Mcm10 functions to isomerize CMG-DNA for replisome bypass of DNA blocks

Replicative helicases of all cell types are rings that unwind DNA by steric exclusion in which the helicase ring only encircles the tracking strand, excluding the other strand outside the ring. Steric exclusion mediated unwinding enables helicase rings to bypass blocks on the strand that is excluded from the central channel. Unlike other replicative helicases, eukaryotic CMG encircles duplex DNA at a forked junction and is stopped by a block on the non-tracking (lagging) strand. This report demonstrates that Mcm10, an essential replication protein unique to eukaryotes, binds CMG and enables the replisome to bypass blocks on the non-tracking strand, implying that Mcm10 isomerizes the CMG-DNA complex to position only one strand through the central channel. A similar CMG-DNA isomerization is needed at the origin for head-to-head CMGs to bypass one another during formation of bidirectional replication forks.

"top" of CMG. The dsDNA appears to be tightly held because if the CMG-dsDNA contact was 51 flexible the DNA would have been averaged out during 3D reconstruction. 52 53 The structural evidence that S. cerevisiae CMG encircles dsDNA during unwinding is 54 supported by recent biochemical experiments using strand specific dual streptavidin blocks 55 that show CMG is halted by a block placed on either the non-tracking (lagging) or the 56 tracking (leading) strand ( Figure 1C) (Langston and O'Donnell, 2017). Interestingly, given a 57 sufficiently long time S. cerevisiae CMG can proceed through the lagging strand block 58 without displacing the streptavidin, suggesting that CMG slowly isomerizes to a steric 59 exclusion mode that only encircles the tracking (leading) strand. Earlier studies in Xenopus 60 extracts demonstrate that replisome progression is not hindered by a dual streptavidin 61 block on the lagging strand and conclude that CMG functions in a steric exclusion mode by 62 only encircling the leading strand (Fu et al., 2011). Considering that isolated S. cerevisiae 63 CMG encircles dsDNA, we proposed that Xenopus extracts contain a factor that facilitates 64 isomerization of CMG-DNA (e.g. such that CMG encircles only ssDNA) enabling it to bypass 65 blocks on the non-tracking strand (Langston and O'Donnell, 2017).

79
The present study identifies Mcm10 as the factor that enables CMG to rapidly bypass a block 80 on the lagging strand. Mcm10 is unique to eukaryotes and is an essential gene product that, The current study demonstrates that Mcm10 forms an isolable stoichiometric complex with 97 CMG that highly stimulates CMG unwinding (up to 30-fold) and the processivity of CMG in 98 unwinding. Furthermore, Mcm10 uniquely promotes CMG bypass of a lagging strand block 99 suggesting that Mcm10 isomerizes the CMG-DNA complex to a steric exclusion mode. 100 Replisomes also stall at a lagging strand block in the absence of Mcm10 and addition of 101 Mcm10 is needed to rescue the replisome for continued fork advance past the block. 102 Interestingly, the rate of DNA synthesis by the replisome is not significantly affected by 103 Mcm10, yet CMG presumably encircles dsDNA since these replisomes are stalled by a 104 lagging strand block. Therefore replisome advance is not impeded by CMG encircling dsDNA 105 until faced with a block on the DNA. Collectively, these studies demonstrate that Mcm10 106 functions at the level of the helicase to move blocked replisomes past a lagging strand 107 obstacle. This function might also explain the role of Mcm10 in origin initiation as described 108 in the Discussion.

110
Results: Mcm10 forms a complex with CMG and greatly stimulates helicase activity. In Figure  113 2A Mcm10 beyond this amount has little additional effect. (Figure 2C).

