Voltar
Artigo Acesso aberto Revisado por pares
BIBTEX RIS

The Three-dimensional Structure of the C-terminal DNA-binding Domain of Human Ku70

2001; Elsevier BV; Volume: 276; Issue: 41 Linguagem: Inglês

10.1074/jbc.m105238200

ISSN

1083-351X

Autores

Ziming Zhang, Lingyang Zhu, Donghai Lin, Fanqing Chen, David J. Chen, Yuan Chen,

Tópico(s)

Enzyme Structure and Function

Resumo

The proteins Ku70 (69.8 kDa) and Ku80 (82.7 kDa) form a heterodimeric complex that is an essential component of the nonhomologous end joining DNA double-strand break repair pathway in mammalian cells. Interaction of Ku with DNA is central for the functions of Ku. Ku70, which is mainly responsible for the DNA binding activity of the Ku heterodimer, contains two DNA-binding domains. We have solved the solution structure of the Ku80-independent DNA-binding domain of Ku70 encompassing residues 536–609 using nuclear magnetic resonance spectroscopy. Residues 536–560 are highly flexible and have a random structure but form specific interactions with DNA. Residues 561–609 of Ku70 form a well defined structure with 3 α-helices and also interact with DNA. The three-dimensional structure indicates that all conserved hydrophobic residues are in the hydrophobic core and therefore may be important for structural integrity. Most of the conserved positively charged residues are likely to be critical for DNA recognition. The C-terminal DNA-binding domain of Ku70 contains a helix-extended strand-helix motif, which occurs in other nucleic acid-binding proteins and may represent a common nucleic acid binding motif. The proteins Ku70 (69.8 kDa) and Ku80 (82.7 kDa) form a heterodimeric complex that is an essential component of the nonhomologous end joining DNA double-strand break repair pathway in mammalian cells. Interaction of Ku with DNA is central for the functions of Ku. Ku70, which is mainly responsible for the DNA binding activity of the Ku heterodimer, contains two DNA-binding domains. We have solved the solution structure of the Ku80-independent DNA-binding domain of Ku70 encompassing residues 536–609 using nuclear magnetic resonance spectroscopy. Residues 536–560 are highly flexible and have a random structure but form specific interactions with DNA. Residues 561–609 of Ku70 form a well defined structure with 3 α-helices and also interact with DNA. The three-dimensional structure indicates that all conserved hydrophobic residues are in the hydrophobic core and therefore may be important for structural integrity. Most of the conserved positively charged residues are likely to be critical for DNA recognition. The C-terminal DNA-binding domain of Ku70 contains a helix-extended strand-helix motif, which occurs in other nucleic acid-binding proteins and may represent a common nucleic acid binding motif. nuclear Overhauser effect root mean square deviation The proteins Ku70 (69.8 kDa) and Ku80 (82.7 kDa) form a heterodimeric complex that is an essential component of the nonhomologous end joining DNA double-strand break repair pathway in mammalian cells (for a recent review, see Ref 1Critchlow S.E. Jackson S.P. Trends Biochem. Sci. 1998; 23: 394-398Abstract Full Text Full Text PDF PubMed Scopus (484) Google Scholar). DNA double-strand breaks are caused by ionizing radiation, V(D)J recombination, and physiological oxygen free-radical damage. The Ku-dependent repair process is the main DNA double-strand break repair mechanism in mammalian cells and is essential for the preservation of genomic stability. Ku is an integral component of the machinery involved in repairing DNA double-strand breaks. The Ku heterodimer binds with high affinity to broken DNA ends (2Blier P.R. Griffith A.J. Craft J. Hardin J.A. J. Biol. Chem. 1993; 268: 7594-7601Abstract Full Text PDF PubMed Google Scholar) and can bridge two proximal DNA ends (3Cary R.B. Peterson S.R. Wang J. Bear D.G. Bradbury E.M. Chen D.J. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 4267-4272Crossref PubMed Scopus (223) Google Scholar). Ku also recruits a 465-kDa DNA-dependent protein kinase catalytic subunit (DNA-PKcs) to the DNA double-strand break sites stimulating the DNA-PKcs kinase activity (4Hammarsten O. Chu G. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 525-530Crossref PubMed Scopus (201) Google Scholar). DNA-PKcs interacts with and phosphorylates a nuclear protein, XRCC4 (5Critchlow S.E. Bowater R.P. Jackson S.P. Curr. Biol. 1997; 7: 588-598Abstract Full Text Full Text PDF PubMed Scopus (398) Google Scholar, 6Leber R. Wise T.W. Mizuta R. Meek K. J. Biol. Chem. 1998; 273: 1794-1801Abstract Full Text Full Text PDF PubMed Scopus (161) Google Scholar). XRCC4 associates tightly with DNA ligase IV through the ligase IV C-terminal extension (5Critchlow S.E. Bowater R.P. Jackson S.P. Curr. Biol. 1997; 7: 588-598Abstract Full Text Full Text PDF PubMed Scopus (398) Google Scholar, 7Koonin E.V. Altschul S.F. Bork P. Nat. Genet. 1996; 13: 266-268Crossref PubMed Scopus (359) Google Scholar, 8Grawunder U. Wilm M. Wu X. Kulesza P. Wilson T.E. Mann M. Lieber M.R. Nature. 1997; 388: 492-495Crossref PubMed Scopus (540) Google Scholar). DNA ligase IV is thought to catalyze DNA-end joining (9Ramsden D.A. Gellert M. EMBO J. 1998; 17: 609-614Crossref PubMed Scopus (246) Google Scholar). Several studies indicate that Ku70 is mainly responsible for the DNA binding activity of the Ku heterodimer. Ku70 contains two DNA-binding domains; one is located at the N terminus before residue 440, and the other one is located at the C terminus (10Chou C.H. Wang J. Knuth M.W. Reeves W.H. J. Exp. Med. 1992; 175: 1677-1684Crossref PubMed Scopus (67) Google Scholar, 11Wu X. Lieber M.R. Mol. Cell. Biol. 1996; 16: 5186-5193Crossref PubMed Scopus (110) Google Scholar). The DNA binding activity of the C-terminal DNA-binding domain is independent of the heterodimer formation with Ku80, whereas the DNA binding activity of the N-terminal domain of Ku70 depends on heterodimer formation with Ku80 (12Wang J. Dong X. Reeves W.H. J. Biol. Chem. 1998; 273: 31068-31074Abstract Full Text Full Text PDF PubMed Scopus (45) Google Scholar). Both Ku70 DNA-binding domains are required for the high affinity binding to DNA as demonstrated in gel shift assays (12Wang J. Dong X. Reeves W.H. J. Biol. Chem. 1998; 273: 31068-31074Abstract Full Text Full Text PDF PubMed Scopus (45) Google Scholar, 13Wang J. Dong X. Myung K. Hendrickson E.A. Reeves W.H. J. Biol. Chem. 1998; 273: 842-848Abstract Full Text Full Text PDF PubMed Scopus (68) Google Scholar). Elimination of either DNA-binding domain resulted in a significant reduction of the DNA binding affinity. This may because of primarily entropic effects. Biochemical studies performed by Reeves and co-workers have shown that the segment spanning residues 536–609 constitutes the C-terminal DNA-binding domain of human Ku70 (10Chou C.H. Wang J. Knuth M.W. Reeves W.H. J. Exp. Med. 1992; 175: 1677-1684Crossref PubMed Scopus (67) Google Scholar, 12Wang J. Dong X. Reeves W.H. J. Biol. Chem. 1998; 273: 31068-31074Abstract Full Text Full Text PDF PubMed Scopus (45) Google Scholar, 13Wang J. Dong X. Myung K. Hendrickson E.A. Reeves W.H. J. Biol. Chem. 1998; 273: 842-848Abstract Full Text Full Text PDF PubMed Scopus (68) Google Scholar). Deletion of the first 25 residues of this domain (residues 536–560) abolished the DNA binding activity of the C-terminal domain (10Chou C.H. Wang J. Knuth M.W. Reeves W.H. J. Exp. Med. 1992; 175: 1677-1684Crossref PubMed Scopus (67) Google Scholar). Deletion of the last 10 amino acid residues (residues 600–609) of Ku70 also significantly reduced the DNA binding activity (13Wang J. Dong X. Myung K. Hendrickson E.A. Reeves W.H. J. Biol. Chem. 1998; 273: 842-848Abstract Full Text Full Text PDF PubMed Scopus (68) Google Scholar). In the study described in this paper, we have expressed the segment encompassing residues 536–609 of human Ku70 and solved the three-dimensional structure of this DNA-binding domain of Ku70. The three-dimensional structure reveals a common helix-extended loop-helix structural motif. Human Ku70 C-terminal fragment 536–609 was cloned into expression vector pET28a (Novagen) and