Artigo Acesso aberto Revisado por pares

Crystal Structure of a Mini-intein Reveals a Conserved Catalytic Module Involved in Side Chain Cyclization of Asparagine during Protein Splicing

2003; Elsevier BV; Volume: 278; Issue: 40 Linguagem: Inglês

10.1074/jbc.m306197200

ISSN

1083-351X

Autores

Yi Ding, Ming‐Qun Xu, Inca Ghosh, Xuehui Chen, Sébastien Ferrandon, Guillaume Lesage, Zihe Rao,

Tópico(s)

Enzyme Structure and Function

Resumo

We have determined the crystal structure of a 154-residue intein derived from the dnaB gene of Synechocystis sp. strain PCC6803 and refined it to a 2.0-Å resolution. The x-ray structure suggests that this intein possesses two catalytic sites that appear to be separately responsible for splicing and cleavage of the N- and C-terminal scissile bonds. The conserved intein block F residues are the important components of a catalytic site for side chain cyclization of the last intein residue, Asn-154. The data suggest that the imidazole ring of His-143 is involved in the activation of the side chain Nδ atom of Asn-154, leading to a nucleophilic attack on the carbonyl carbon of Asn-154. Substitution of His-143 with Ala or Gln resulted in the inhibition of C-terminal cleavage. His-153, Asp-136, and a water molecule appear to constitute an oxyanion binding site by contacting the carbonyl oxygen of Asn-154 to stabilize the transition state. The structure and mutagenesis data also support that the close contact between the hydroxyl groups of Thr-138 and Ser-155, whose side chain participates in an S → O acyl shift, plays an important role in the nucleophile orientation. Our structural modeling suggests that this catalytic module is conserved in the C-terminal subdomains of inteins from diverse organisms. We have determined the crystal structure of a 154-residue intein derived from the dnaB gene of Synechocystis sp. strain PCC6803 and refined it to a 2.0-Å resolution. The x-ray structure suggests that this intein possesses two catalytic sites that appear to be separately responsible for splicing and cleavage of the N- and C-terminal scissile bonds. The conserved intein block F residues are the important components of a catalytic site for side chain cyclization of the last intein residue, Asn-154. The data suggest that the imidazole ring of His-143 is involved in the activation of the side chain Nδ atom of Asn-154, leading to a nucleophilic attack on the carbonyl carbon of Asn-154. Substitution of His-143 with Ala or Gln resulted in the inhibition of C-terminal cleavage. His-153, Asp-136, and a water molecule appear to constitute an oxyanion binding site by contacting the carbonyl oxygen of Asn-154 to stabilize the transition state. The structure and mutagenesis data also support that the close contact between the hydroxyl groups of Thr-138 and Ser-155, whose side chain participates in an S → O acyl shift, plays an important role in the nucleophile orientation. Our structural modeling suggests that this catalytic module is conserved in the C-terminal subdomains of inteins from diverse organisms. Protein splicing is a posttranslational processing event that involves the precise removal of an intervening sequence, an intein, from a protein precursor with concomitant ligation of the flanking protein sequences (N and C exteins) via a native peptide bond (1Kane P.M. Yamashiro C.T. Wolczyk D.F. Neff N. Goebl M. Stevens T.H. Science. 1990; 250: 651-657Crossref PubMed Scopus (385) Google Scholar, 2Hirata R. Ohsumk Y. Nakano A. Kawasaki H. Suzuki K. Anraku Y. J. Biol. Chem. 1990; 265: 6726-6733Abstract Full Text PDF PubMed Google Scholar, 3Perler F.B. Davis E.O. Dean G.E. Gimble F.S. Jack W.E. Neff N. Noren C.J. Thorner J. Belfort M. Nucleic Acids Res. 1994; 22: 1125-1127Crossref PubMed Scopus (316) Google Scholar). In vitro splicing of inteins in heterologous proteins suggests that inteins with the first C-extein residue contain sufficient information needed for the autocatalytic splicing process without the requirement of exogenous energy or protein co-factor (4Paulus H. Annu. Rev. Biochem. 