insensitivity of ssb1-R339H, ssb1-D223Y, ssb1-G78E, and ssb3∆ to acute HU treatment. 209 Since the DRC remains functional in the RPA mutants, we wanted to investigate their HU  (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.06.531248 doi: bioRxiv preprint Destabilized Ssb1 in the R339H, D223Y, and G78E mutants. RPA is essential for cell 227 growth and perturbation of its protein level causes genome instability or even cell death (50).

228
To better understand the drug sensitivities of the RPA mutants, we generated an antibody 229 against Ssb1. The specificity of the antibody was confirmed by Western blotting of the S.  telomere maintenance (35), which suggests that our screening is extensive (see Discussion).

295
. CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.06.531248 doi: bioRxiv preprint We then examined the sensitivity of the primary mutants to acute HU treatment by spot 296 assay. As shown in Fig. 2C, all mutants except the #4, #5 and #13 mutants showed a growth 297 defect. Among the thirteen mutants, the #1, #2, and #7 mutants showed noticeable sensitivities, 298 although the sensitivities were lower than cds1∆ cells. When the Ssb1 levels were examined in 299 these mutants (Fig. 2D), we found that Ssb1 in #2, #4, #7, and #9 mutants was reduced to ≤ 300 40% of the wild-type level and the rest of mutants showed a moderately reduced or slightly 301 increased Ssb1. The increased Ssb1 likely compensates for the functional loss caused by the 302 mutations.

303
Next, we examined the Rad3 signaling in the thirteen mutants by Western blotting. As (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.06.531248 doi: bioRxiv preprint that the screen is extensive or near exhaustion for indentifying the non-lethal ssb1 mutants with 342 defects in checkpoint signaling or other functions (see Discussion).

