Nucleosomes around a mismatched base pair are excluded via an Msh2-dependent reaction with the aid of SNF2 family ATPase Smarcad1

Here, Terui et al. studied the mechanisms underlying chromatin remodeling that occurs during MMR. They show that the eukaryotic MMR system has an ability to exclude local nucleosomes and identify Smarcad1/Fun30 as an accessory factor for the MMR reaction.


Cloning and plasmids
Construction of pMM1 was described previously (Kawasoe et al. 2016).
Construction of pMM3 was performed as follows: A linker DNA fragment was amplified by PCR with primers 1079 and 1158 using fission yeast genomic DNA as a template. The DNA fragment was digested with PstI and BspQI (New England Biolabs) and inserted between the same sites in pMM1, resulting in pMM3.
Cloning of Xenopus laevis smarcad1 gene was performed as follows: A BLAST search using the Xenopus tropicalis Smarcad1 sequence identified two Xenopus laevis EST clones, TC422950 and TC460920. Based on these EST sequences, we designed two primers, 900 and 887, and amplified the smarcad1 gene by PCR from Xenopus egg cDNA. The smarcad1 gene fragment was digested with NdeI and BamHI-HF (New England Biolabs) and cloned into pDE1a, a derivative of the pDONR201 vector (Life Technologies, Carlsbad, CA, USA) carrying NdeI and BamHI sites between attL1 and L2 sites. Sequencing of cloned genes revealed that two distinct isoforms, which we named smarcad1a and smarcad1b, were present (Plasmids: pDE1a-SMARCAD1A and pDE1a-SMARCAD1B). Smarcad1a and Smarcad1b were 90% identical and 95% similar with respect to their amino acid sequences. The smarcad1a gene was used for all subsequent construction and experiments, and therefore the gene product was called simply Smarcad1, unless otherwise indicated. To introduce the lysine 503 to alanine substitution in the Walker A motif, the gene fragment was amplified by PCR using primer pairs, 887 and 955, and 900 and 956, using pDE1a-SMARCAD1A as a template, and the two PCR fragments were fused by overlap-extension PCR with primers 887 and 900. The resulting smarcad1a K503A fragment was digested with NdeI and EcoRI (New England Biolabs), and cloned into the same sites in pDE1a-SMARCAD1A, resulting in pDE1a-SMARCAD1A-K503A. To add two tandem FLAG tags to the Nterminus of Smarcad1, a synthetic linker prepared by annealing of 5′phosphorylated oligonucleotides 60 and 61 was inserted in the NdeI sites in pDE1a-SMARCAD1A and pDE1a-SMARCAD1A-K503A, resulting in pDE1a-FLAG-SMARCAD1A and pDE1a-FLAG-SMARCAD1A-K503A, respectively.
Baculoviruses for expression of FLAG-Smarcad1 and FLAG-Smarcad1-K503A were prepared by transferring the FLAG-smarcad1a and FLAG-smarcad1a K503A genes into BaculoDirect C-term Linear DNA (Life Technologies) using the Gateway LR reaction. cDNAs of Xenopus laevis spt16 and ssrp1 genes were kind gifts from Haruhiko Takisawa, Yumiko Kubota, and Masato Kanemaki. The spt16 gene was amplified by two-step PCR using primers 798 and 799, and then primers 344 and 345, and cloned into the pDONR201 vector using the Gateway BP reaction, resulting in pDONR-SPT16. The ssrp1 gene was amplified by PCR using primers 770 and 771, digested with NcoI (New England Biolabs) and Sse8387I (Takara, Kusatsu, Japan), and cloned into the same sites in a modified pDE1a vector, resulting in pDONR-SSRP1. For protein expression in Escherichia coli, the gene fragments on the Gateway entry vectors were transferred into pET-HSD, a derivative of the pETDuet-1 vector (Merck Millipore, Billerica, MA, USA, Cat#71146-3CN) carrying a Gateway recombination cassette and a His-tag for Nterminal fusion, by the Gateway LR reaction, resulting in pET-HSD-SPT16 and pET-HSD-SSRP1, respectively. The N-terminally His 6 -FLAG-tagged spt16 gene was amplified by two-step PCR using primers 799 and 827, and then primers 799 and 81, digested with NcoI, and cloned into pDONR-SPT16, resulting in pDONR-His 6 -FLAG-SPT16. Baculoviruses for expression of His 6 -FLAG-Spt16 and Ssrp1 were constructed by transferring the His 6 -FLAG-spt16 and ssrp1 genes into BaculoDirect C-term Linear DNA by the Gateway LR reaction.
Cloning of the Xenopus laevis msh3 gene was performed as follows: A BLAST search using the Xenopus tropicalis Msh3 sequence identified a partial Xenopus laevis EST clone, CA988114. The missing 5′ and 3′ portions of the msh3 cDNA were cloned by 5′ and 3′ RACE using the SMARTer RACE cDNA Amplification kit (Clontech, CA, USA) with primers 784 and 780, respectively. The full-length msh3 ORF was then PCR-amplified from Xenopus laevis egg cDNA by using primers 957 and 958, and then 344 and 355, and cloned into pDONR201 by the Gateway BP reaction, resulting in pDONR-MSH3. For protein expression in E. coli, the msh3 gene was transferred into pET-HSD by the Gateway LR reaction, resulting in pET-HSD-MSH3.
The budding yeast fun30-K603A mutant gene in which lysine 603 in the Walker A motif was replaced with alanine was prepared by overlap-extension PCR with primers 1564, 1565, 1566, and 1567 using BY4741 genomic DNA as templates. The resulting fragment was digested with EcoRI and BamHI, and cloned into YIplac211, resulting in YIpURA3-fun30-K603A.
The lys2::insE-A14 gene was constructed as follows: Two partially overlapping fragments of the lys2::insE-A14 gene were separately prepared by two-step PCR with following primer pairs: the 5′ half of the fragment, 1296 and 1298, and 1296 and 1426; the 3′ half of the fragment, 1297 and 1301, and 1297 and 1300. Two fragments were then simultaneously inserted into pBluescript II KS(-) linearised by PCR with primers 1294 and 1295 by the Gibson assembly reaction (New England Biolabs), resulting in pBS-lys2::insE-A14. The PvuII-PstI fragment of pBS-lys2::insE-A14 was subcloned between the PstI and SmaI sites in YIplac211, resulting in YIpURA3-lys2::insE-A14.
The spt16-d922 mutant gene was prepared by two-step overlap-extension PCR with primers 1571, 1572, 1573, and 1574 using BY4741 genomic DNA as templates. The fragment was digested with BamHI and HindIII (New England Biolabs), and cloned into YIplac211, resulting in YIpURA3-spt16-d922.
Construction of pDONR-xMLH1 was described previously (Kawasoe et al. 2016). For protein expression in E. coli, the mlh1 gene was transferred into pDEST17 (Life Technologies) by the Gateway LR reaction, resulting in pDEST17-