2000; 69: 447-496Crossref PubMed Scopus (380) Google Scholar, 5Xu M.Q. Southworth M.W. Mersha F.B. Hornstra L.J. Perler F.B. Cell. 1993; 75: 1371-1377Abstract Full Text PDF PubMed Scopus (184) Google Scholar). Among the more than 130 inteins identified so far, the majority contain two discrete functional domains, the homing endonuclease domain and the splicing domain (Fig. 1A) (InBase, the New England Biolabs Intein Data Base (6Perler F.B. Nucleic Acids Res. 2000; 28: 344-345Crossref PubMed Scopus (70) Google Scholar)). The protein splicing function of the intein does not depend on their homing endonuclease activity since splicing-proficient minimal inteins (mini-inteins) occur naturally (7Chong S. Xu M.Q. J. Biol. Chem. 1997; 272: 15587-15590Abstract Full Text Full Text PDF PubMed Scopus (95) Google Scholar, 8Derbyshire V. Wood D.W. Wu W. Dansereau J.T. Dalgaard J.Z. Belfort M. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 11466-11471Crossref PubMed Scopus (111) Google Scholar, 9Shingledecker K. Jiang S.Q. Paulus H. Gene (Amst.). 1998; 207: 187-195Crossref PubMed Scopus (72) Google Scholar) and have been generated by deleting the centrally located endonuclease domain. The splicing domain of a dual function intein appears to be split by the endonuclease domain into two segments, the N- and C-terminal subdomains. The N-terminal subdomain (in the range of 100-150 residues) contains the conserved intein blocks A, B, N2, and N4, whereas the C-terminal subdomain of ∼35-50 residues contains blocks F and G (6Perler F.B. Nucleic Acids Res. 2000; 28: 344-345Crossref PubMed Scopus (70) Google Scholar, 10Pietrokovski S. Protein Sci. 1998; 7: 64-71Crossref PubMed Scopus (131) Google Scholar). The N-terminal subdomain of inteins not only exhibits sequence homology with the hedgehog proteins of eucaryotes but also shares an acyl rearrangement mechanism of breaking the peptide bond preceding a nucleophile-containing residue (11Hall T.M. Porter J.A. Young K.E. Koonin E.V. Beachy P.A. Leahy D.J. Cell. 1997; 91: 85-97Abstract Full Text Full Text PDF PubMed Scopus (229) Google Scholar). This raises the question of whether inteins share a common ancestor with other autoprocessing proteins and have evolved by acquiring a new catalytic capacity imbedded in their C-terminal subdomain. Biochemical and mutational studies reveal that protein self-splicing consists of a series of concerted nucleophilic replacements (4Paulus H. Annu. Rev. Biochem. 2000; 69: 447-496Crossref PubMed Scopus (380) Google Scholar, 12Xu M.Q. Perler F.B. EMBO J. 1996; 15: 5146-5153Crossref PubMed Scopus (253) Google Scholar, 13Xu M.Q. Comb D.G. Paulus H. Noren C.J. Shao Y. Perler F.B. EMBO J. 1994; 13: 5517-5522Crossref PubMed Scopus (106) Google Scholar) (Fig. 1B). The splicing reaction is dependent on a set of highly conserved residues at both splice junctions (12Xu M.Q. Perler F.B. EMBO J. 1996; 15: 5146-5153Crossref PubMed Scopus (253) Google Scholar, 14Hirata R. Anraku Y. Biochem. Biophys. Res. Commun. 1992; 188: 40-47Crossref PubMed Scopus (56) Google Scholar, 15Cooper A.A. Chen Y.J. Lindorfer M.A. Stevens T.H. EMBO J. 1993; 12: 2575-2583Crossref PubMed Scopus (111) Google Scholar, 16Davis E.O. Jenner P.J. Brooks P.C. Colston M.J. Sedgwick S.G. Cell. 1992; 71: 201-210Abstract Full Text PDF PubMed Scopus (140) Google Scholar). The vast majority of inteins possess an asparagine residue at their C termini and a hydroxyl- or thiol-containing residue after each splice junction. Splicing in these inteins is initiated by an acyl rearrangement involving the side chain of serine or cysteine at the N-terminal splice junction, resulting in a (thio)ester bond (12Xu M.Q. Perler F.B. EMBO J. 1996; 15: 5146-5153Crossref PubMed Scopus (253) Google Scholar, 17Shao Y. Xu M.Q. Paulus H. Biochemistry. 