344
Secondary mutations in the #7 and the #24 primary mutants. So far, our extensive 345 mutational analysis has identified a number of ssb1 non-lethal mutants such as #1, #7, #17, and 346 #24 mutants that might be defective in checkpoint signaling (Fig. 4B). These four mutants as 347 well as the #10 and the #19 mutants that showed the checkpoint defect to a lesser degree (  (Fig. S6). These mutant integrants were hereafter renamed as ssb1-1, ssb1-7, ssb1-10, 352 ssb1-17, ssb1-19, and ssb1-24. When drug sensitivities of the six integrants were compared 353 with their primary mutants (Fig. 5A), we found that ssb1-1, ssb1-10, ssb1-17, and ssb1-19   354 showed similar sensitivities as their primary mutants, the ssb1-24 integrant was less sensitive to 355 both HU and MMS. Although ssb1-7 showed a similar sensitivity to HU, it was less sensitive to 356 MMS than its primary mutant. These results show that the #7 and the #24 primary mutants 357 likely carry a secondary mutation. This is a surprise as the mutations were generated by precise 358 allele replacement, and the primary mutants have been backcrossed with wild-type S. pombe at 359 least once. We believe that the near-saturation screening is a contribtion factor of the secondary (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.06.531248 doi: bioRxiv preprint resulting asci were analyzed by tetrad dissection (Fig. S7B). The 2:2 ratios of the ura+ spores 365 and Western blotting with anti-HA antibody (Fig. S7C) confirmed the tagging of the ssb1 366 mutant. However, the hus phenotype varied among the ura+ spores, showing unequal 367 segregation (Fig. S7B). Interestingly, some ura-spores also showed the hus phenotype. When 368 the spores were individually analyzed by Western blotting, all hus spores expressed Ssb1 369 similar to the wild-type level (Fig. S7C), suggesting that the varied hus phenotype is unrelated 370 to Ssb1 levels. When the hus phenotype was assessed by spot assay, we found while all ura+ 371 colonies were sensitive to both HU and MMS, the ura-spores were sensitive to HU but not 372 MMS (Fig. S7D). These results show that the #24 mutant carries a secondary mutation likely in 373 a metabolic pathway that sensitizes the cells to chronic exposure to HU, but not MMS, nor   -1 and ssb1-10. We then examined the checkpoint 385 signaling defects in the six ssb1 mutants whose mutations have been confirmed by the genomic 386 integration. When phosphorylation of Mrc1 was examined in the presence of HU, we found 387 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.06.531248 doi: bioRxiv preprint that the phosphorylation was reduced in ssb1-1 and ssb1-10 and unaffected in ssb1-7, ssb1-17, 388 ssb1-19, and ssb1-24 mutants (Fig. 5B and S8A). When Cds1 phosphorylation was examined, 389 we found that it was more significantly reduced in ssb1-1 and ssb1-10 than the rest four 390 mutants ( Fig. 5C and S8B). When Chk1 phosphorylation was examined by the phospho-391 specific antibody and the mobility shift assay, we found that the six mutants showed either 392 increased (ssb1-1, ssb1-17, ssb1-19, and ssb1-24) or slightly reduced phosphorylation (ssb1-7 393 and ssb1-10), suggesting a functional DDC (Fig. 5D and E and Fig. S8C and D). In the 394 presence of DNA damage or replication stress, Rad9 of the 911 complex is phosphorylated by 395 Rad3 to promote Chk1 and Cds1 activation although Tel1 also contributes to the 396 phosphorylation at a basal level. We then examined Rad9 phosphorylation using a phospho-397 specific antibody for Rad9-pT412 (40, 55). In the presence of HU, Rad9 phosphorylation was 398 reduced in ssb1-1, ssb1-7, and ssb1-10 while it was moderately reduced in ssb1-17, ssb1-19, 399 and ssb1-24 mutants ( Fig. 5F and S8E). When treated with MMS, ssb1-1, ssb1-7, and ssb1-10 400 showed a moderately reduced Rad9 phosphorylation, whereas the phosphorylation in ssb1-17, 401 ssb1-19, and ssb1-24 was at the wild-type level or slightly higher ( Fig. 5G and S8F). Together, 402 these results suggest that while ssb1-7 has a minor defect in the DRC, ssb1-1 and ssb1-10 have   in ssb1-1 and ssb1-10. To confirm the DRC defect in 407 ssb1-1 and ssb1-10, particularly ssb1-1, we first examined the sensitivities of the mutants to 408 acute treatment with HU and MMS by spot assay (Fig. 6A). The results showed that while 409 ssb1-1 and ssb1-10 were slightly sensitive to HU, the ssb1-7, ssb1-17, ssb1-19, and ssb1-24 410 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.06.531248 doi: bioRxiv preprint were relatively insensitive. To confirm the acute HU sensitivity, we performed the colony 411 recovery assay (Fig. 6B) and found that while ssb1-7 was insensitive, ssb1-1 and ssb1-10 were 412 sensitive although the sensitivities were much lower than cds1∆ cells, consisitent with the 413 partial DRC defect. All six mutants were highly sensitive to acute MMS treatment (Fig. 6A). 414 Interestingly, except ssb1-7, the acute MMS sensitivities were even higher than cells lacking 415 Chk1, suggesting a defect in DNA repair (see below).

416
Both ssb1-1 and ssb1-10 mutants have a growth defect, as evidenced by their different 417 and overall smaller sizes of colonies of (Fig. S9A), which may indirectly affect the DRC. To 418 preclude this possibility, we examined Mrc1 phosphorylation every hour during the HU 419 treatment ( Fig. 6C and S9B). Unlike the ssb1-7 mutant in which Mrc1 phosphorylation was 420 slightly reduced during the six hours of HU treatment, the phosphorylation was significantly 421 reduced in ssb1-1 and ssb1-10 cells, particularly during the first three hours of HU treatment.

422
When the cell cycle progression was monitored in the presence of HU, most of the wild-type 423 and rad3∆ cells were arrested at the S phase in ~ 3 h (Fig. 6D). However, unlike wild-type cells  (Fig. S7E). Furthermore, ssb1-7 cells finished the bulk DNA synthesis almost like the wild-type 428 cells, whereas the DNA synthesis in ssb1-1 and ssb1-10, particularly ssb1-1, was slightly 429 slower than in wild-type cells. These results are consistent the partial DRC defect in ssb1-1 and 430 ssb1-10 and the defect is not due to their growth defect.