transfected intoEscherichia coli strain BL21 (DE3). The resulting protein product is a recombinant fusion protein starting with a 6-histidine tag and a thrombin protease site. Uniformly 15N-labeled protein and 15N- and 13C-labeled protein were produced by growing the expression strain in M9 medium with15NH4Cl as the sole nitrogen source and with15NH4Cl and [13C]glucose as the sole nitrogen and carbon source, respectively. The protein was purified using Ni2+ affinity chromatography. The NMR samples contained 1.0 mm protein in 100 mm phosphate buffer at pH 6.0 and 5 mm dithiothreitol in 95% H2O/5% 2H2O or 100%2H2O. All NMR experiments were performed at 20 °C on a Varian Unity-plus 500 NMR spectrometer equipped with a triple-resonance probe, pulsed-field gradient, and pulse-shaping capabilities. Felix 98 (Molecular Simulations Inc., San Diego, CA) was used for all NMR data processing.1H, 15N, and 13C resonance assignments were obtained from the following experiments: HNCACB, C(CO)NH, CBCA(CO)NH, HCCH-TOCSY, 15N-edited TOCSY-HSQC, H(CCO)NH, and HBHA(CO)NH (for reviews, see Refs. 14Bax A. Vuister G.W. Grzesiek S. Delaglio F. Wang A.C. Tschudin R. Zhu G. Methods Enzymol. 1994; 239: 79-105Crossref PubMed Scopus (381) Google Scholar and 15Clore G.M. Gronenborn A.M. Methods Enzymol. 1994; 239: 349-362Crossref PubMed Scopus (253) Google Scholar). Stereospecific assignments of β-methylene protons were obtained by the analysis of 3JHα-Hβ and3JHN-Hβ coupling constants extracted from the semi-quantitative interpretation of relative peak intensities in the three-dimensional 15N-separated TOCSY-HSQC of 30-ms mixing time (16Clore G.M. Bax A. Gronenborn A.M. J. Biomol. NMR. 1991; 1: 13-22Crossref PubMed Scopus (102) Google Scholar) and the three-dimensional HNHB (17Archer S.J. Ikura M. Torchia D. J. Magn. Reson. 1991; 95: 636-641Google Scholar) spectra, respectively. The dihedral angle φ restraints were obtained based on3JHN-Hα coupling constants measured in an HNHA experiment (18Kuboniwa H. Grzesiek S. Delaglio F. Bax A. J. Biomol. NMR. 1994; 4: 871-878Crossref PubMed Scopus (336) Google Scholar). Distance restraints were obtained from15N-separated (mixing time, 100 ms) and13C-separated (mixing time, 100 ms) three-dimensional NOESY spectra. The distance restraints were grouped into the following ranges: 3.0, 4.0 (3.7 if HN, Hα, or Hβ is involved), and 5.0 Å. Structures were obtained from simulated annealing calculations using Dyana (19Guntert P. Mumenthaler C. Wuthrich K. J. Mol. Biol. 1997; 273: 283-298Crossref PubMed Scopus (2553) Google Scholar) and CNS (36Brunger A.T. Adams P.O. Clore G.M. Delano W.L. Gros P. Grosse-Kunsleve R.W. Jiang J.-S. Kuszewski J. Nilges M. Pannu N.S. Read R.F. Rice L.M. Simonson T. Warren G.L. . 1998; D54: 905-921Google Scholar). Quality of the NMR structures was evaluated using the program PROCHECK (20Raaijmakers H. Vix O. Toro I. Golz S. Kemper B. Suck D. EMBO J. 1999; 18: 1447-1458Crossref PubMed Scopus (111) Google Scholar). A sample of 0.5 mm15N-labeled Ku70 domain was titrated to a sample of 10 mm unlabeled DNA to a molar ratio of 1:1. The titration was monitored using one-dimensional and two-dimensional HSQC NMR spectra. The final concentration of the protein-DNA complex is ∼0.35 mm. Because Ku binds to the double-strand break sites of DNA, a palindromic DNA sequence was used to ensure that the binding of Ku70 to each end of the DNA forms the same complex. The DNA molecule is a duplex of 16 base pairs with the upper strand sequence being 5′-GCTATGGATCCATAGC-3′. The structure of the Ku70 C-terminal domain was solved with 1144 structural constraints derived from NMR measurements. An ensemble of 20 structures of Ku70 C-terminal domain with the lowest NMR constraint violations and lowest XPLOR energies were used for detailed analysis of the structure. The structural statistics are given in Table I. All experimental NMR constraints are well satisfied. There are no NOE1 constraints violated more than 0.5 Å, no J-coupling constraints violated more than 1 Hz, and no dihedral constraints violated more than 