The Many Interfaces of Mre11
1998; Cell Press; Volume: 95; Issue: 5 Linguagem: Inglês
10.1016/s0092-8674(00)81626-8
ISSN1097-4172
Autores Tópico(s)Genomics and Chromatin Dynamics
ResumoOne of the most memorable parody advertisements from the television program "Saturday Night Live" touted a product that was both a dessert topping and a floor wax. Claims like this pile up for some notable proteins, too. Assertions about p53 would give it a hand in virtually every aspect of DNA and RNA metabolism. One might have similar concerns about the incredible list of jobs assigned to Mre11p, Rad50p, and p95/Xrs2p, were it not for the fact that all of them have been demonstrated in vivo. These proteins are important for two different processes of repairing of double-strand chromosome breaks, homologous recombination and nonhomologous end joining, and they play a key role in telomere maintenance (Figure 1). In addition, this trio is implicated in the cell's checkpoint response to the presence of double-strand breaks (DSBs). But there's more: in meiosis, these three proteins are required not only for the resection of DSBs, but to create the meiotic DSBs in the first place. Deletion of any of these three genes in yeast causes cells to grow slowly, and in mouse cells, the deletion of MRE11 is lethal (27Xiao Y Weaver D.T Nucleic Acids Res. 1997; 25: 2985-2991Crossref PubMed Scopus (248) Google Scholar). A recent flood of papers have established the basis of some of the complicated roles that Mre11, Rad50, and p95/Xrs2 proteins perform. Mre11p acts as an exonuclease and as an endonuclease, but has other, perhaps noncatalytic, attributes as well. In humans and in yeast, Mre11p, Rad50p, and p95 assemble into large complexes. In human cells, 23Trujillo K.M Yuan S.S Lee E.Y Sung P J. Biol. Chem. 1998; 273: 21447-21450Crossref PubMed Scopus (316) Google Scholar find complexes containing only the three proteins, but 5Carney J.P Maser R.S Olivares H Davis E.M Le Beau M Yates III, J.R Hays L Morgan W.F Petrini J.H Cell. 1998; 93: 477-486Abstract Full Text Full Text PDF PubMed Scopus (979) Google Scholar report the association of additional, large proteins. In yeast, Mre11p, Rad50p, and Xrs2p are associated, but in meiotic cells the complex contains two other proteins similar in size to two meiosis-specific proteins required to make DSBs in meiosis (Usui et al., 1998 [this issue of Cell]) (Figure 1). The stoichiometry of these complexes has not been established. That Mre11p should be a nuclease was predicted by 22Sharples G.J Leach D.R Mol. Microbiol. 1995; 17: 1215-1217Crossref PubMed Scopus (188) Google Scholar, who noted that the E. coli nuclease subunits SbcC and SbcD are homologous to Rad50p and Mre11p, respectively. Rad50p is a coiled-coil protein with ATP-dependent DNA binding activity and shares homology with SMC proteins involved in chromosome condensation and sister-chromatid cohesion (21Raymond W.E Kleckner N Nucleic Acids Res. 1993; 21: 3851-3856Crossref PubMed Scopus (98) Google Scholar). Several labs have found that human and yeast Mre11p act as both Mn2+-dependent double-stranded DNA 3′-to-5′ exonuclease and as single-stranded endonuclease (7Furuse M Nagase Y Tsubouchi H Murakami-Murofushi K Shibata T Ohta K EMBO J. 1998; 17: 6412-6425Crossref PubMed Scopus (213) Google Scholar, 19Paull T.T Gellert M Mol. Cell. 1998; 1: 969-980Abstract Full Text Full Text PDF PubMed Scopus (672) Google Scholar, 23Trujillo K.M Yuan S.S Lee E.Y Sung P J. Biol. Chem. 1998; 273: 21447-21450Crossref PubMed Scopus (316) Google Scholar, 25Usui T Ohta T Oshiumi H Tomizawa J Ogawa H Ogawa T Cell. 1998; 95 (this issue,): 