Protein expression and purification
Purification of Xenopus laevis MutSα was carried out as described previously (Kawasoe et al. 2016 protein was eluted from the FLAG-M2 resin with 50 μg/mL FLAG-peptide (Sigma Aldrich) in buffer S containing 0.1x cOmplete EDTA-free. Peak fractions were pooled and three-fold diluted with buffer A (20 mM Tris-HCl, 5% glycerol, 5 mM 2mercaptoethanol, 1 mM EDTA, pH 7.4) containing 0.1x cOmplete EDTA-free, loaded on a MonoQ 5/50 GL column (GE Healthcare, Little Chalfont, UK), and the column was developed with a 0-1 M NaCl linear gradient in buffer A containing 0.1x cOmplete EDTA-free. Peak fractions were pooled and loaded on a Hi Load 16/60 Superdex 200 prep grade column (GE Healthcare), and the column was developed with buffer A containing 0.14 M NaCl. Fractions corresponding to the molecular mass of 2.5-5.0 × 10 5 (FLAG-Smarcad1: M r = 1.19 × 10 5 ) were pooled, concentrated using Amicon Ultra (Merck Millipore), and frozen in liquid nitrogen as small aliquots.
Purification of the X. laevis FACT heterodimer was performed as follows: Purification of the N-terminally His 6 -tagged, full-length X. laevis Spt16 protein was performed as follows: Spt6-containing inclusion bodies were purified by the method essentially the same as that for Msh3, except that protein expression was induced for 2 h. The inclusion bodies were resuspended in 0.5× buffer SO containing 0.5 mM PMSF, 1 mM benzamidine, 7 M urea, 2 M thiourea, 100 mM DTT. 4× Laemmli's SDS sample buffer was also added to final 1× concentration. The sample was incubated for 20 min at 37°C and centrifuged at 15,000 rpm for 20 min in TA-24BH to remove insoluble debris. The Spt16 protein was then purified by SDS-PAGE followed by electroelution.
Purification of the N-terminally His 6 -tagged, full-length X. laevis Ssrp1 protein was performed as follows: The method for protein expression and preparation of bacterial lysate were essentially the same as that for Msh3, except that protein expression was induced at 20°C for 20 h. The lysate was centrifuged at 81,800 ×g (30,000 rpm) for 30 min in Beckman 50.2Ti. The His-Ssrp1 protein in the cleared lysate was bound to the TALON metal affinity resin (Clontech) for 1 h at 4°C and eluted with 100 mM imidazole in buffer W (20 mM Na-phosphate, 500 mM NaCl, 0.1% Triton X-100, pH 8.0) containing 0.1 mM PMSF and 0.2 mM benzamidine. The eluate was diluted four-fold with buffer B (50 mM Na-phosphate, 5% glycerol, pH 6.8), loaded on a HiTrap Q-HP 1-mL column, and the column was developed with a 0-1 M NaCl linear gradient in buffer B. Peak fractions were pooled, diluted four-fold with buffer B, loaded on a HiTrap SP-HP 1-mL column (GE Healthcare), and the column was developed with a 0-1 M NaCl linear gradient in buffer B.
The E. coli BL21 codon plus (DE3) cells carrying pET28c-xHIRA was a kind gift from Masato Kanemaki. The method for expression and purification of the X. laevis HIRA protein was essentially the same as that for Msh3, except that protein expression was induced for 7 h at 37°C.
Purification of the N-terminally His 6 -tagged, full-length X. laevis Mlh1 protein was performed as follows: The method for protein expression and preparation of bacterial lysate were essentially the same as that for Msh3, except that protein expression was induced for 5 h. Inclusion bodies containing the Mlh1 protein were resuspended in wash buffer (50 mM Na-phosphate, 1 M NaCl, 0.1% Triton X-100, pH 8.0) and centrifuged at 13,000 rpm for 20 min in TA-24BH. The pellet was resuspended in wash buffer containing 1 M urea, centrifuged again at 13,000 rpm for 20 min in TA-24BH, and these procedures were repeated three times. The Mlh1 protein was dissolved in Laemmli's SDS sample buffer containing 4 M urea and purified by SDS-PAGE followed by electroelution.

