Purified Human MSH2 Protein Binds to DNA Containing Mismatchednucleotid&
Total Page:16
File Type:pdf, Size:1020Kb
[CANCERRESEARCH54,5539—5542,November1,19941 Advances in Brief Purified Human MSH2 Protein Binds to DNA Containing MismatchedNucleotid& Richard Fishel,2 Amy Ewel, and Mary Kay Lescoe Department ofMicrobiology and Molecular Genetics, Markey Centerfor Molecular Genetics, University of Vermont School ofMedicine, Burlington, Vermont 05405 Abstract context of the mutator hypothesis, it is proposed that mutator genes exist which increase the mutation rate of a preneoplastic cell, allowing The human hMSH2 protein is a member of a highly conserved family multiple mutations to accumulate (5). A role for hMSH2 in alteration of postreplication mismatch repair components found from bacteria to of mutation rates is clear if its biochemical functions are similar to the humai@. Alterafions of the gene coding for this protein cosegregate with, bacterial and yeast mismatch repair proteins. and are the likely cause of, chromosome 2.Iinked hereditary nonpolyposis colon cancer. PostrepUcadon mismatch repair has been found to faithfully Based on its similarity to the bacterial MutS protein (5, 27, 28), it replace misincorporated nucleotides, thereby increasing the overall fidel is anticipated that hMSH2 is responsible for the initial recognition of Ity of DNA replication. Loss of postreplicadon mismatch repair function a mismatched DNA substrate. In this report, we demonstrate that leads to a mutator phenotype, which is proposed to account for the purified hMSH2 binds specffically to mismatched nucleotides, multiple mutations required for multistep carcinogenesis. Although the thereby providing a target for the mismatch repair process. fUnctions of hMSH2 can be anticipated based on Its similarity to well characterized bacterial and yeast proteins, proof of its functions has not Materials and Methods been established. Here we demonstrate that purified hMSH2 binds specifically to mismatched nucleotides, providing a target for the excision Chemicals and Enzymes. Ultrapure Tris (acid and base), EDTA, MgCI2, repair processes characteristic of postreplication mismatch repair. MgSO4, NaC1,phenylmethylsulfonylfluoride, and analyticalgrade sodium citrate, KC1,potassium phosphate monobasic (KH2PO4),and potassium phos Introduction phatedibasic(K2HPO4)wereobtainedfromAmresco(Solon, OH). Ultrapure glycerol was obtained from Mallinckrodt,Inc. (Paris, KY). Pepstatinand HNPCC3 may affect up to 1 in 200 people in industrialized nations leupeptinwerepurchasedfromSigmaChemicalCo. (St. Louis,MO).AlP was (1—4).Fourgenes have been identifiedthatcosegregatewith, andare purchasedfrom PharmaciaLKB Biotechnology, Inc. (Stockholm, Sweden). the likely cause of HNPCC susceptibility in at least 50% of the Restriction endonucleases, polynucleotide kinase, and Factor Xa were identified kindredS. Within these kindreds, hMSH2 on chromosome purchasedfromNew EnglandBiolabs, Inc. (Beverly, MA). DTT and isopro pylthio-@3-galactosidewaspurchasedfromGIBCO-BRL(Gaithersburg,MD). 2p21—22appears to account for 60% of the cases (5), hMLHi on Purification of hMSH2 PrOtein. The overproduction of hMSH2 was chromosome3p21 appearsto accountfor 30% of the cases (6), and achieved by insertion of a recombinant hMSH2 complementary DNA se PMS1 on chromosome 2q31—33and PMS2 on Chromosome 7p2l quenceconsistingof the entirehMSH2gene (5) with an initiatormethionine appearsto each accountfor 5% of the cases(7). (containing an engineered NcoI site), six histidine residues, and a factor Xa hMSH2, hMLH1, hPMS1, and hPMS2 are members of two highly cleavage site (IleGluGlyArg)fused to the normalhMSH2NH2-terminalme conserved families of postreplication mismatch repair genes, MutS thionine and cloned into the NcoI site of the pET1id expression vector and MaiL, that are involved in increasing the fidelity of replication by (Novogen, Madison,WI). Inductionwith 0.1 M isopropylthio-(3-galactoside specific repair of DNA polymerase misincorporation errors (8, 9). resulted in expression of hMSH2 (Mr 106,000) in pLysS bacteria. Four liters Ablation of MutS, MutL, or their homologues in bacteria and yeast of induced bacterialcells were pelleted and washed with 150 ml 20 mM increases the rate of spontaneous mutation resulting in a mutator potassium phosphate(pH 8), 150 mMNaCl, 10 mM EDTA plus protease inhibitors (0.5 mM phenytmethylsulfonyl fluoride, 0.8 @g/mlpepstatin, 0.8 phenotype (10—12).The most commonly recognized mutator pheno @xg/mlleupeptin), then washed twice with 150 ml 20 mM potassium phosphate type is exhibited as replication errors of simple repetitive DNA that (jH 8), 150 mMNaCl plus proteaseinhibitorsandthenresuspendedin 20 ml results in microsatellite instability (i3—15).A similar microsatellite of