5°. These structures also display only small deviations from idealized covalent geometry (Table I). All peptide bonds are trans. Almost all residues in the structurally well defined regions have Φ and Ψ dihedral angles in the most favorable regions of the Ramachandran plot. The residues in the generally allowed regions and disallowed regions of the Ramachandran plot are located mainly in the loop between residues 588–595, where the structure is less well defined.Table IStructural statistics of the Ku70 C-terminal DomainTotal restraints used1144Total NOE restraints1087 Intraresidue385 Sequential (‖I − j‖ = 1)230 Medium (1 < ‖I − j‖ ≤ 4)250 Long range (‖I − j‖ > 4)222Dihedral φ Angles44 χ1 Angles13r.m.s.d. from experimental distance restrains Distance restraints (Å)0.0088 Dihedral angles (degrees)0.2476r.m.s.d. from idealized covalent geometry Bonds (Å)0.0014 Bond angles (degrees)0.3805 Improper torsions (degrees)0.1646r.m.s.d. from average structure (Å) Backbone atoms in helical regions1-aHelical regions include residues 562–570, 578–587, and 596–606.0.36 All backbone atoms (residues 561–609)0.69 All heavy atoms in helical regions1-aHelical regions include residues 562–570, 578–587, and 596–606.1.08 All heavy atoms (residues 561–609)1.38Procheck analysis Residues in most favorable region (%)73.0 Residues in additional allowed regions (%)19.1 Residues in generously allowed regions (%)6.8 Residues in disallowed regions (%)1.11-a Helical regions include residues 562–570, 578–587, and 596–606. Open table in a new tab The NMR structure is of high precision. The average root mean square deviation (r.m.s.d.) from the average structure for backbone heavy atoms (C′, Cα, N) of residues in α-helices is 0.36 Å. The average r.m.s.d. from the average structure for backbone heavy atoms of all residues from 561 to 609 is 0.58 Å. The overall high precision of the NMR structure is clearly correlated to the high density of experimental constraints. In the structured region, more than 26 NMR constraints per residue have been identified for structural calculation. The C-terminal DNA-binding domain of Ku70 consists of a well structured region and a highly random and flexible N terminus. Residues 536–560, which include the nuclear localization domain (21Koike M. Ikuta T. Miyasaka T. Shiomi T. Exp. Cell. Res. 1999; 250: 401-413Crossref PubMed Scopus (49) Google Scholar), are completely random and flexible and do not form any defined structure in solution. This is evident by much narrower resonances, absence of any medium and long range NOEs, and chemical shifts that are similar to random coil values. Significant information is now available to correlate primary sequences of proteins to their secondary structures, and secondary structure prediction from amino acid sequence is successful in most cases (22Rost B. Methods Enzymol. 1998; 266: 525-539Crossref Google Scholar). Secondary structural prediction of Ku70 using the program PHD (22Rost B. Methods Enzymol. 1998; 266: 525-539Crossref Google Scholar) shows that residues 530–560 of Ku70 do not contain helices or β-strands but form a loop. This is consistent with the observation that this region is unstructured in solution. These results suggest that the nuclear localization domain of human Ku70, like the nuclear localization domains of many proteins, is unstructured in solution. Residues 561–609 of Ku70 form a well defined structure. This region contains 3 α-helices. Superposition of the backbone Cα atoms of residues in this region is shown in Fig.1A, and the ribbon diagram of a representative structure is shown in Fig. 1B. On the basis of backbone dihedral angles and characteristic NOEs, the three α-helices encompass residues 562–570, 578–587, and 596–606. Helices Hb and Hc are nearly parallel and connected by an extended loop region. This interhelical loop (residues 588–595) is less well defined than the helical regions and appears to be more flexible. This region has fewer NOE constraints per residue than average. Helices Hb and Hc and the loop connecting them form a unique helix-extended