Portal de Periódicos da CAPES

MutLα: At the Cutting Edge of Mismatch Repair

2006; Cell Press; Volume: 126; Issue: 2 Linguagem: Inglês

10.1016/j.cell.2006.07.003

1097-4172

Josef Jiricny,

Genomics and Rare Diseases

The mismatch repair process corrects errors in newly synthesized DNA. In this issue, Modrich and colleagues (Kadyrov et al., 2006Kadyrov F.A. Dzantiev L. Constantin N. Modrich P. Cell. 2006; (this issue)PubMed Google Scholar) show that a component of the human mismatch repair machinery, MutLα, has endonuclease activity. MutLα introduces single-strand breaks near the mismatch and thus generates new entry points for the exonuclease EXOI to degrade the strand containing the mismatch. The mismatch repair process corrects errors in newly synthesized DNA. In this issue, Modrich and colleagues (Kadyrov et al., 2006Kadyrov F.A. Dzantiev L. Constantin N. Modrich P. Cell. 2006; (this issue)PubMed Google Scholar) show that a component of the human mismatch repair machinery, MutLα, has endonuclease activity. MutLα introduces single-strand breaks near the mismatch and thus generates new entry points for the exonuclease EXOI to degrade the strand containing the mismatch. Removal of errors from newly synthesized DNA by the mismatch repair (MMR) machinery increases the fidelity of the replication process by up to three orders of magnitude. Moreover, the inactivation of genes involved in MMR by mutation or epigenetic silencing leads to a substantial increase in mutation frequency, and in mammals, loss of MMR promotes cancer of the colon and other organs (Jiricny, 2006Jiricny J. Nat. Rev. Mol. Cell Biol. 2006; 7: 335-346Crossref PubMed Scopus (841) Google Scholar). Given the importance of the MMR system in the maintenance of genomic stability, the study of the molecular mechanisms underpinning this process is of substantial interest. The human MMR process has been recently reconstituted in vitro using purified recombinant constituents. This work has shown that the degradation of the mismatch-containing strand from a preexisting nick or gap is bidirectional—that is, it can take place irrespective of whether this break is situated 5′ or 3′ to the mismatch. This observation has been puzzling because the reconstituted system uses only a single exonuclease, EXOI, which degrades DNA in a 5′-to-3′ direction. Thus, it has been difficult to explain how EXOI degrades the mismatch-containing strand when the mismatch is 5′ to the preexisting nick or gap. Kadyrov et al., 2006Kadyrov F.A. Dzantiev L. Constantin N. Modrich P. Cell. 2006; (this issue)PubMed Google Scholar now show how this can occur: A component of the MMR machinery, MutLα, possesses an endonuclease activity that introduces single-strand breaks selectively into the discontinuous strand. These new breaks then serve as targets for EXOI degradation. This allows the MMR machinery to remove the mismatch, regardless of its placement with respect to the preexisting break. The existence of MMR was first invoked more than 30 years ago to explain cosegregation of closely spaced genetic markers in phage lambda. Subsequent efforts identified the genes required for mismatch repair in the Escherichia coli genome: mutH, mutL, mutS, and mutU (uvrD). In 1989, the E. coli MMR system was reconstituted from purified proteins. The reconstituted system revealed that the repair process begins by binding of the MutS protein to the mispair; MutS then recruits MutL and activates the endonucleolytic activity of MutH. MutH is the strand-discrimination factor that directs the repair process to the newly synthesized strand by nicking the nearest unmethylated GATC site in this strand, irrespective of whether it is situated 5′ or 3′ to the mismatched nucleotide. The complex then loads the MutU (UvrD) DNA helicase as well as one of several exonucleases (which degrade the error-containing strand in the 5′-to-3′ or 3-to-5′ direction, depending on the position of the MutH-mediated incision) until the mismatch has been removed. Polymerase III then fills in the resulting single-stranded gap, and the repair process is completed by DNA ligase (Modrich and Lahue, 1996Modrich P. Lahue R. Annu. Rev. Biochem. 