The FANCM Ortholog Fml1 Promotes Recombination at Stalled Replication Forks and Limits Crossing Over during DNA Double-Strand Break Repair

Summary The Fanconi anemia (FA) core complex promotes the tolerance/repair of DNA damage at stalled replication forks by catalyzing the monoubiquitination of FANCD2 and FANCI. Intriguingly, the core complex component FANCM also catalyzes branch migration of model Holliday junctions and replication forks in vitro. Here we have characterized the ortholog of FANCM in fission yeast Fml1 in order to understand the physiological significance of this activity. We show that Fml1 has at least two roles in homologous recombination—it promotes Rad51-dependent gene conversion at stalled/blocked replication forks and limits crossing over during mitotic double-strand break repair. In vitro Fml1 catalyzes both replication fork reversal and D loop disruption, indicating possible mechanisms by which it can fulfill its pro- and antirecombinogenic roles.

Logarithmically growing strains were transformed with either 4 µg of NcoI-linearized pAN1 or uncut pAN1 essentially as described by Keeney and Boeke (1994) except that carrier DNA was omitted. Cells were plated onto EMMG plates lacking uracil (EMMGu) and incubated at 30˚C for 5 -8 days. The numbers of transformants were then counted to determine the relative transformation efficiency of cut versus uncut plasmid. This provides a measure of how well the cut plasmid is repaired. The cut plasmid transformants were then patched onto EMMG-u, incubated for 3 days at 30˚C, and replica plated onto EMMG lacking both adenine and uracil (EMMG-u-a) to score the number of Ade + recombinants amongst the Ura + transformants. The patched transformants on EMMG-u were also replica plated onto YELA plates, which after 3 days at 30˚C were re-replica plated onto YELA. Following these two rounds of growth on non-selective media the YELA patches were replica plated onto complete media containing 5'-fluoroorotic acid (FOA), which counter-selects Ura + cells. Growth on FOA indicates that ura4 + has not integrated into the chromosome (noncrossover) and therefore is readily lost without selection, whereas poor growth on FOA indicates that ura4 + is not easily lost due to its integration into the chromosome via crossover recombination. Assays were repeated at least three times for each strain. In each assay 300 Ura + transformants were patched and assessed for Ade + and crossover status by the replica plating protocol described above.

Proteins
RuvA and RuvB were a gift from Robert Lloyd (University of Nottingham), and RecA was from New England Biolabs. The purification of Fml1∆C and RuvC are described below. Amounts of protein are expressed in moles of monomer, and were estimated using a Bio-Rad protein assay kit with bovine serum albumin as the standard.

Purification of RuvC
RuvC was overexpressed from plasmid pGS775 in BL21 (DE3) pLysS as described (Dunderdale et al., 1994). Cell lysis, and the precipitation of RuvC and its subsequent redissolving were all essentially as described (Dunderdale et al., 1994). However, the redissolved RuvC was loaded onto a 5 ml HiTrap Blue column (Pharmacia Biotech) preequilibrated in R buffer (20 mM Tris-HCl, pH 8.0, 1 mM EDTA, 1 mM dithiothreitol (DTT) and 10% (v/v) glycerol) containing 0.5 M KCl. RuvC eluted from this column between 0.6 -0.7 M KCl, and was then dialysed against R buffer containing 0.3 M KCl before loading onto a 1 ml HiTrap Heparin column (Pharmacia Biotech). The peak of RuvC (1.5 ml) eluted from this column at ~750 mM KCl. This sample was mixed with 1.5 ml of 100% (v/v) glycerol, and stored as aliquots at -80˚C.
Purification of Fml∆C 2-litre cultures of E. coli Rosetta(DE3)pLysS (Novagen) containing pSN3 were grown with aeration at 30 °C in Luria-Bertani broth containing 125 µg/ml ampicillin and 34 µg/ml chloramphenicol. At a cell density corresponding to an A 600 of 0.6, Fml1∆C was induced by adding isopropyl-1-thio-β-D-galactopyranoside to a final concentration of 0.5 mM, following which the cells were incubated for a further 7 hrs. The cells were then harvested by centrifugation, resuspended in Buffer H (50 mM potassium phosphate, pH 8.0, 0.3 M NaCl, 10% glycerol), and frozen at -80 °C until required. All of the subsequent steps were at 4 °C. The defrosted cells were mixed with 1% Triton X-100, 10 mM β-mercaptoethanol and protease inhibitors before passage through a French pressure cell at 30,000 p.s.i.. Cell debris was then removed by centrifugation at 43,700 x g for 50 min, and the supernatant was loaded directly onto a 2 ml nickel-nitrilotriacetic acid (Ni-NTA) Superflow column (Qiagen) that was washed with 60 ml of Buffer H plus 20 mM imidazole before eluting bound Fml1∆C with Buffer H plus 200 mM imidazole into 2 ml fractions. The second 2 ml fraction contained the peak of Fml1∆C and was loaded directly onto a HiLoad 16/60 Superdex 200 gel filtration column (Amersham Biosciences), which was then developed with 120 ml of Buffer A (50 mM Tris-HCl, pH8.0, 1 mM EDTA, 1 mM DTT, 10% glycerol) plus 0.3 M NaCl. 2 ml fractions were collected and the peak of Fml1∆C eluted between fractions 33 -38. These fractions were pooled, diluted with an equal volume of Buffer A, and loaded onto a 1 ml Hi-Trap Heparin column (GE Healthcare). The column was then washed with 5 ml of Buffer A plus 0.1 M NaCl before eluting bound protein with an 18 ml gradient from 0.1 to 1.0 M NaCl. The peak of Fml1∆C eluted between 0.41 -0.44 M NaCl, and these fractions were pooled and stored as aliquots at -80 °C.