705-716Abstract Full Text Full Text PDF PubMed Scopus (297) Google Scholar, 14Moreau S Ferguson H.R Symington L.S Mol. Cell. Biol. 1999; in pressGoogle Scholar). In yeast, Mre11p also has 3′ to 5′ exonuclease activity on single-stranded DNA (25Usui T Ohta T Oshiumi H Tomizawa J Ogawa H Ogawa T Cell. 1998; 95 (this issue,): 705-716Abstract Full Text Full Text PDF PubMed Scopus (297) Google Scholar). The spectrum of activities of the entire complex (23Trujillo K.M Yuan S.S Lee E.Y Sung P J. Biol. Chem. 1998; 273: 21447-21450Crossref PubMed Scopus (316) Google Scholar) and of purified Mre11p (19Paull T.T Gellert M Mol. Cell. 1998; 1: 969-980Abstract Full Text Full Text PDF PubMed Scopus (672) Google Scholar, 25Usui T Ohta T Oshiumi H Tomizawa J Ogawa H Ogawa T Cell. 1998; 95 (this issue,): 705-716Abstract Full Text Full Text PDF PubMed Scopus (297) Google Scholar, 14Moreau S Ferguson H.R Symington L.S Mol. Cell. Biol. 1999; in pressGoogle Scholar) are similiar. In yeast, mutations in amino-terminal residues conserved among Mre11 homologs (Figure 2) eliminate in vitro nuclease activity and/or reduce 5′-to-3′ resection in vivo (7Furuse M Nagase Y Tsubouchi H Murakami-Murofushi K Shibata T Ohta K EMBO J. 1998; 17: 6412-6425Crossref PubMed Scopus (213) Google Scholar, 24Tsubouchi H Ogawa H Mol. Cell. Biol. 1998; 18: 260-268Crossref PubMed Scopus (180) Google Scholar, 25Usui T Ohta T Oshiumi H Tomizawa J Ogawa H Ogawa T Cell. 1998; 95 (this issue,): 705-716Abstract Full Text Full Text PDF PubMed Scopus (297) Google Scholar, 14Moreau S Ferguson H.R Symington L.S Mol. Cell. Biol. 1999; in pressGoogle Scholar). These and other mutations (4Bressan D.A Olivares H.A Nelms B.E Petrini J.H Genetics. 1998; 150: 591-600Crossref PubMed Google Scholar) are defective in DSB repair after ionizing radiation. It is surprising that the Mre11 complex has in vitro 3′-to-5′ exonuclease activity on dsDNA, in view of the effects of deleting the Saccharomyces MRE11, RAD50, or XRS2 genes on the processing of DSBs in vivo. DSBs induced in mitotic cells by the HO endonuclease are almost exclusively resected in a 5′-to-3′ fashion, and this degradation of DSB ends is markedly retarded when MRE11, RAD50, or XRS2 is deleted (10Ivanov E.L Sugawara N White C.I Fabre F Haber J.E Mol. Cell. Biol. 1994; 14: 3414-3425Crossref PubMed Scopus (201) Google Scholar, 12Lee S.-E Moore J.K Holmes A Umezu K Kolodner R Haber J.E Cell. 1998; 94: 399-409Abstract Full Text Full Text PDF PubMed Scopus (625) Google Scholar, 24Tsubouchi H Ogawa H Mol. Cell. Biol. 1998; 18: 260-268Crossref PubMed Scopus (180) Google Scholar). The absence of 3′-to-5′ degradation from DSBs such as those created by HO endonuclease might be rationalized by the finding in vitro that Mre11p does not act on 3′-overhanging ends, such as those produced by HO (25Usui T Ohta T Oshiumi H Tomizawa J Ogawa H Ogawa T Cell. 1998; 95 (this issue,): 705-716Abstract Full Text Full Text PDF PubMed Scopus (297) Google Scholar), but one still must explain why the mutants affect 5′-to-3′ degradation. In meiosis, DSBs are created by Spo11p, a type II topoisomerase that, in mutant backgrounds, remains attached to the 5′ ends of DSBs (reviewed in 8Haber J.E Cell. 1997; 89: 163-166Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar). Rad50p (1Alani E Padmore R Kleckner N Cell. 1990; 61: 419-436Abstract Full Text PDF PubMed Scopus (466) Google Scholar), Mre11p (15Nairz K Klein F Genes Dev. 1997; 11: 2272-2290Crossref PubMed Scopus (211) Google Scholar, 25Usui T Ohta T Oshiumi H Tomizawa J Ogawa H Ogawa T Cell. 1998; 95 (this issue,): 705-716Abstract Full Text Full Text PDF PubMed Scopus (297) Google Scholar), and probably Xrs2p and the Com1/Sae2 protein (reviewed in 8Haber J.E Cell. 1997; 89: 163-166Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar), are apparently required to excise Spo11p (effectively resecting the DNA end in a 5′-to-3′ direction). The removal of Spo11p attached to a 5′ end and the creation of long 3′-ended single strands at DSBs in meiosis again argues that the predominant resection activity in vivo moves 5′ to 3′. This apparent contradiction can be explained if Mre11/Rad50/Xrs2 acts similarly to the E. coli RecBCD enzyme, which acts as a helicase to unwind the DNA end and then as an endonuclease to clip off the 5′-ended strand. As isolated, mammalian Mre11p or Mre11/Rad50/p95 complex does not require ATP for its nuclease activities (19Paull T.T Gellert M Mol. Cell. 1998; 1: 969-980Abstract Full Text Full Text PDF PubMed Scopus (672) Google Scholar, 25Usui T Ohta T Oshiumi H Tomizawa J Ogawa H Ogawa T Cell. 1998; 95 (this issue,): 705-716Abstract Full Text Full Text PDF PubMed Scopus (297) Google Scholar), which might argue that this complex does not have the requisite ATP-dependent helicase activity and may need to associate with an as-yet unknown helicase. This is distinctly different from the bacterial homolog: SbcC/SbcD exonuclease activity requires ATP, although endonuclease activity is ATP independent (6Connelly J.C Kirkham L.A Leach D.R.F Proc. Natl. Acad. Sci. USA. 1998; 95: 7969-7974Crossref PubMed Scopus (187) Google Scholar). The creation of double-strand breaks in yeast meiosis represents a distinct ability of the Mre11 complex. Meiotic DSB formation requires Spo11, Mre11/Rad50/Xrs2, and at least five other proteins, two of which may associate with the Mre11 complex (25Usui T Ohta T Oshiumi H Tomizawa J Ogawa H Ogawa T Cell. 1998; 95 (this issue,): 705-716Abstract Full Text Full Text PDF PubMed Scopus (297) Google Scholar; Figure 1). In the absence of Mre11p, Rad50p, or Xrs2p, no meiotic DSBs are created. However, a separation-of-function mutation, rad50S, an amino acid substitution in the ATP-binding domain, allows DSB formation (with Spo11 attached to the ends) but prevents their recombination (1Alani E Padmore R Kleckner N Cell. 1990; 61: 419-436Abstract Full Text PDF PubMed Scopus (466) Google Scholar). The use of rad50S permitted several labs to define the genetic and chromosomal conditions necessary for the formation of DSBs at meiotic "hot spots," which are characterized as micrococcal nuclease and DNase I hypersensitive sites, nearly always in promoter regions. In some cases there are additional changes in chromatin structure at the time that meiotic DSBs are generated and these are affected by Mre11p, Rad50p, and Xrs2p (17Ohta K Nicolas A Furuse M Nabetani A Ogawa H Shibata T Proc. Natl. Acad. Sci. USA. 1998; 95: 646-651Crossref PubMed Scopus (95) Google Scholar). Recently, a number of Mre11 separation-of-function mutations have been created. 15Nairz K Klein F Genes Dev. 1997; 11: 2272-2290Crossref PubMed Scopus (211) Google Scholar identified an mre11S mutant that, like rad50S, permitted DSB formation but prevented their resection. They also found several opposite separation-of-function mutants defective in making meiotic DSBs but quasi-normal for other functions. The mutations, including mre11-T10 (Figure 2), disrupt the C terminus of Mre11p. However, mre11-T10 can still process breaks. An mre11-T10/mre11S heterozygote can make DSBs (courtesy of mre11S) and process them (presumably by mre11-T10). Whether mre11S/mre11-T10 complement each other within a single complex or act via several distinct Mre11p-containing machines (favored