loop-helix motif. The three-dimensional structure suggests that conserved hydrophobic residues are all located in the hydrophobic core of the protein. Fig.2A shows the sequence alignment of the Ku70 C-terminal domain from known species. The following hydrophobic residues are highly conserved throughout these species: 564, 568, 573, 576, 578, 581, 590, 599, 600, 603, and 607. All of these residues are in the hydrophobic core of the Ku70 C-terminal domain structure and are indicated in green in Fig. 2A and shown with their sidechains in Fig. 2B. The extensive hydrophobic core of the Ku70 C-terminal domain is formed with nearly all conserved hydrophobic residues. The structure of the C-terminal DNA-binding domain is consistent with previous studies (10Chou C.H. Wang J. Knuth M.W. Reeves W.H. J. Exp. Med. 1992; 175: 1677-1684Crossref PubMed Scopus (67) Google Scholar, 23Reeves W.H. Pierani A. Chou C. Ng T. Nicastri C. Roeder R.G. Sthoeger Z.M. J. Immunol. 1991; 146: 2678-2686PubMed Google Scholar), which suggested that the fragment from residues 560–609 forms an intact structure based on reactions with anti-Ku70 antibodies. Antibodies against full-length Ku70 and Ku80 proteins were used to react with E. coli-expressed recombinant Ku70 and Ku80 fragments. The fragment consisting of residues 560–609 of human Ku70 was the minimum length for a reaction with these antibodies (23Reeves W.H. Pierani A. Chou C. Ng T. Nicastri C. Roeder R.G. Sthoeger Z.M. J. Immunol. 1991; 146: 2678-2686PubMed Google Scholar). Fragments that contained this region (e.g. 419–609, 536–609, and 560–609) reacted well with the antibodies. Fragments that do not entirely include this region, such as residues 560–600, 536–600, and 589–609, did not react with the antibodies. Because the fragment containing residues 560–609 can be recognized by antibodies generated by the full-length Ku proteins, this would suggest that the structure of the E. coli-expressed fragment encompassing residues 560–609 has the same structural characteristic as the fragment in the full-length protein. Interaction of the Ku70 C-terminal domain with DNA has been investigated. 15N-Labeled Ku70 C-terminal domain (0.5 mm) was titrated to the solution of 10 mmunlabeled 16-base pair palindromic DNA with the sequence described under “Experimental Procedures.” Chemical shifts of the DNA in the complex were monitored by one-dimensional jump return (24Hore P.J. J. Magn. Reson. 1983; 55: 283-300Google Scholar) and one-dimensional 15N-filtered spectra (25Ikura M. Bax A. J. Am. Chem. Soc. 1992; 114: 2433-2440Crossref Scopus (284) Google Scholar). Chemical shift changes of the protein were monitored using1H–15N HSQC spectra. Fig.3 shows the superposition of the1H–15N HSQC spectra of the C-terminal domain of Ku70 free and in complex with DNA. Overall, chemical shift perturbation has been observed for many residues in Ku70. In addition, the HSQC spectra were essentially identical at the protein:DNA molar ratio of 1:2 and 1:1. This suggests that the percentages of the Ku70 domain in the complex were the same at the protein:DNA molar ratio of 1:2 and 1:1 and therefore were close to 100%. Therefore, the dissociation constant should be in the micromolar range or smaller. The DNA signals as monitored by one-dimensional experiments during the titration also suggest that the exchange rate between the free and complex states is slow on the chemical shift time scale. Chemical shift perturbation was used to identify regions of the protein that are involved in the protein-DNA interaction, an approach used successfully in our, as well as other, laboratories (26Chen Y. Reizer J. Saier Jr., M.H. Fairbrother W.J. Wright P.E. Biochemistry. 