123
To determine if Mcm10 binds to CMG in a stable fashion, we mixed a 3-fold molar excess of 124 Mcm10 (as monomer) with FLAG-tagged CMG and then isolated the CMG-Mcm10 complex 125 using FLAG antibody magnetic beads ( Figure 2D). The reaction was washed twice with 126 buffer containing 300 mM NaCl and CMG-Mcm10 complex was eluted using FLAG peptide. 127 The eluted material was analyzed by SDS/PAGE, and stoichiometric levels of Mcm10 were 128 clearly visible with CMG (lane 3). Mcm10 was not visible in a control reaction in which 129 Mcm10 was present with beads in the absence of CMG (lane 2). We also found that the CMG-130 Mcm10 complex could be isolated using a MonoQ column and elutes at ~400 mM NaCl 131 (Figure 2 -Supplement 1). The reconstituted complex is functional in unwinding assays 132 and shows much greater activity on the forked substrate than CMG alone (Figure 2 -133 Supplement 2), comparable to that seen when adding Mcm10 directly to the unwinding 134 assay in Figure 2. Mcm10 enhances the processivity of CMG helicase. The stimulation of unwinding by 153 Mcm10 observed in Figure 2 could be attributable to more efficient loading of CMG onto the 154 substrate, faster unwinding, and/or greater processivity of unwinding. To distinguish 155 among these possibilities, we compared CMG unwinding of a fork with a longer, 160 bp 156 duplex region to that of the fork with a 50 bp duplex (as in Figure 2) in the presence and 157 absence of Mcm10 (Figure 3). The two substrates have identical 3' and 5' ssDNA tails, so 158 loading of CMG onto the forks should be the same and any differences in unwinding should 159 be attributable to differences in CMG unwinding activity over the different lengths of duplex. 160 For these experiments, CMG was pre-incubated with the DNA substrate for 10' followed by 161 addition of ATP ± Mcm10 to start the reaction (reaction scheme in Figure 3A). In the 162 absence of Mcm10, substantial differences were observed in unwinding of the two 163 substrates that suggest limited processivity of the helicase, consistent with reports of low in Figure 3D) compared to 24% for the shorter 50 bp duplex fork at 10' (Figure 3C). Even 167 after 30' only 8% of the 160 bp duplex has been unwound by CMG indicating that the 168 difference in activity between the two substrates is attributable to low processivity of CMG 169 rather than simply the additional time it takes to unwind the longer substrate.

171
The experiments were repeated in the presence of Mcm10 at a 2:1 ratio to CMG (as 172 determined in Figure  ( Figure 3B and C, lanes 11-19 and graph in Figure 3D). This result suggests that Mcm10 175 enhances the processivity of CMG and possibly also stimulates the rate of CMG unwinding. 176 Examination of the first 5' of unwinding in the presence of Mcm10 shows a short delay in 177 the appearance of products using the 160 bp duplex compared to the 50 bp duplex ( Figure  178 3E). We took advantage of this delayed appearance of the longer products to design an 179 experiment that approximates the average rate of unwinding by CMG-Mcm10 ( from unwinding the DNA before assembly of the replisome. Pol e, RFC, PCNA and 2 dNTPs 207 are then added and incubated a further 4' to assemble the core leading strand replisome and 208 the reaction is started upon addition of the remaining dNTPs and ATP along with RPA. 209 Analysis of the autoradiogram in Figure 4B shows a modest enhancement of CMG- slightly slowed the replisome ( Figure 4C, lanes 11-15). Thus, the improved rate effect of 248 Mcm10 on the core leading strand replisome observed in Figure 4B is eclipsed by the rate 249 enhancement provided by the MTC complex. In contrast to studies with pure CMG, studies of replication in Xenopus egg extracts show 282 that two adjacent streptavidin blocks on the non-translocating (lagging) strand are quickly 283 bypassed by the replisome (Fu et al., 2011). The fact that isolated CMG is strongly inhibited 284 by the same dual streptavidin block, suggests that some other factor in the complete extract 285 helps CMG bypass lagging strand blocks. To test whether Mcm10 may be the factor that 286 facilitates CMG bypass of a lagging strand block, we used a forked DNA substrate with two 287 biotinylated nucleotides on the duplex portion of the lagging strand template (see Figure  288 5A), similar to that used in the Xenopus study that demonstrated replisome bypass of Having shown that the MTC complex greatly stimulates progression of the leading strand 301 replisome even in the absence of Mcm10 (i.e. in Figure 4), we tested the MTC complex in 302 CMG helicase assays to see if it enhances the ability of CMG to bypass blocks on DNA ( Figure  303 5C). In contrast to Mcm10, MTC was unable to promote CMG bypass of the block (compare 304 lanes 11-13 in Figure 5C to lanes 11-13 in Figure 5B). We also note that MTC did not 305 stimulate CMG unwinding even in the absence of the block ( Figure 5C  2014). MTC has little or no effect on block bypass by the replisome (Figure 5 -Supplement  329  2 (B) lanes 7-9), and is comparable to Pol e-PCNA-CMG in lanes 1-3. By contrast, addition of 330 Mcm10 allows the replisome to efficiently and rapidly proceed past these blocks as 331 evidenced by the appearance of full length 141 bp product in lanes 4-6. These results 332 confirm that Mcm10 is needed for replisome bypass of a block and they also indicate that 333 neither MTC nor Pol e-PCNA is able to drive the forward movement of a stalled CMG in the 334 absence of Mcm10.