by 25Usui T Ohta T Oshiumi H Tomizawa J Ogawa H Ogawa T Cell. 1998; 95 (this issue,): 705-716Abstract Full Text Full Text PDF PubMed Scopus (297) Google Scholar) is not yet known. Recently, three other labs have made similar mutations in MRE11 (see Figure 2) and used them to learn more about how Mre11p works. Terminal truncation mutations analogous to mre11-T10 (e.g., mre11-ΔC49 or mre11–5) have nuclease activity, but are unable to make DSBs. They lack one of two identified DNA-binding sites within the protein, but coimmunoprecipitate with Rad50p. In contrast, the nuclease-defective, but meiotic DSB-proficient mutant mre11–58, associates only weakly with Rad50p. Because this mutant allows meiotic DSB formation, 25Usui T Ohta T Oshiumi H Tomizawa J Ogawa H Ogawa T Cell. 1998; 95 (this issue,): 705-716Abstract Full Text Full Text PDF PubMed Scopus (297) Google Scholar argue that Mre11–58p exists in a "loose" complex with Rad50p that is competent for meiotic DSB formation, but otherwise is found as a "tight" complex of Mre11p and Rad50p when it acts as a nuclease. However, whether it associates with Mre11–58p or not, Rad50p is essential for meiotic DSB formation and is almost certainly present at the sites where Spo11p acts. The idea that Mre11p and Rad50p might be operating in a less concerted fashion during DSB formation is supported by analyzing chromatin structural changes at the ARG4 hot spot. Both the mre11Δ deletion and the mre11-ΔC49 truncation mutants fail to elicit the increase in DNase I hypersensitivity that normally coincides with DSB formation (7Furuse M Nagase Y Tsubouchi H Murakami-Murofushi K Shibata T Ohta K EMBO J. 1998; 17: 6412-6425Crossref PubMed Scopus (213) Google Scholar). In contrast, rad50Δ or xrs2Δ deletion mutants enhance these chromatin changes despite the absence of breaks. These observations suggest different roles for Mre11p and for Rad50p and Xrs2p in creating meiotic DSBs. Most of what we know about the role of the Mre11 complex in the repair of DSBs comes from work in Saccharomyces. There are two distinct types of end joining: the simple end joining (religation) of complementary DSB ends and nonhomologous end joining (NHEJ), in which the rejoined DNA has small deletions or insertions. Simple religation is efficient, and most cells survive the creation of a DSB with 4 bp overhanging complementary 3′ ends. But when the ends are noncomplementary or when the nuclease that created the DSB is continuously expressed, DSB ends are fused imprecisely by NHEJ. NHEJ is inefficient in yeast: only about one in 1000 cells succeeds. In the absence of Mre11/Rad50/Xrs2, both religation and NHEJ are severely impaired in Saccharomyces, while homologous recombination is only modestly affected. There are actually two pathways of NHEJ (13Moore J.K Haber J.E Mol. Cell. Biol. 1996; 16: 2164-2173Crossref PubMed Scopus (576) Google Scholar). The more efficient mechanism mostly produces small insertions at the site of the DSB, by the misalignment and filling-in of the DNA ends. These events depend on the Mre11 complex and are restricted to cells either in S or G2 phases of the cell cycle. A less efficient deletion pathway is cell cycle independent and largely Mre11/Rad50/Xrs2 independent. Both pathways depend on the DNA end–binding Ku proteins and require ligase 4 and the yeast XRCC4 homolog (Figure 1). Exactly what Mre11p does in DNA end joining is, at the moment, confusing. The absence of Mre11p, Rad50p, or Xrs2p severely impairs end joining in Saccharomyces, but what remains unresolved is whether these proteins act "structurally" (perhaps holding ends together) or nucleolytically. 