431
When treated with HU, the DRC mutants undergo premature mitosis, generating a so-432 called cut (cells untimely torn) phenotype (56) in S. pombe that can be examined under 433 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.06.531248 doi: bioRxiv preprint 20 microscope after staining the cells with Hoechst for genomic DNA and Blankophor for septum. 434 As shown in Fig. S10, after the HU treatment for six hours in liquid cultures, wild-type cells 435 were elongated and mononuclear whereas the rad3∆ cells were all short and most of the cells 436 showed the cut phenotype (arrows). Unlike the rad3∆ cells, the cds1∆ cells were elongated in 437 HU because the DDC remains is activated in the presence of collapsed forks. However, >30% 438 of cds1∆ cells showed the cut phenotype due to the lack of DRC (Fig. S10, bottom right). In  breaks (59). We found that all six mutants were sensitive to UV, and the sensitivities were 455 lower than rad3∆ but higher than cds1∆ and chk1∆ cells (Fig. 6E). The six mutants, except 456 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.06.531248 doi: bioRxiv preprint 21 ssb1-7, were highly sensitive to CPT and the sensitivity was comparable to that in chk1∆. 457 Remarkably, all mutants were more sensitive to bleomycin than rad3∆ cells that lack 458 checkpoints. These results strongly indicate that the six ssb1 mutants, including ssb1-1 and 459 ssb1-10, are defective in DNA repair, particularly the pathways for repairing strand breaks.   (Fig. 6D). Second, consistent with the reduced Rad3 phosphorylations in the DRC 482 pathway, the two mutants show moderate to minimal sensitivities to acute treatment with HU 483 (Fig. 6A and B). Third, although the numbers are low, the two mutants show cut cells in the 484 presence of HU (Fig. S10). The low numbers of cut cells and the cell elongation in HU 485 observed in the two mutants are likely due to their functional DDC pathway because Chk1 486 phosphorylation is largely unaffected (Fig. 5D and E and Fig. S8C and D). Finally, the protein 487 levels of Ssb1 are normal or slightly increased in the two mutants (Fig. 2D) although they all 488 show a growth defect (Fig. S9A). We believe that the growth defect of the two mutants is 489 unrelated to their partial DRC defect, and the time course analysis of Mrc1 phosphorylation 490 (Fig. 6C) and the flow cytometry data (Fig. 6D) support this conclusion.

491
Although the two ssb1 mutants with only a partial checkpoint defect in the DRC 492 pathway are identified, this targeted screening is likely extensive or near exhaustion, because 493 (1) the mutated residues in the two previously reported ssb1 mutants rad11A (ssb1-R339H) and 494 ssb1-D223Y were identified at least once by the screening (Fig. 4A). The Gly 78 residue in ssb1-495 G78E was not identified likely due to its moderate sensitivities to HU and MMS (Fig. 1A).

496
(2) >40% of the mutations were individually identified at least two times (Fig. 4B). And (3) 497 ~31% amino acid residues in the N-terminal region and ~4.4% in the rest of the Ssb1 molecule 498 were mutated (Fig. 4A). We believe that our screened ssb1-1 and ssb1-10 mutants, particularly 499 the former, have maximally eliminated the checkpoint function that can be genetically 500 separated in Ssb1. The partial checkpoint defect can be explained by at least three possiblities.

501
. CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made First, the amino acid residues that function in checkpoint signaling in Ssb1 are also required for 502 cell surival. Second, although this screen focuses on Ssb1, Ssb2 may also contribute to 503 checkpoint signaling. Finally, RPA may function redundantly with an unknown factor in Rad3-504 mediated checkpoint initiation in fission yeast.

505
The results in Fig. 6E indicate that ssb1-1 and ssb1-10 are also defective in strand break 506 repair, which echoes the major homologous recombination defect in the budding yeast rfa1-t11 507 mutant (25, 26, 32). The rest four ssb1 mutants are also more sensitive to bleomycin than 508 rad3∆ cells, suggesting an important role of Ssb1 in strand break repair. Surprisingly, none of 509 the identifed mutants show a significant defect in Chk1 phosphorylation of the DDC pathway 510 because strand break repair mainly occurs at G2 where the DDC is highly functional. Since 511 more repair mutants were identified than the checkpoint mutant by this screen, it is possible 512 that the repair function of Ssb1 can be more readily separated genetically from its essential 513 function or it plays an more important role in DNA break repair than the checkpoint. The 514 specific DRC defect in ssb1-1 and ssb1-10 described here is similar to the S phase checkpoint 515 defect of the budding yeast mutant rfa1-M2 (27), but not the DNA damage checkpoint defect in 516 rfa1-t11 (28, 32). As mentioned above, Chk1 phosphorylation in the DDC pathway is 517 commonly monitored by mobility shift assay. Using this assay, we found that all twenty-five 518 primary ssb1 mutants did not show a significant defect in Chk1 phosphorylation. Since Rad3 519 also phosphorylates other residues on Chk1, we were concerned with the non-essential 520 phosphorylation events that might affect the sensitivity of the mobility shift assay leading to a 521 wrong conlusion. To eliminate this concern, we generated a phospho-specific antibody for 522 phosphorylated Chk1-Ser 345 . Although the antibody is highly specific and sensitive, we found 523 that Western blottings using the antibody show significant variations among experimental 524 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.06.531248 doi: bioRxiv preprint repeats. Nonetheless, although the differences are noticeable, the experimental results obtained 525 by using the phospho-specific antibody are generally consistent with that from the mobility 526 shift assay. Some of the differences in the results with the primary mutants by the two methods 527 are likely technical as mentioned above or due to the complication of secondary mutations.