1996; 35: 3810-3815Crossref PubMed Scopus (70) Google Scholar). Next, a branched intermediate is formed as a result of transesterification by the sulfhydryl- or hydroxyl-containing side chain of cysteine, serine, or threonine at the C-terminal splice junction (5Xu M.Q. Southworth M.W. Mersha F.B. Hornstra L.J. Perler F.B. Cell. 1993; 75: 1371-1377Abstract Full Text PDF PubMed Scopus (184) Google Scholar, 13Xu M.Q. Comb D.G. Paulus H. Noren C.J. Shao Y. Perler F.B. EMBO J. 1994; 13: 5517-5522Crossref PubMed Scopus (106) Google Scholar). The last catalytic step involves the excision of the intein by side chain cyclization of the C-terminal asparagines (13Xu M.Q. Comb D.G. Paulus H. Noren C.J. Shao Y. Perler F.B. EMBO J. 1994; 13: 5517-5522Crossref PubMed Scopus (106) Google Scholar, 18Shao Y. Xu M.Q. Paulus H. Biochemistry. 1995; 34: 10844-10850Crossref PubMed Scopus (63) Google Scholar). This yields a transient (thio)ester intermediate that can undergo a spontaneous acyl rearrangement to form a native amide bond between the two exteins. It has been observed that the splicing reaction could be suppressed by mutations, often leading to cleavage at N-terminal and/or C-terminal splice junctions (12Xu M.Q. Perler F.B. EMBO J. 1996; 15: 5146-5153Crossref PubMed Scopus (253) Google Scholar, 15Cooper A.A. Chen Y.J. Lindorfer M.A. Stevens T.H. EMBO J. 1993; 12: 2575-2583Crossref PubMed Scopus (111) Google Scholar, 19Mathys S. Evans T.C. Chute I.C. Wu H. Chong S. Benner J. Liu X.Q. Xu M.Q. Gene (Amst.). 1999; 231: 1-13Crossref PubMed Scopus (187) Google Scholar, 20Chong S. Shao Y. Paulus H. Benner J. Perler F.B. Xu M.Q. J. Biol. Chem. 1996; 271: 22159-22168Abstract Full Text Full Text PDF PubMed Scopus (213) Google Scholar). Therefore, it is of interest to understand how the spatial arrangement of an intein active site could account for independent scission of the peptide bonds flanking an intein. The x-ray structures of the intein of the vacuolar ATPase subunit from Saccharomyces cerevisiae (Sce VMA intein) (21Duan X. Gimble F.S. Quiocho F.A. Cell. 1997; 89: 555-564Abstract Full Text Full Text PDF PubMed Scopus (234) Google Scholar, 22Poland B.W. Xu M.Q. Quiocho F.A. J. Biol. Chem. 2000; 275: 16408-16413Abstract Full Text Full Text PDF PubMed Scopus (125) Google Scholar, 23Mizutani R. Nogami S. Kawasaki M. Ohya Y. Anraku Y. Satow Y. J. Mol. Biol. 2002; 316: 919-929Crossref PubMed Scopus (65) Google Scholar, 24Hu D. Crist M. Duan X. Quiocho F.A. Gimble F.S. J. Biol. Chem. 2000; 275: 2705-2712Abstract Full Text Full Text PDF PubMed Scopus (30) Google Scholar), the archaeal PI-PfuI intein from Pyrococcus furiosus (PIPfuI intein) (25Ichiyanagi K. Ishino Y. Ariyoshi M. Komori K. Morikawa K. J. Mol. Biol. 2000; 300: 889-901Crossref PubMed Scopus (97) Google Scholar), and the mini-intein of the GyrA protein from Mycobacterium xenopi (Mxe GyrA intein) (26Klabunde T. Sharma S. Telenti A. Jacobs Jr., W.R. Sacchettini J.C. Nat. Struct. Biol. 1998; 5: 31-36Crossref PubMed Scopus (203) Google Scholar) have revealed that inteins contain a horseshoe-like β-strand scaffold termed the Hint (Hedgehog, intein) module (11Hall T.M. Porter J.A. Young K.E. Koonin E.V. Beachy P.A. Leahy D.J. Cell. 1997; 91: 85-97Abstract Full Text Full Text PDF PubMed Scopus (229) Google Scholar). Mutation of catalytic residues at the N- and C-terminal splice junctions of the Sce VMA