336
Discussion 337 We have shown that a stoichiometric CMG-Mcm10 complex can be reconstituted and 338 isolated ( Figure 2D and Figure 2 -Supplement 1) and that Mcm10 greatly stimulates 339 CMG unwinding and processivity (Figures 2 and 3). Notably, we find that Mcm10 enables 340 CMG and the replisome to overcome lagging strand blocks that otherwise bring unwinding 341 and leading strand synthesis to a halt (Figure 5 and Figure 5 -Supplement 2). 342 Interestingly, the effect of Mcm10 on fork progression per se is relatively insignificant in the 343 absence of blocks, indicating that CMG encircling dsDNA does not slow the rate of fork 344 progression (Figure 4) appears to bind to the N-tier region of the CMG complex that faces the forked junction 358 (Figure 2  (as in Figure 1A); right, with the lagging strand freed from the interior of the Mcm2-7 ring, 375 CMG can bypass the impediment on the DNA and proceed. 376  Figure  395 6) so they no longer collide and can bypass one another to produce bidirectional replication 396 forks.

398
Is complex from beads after high salt washes (Figure 2D) Pierce/Thermo Scientific. 5 mg of streptavidin powder was resuspended in 0.5 ml distilled water to make a 10 mg/ml stock. Protein concentrations were determined using the Bio-428 Rad Bradford Protein stain using BSA as a standard. S. cerevisiae CMG, Pol e, RFC, PCNA and 429 RPA were overexpressed and purified as previously described (Georgescu et  collecting 1.5 ml fractions. The elution buffer was applied to the column in two 6 ml 467 increments pausing 30' after each increment followed by 3 ml increments with 30' pauses 468 until the elution was complete. Eluted material was aliquoted, flash frozen in liquid 469 nitrogen, and stored at -80°C.

471
Mcm10-Tof1-Csm3 complex. The 3 subunits of the MTC complex (Mrc1-Tof1-Csm3) were 472 co-expressed in yeast. Genes were integrated into the chromosome of strain OY01 (ade2-1 473 ura3-1 his3-11,15 trp1-1 leu2-3,112 can1-100 bar1Δ MATa pep4::KANMX6) with 474 Mrc1 Flag integrated at the Ade2 locus, untagged Tof1 at the His3 locus, and His Csm1 at the 475 Leu2 locus, each under control of the Gal1/10 promoter. Cells were grown under selection 476 at 30 °C in SC glucose, then split into 18L YP-glycerol and grown to OD600 of 0.7 at 30 °C 477 before induction for 6 h upon addition of 20 g of galactose/L. After 6 h, cells were harvested 478 by centrifugation, resuspended in a minimal volume of 20 mM HEPES, pH 7.6, 1.2% 479 polyvinylpyrrolidone, and protease inhibitors and frozen by dripping into liquid nitrogen. 480 Purification of MTC was performed by lysis of 18 L equivalent of frozen cells with a SPEX 481 cryogenic grinding mill. Ground cell powder was thawed in the cold room and resuspended 482 to 25 ml final volume with 5X FLAG binding buffer (1x is 250 mM K glutamate, 50 mM 483 HEPES pH 7.5, 1 mM EDTA pH 8.0) plus protease inhibitors and stirred slowly for 30'. Cell 484 debris was removed by centrifugation (19,000 r.p.m. in a SS-34 rotor for 1 h at 4 °C) and the 485 supernatant was collected and mixed with 1.5 ml anti-Flag M2 affinity resin (Sigma) 486 equilibrated in 1X FLAG binding buffer with 10% glycerol. The mixture was rotated on an 487 orbital platform in the cold room at 30 rpm for 1 h. To collect the bound proteins, anti-FLAG 488 resin was pelleted at 1000 X g in 50 ml conical tubes and washed 5 times with 5 ml of FLAG 489 binding buffer followed by centrifugation. After the final wash step, the anti-Flag affinity 490 resin was resuspended in 2 ml of FLAG binding buffer with 10% glycerol, loaded onto a 491 gravity column and washed twice with 7.5 ml of FLAG binding buffer containing 10% 492 glycerol. Bound protein was eluted with the same buffer containing 0.2 mg/ml 3X FLAG 493 peptide (EZ Biolab, Carmel, Indiana USA). Eluted protein was concentrated to 0.75 ml of 1.5 494 mg/ml protein and injected onto a 24 ml Superose 12 gel filtration column equilibrated in 495 2X PBS with 10% glycerol in two separate runs of 0.5 ml and 0.25 ml. Fractions were 496 analyzed on a 7.5% SDS-PAGE gel and MTC-containing fractions were pooled, aliquoted, 497 flash frozen in liquid nitrogen, and stored at -80°C. 498 499 Helicase substrates. For all radiolabeled oligonucleotides, 10 pmol of oligonucleotide was 500 labeled at the 5' terminus with 0.05 mCi [γ-32 P]-ATP using T4 Polynucleotide Kinase (New 501 England Biolabs) in a 25 μl reaction for 30' at 37°C according to the manufacturer's 502 instructions. The kinase was heat inactivated for 20' at 80°C. For annealing, 4 pmol of the 503 radiolabeled strand was mixed with 6 pmol of the unlabeled complementary strand, NaCl 504 was added to a final concentration of 200 mM, and the mixture was heated to 90°C and then 505 cooled to room temperature over a time frame of >1 h. DNA oligonucleotides used in this 506 study are listed in Table I. 507 508 Helicase assays with forked DNA substrates. For the assays in Figures 2 and 3C, the 509 forked DNA was formed using the following two oligos (Table 1): 50duplex LEAD and 5'-32 P-510 50duplex LAG. For the assays in Figure 3B, the forked DNA was formed using unlabeled 511 160mer duplex LEAD and 5'-32 P-160mer duplex LAG. Oligos were annealed as described 512 above.