528
Indeed, after the secondary mutation was removed from the #24 primary mutant, Chk1 529 phosphorylation was increased in ssb1-24 to a level similar to or higher than in wild-type cells 530 (compare Fig. S5C and D with S8C and D). We believe that two assays used here are sensitive (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.06.531248 doi: bioRxiv preprint screening has identified several mutants that are defective more specifically in the DRC 548 pathway (36, 37, 38). It would be interesting to investigate how the ssb1 mutants interact 549 genetically with those hus mutants. Together, the genetic data from both yeasts strongly 550 suggest that the molecular mechanisms by which ATR initiates the checkpoint signaling, 551 particularly at the replication fork, remain to be fully understood.

552
The remaining nineteen ssb1 primary mutants show a minimal or uncompromised 553 checkpoint defect and are likely defective in other cellular processes. Their sensitivities to 554 various DNA-damaging agents strongly suggest that at least some of them are defective in 555 DNA repair. Indeed, as mentioned above, the ssb1-1, ssb1-10, ssb1-17, ssb1-19, and ssb1-24 556 mutants are more sensitive to strand breaks induced by bleomycin than rad3∆ cells (Fig. 6E).

557
Preliminary data have also shown significantly shorter or complete loss of telomeres in some of 558 the ssb1 primary mutants (data not shown), which support its important role in telomere 559 maintenance (35). Further studies are needed to eliminate the secondary mutations from the 560 remaining nineteen mutants and dissect the versatile functions of RPA in genome maintenance.

561
The previously uncharacterized ssb1 mutants described in this study provide a valuable tool for 562 future investigations in fission yeast.  The genetic screen of ssb1 mutants. The ssb1 mutants were screened by the targeted forward 572 genetic approach (52). The ssb1 expression cassette was cloned into the S. pombe pJK210 573 integrating vector that carries the ura4 marker (53) (see Fig. S3A). To facilitate the cloning, 574 NdeI, and XmaI sites were engineered into the vector before and after the ORF, respectively. A with BstBI (see diagram in Fig. 2A). For screening the N-terminus 154 amino acid region, 581 random mutations were made by PCR between NdeI and PstI sites (see diagram in Fig. 2A).

582
The linearized library DNAs were transformed into wild-type S. pombe lacking ura4. The 583 transformed cells were selected by sequential cultures in EMM6S[ura-] medium during the first 584 pop-in step (Fig. S3A, step 1 and 2). In the next pop-out step, the cells with integrated ura4 585 marker were grown up in 150 ml YE6S media to saturation to lose the ura4 marker. The cells 586 were then spread onto FOA plates to counter-select the cells that had lost the ura4 gene. The 587 colonies formed on FOA plates were replicated onto YE6S plates containing 5 mM HU. The 588 colonies with hus phenotype were selected, streaked out into single colonies, and the 589 sensitivities to HU and MMS were assessed by spot assay. The selected mutants were 590 backcrossed at least once before DNA sequencing to identify the mutations. The backcrossed 591 mutants were also used for analyzing drug sensitivities ( Fig. 2A and S4A) and checkpoint 592 signaling defects (Fig. 3 and S5).

593
. CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made Integration of ssb1 mutations at the genomic locus. ssb1 with the identified mutations were 595 cloned into a integration vector with a kanR gene (Table S2,     (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made  CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted March 6, 2023. ; https://doi.org/10.1101/2023.03.06.531248 doi: bioRxiv preprint