inteins resulted in the trapping of the inactive precursor proteins for crystallization studies (22Poland B.W. Xu M.Q. Quiocho F.A. J. Biol. Chem. 2000; 275: 16408-16413Abstract Full Text Full Text PDF PubMed Scopus (125) Google Scholar, 23Mizutani R. Nogami S. Kawasaki M. Ohya Y. Anraku Y. Satow Y. J. Mol. Biol. 2002; 316: 919-929Crossref PubMed Scopus (65) Google Scholar). These investigations indicated that the conserved block B residues and the residues at the N-terminal splice site form an active center that facilitates cleavage of the N-terminal scissile peptide bond. In addition, it has been suggested that the penultimate histidine residue plays a key role in succinimide formation because of its proximity to the side chain of the C-terminal Asn (26Klabunde T. Sharma S. Telenti A. Jacobs Jr., W.R. Sacchettini J.C. Nat. Struct. Biol. 1998; 5: 31-36Crossref PubMed Scopus (203) Google Scholar). Other interactions may exist to assist the cyclization of Asn in the more than 20 inteins known to lack the penultimate histidine (27Wu H. Xu M.Q. Liu X.Q. Biochim. Biophys. Acta. 1998; 1387: 422-432Crossref PubMed Scopus (153) Google Scholar, 28Southworth M.W. Benner J. Perler F.B. EMBO J. 2000; 19: 5019-5026Crossref PubMed Scopus (76) Google Scholar, 29Chen L. Benner J. Perler F.B. J. Biol. Chem. 2000; 275: 20431-20435Abstract Full Text Full Text PDF PubMed Scopus (42) Google Scholar, 30Wang S. Liu X.Q. J. Biol. Chem. 1997; 272: 11869-11873Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar). Mutation of the highly conserved block F residues impaired splicing (31Ghosh I. Sun L. Xu M.Q. J. Biol. Chem. 2001; 276: 24051-24058Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar); however, the structural and catalytic roles of these residues have not been elucidated. Furthermore, the mechanisms for generating a nucleophilic potential at the intein N terminus and at the first C-extein position have yet to be determined. A complete examination of a pre-splicing structure would lead to further understanding of the chemical mechanism of protein splicing and the evolution pathway of autocatalytic inteins. To explore the outstanding questions concerning the structure and evolution of inteins, we conducted a crystal structure study and mutational analysis of a mini-intein precursor. This mini-intein was derived from a 429-amino acid intein encoded in the DNA helicase gene of the cyanobacterium Synechocystis sp. strain PCC6803 (Ssp DnaB intein) (27Wu H. Xu M.Q. Liu X.Q. Biochim. Biophys. Acta. 1998; 1387: 422-432Crossref PubMed Scopus (153) Google Scholar, 32Pietrokovski S. Trends Genet. 1996; 12: 287-288Abstract Full Text PDF PubMed Scopus (40) Google Scholar). Deletion of a centrally located 275-amino acid sequence encoding an endonuclease domain resulted in a splicing-proficient minimal intein (Ssp DnaB mini-intein) (27Wu H. Xu M.Q. Liu X.Q. Biochim. Biophys. Acta. 1998; 1387: 422-432Crossref PubMed Scopus (153) Google Scholar). Our investigation determined that two catalytic sites within the structure independently catalyze cleavage of the peptide bonds flanking the intein. The C-terminal catalytic site, formed by the conserved residues of blocks F and G, constitutes a charge relay system and an oxyanion binding site, which may play vital roles in catalyzing the cyclization of Asn-154. This intein structure also reveals that the peptide bond between Asn-154 and the first C-extein residue, Ser-155, is in a distorted trans conformation. The implications of the existence of a catalytic site in each intein subdomain are discussed with respect to intein evolution. Protein Expression and Purification—The pTWIN vector (33Evans Jr., T.C. Xu M.Q. Biopolymers. 