thesamebuffer.Thesuspensionwas frozenandthawedon ice to inducecell instability mutator phenotype has been observed in tumors from lysis, and the cell debri was pelleted by centrifugation at 100,000 X g (Fraction HNPCC patients (16), a wide variety of sporadic tumors (15, 17—20), I, 220 mg). To thesupernate,saturatedammoniumsulfatewas addedon ice to and several cell lines derived from tumors (21). A proposed mecha a final concentrationof45%(v/v), andthe solutionwas allowed to stiron ice nism for the microsatellite instability mutator phenotype involves for an additional60 min. The precipitatewas collected by centrifugationand misrecognition of replication-induced errors that result in mismatched resuspendedin4 ml of 20 mMpotassiumphosphate(jH 8), 300 mMNaC1plus nucleotides within these repeat sequences (22). These mismatched proteaseinhibitorsand dialyzed 8 h against200 volumes of the same buffer nucleotides are subsequently fixed into the genome by a second round (Fraction II, 46 mg). Fraction II was applied to a Ni/NTA agarose column (0.5 of replication,resultingin localized genome instability. cm X 5 cm; Qiagen, Chatsworth, CA) equilibrated 20 mMpotassium phos phate(pH 8), 300 mMNaClplus proteaseinhibitors.Thecolumnwas washed A mutator phenotype has been proposed to account for the multiple with 10 volumes of the equilibration buffer and developed with pH step mutations required for multistage carcinogenesis (23—26).Within the gradients (pH 7, pH 6, pH 5.5, pH 4.5, and pH 4). hMSH2 eluted in the pH 4.5 step (Fraction III, 2 mg). Fraction III was diluted two fold with 20 mM Received 9/13/94; accepted 9/19/94. potassium phosphate (pH 8), 2 m@tDTT, 0.2 m@iEDTA plus protease inhib Thecostsof publicationofthisarticleweredefrayedinpartbythepaymentofpage itors and applied directly to a (0.5 cm X 8 cm) hydroxyappetite column charges.Thisarticlemustthereforebeherebymarkedadvertisementinaccordancewith 18U.S.CSection1734solelyto indicatethisfact. (Bio-Rad,Hercules,CA) equilibratedin20 mMpotassiumphosphate,100 mM @ I work was supported by NIH Grant CA 56542 and the Lake Champlain Cancer NaCl, i mMDTF, 0.1 mMEDTA plus protease inhibitors(buffer A). The ResearchOrganization. column was washed in buffer A except the phosphate concentration was 2 To whom requests for reprints should be addressed, at Department of Microbiology increasedto50 mM,followedby a lineargradientto0.25 Mphosphate.hMSH2 and Molecular Genetics, University of Vermont College of Medicine, Stafford Hall, Burlington,VT05405-0068. eluted in a broad band between 0.2 and 0.25 Mphosphate (Fraction IV, 0.7 3 The abbreviations used are: HNPCC, hereditary nonpolyposis colon cancer; D'IT, mg). Fraction IV was diluted 2-fold with buffer A minus the potassium dithiothreitol; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis. phosphateandapplieddirectlyto a (0.5 cm X 6 cm) heparin-Sepharosecolumn 5539 Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1994 American Association for Cancer Research. BINDING OF hMSH2 TO MISMATCHED NUCLEOTIDES AB C 1 23456 50 ,@ 200— 40@ Fig. 1. Purification and mismatch binding activ ity of hMSH2. A. Silver-stained SDS-PAGE of Frac tion V hMSH2 protein (see “MaterialsandMeth 116— ods―).B, mismatch binding gel-shift assay using 97— Fraction V hMSH2. Lane 1, no protein; Lane 2, 24 30@ ng hMSH2;Lane 3, 74 ng hMSH2;Lane4, 239 ng counts 67— hMSH2; Lane 5, 399 ng hMSH2; Lane 6, 798 ng I mm2 hMSH2. C. quantitationof the binding activity kDa (x103) 20 shown in (B). Countsper mm2of the shifted mate 45— rialwas determinedusingthe MolecularAnalyst2.0 software associated with the BiOradphosphoimager (Bio-Rad). 10 31 — @ 21 — 0 • • I I I I 0 200 400 600 800 MSH2(ng) (Pharmacia) equilibrated in buffer A, washed with 10 column volumes, and tains a G-T mismatch at position 19 within the context of a 39-mer eluted with 0.1 MNaCl step gradients. hMSH2 eluted in the 0.5 MNaC1step oligonucleotide, which may be bound by a mismatch binding proteinand (Fraction V, 0.3 mg) and was homogeneous as judged by SDS-PAGE visualized by its retardation relative to the unbound material following following silver stainingfor protein(Fig. IA). gel electrophoresis (29). The binding of purified hMSH2 to the G-T Mismatch Binding Assay. Complementary strands of a 39-mer were syn mismatch-containing oligonucleotide is shown in Fig. 1B, and quantita thesized (Applied Biosystems 392, Foster City, CA). The top strand always contained the sequence 5'-TGG GCG CFA GAG GCG TGG GAA AGA GTG lion of this binding activity is shown in Fig. 1C. The binding appears to CCCTGA CCATCC.The bottomstrandforthe homoduplexoligonucleotide be cooperative and reaches 80% saturation at a protein:DNA ratio of was 5'-GGA TGG TCA 000 CAC TCT UC CCA CGC CFC TAG CGC approximately 4:1. These results are qualitatively similar to those