1993; 32: 32-37Crossref PubMed Scopus (135) Google Scholar). Large chemical shift changes (Δδ1H > 0.08 ppm and/or Δδ15N > 0.2 ppm) have been observed for residues 540–544, 554–562, 573–577, 581–588, and 590–604. The first two segments are located in the N-terminal flexible region. The next three segments are in the structured region. These regions are indicated in yellow in the 3-dimensional structure of the Ku70 domain shown in Fig.4A. The residues in the His tag did not form any defined structures as no non-sequential NOE has been observed involving the residues in the His tag. Furthermore, the His tag did not form any interactions with the DNA either, because no significant chemical shift change has been observed for resonances of the His tag residues. The largest chemical shift changes occur in regions where most of the conserved Arg or Lys residues are located and have the strongest positive electrostatic potential. Several residues within the C-terminal DNA-binding domain of Ku70 are conserved as Arg or Lys as indicated in blue in Fig. 2A. These residues are 539, 542–544, 553, 554, 556, 565, 575, 582, 586, 591, 595, and 596. Except residue 565, all other conserved positively charged residues are located on the surface, which has a strong positive electrostatic potential (Fig. 4, A and B). In particular, residues 582, 586, 591, 595, and 596 are clustered together forming the strongest positive surface of the structured domain. Residue 582 and 586 are located in α-helix 72 b, and residues 591, 595, and 596 are located in an adjacent flexible loop. Most of these residues are located in the segments where large chemical shift perturbation has been observed (Fig. 4A). The positively charged amino acid residues often play key roles in the binding of nucleic acids. Conservation of these residues suggests their importance in DNA binding. Smaller chemical shift changes in other segments in the structured region may be because of additional structural perturbation as a result of the protein-DNA interaction. The surface shape of the proposed DNA-binding site in the structured region (Hb and the loop containing residues 590–596) suggests that this surface may interact with the major groove of DNA and the phosphate backbone. The size of an α-helix matches that of the major groove of DNA, and α-helices often insert into the major groove of DNA in many DNA-binding proteins. Thus it is possible that the helix, where the conserved residues 582 and 586 are located, may interact with the major groove of DNA. Then the adjacent loop, where the conserved residues 591, 595, and 596 are located, may interact with the phosphate backbone. This is consistent with a previous study using a photocross-linking method that shows that Ku70 binds in the major groove near double-helical ends (27Yoo S. Kimzey A. Dynan W.S. J. Biol. Chem. 1999; 274: 20034-20039Abstract Full Text Full Text PDF PubMed Scopus (75) Google Scholar). Further biochemical studies are needed to define the binding site of the flexible N terminus of the Ku70 C-terminal domain on the DNA. The structure of the C-terminal region of Ku70 does not resemble the helix-turn-helix DNA binding motif or other common sequence-specific DNA-binding domains, such as zinc finger, leucine zipper, or helix-loop-helix motif. It contains a helix-extended loop-helix structure that has not been observed in known sequence-specific DNA-binding domains. We have used the program DALI (28Holmes L. Sander C. J. Mol. Biol. 1993; 233: 123-138Crossref PubMed Scopus (3561) Google Scholar) to search for structures in the Protein Data Bank that share structural similarities to the C-terminal DNA-binding domain of Ku70. The search resulted in the following three hits: T4 endonuclease VII (29Laskowski R.A. Rullmann J.A.C. MacArthur M.W. Kaptein R. Thornton J. J. Biomol. NMR. 1996; 8: 477-486Crossref PubMed Scopus (4401) Google Scholar), transcription termination factor rho (30Allison T.J. Wood T.C. Briercheck D.M. Rastinejad F. Richardson J.P. Rule G.S. Nat. Struct. Biol. 1998; 5: 352-356Crossref PubMed Scopus (67) Google Scholar, 31Briercheck D.M. Wood T.C. Allison T.J. Richardson J.P. Rule G.S. Nat. Struct. Biol. 1998; 5: 393-398Crossref PubMed Scopus (60) Google Scholar), and lysyl-tRNA synthetase (32Onesti S. Miller A.D. Brick P. Structure. 