19Paull T.T Gellert M Mol. Cell. 1998; 1: 969-980Abstract Full Text Full Text PDF PubMed Scopus (672) Google Scholar suggest that the nucleolytic role might be important, based on an in vitro experiment in which they added hMre11p to a mixture of blunt-ended DNA and DNA ligase I and found that the presence of the endo- and exonuclease activities improved ligation, producing joints at microhomologies of 1 to 5 bp. Whether this happens in vivo is still unclear, because, in Saccharomyces, some nuclease-negative mutations of yeast Mre11p are still end joining proficient (25Usui T Ohta T Oshiumi H Tomizawa J Ogawa H Ogawa T Cell. 1998; 95 (this issue,): 705-716Abstract Full Text Full Text PDF PubMed Scopus (297) Google Scholar, 14Moreau S Ferguson H.R Symington L.S Mol. Cell. Biol. 1999; in pressGoogle Scholar). It is possible that Mre11p plays still another role in end joining. 19Paull T.T Gellert M Mol. Cell. 1998; 1: 969-980Abstract Full Text Full Text PDF PubMed Scopus (672) Google Scholar have shown that Mre11p is also an endonuclease that cleaves hairpin DNA ends of the sort that appear following Rag1/Rag2-mediated cleavage of signal sequences in the rearrangement of immunoglobulin genes. Hairpin cutting also is seen with SbcD/SbcD (6Connelly J.C Kirkham L.A Leach D.R.F Proc. Natl. Acad. Sci. USA. 1998; 95: 7969-7974Crossref PubMed Scopus (187) Google Scholar). In this view, the cut hairpins would then undergo NHEJ, again requiring the Mre11 complex. In homologous recombination, the importance of the Mre11 complex depends on the assay that one employs. Initially RAD50, MRE11, and XRS2 were grouped with RAD51, RAD52, and other genes in one X-ray sensitivity group, but it is now clear that there are three distinct subgroups. RAD52 is required for essentially all homologous recombination, whereas RAD51 (the RecA homolog of eukaryotic cells), RAD54, RAD55, and RAD57 form a second group that is surprisingly not required for spontaneous recombination or for some DSB-induced events. Mre11p, Rad50p, and Xrs2p form a third set that is nearly dispensable for several homologous recombination processes requiring both RAD52 and RAD51, but is important for others. Thus, during HO endonuclease–induced switching of the MAT locus, the absence of Mre11p, Rad50p, or Xrs2p significantly delays, but only slightly reduces, the completion of gene conversion, while both Rad51p and Rad52p are essential (10Ivanov E.L Sugawara N White C.I Fabre F Haber J.E Mol. Cell. Biol. 1994; 14: 3414-3425Crossref PubMed Scopus (201) Google Scholar, 24Tsubouchi H Ogawa H Mol. Cell. Biol. 1998; 18: 260-268Crossref PubMed Scopus (180) Google Scholar). rad50S has no mitotic defect in resection of DNA ends (24Tsubouchi H Ogawa H Mol. Cell. Biol. 1998; 18: 260-268Crossref PubMed Scopus (180) Google Scholar). However, rad50 mutants are as sensitive to hydroxyurea as rad51 and rad52 mutants (2Allen J.B Zhou Z Siede W Friedberg E.C Elledge S.J Genes Dev. 1994; 8: 2416-2428Crossref PubMed Scopus (330) Google Scholar). This could reflect the need for Rad50p and its partners to effect sister-chromatid repair or the necessity to cleave hairpin structures that might accumulate in stalled replication forks. Alternatively, the HU-sensitivity of rad50Δ may reflect a role in checkpoint regulation. In some cases the absence of Mre11p, Rad50p, and Xrs2p increases recombination. Spontaneous interchromosomal recombination between two alleles of a nutritional marker increases 10-fold in mre11, rad50 and xrs2 mutants, but decreases 5- to 10-fold in rad51 mutants, and decreases 1000-fold in rad52 mutants (9Ivanov E.L Korolev V.G Fabre F Genetics. 