1999; 51: 333-342Crossref PubMed Google Scholar) was used for expression and purification of the Ssp DnaB mini-intein in Escherichia coli. Purification of the Ssp DnaB intein was conducted by an intein-mediated cleavage method, as described previously (34Xu M.Q. Evans Jr., T.C. Methods. 2001; 24: 257-277Crossref PubMed Scopus (116) Google Scholar). The protein sample was then applied to a Resource Q column (Amersham Biosciences) and eluted with a linear gradient of NaCl. The peak containing the Ssp DnaB mini-intein was loaded onto a Superdex 75 column (Amersham Biosciences) pre-equilibrated with 10 mm Tris-HCl, 25 mm NaCl, pH 7.5, 10 mm β-mercaptoethanol and eluted with the same buffer. The final protein concentration was 24 mg/ml before crystallization. Dithiothreitol was added to the protein to a final concentration of 5 mm (35Chen X. Xu M.Q. Ding Y. Ferrandon S. Rao Z. Acta Crystallogr. Sect. D Biol. Crystallogr. 2002; 58: 1201-1203Crossref PubMed Scopus (4) Google Scholar). Mutagenesis and C-terminal Cleavage Activity Assay—The gene encoding the Ssp DnaB intein with the C1A mutation was cloned into the XhoI and AgeI sites of the LITMUS 28 vector (New England Biolabs, Inc.). Kunkel mutagenesis was used to generate the point mutations in the intein using the following sense strand primers (36Kunkel T.A. Proc. Natl. Acad. Sci. U. S. A. 1985; 82: 488-492Crossref PubMed Scopus (4897) Google Scholar, 37Kunkel T.A. Roberts J.D. Zakour R.A. Methods Enzymol. 1987; 154: 367-382Crossref PubMed Scopus (4558) Google Scholar): V134A, 5′-ACTGGAGTCGAAGAGGCTTTTGATTTGACTGTGC-3′; F135A, 5′-ACTGGAGTCGAAGAGGTTGCTGATTTGACTGTGC-3′; F135Y, 5′-ACTGGAGTCGAAGAGGTTTATGATTTGACTGTGCCA-3′; D136A, 5′-GTCGAAGAGGTTTTTGCTTTGACTGTGCCAGGA-3′; H143A, 5′-ACTGTGCCAGGACCAGCTAACTTTGTCGCGAA TG-3′; H143Q, 5′-ACTGTGCCAGGACCAGAGAACTTTGTCGCGAATG-3′; N144A, 5′-ACTGTGCCAGGACCACATGCCTTTGTCGCGAATGAC-3′; F145A, 5′-CAGGACCACATAACGCTGTCGCGAATGACATCAT-3′; F145Y, 5′-CCAGGACCACATAACTATGTCGCGAATGACATCAT-3′; N144A, 5′-ACTGTGCCAGGACCACATGCCTTTGTCGCGAATGAC-3′. The mutants were verified by DNA sequencing, and the mutated intein was transferred to the vector pMSL to replace the wild type intein sequence using XhoI and AgeI (19Mathys S. Evans T.C. Chute I.C. Wu H. Chong S. Benner J. Liu X.Q. Xu M.Q. Gene (Amst.). 1999; 231: 1-13Crossref PubMed Scopus (187) Google Scholar). The pMSL vector expresses a tripartite fusion protein with maltose-binding protein (MBP) 1The abbreviations used are: MBP, maltose-binding protein; I, intein; MI, MBP intein; CBD, chitin binding domain. at the N terminus followed by the Ssp DnaB intein (17 kDa) containing the C1A mutation and T4 DNA ligase (58 kDa). The plasmids were transformed into E. coli ER2566, and a colony was inoculated in LB medium containing 100 μg/ml ampicillin. The cells were grown at 37 °C until A 600 of 0.5 was reached and induced for 3 h at 30 °C with the addition of 0.3 mm isopropyl β-d-thiogalactoside. In vivo cleavage activities of the various mutants were examined by Coomassie Blue staining of 12% SDS-PAGE. The identity of the precursor and cleavage products was confirmed by Western blot analysis using antibodies against MBP and the Ssp DnaB intein. Amino acid sequencing confirmed that the MI species contained the sequence expected for the N terminus of MBP (MKTEEGKLV), and the L species contained the N terminus of the expected C-extein (SIEQDTGMLP). The T138A or T138S mutants were examined in a construct (pMIB) in which the mutated DnaB intein sequence was inserted between the E. coli MBP and the chitinase A1 chitin binding domain (CBD) of Bacillus circulans (31Ghosh I. Sun L. Xu M.Q. J. Biol. Chem. 2001; 276: 24051-24058Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar). Expression of the wild type or mutant MIB fusion proteins was induced at 37 °C for 3 h. In vivo splicing activity was analyzed by Coomassie Blue staining of a 10-20% SDS-PAGE. The identity of the splicing and cleavage products was confirmed by Western blot analysis with antibodies against MBP, the DnaB intein, and CBD. Crystallization and Data Collection—Crystals were grown in hanging drops equilibrated by vapor diffusion against reservoir solutions of 100 mm Tris buffer at pH 7.5-8.0, 16-20% polyethylene glycol 4000, and 4% (v/v) ethyl glycerol (35Chen X. Xu M.Q. Ding Y. Ferrandon S. Rao Z. Acta Crystallogr. Sect. D Biol. Crystallogr. 