1995; 3: 163-176Abstract Full Text Full Text PDF PubMed Scopus (128) Google Scholar). The superpositions of the C-terminal domain of Ku70 with these proteins are shown in Fig.5, A–C. T4 endonuclease VII binds to and cleaves unusual DNA-structures such as Holiday structures, three-way junctions, single-strand overhangs, heteroduplex loops, base mismatches, bulky adducts, and curved DNA (29Laskowski R.A. Rullmann J.A.C. MacArthur M.W. Kaptein R. Thornton J. J. Biomol. NMR. 1996; 8: 477-486Crossref PubMed Scopus (4401) Google Scholar). The region of T4 endonuclease VII that shares structural homology to the Ku70 C-terminal domain is located at the C terminus, which is critical for interactions with DNA (33Golz S. Christoph A. Birkenkamp-Detroder K. Kemper B. Eur. J. Biochem. 1997; 245: 573-580Crossref PubMed Scopus (26) Google Scholar, 34Birkenbihl R.P. Kemper B. EMBO J. 1998; 17: 4527-4534Crossref PubMed Scopus (17) Google Scholar). rho is a transcriptional terminator in most eubacterial species (30Allison T.J. Wood T.C. Briercheck D.M. Rastinejad F. Richardson J.P. Rule G.S. Nat. Struct. Biol. 1998; 5: 352-356Crossref PubMed Scopus (67) Google Scholar, 31Briercheck D.M. Wood T.C. Allison T.J. Richardson J.P. Rule G.S. Nat. Struct. Biol. 1998; 5: 393-398Crossref PubMed Scopus (60) Google Scholar). rho functions as a hexamer and binds either single-stranded DNA or RNA. The α-helical domain within the RNA-binding domain shares structural similarity with the C-terminal domain of Ku70. It is not clear whether the N-terminal domain of rho is involved in RNA binding. Lysyl-tRNA synthetase catalyzes the covalent conjugation of lysine to the cognate tRNA (32Onesti S. Miller A.D. Brick P. Structure. 1995; 3: 163-176Abstract Full Text Full Text PDF PubMed Scopus (128) Google Scholar). The complex of an Asp-tRNA synthetase, homologous to the lysyl-tRNA synthetase, in complex with tRNA has been solved (35Cavarelli J. Rees B. Ruff M. Thierry J.C. Moras D. Nature. 1993; 362: 181-184Crossref PubMed Scopus (273) Google Scholar). Homology modeling shows that the region in lysyl-tRNA synthetase that shares structural similarity to the Ku70 C-terminal domain is located at one end of double-helical arm of tRNA and may form interactions with tRNA. Sequence similarity among the four proteins has been observed. Fig.5D shows the sequence alignment comparing the Ku70 C-terminal domain with T4 endonuclease VII, rho, and lysyl-tRNA synthetase. The sequence alignment is based on the structural superposition. Most of the conserved hydrophobic residues in Ku70, which are located in the hydrophobic core, are also conserved in the structurally homologous domains of the three other proteins. In particular, the sequence similarity is higher in the helix-extended loop-helix (the second and third helices) region. This region has a high structural similarity among the four proteins. This would suggest that the conserved hydrophobic core is responsible for stabilizing these similar structures. The structural motif represented by the C-terminal DNA-binding domain of Ku70 occurs in at least four proteins including itself as discussed here. These proteins are involved in interactions with unusual structures of nucleic acids. The regions that share structural similarity with the Ku70 C-terminal domain in these three proteins are known to interact with nucleic acid or may form interactions with nucleic acids. Thus this motif may be involved in supporting or enhancing binding to nucleic acid by proteins that recognize unusual nucleic acid structures. In conclusion, this is the first study describing the atomic resolution three-dimensional structure of the Ku proteins. Interaction of the Ku with DNA is central for the functions of Ku. The three-dimensional structure of the C-terminal DNA-binding domain of Ku70 indicates that all conserved hydrophobic residues are in the hydrophobic core and are therefore likely to be important for structural integrity. Most of the conserved positively charged residues are likely to be critical for DNA recognition. The C-terminal domain of human Ku70 may represents a common structural motif that is involved in the recognition of unusual nucleic acid structures.

Referência(s)