1992; 132: 651-664Crossref PubMed Google Scholar, 20Rattray A.J Symington L.S Genetics. 1995; 139: 45-56Crossref PubMed Google Scholar, 24Tsubouchi H Ogawa H Mol. Cell. Biol. 1998; 18: 260-268Crossref PubMed Scopus (180) Google Scholar). There are several possible explanations for this unexpected phenotype. First, the cell might normally produce DSBs that are easily religated, but if religation were blocked in a rad50 or mre11 mutant, then these breaks would enter a recombination pathway and cells would appear hyperrecombinational. But if this were this the case, yeast cells lacking Ku proteins should also be hyperrecombinational, and this is not observed. Alternatively, DSBs in G2 may normally be repaired using a sister chromatid as a partner, and the Mre11 complex may be important for this type of repair. X-ray survival data support this idea, since rad50 and xrs2 mutants, and presumably mre11, are more defective in sister-chromatid repair of G2 haploid cells than in interhomolog repair in G1 diploids (9Ivanov E.L Korolev V.G Fabre F Genetics. 1992; 132: 651-664Crossref PubMed Google Scholar). Thus, an mre11 or rad50 diploid mutant would engage less often in sister chromatid recombination, thus increasing interhomolog heteroallelic recombination. A third possiblity reflects the reduced 5′-to-3′ resection of DNA ends in the absence of these proteins. Most heteroallelic recombination occurs when strand invasion creates heteroduplex DNA covering one of the two alleles, and a prototrophic recombinant will result after mismatch repair. If heteroduplex DNA covers both alleles, they are usually coordinately mismatch repaired, so that one mutant allele is simply converted to the other mutant. The reduced 5′-to-3′ resection seen in mre11, rad50, or xrs2 mutants would therefore increase the probability that heteroduplex DNA would cover only one allele and thus more prototrophic recombinants would be obtained. These explanations are not mutually exclusive. Deleting any of the Mre11/Rad50/Xrs2 proteins creates a yeast strain with shortened telomeres and, in some strain backgrounds, cells become senescent as they do without telomerase (11Kironmai K.M Muniyappa K Genes Cells. 1997; 2: 443-455Crossref PubMed Scopus (114) Google Scholar, 3Boulton S.J Jackson S.P EMBO J. 1998; 17: 1819-1828Crossref PubMed Scopus (539) Google Scholar, 18Nugent C.I Bosco G Ross L.O Evans S.K Salinger A.P Moore J.K Haber J.E Lundblad V Curr. Biol. 1998; 8: 657-660Abstract Full Text Full Text PDF PubMed Google Scholar). Genetic analysis suggests that the Mre11 complex works alongside of telomerase and its associated factors (hence double mutants among these components do not have a more severe phenotype than telomerase mutants themselves). A possible role for the Mre11 complex that relates to its apparent involvement in sister-chromatid repair would be that these proteins "tell" the telomerase complex that DNA replication has reached the chromosome end and that it is time to add new telomere sequences to both sisters, perhaps simultaneously. Alternatively, the Mre11 complex may remove part of the C-rich strand from the terminus. Normal telomere maintenance requires both the Mre11 complex and the Ku proteins. A double mutant lacking yKu70p and Mre11p is more temperature sensitive than either single mutant, implying that the Ku and Mre11 families act synergistically, and not epistatically as they do in NHEJ. Again, we don't know if the role of the Mre11 complex is structural or nucleolytic, because the nuclease-deficient mre11-H125N mutant has normal telomere length (14Moreau S Ferguson H.R Symington L.S Mol. Cell. Biol. 