Immunological methods
Production and usage of Msh2R1, Msh6, Mlh1 (Kawasoe et al. 2016), and Cdc7 antibodies (Takahashi and Walter 2005) were described previously. The rabbit Msh2pep antiserum was raised against peptide NH 2 -CLAKNNRFVSEVISRTKTGL-COOH, corresponding to residues 914-932 of Msh2. The rabbit Msh2R2 antiserum was raised against N-terminally His 6 -tagged and C-terminally Strep-II-tagged fulllength Msh2 expressed in E. coli. The rabbit Msh3 antiserum was raised against Nterminally His 6 -tagged, full-length Msh3 expressed in E. coli. The rabbit HIRA antiserum was raised against N-terminally His 6 -tagged, full-length HIRA expressed in E. coli. The rabbit Spt16 antiserum was raised against N-terminally His 6 -tagged, full-length Spt16 expressed in E. coli. The rabbit Ssrp1 antiserum was raised against N-terminally His 6 -tagged, full-length Ssrp1 expressed in E. coli. The rabbit xH2B antiserum was raised against peptide NH 2 -CAKHAVSEGTKAVTKYTSAK-COOH, corresponding to residues 108-126 of H2B. The rabbit xH3 antiserum was raised against peptide NH 2 -ARTKQTARKSTGGKAC-COOH and NH 2 -CPKDIQLARRIRGERA-COOH, corresponding to residues 1-15 and 121-135 of H3, respectively. The rabbit Smarcad1 antiserum was raised against peptide NH 2 -CDEGTIPLDMATLLKTSLGL-COOH, corresponding to residues 983-1001 of Smarcad1a. This peptide is 100% conserved between Smarcad1a and Smarcad1b, and therefore the resulting antibodies should recognize both isoforms. The rabbit xCAF-1 antiserum was raised against peptide NH 2 -CSSADKPSGSDQTNK-COOH and NH 2 -CFDEIKKRKPRKMG-COOH, corresponding to residues 555-569 of xCAF-1 p60 and 450-452 of xCAF-1 p150, respectively. All antibodies except for Mlh1, Spt16, Ssrp1, and CAF-1 were affinity-purified using corresponding antigens. The  To obtain the p-values, the number of revertants obtained by the same procedure was normalized by using viable cell counts, and compared by Mann-Whitney's U-test. Calculation was performed using Graphpad Prism 6 (Graphpad Software, La Jolla, CA, USA).

Repeatability
For supercoiling assays, mismatch-DNA binding assays, immunoprecipitations, and micrococcal nuclease digestion experiments, representative results, out of at least three independent experiments using at least two different preparations of NPE, are shown. Immunoblots for evaluation of depletion efficiencies were carried out once for each single depletion experiment. Spectral counting by mass spectrometry was carried out three times using three independent samples. Because there was no reliable method to merge spectral counts obtained from different experiments, two representative data were presented.

Code availability
Calculation of the MSS maximum likelihood function was performed by using a custom software coded by using Swift2.0 on Xcode (Apple Computer, Cupertino, CA, USA). The source code and the compiled software will be provided upon request.

Data availability
The GenBank accession numbers for sequences of Xenopus smarcad1a, smarcad1b, and msh3 mRNA reported in this paper are LC183875, LC183876, and LC183877, respectively. All data supporting the findings of this study are available from the corresponding author on request.

Supplemental Figure S3. Characterization of Smarcad1 and FACT antisera
The indicated amount of low-speed supernatant (LSS), NPE, or recombinant proteins was separated by SDS-PAGE and transferred onto PVDF membranes.
Each membrane strip was probed with either the indicated antiserum or the pre-immune serum (PI) from the same rabbit. The same exposure sets are presented for each pair of PI and antiserum. Smarcad1 (M r = 1.14 × 10 5 ), Spt16 (M r = 1.18 × 10 5 ), and Ssrp1 (M r = 0.79 × 10 5 ) were detected as nearly a single band in NPE.
Each band was specifically immunoprecipitated by the corresponding antibody (see         Figure S7 Supplemental