1996; 65: 101-133Crossref PubMed Scopus (1286) Google Scholar) (Figure 1). The MMR system is highly conserved in evolution, and homologs of MutS and MutL (MSH and MLH, respectively) are found in all organisms except some archaea; the mechanism of MMR also appears to be broadly similar. Thus, mismatch recognition in eukaryotes is mediated by one of two heterodimers, composed of the MutS homologs MSH2 and MSH6 (MutSα) or MSH2 and MSH3 (MutSβ). MutSα is more abundant and participates in the repair of base/base mismatches and small strand misalignments of 1–2 nucleotides, whereas MutSβ is responsible for initiating the repair of larger loops that arise during replication. In the subsequent step, the mismatch bound MutSα or MutSβ recruits a heterodimer composed of two MutL homologs, MLH1 and PMS2 (PMS1 in yeast), referred to as MutLα. The MutSα/MutLα complex is then thought to translocate along the DNA contour in an ATP-dependent manner until it encounters a strand break, where it can load the exonuclease required for the degradation of the error-containing strand. This is where eukaryotic and E. coli MMR differ. Only Gram-negative bacteria have an endonuclease homologous to MutH. In other organisms, it has been suggested that MMR is directed to the newly synthesized DNA strand by preexisting termini, such as gaps between Okazaki fragments in the lagging strand or the 3′ terminus of the primer on the leading strand. This presupposes that the MMR process is closely coupled to replication. This would appear to be the case, as both MSH6 and MSH3 interact with proliferating cell nuclear antigen (PCNA) (Jiricny, 2006Jiricny J. Nat. Rev. Mol. Cell Biol. 2006; 7: 335-346Crossref PubMed Scopus (841) Google Scholar), which is the processivity factor of replicating DNA polymerases. The first evidence of nick-directed MMR was obtained when mismatch-containing heteroduplexes were incubated with extracts of human cells. Covalently closed circular DNA molecules were refractory to mismatch correction. However, if a nick was introduced within roughly 1 kilobase either 5′ or 3′ to the mispair, the extracts converted the heteroduplex to a homoduplex by degrading the stretch between the mismatch and the nick and then filling in the resulting single-stranded gap (Holmes et al., 1990Holmes J.J. Clark S. Modrich P. Proc. Natl. Acad. Sci. USA. 1990; 87: 5837-5841Crossref PubMed Scopus (325) Google Scholar). Recently, the laboratories of Paul Modrich (Constantin et al., 2005Constantin N. Dzantiev L. Kadyrov F.A. Modrich P. J. Biol. Chem. 2005; 280: 39752-39761Crossref PubMed Scopus (174) Google Scholar) and Guo-Min Li (Zhang et al., 2005Zhang Y. Yuan F. Presnell S.R. Tian K. Gao Y. Tomkinson A.E. Gu L. Li G.M. Cell. 2005; 122: 693-705Abstract Full Text Full Text PDF PubMed Scopus (275) Google Scholar) succeeded in reconstituting the minimal human MMR system from purified recombinant MutSα or MutSβ, MutLα, EXOI, PCNA, replication factor C (RFC, which loads PCNA onto DNA), the single-strand binding factor replication protein A (RPA), polymerase δ, and DNA ligase I. The high mobility group protein B1 (HMGB1) could substitute for RPA. In this system, the 5′-to-3′ mismatch-directed strand excision required only MutSα, EXOI, and RPA, whereas processing of heteroduplex substrates carrying a 3′ nick required also MutLα, PCNA, and RFC (Dzantiev et al., 2004Dzantiev L. Constantin N. Genschel J. Iyer R.R. Burgers P.M. Modrich P. Mol. Cell. 2004; 15: 31-41Abstract Full Text Full Text PDF PubMed Scopus (193) Google Scholar). The surprising finding was that the latter system required no 3′-to-5′ exonuclease. Because mutations in the active site of EXOI abolished the repair reaction, the latter enzyme was suggested to possess a cryptic 3′-to-5′ exonuclease activity that might be activated upon formation of the multiprotein complex at the site of the 3′ nick. However, subsequent careful examination of the denaturing gels used to follow the progress of the strand degradation revealed that the nicked strand, rather than being degraded by an exonuclease into mononucleotides in the 3′-to-5′ direction, was cleaved into short single-stranded oligonucleotides (Kadyrov et al., 2006Kadyrov F.A. Dzantiev L. Constantin N. Modrich P. Cell. 2006; (this issue)PubMed Google Scholar). This finding suggested that new incisions were introduced into the nicked strand and triggered the search for a polypeptide in the reconstituted MMR system that harbors an endonuclease activity. Such an activity has now been identified in MutLα. Using Fenton chemistry, the active site of the enzyme was located in the PMS2 subunit, and site-directed mutagenesis showed that substitution of either acidic residue in the active-site motif DQHA(X)2E(X)4E abolished not only the endonuclease activity of MutLα but also the mismatch-dependent excision on the 3′ substrate. This finding explains why the reconstituted system does not require a 3′-to-5′ exonuclease: If the MutLα-catalyzed