2002; 58: 1201-1203Crossref PubMed Scopus (4) Google Scholar). The crystals were mounted on a nylon-fiber loop and flash-frozen in a nitrogen-gas stream at 100 K before x-ray data analysis. Diffraction data were collected using a Mar345 Image Plate with a Rigaku rotating copper anode x-ray generator at 48 kV and 98 mA (λ = 1.5418 Å). All intensity data were processed and scaled using the programs DENZO and SCALEPACK (38Otwinowski Z. Minor W. Methods Enzymol. 1997; 276: 307-326Crossref PubMed Scopus (38526) Google Scholar). Structure Determination and Refinement—The structure was determined by molecular replacement using AMORE from the CCP4 software package. The crystal structure of the GyrA intein from M. xenopi (PDB entry 1AM2) was used as the search model. The amino acid sequence similarity between the Mxe GyrA intein and the Ssp DnaB mini-intein is 41%. The two α-helices were omitted from the model, and the nonhomologous residues in the model were substituted with alanine. The rotation solution was obtained using 8.0-4.0-Å data, and the highest peak appeared with α = 2.32°, β = 105.75°, γ = 141.58°. Translation searches were carried out in the two possible space groups P3121 and P3221, the latter giving a better solution (x = 0.3328, y = 0.3246, z = 0.2620), which was significantly above the noise level. The correlation coefficient and R-value were 0.186 and 0.519, respectively. After rigid body refinement, the R-factor was reduced to 0.496. The model was rebuilt with O (39Jones T.A. Zou J.Y. Cowan S.W. Kjeldgaard M. Acta Crystallogr. Sect. A. 1991; 47: 110-119Crossref PubMed Scopus (13009) Google Scholar) and refined with CNS (40Brunger A.T. Adams P.D. Clore G.M. DeLano W.L. Gros P. Grosse-Kunstleve R.W. Jiang J.S. Kuszewski J. Nilges M. Pannu N.S. Read R.J. Rice L.M. Simonson T. Warren G.L. Acta Crystallogr. Sect. D Biol. Crystallogr. 1998; 54: 905-921Crossref PubMed Scopus (16957) Google Scholar). No data truncation was applied in the refinement. 10% of the data was set aside to calculate the free R-factor. Because of the low sequence similarity between the GyrA intein and Ssp DnaB intein, significant model building work was done manually using O involving mutations, deletions, and insertions. Using the data of 40-2.0 Å, 2|F o|-|F c| and|F o|-|F c| electron density maps were calculated and examined with O. The refinement was completed by alternating between manual building and minimization using data in the resolution range 40-2.0 Å. Group B-factor refinement was then used to refine the temperature factors in the model. In each refinement step, initial anisotropic overall B-factor correction with a lower resolution limit of 6 Å and a bulk-solvent correction were applied to the data. The 52 water molecules were located in difference electron density peaks only if the peaks were above 3.0 σ and if acceptable hydrogen bonds to atoms in the model could be made. The final protein model and the waters were checked using O. The highly ordered water molecule, Wat-30, carries a B-factor of ∼26 Å2, lower than the average of solvent molecules. The model was refined to an R of 0.210 and an R free of 0.263 at 2.0-Å resolution. No residues are in the