DNA substrates
Oligonucleotides 1 -8, 10, 15 -19 and 22, used to make the X-junctions, part X-junctions fork substrates, and static D-loops are listed in Table S3. These oligonucleotides, together with the procedures for substrate preparation, have been described previously (Doe et al., 2002;Osman et al., 2003;Whitby and Dixon, 1998). The X-junctions, part X-junctions and fork substrates are 32 P-labelled at the 5′ end of oligonucleotide 2. The static D-loops are 32 P-labelled at the 5' end of oligonucleotide 16. Protocols for the construction of plasmid-based D-loops have been described previously (McIlwraith et al., 2001). In brief, 5' 32 P-labelled oligonucleotide oMW592 (5' -TGCCGAATTCTACCAGTGCACGCCTCCATCCAGTCTATTAATTGTTGCCGGGA AGCTAGAGTAAGTAGTTCGCCAGTTAATAGTTTGCGCAACGTTGTTG -3') was incubated with RecA for 5 minutes at 37˚C before the addition of supercoiled pBR322 plasmid to initiate strand invasion. After 5 minutes the reaction was terminated by the addition of stop mix and incubating for a further 30 min at 37 °C to deproteinize the mixture. To purify D-loops the deproteinized reaction mixture was applied to a 3.5 ml sepharose CL-2B column, which was developed with 20 mM Tris-HCl (pH 8.0) and 0.5 mM MgCl 2 . Substrate concentrations were determined with reference to the specific activity of the radiolabelled oligonucleotide. χ, χ Kpn and χ Sma substrates were made as described (McGlynn and Lloyd, 2000).

EMSA
Reaction mixtures (20 µl) contained 0.5 nM labeled substrate DNA in Binding Buffer (50 mM Tris-HCl, pH 8.0, 1 mM DTT, 100 µg/ml BSA, 6% glycerol). Reactions were started typically by the addition of Fml1∆C, held on ice for 15 min, and then loaded immediately onto a pre-equilibrated 4% native polyacrylamide gel in low ionic strength buffer (6.7 mM Tris-HCl, pH 8.0, 3.3 mM sodium acetate, 2 mM EDTA). Samples were run into the gel typically for 1hr and 30 mins at 160 V with buffer recirculation occurring throughout. Both buffer and gel were pre-cooled at 4°C, but electrophoresis was at room temperature. Gels were dried on 3 MM Whatman paper, and analysed with a Fuji FLA3000 PhosphorImager.   Repair of ade6 can occur with or without crossing over. Crossing over results in the integration of the plasmid into the chromosome and therefore stability of the Ade + Ura + plasmid markers, whereas for noncrossovers the plasmid is maintained extrachromosomally due to the autonomously replicating sequence ars1, and therefore can be lost if selection for it is not maintained (unstable Ade + Ura + ). The position of the M26 mutation is indicated by the filled circle and is outside the position of the double-strand gap. Restriction sites, DNA probes, and fragment sizes, relevant to the southern blot analysis in (C), are indicated. (B) An example of the analysis of plasmid gap repair assay transformants. (I) Ura + transformants are first patched onto minimal media lacking uracil (EMMG-u). (II) The number of ura + transformants that are ade + is assessed by replica plating the EMMG-u plate onto minimal media lacking uracil and adenine (EMMG-u-a). (III and IV) Two rounds of replica plating onto non-selective media (YELA) are performed to allow loss of non-integrated plasmid DNA. (V) The second YELA plate is replica plated onto FOA to identify stable and unstable transformants. Stable transformants remain ura + and therefore are unable to grow on FOA (examples indicated by black arrowheads), whereas, unstable transformants lose the ura4 + gene and are therefore able to grow on FOA. (C) Southern blot analysis of eight crossover transformants and eight noncrossover transformants identified by the protocol described in (B). Genomic DNA from eight stable Ura + Ade + and eight unstable Ura + Ade + transformants was digested with EcoRV, run on a agarose gel, southern blotted and probed with a fragment of ade6 DNA as indicated in (A). In all cases the pattern of bands confirms the designation of transformants as crossover or noncrossover by the protocol in (B). To see whether Fml1∆C retains activity in vivo we compared full-length and truncated Fml1 for their ability to complement the MMS hypersensitivity of the fml1∆ mutant. The proteins were expressed from the thiamine-repressible nmt promoter in pREP41, and in the absence of thiamine both fully complemented the fml1∆ mutant. Even in the presence of thiamine, the low-level of Fml1 that leaks from the repressed nmt promoter is sufficient to fully complement fml1∆. In contrast, the repressed level of Fml∆C only partially complements. These data indicate that Fml1∆C retains its core biological function, albeit it needs to be over expressed to be fully effective in vivo. Possibly the non-conserved C-terminal domain, which is deleted in Fml1∆C, promotes protein stability or efficient targeting of substrates.        Each DNA substrate is made from the oligonucleotides indicated by the number on each representative schematic. The number is positioned at the 5'-end of its respective oligonucleotide. Note that the 3-strand junction is called F9 in the paper.