514
Reactions in Figure 2A  40 μg/ml BSA). Reactions were mixed on ice and started by placing in a water bath at 30° C. 518 1' after starting the reaction, 25 nM unlabeled 50duplex LAG oligo was added as a trap to 519 prevent re-annealing of unwound radiolabeled DNA. At the indicated times, 12 μl aliquots 520 were removed, stopped with buffer containing 20 mM EDTA and 0.1% SDS (final 521 concentrations), and flash frozen in liquid nitrogen. Frozen reaction products were thawed 522 quickly in water at room temperature and separated on 10% native PAGE minigels in TBE 523 buffer. Gels were washed in distilled water, mounted on Whatman 3MM paper, wrapped in 524 plastic and exposed to a phosphor screen that was scanned on a Typhoon 9400 laser imager 525 (GE Healthcare). Scanned gels were analyzed using ImageQuant TL v2005 software (e.g. for 526 Figure 2B and 2C). For all quantitations of helicase assays, the small % background of 527 unannealed radiolabeled primer in the "No CMG" lane was subtracted from the % unwound 528 at each time point. 529 Figure 3 were similar to those in Figure 2A but using 20 nM CMG 531

Reaction conditions in
and 40 nM Mcm10 (where indicated). CMG was mixed with the substrate on ice in the 532 absence of ATP and placed at 30° C for 10' to allow CMG to load onto the substrates without 533 unwinding. To start the reaction, ATP was added with or without Mcm10 (as indicated). 1' 534 after starting the reaction, 50 nM unlabeled lagging strand oligo was added as a trap to 535 prevent re-annealing of unwound radiolabeled DNA. Total reaction volumes were 126 μl, 536 and 11 μl aliquots were stopped at the indicated times after addition of ATP and processed 537 as described for the assays of Figure 2A.