incisions are introduced 5′ to the nick and beyond the mismatch, EXOI can be loaded at these sites and can degrade the single strand in a 5′-to-3′ direction, concurrently removing the mismatch (Figure 1). In human cells, MLH1 also forms heterodimers with PMS1 and MLH3, which are called MutLβ and MutLγ, respectively. MutLβ has no detectable MMR activity, and MutLγ is predominantly involved in meiotic recombination but may play a minor role in MMR (Jiricny, 2006Jiricny J. Nat. Rev. Mol. Cell Biol. 2006; 7: 335-346Crossref PubMed Scopus (841) Google Scholar). Interestingly, the consensus endonuclease sequence is conserved in MLH3, but not in PMS1, which helps explain why MutLβ has no activity in MMR. The sequence is also not conserved in MutL homologs from organisms that have MutH activity. However, the MutLα endonuclease is not a functional equivalent of MutH, as it cannot incise covalently closed DNA molecules. The discovery of an endonuclease activity associated with MutLα solves the mystery of how a reconstituted bidirectional DNA repair process can operate with just a single exonuclease that degrades only in the 5′-to-3′ direction. What this elegant work fails to explain is how the endonuclease distinguishes between the intact and the nicked strands when the discontinuity can be several hundred nucleotides away. Is the MMR complex loaded at the nick or gap in a directional manner? If so, how can it retain this directionality while following the contour of the DNA through tens of helical turns? The present finding also raises a number of other questions. The first is linked to the role of MutLα in MMR in vivo. Cells lacking MLH1 or MSH2 both have similarly high rates of mutation. If, as implied from the reconstituted system, MutLα is not required for 5′-to-3′ MMR, the high rate of mutation in cells lacking MLH1 can only be explained if the majority of MMR events in vivo proceed in the 3′-to-5′ direction. However, if this is the case, why is the phenotype of cells that lack EXOI not as pronounced as that of cells lacking MLH1 (Wei et al., 2003Wei K. Clark A.B. Wong E. Kane M.F. Mazur D.J. Parris T. Kolas N.K. Russell R. Hou Jr., H. Kneitz B. et al.Genes Dev. 2003; 17: 603-614Crossref PubMed Scopus (243) Google Scholar)? Do other exonucleases substitute for EXOI? And if so, which ones? There are other aspects of MMR to consider. MMR proteins also correct mispairs that arise during DNA recombination. Indeed, most of the S. cerevisiae strains harboring mutations or deletions in MMR gene loci were first identified as having defects in postmeiotic segregation (hence the term PMS). Interestingly, mice lacking Mlh1, Pms2 (Prolla et al., 1998Prolla T.A. Baker S.M. Harris A.C. Tsao J.L. Yao X. Bronner C.E. Zheng B. Gordon M. Reneker J. Arnheim N. et al.Nat. Genet. 1998; 18: 276-279Crossref PubMed Scopus (289) Google Scholar), Mlh3 (Lipkin et al., 2002Lipkin S.M. Moens P.B. Wang V. Lenzi M. Shanmugarajah D. Gilgeous A. Thomas J. Cheng J. Touchman J.W. Green E.D. et al.Nat. Genet. 2002; 31: 385-390Crossref PubMed Scopus (270) Google Scholar), or ExoI (Wei et al., 2003Wei K. Clark A.B. Wong E. Kane M.F. Mazur D.J. Parris T. Kolas N.K. Russell R. Hou Jr., H. Kneitz B. et al.Genes Dev. 2003; 17: 603-614Crossref PubMed Scopus (243) Google Scholar) also display defects in meiosis. How are these factors involved in the processing of intermediates of meiotic recombination? And what about mitotic events in which MMR has been implicated, such as gene conversion, homologous or class switch recombination, somatic hypermutation, or nonhomologous end-joining (Jiricny, 2006Jiricny J. Nat. Rev. Mol. Cell Biol. 2006; 7: 335-346Crossref PubMed Scopus (841) Google Scholar)? The present findings from the Modrich laboratory move the study of MMR to a new level and will almost certainly help us find answers to these fundamental questions. Endonucleolytic Function of MutLα in Human Mismatch RepairKadyrov et al.CellJuly 28, 2006In BriefHalf of hereditary nonpolyposis colon cancer kindreds harbor mutations that inactivate MutLα (MLH1•PMS2 heterodimer). MutLα is required for mismatch repair, but its function in this process is unclear. We show that human MutLα is a latent endonuclease that is activated in a mismatch-, MutSα-, RFC-, PCNA-, and ATP-dependent manner. Incision of a nicked mismatch-containing DNA heteroduplex by this four-protein system is strongly biased to the nicked strand. A mismatch-containing DNA segment spanned by two strand breaks is removed by the 5′-to-3′ activity of MutSα-activated exonuclease I. Full-Text PDF Open Archive