1999; in pressGoogle Scholar). A rad50S strain has slightly longer-than-normal telomeres (3Boulton S.J Jackson S.P EMBO J. 1998; 17: 1819-1828Crossref PubMed Scopus (539) Google Scholar). This year, the Mre11 complex has attracted a new group of aficionados interested in cancer and in checkpoint regulation. In humans, the p95 component is mutated in Nijmegen breakage syndrome (NBS), a condition that has similarities with ataxia telangiectasia, including ionizing radiation-sensitivity, cancer predisposition, and a failure to arrest at G1/S in response to DNA damage (5Carney J.P Maser R.S Olivares H Davis E.M Le Beau M Yates III, J.R Hays L Morgan W.F Petrini J.H Cell. 1998; 93: 477-486Abstract Full Text Full Text PDF PubMed Scopus (979) Google Scholar, 26Varon R Vissinga C Platzer M Cerosaletti K.M Chrzanowska K.H Saar K Beckmann G Seemanova E Cooper P.R Nowak N.J et al.Cell. 1998; 93: 467-476Abstract Full Text Full Text PDF PubMed Scopus (833) Google Scholar). Possibly, Mre11/Rad50/NBS binds to DSB ends and signals the presence of that damage. An alternative possibility is that the signal of DNA damage is the extent of single-stranded DNA produced by the exonuclease activity of the Mre11 complex. In Saccharomyces, evidence supporting this idea has come not from studying the ability of cells to arrest after DNA damage, but from their capacity to adapt and resume growth when damage persists. Two DSBs are sufficient to discourage a wild-type yeast cell from adapting to checkpoint-induced G2/M arrest, but this permanent arrest is suppressed by an mre11 or rad50 deletion that reduces the extent of single-stranded DNA (12Lee S.-E Moore J.K Holmes A Umezu K Kolodner R Haber J.E Cell. 1998; 94: 399-409Abstract Full Text Full Text PDF PubMed Scopus (625) Google Scholar). There is no doubt that Mre11p is attracted to the sites of DNA damage, for repair and possibly as part of the damage-signaling apparatus. Following irradiation, foci containing both Mre11p and Rad50p have been seen in mammalian cells and in yeast. A stunning series of micrographs illustrating this point were published by 16Nelms B.E Maser R.S MacKay J.F Lagally M.G Petrini J.H Science. 1998; 280: 590-592Crossref PubMed Scopus (410) Google Scholar, who examined mammalian nuclei irradiated with ultrasoft X-rays passed through a grid that produced stripes of DNA damage. The remarkable finding was not that hMre11p was localized within the irradiated regions, but that in repair-defective cells these stripes persisted for hours, suggesting that the damaged DNA was not diffusing around the nucleus. Interestingly, these foci do not attract Rad51p. If DNA ends were being resected, then the 3′-ended single-stranded regions ought to be attractive sites for the assembly of Rad51p filaments that is the first step in repairing DSBs by homologous recombination. However, no such foci containing both hMre11p and hRad51p were seen. A similar lack of colocalization has been seen in yeast cells that cannot complete meiosis and in irradiated mitotic cells (25Usui T Ohta T Oshiumi H Tomizawa J Ogawa H Ogawa T Cell. 1998; 95 (this issue,): 705-716Abstract Full Text Full Text PDF PubMed Scopus (297) Google Scholar). The Mre11 complex is attracted to DSBs and can participate in several different pathways of repair and recombination. It remains "only" to figure out how Mre11p, Rad50p, Xrs2p, and the proteins with which they interact enable the cell to perform these many tasks. Given that almost half the papers on Mre11p have been published in 1998, it is likely we will soon learn much more about the many facets of this multitalented protein.
Referência(s)