539
Helicase assays using a dual biotin fork DNA: For the assays in Figure 5, the forked DNA 540 was formed by annealing 50 duplex LAG dual biotin and 5'-32 P-50duplex LEAD (see Table I). 541 The biotinylated dT nucleotides are 13 and 20 bases from the forked junction. Oligos were 542 annealed as described above. Reaction conditions were similar to those in Figure 3 except 543 that the final reaction volume was 45 µl and 4 μg/ml streptavidin was added (where 544 indicated) during the 10' CMG pre-incubation. CMG was at 25 nM and Mcm10 or MTC was at 545 50 nM (final concentrations) when present. In these assays, the trap oligo was 50 nM 546 unlabeled 50duplex LEAD oligo. 12 μl aliquots were removed at the indicated times after 547 addition of ATP, terminated with EDTA/SDS stop buffer, flash frozen and processed as 548 above. 549 550 Mcm10 binding to CMG. To determine if Mcm10 binds to CMG in a stable fashion ( Figure  551 2D), we mixed 40 pmol of FLAG-CMG with 120 pmol Mcm10. The mixture was incubated for 552 15 minutes on ice and then spun in a microcentrifuge at 15,000 rpm for 10' at 4°C. The 553 volume of the protein solution was adjusted to 150 µl with binding buffer (25 mM Hepes, pH 554 7.5; 10% glycerol; 0.01% Nonidet P-40; 300 mM NaCl) and mixed with 25 µl anti-FLAG M2 555 magnetic beads (50% suspension; Sigma-Aldrich). The protein-bead mixture was incubated 556 on ice for 1 h and then the beads were collected with a magnetic separator and the 557 supernatant (containing unbound proteins) was removed. The beads were washed three 558 times with 250 µl binding buffer and bound proteins were eluted by incubating in 62.5 µl of 559 the same buffer supplemented with 0.2 mg/ml 3X FLAG peptide on ice for 30'. The beads 560 were collected with a magnetic separator and eluted proteins were collected and analyzed 561 in an 8% SDS-polyacrylamide gel stained with Denville Blue. 562 563 Replication Assays with 2.8 kb duplex substrate. Leading strand replication experiments 564 in Figure 4 used a singly primed 2.8 kb forked linear DNA substrate that was previously 565 described . The duplex portion of the DNA substrate is linearized 566 pUC19 DNA to which a synthetic fork junction has been ligated to one end of the duplex. The 567 fork is primed for leading strand DNA replication with 5'-32 P-C2 oligo (Table 1) Tris Acetate pH 7.5, 5% glycerol, 40 μg/ml BSA, 3 mM DTT, 2mM TCEP, 10 mM magnesium 571 acetate, 50 mM K glutamate, 0.1 mM EDTA, 5 mM ATP, and 120 μM of each dNTP. 572 Replication assays were performed by first incubating CMG (and the indicated amount of 573 Mcm10 and/or MTC, where indicated) with linear forked template for 5' at 30° C, followed 574 by addition of RFC, PCNA, and Pol e for 4' in the presence of dATP and dCTP to support 575 clamp loading and polymerase binding while preventing 3'-5' exonuclease activity on the 576 primer. Reactions were started by addition of ATP, RPA, and the withheld nucleotides (dGTP 577 and dTTP). The reactions proceeded for the indicated amount of time at 30ºC and were 578 stopped with an equal volume of 2X stop solution (40 mM EDTA and 1% SDS). Reaction 579 products were analyzed on 1.3% alkaline agarose gels at 35 V for 17 h, backed with DE81 580 paper, and dried by compression. Gels were exposed to a phosphorimager screen and 581 imaged with a Typhoon FLA 9500 (GE Healthcare). 582 583 Stalled Replisome Assays: For the experiments in Figure 5 - analyzed by PAGE in a 10% Urea gel. Gels were washed in distilled water, mounted on 597 Whatman 3MM paper, wrapped in plastic and exposed to a storage phosphor screen that 598 was scanned on a Typhoon 9400 laser imager (GE Healthcare). 599 600 Table I. Oligonucleotides Used in this Study. All oligonucleotides used in this study were ordered from IDT with the indicated modifications. mM ATP and 40" later unlabeled 50duplex LAG oligo was added (to 50 nM) as a trap for 748 unwound DNA. The total reaction volume was 78 μl and at the indicated times, 11 μl 749 aliquots were removed, stopped with buffer containing 20 mM EDTA and 0.1% SDS (final 750 concentrations), and flash frozen in liquid nitrogen. Frozen reaction products were thawed 751 quickly in room temperature water and separated on 10% native PAGE minigels. Gels were 752 washed in distilled water, mounted on Whatman 3MM paper, wrapped in plastic and 753 exposed to a storage phosphor screen that was scanned on a Typhoon 9400 laser imager 754 (GE Healthcare). The scanned gel was analyzed using ImageQuant TL v2005 software to 755 obtain the quantitations shown in the graph below the gel.   Figure 5B except that the 771 substrate contained a dual biotin-streptavidin block on the leading strand (50duplex LEAD 772 dual biotin); the lagging strand (50duplex LAG) was radiolabeled. Oligo sequences are in 773 Table I. 774