disallowed regions of the main-chain φ-ψ plot. Three N-terminal extein residues (from -5 to -3), three C-terminal extein residues (from 157 to 159), and the loop between residues 99 and 115 were found to be disordered and have been omitted from the model. A model of the wild type DnaB mini-intein precursor was generated by replacing Ala-1 and Ala-154 in the determined intein structure with Cys and Asn, respectively. The model was imported into the CNS (40Brunger A.T. Adams P.D. Clore G.M. DeLano W.L. Gros P. Grosse-Kunstleve R.W. Jiang J.S. Kuszewski J. Nilges M. Pannu N.S. Read R.J. Rice L.M. Simonson T. Warren G.L. Acta Crystallogr. Sect. D Biol. Crystallogr. 1998; 54: 905-921Crossref PubMed Scopus (16957) Google Scholar) program for energy minimization. The Cys-1 and Asn-154 side chain atoms were initially modeled with their χ1 angles equal to the most favored conformation, Cys χ1 = -60° and Asn χ1 = -180°. After energy minimization, the angles were only slightly changed. The backbone angle N-Cα-C (τ) at the C-terminal scissile bond (Ser-155) measures 122° both before and after energy minimization. Identification of Two Catalytic Sites—The Ssp DnaB mini-intein is comprised of an N-terminal segment of 106 residues and a C-terminal segment of 48 residues (27Wu H. Xu M.Q. Liu X.Q. Biochim. Biophys. Acta. 1998; 1387: 422-432Crossref PubMed Scopus (153) Google Scholar). The precursor protein used in this study possesses five native N and C extein residues flanking the 154-residue mini-intein. The first and last intein residues (Cys-1 and Asn-154) were substituted with alanine to capture the precursor in a presplicing state. The crystal structure of this intein precursor was determined by the molecular replacement method and refined to 2.0 Å resolution. The statistics of diffraction data collection and structure refinement are given in Table I. The structure exhibits a typical horseshoe-like 12 β-strand Hint domain (Fig. 2A). The N-terminal extein residues Ser-2 and Gly-1, and C-terminal extein residues Ser-155 and Ile-156 have well defined electron density maps (Fig. 2, B and C). The N- and C-terminal intein subdomains are tightly associated, with numerous hydrogen-bonding contacts and hydrophobic packing interactions. Superposition of the available crystal structures indicates that inteins share overall structural similarity. The root mean square difference for the Cα atoms of the Hint domain is ∼1.4 Å between the Ssp DnaB and Sce VMA inteins (21Duan X. Gimble F.S. Quiocho F.A. Cell. 1997; 89: 555-564Abstract Full Text Full Text PDF PubMed Scopus (234) Google Scholar, 22Poland B.W. Xu M.Q. Quiocho F.A. J. Biol. Chem. 2000; 275: 16408-16413Abstract Full Text Full Text PDF PubMed Scopus (125) Google Scholar, 23Mizutani R. Nogami S. Kawasaki M. Ohya Y. Anraku Y. Satow Y. J. Mol. Biol. 2002; 316: 919-929Crossref PubMed Scopus (65) Google Scholar, 24Hu D. Crist M. Duan X. Quiocho F.A. Gimble F.S. J. Biol. Chem. 2000; 275: 2705-2712Abstract Full Text Full Text PDF PubMed Scopus (30) Google Scholar), 1.2 Å between the Ssp DnaB and Mxe GyrA inteins (26Klabunde T. Sharma S. Telenti A. Jacobs Jr., W.R. Sacchettini J.C. Nat. Struct. Biol. 1998; 5: 31-36Crossref PubMed Scopus (203) Google Scholar), and 1.2 Å between the Ssp DnaB and P. furiosus Pol inteins (25Ichiyanagi K. Ishino Y. Ariyoshi M. Komori K. Morikawa K. J. Mol. Biol. 2000; 300: 889-901Crossref PubMed Scopus (97) Google Scholar).Table IDiffraction data collection and structure refinement statisticsSpace groupP3221Unit-cell dimensions (Å)a = b = 58.2c = 70.3Resolution range (Å)40.0-2.0 ÅNumber of observations107,903Number of unique reflections9,307Completeness (%)97.6 (75.5)aNumbers in parentheses are for the resolution range 2.07-2.00 Å.Average I/σ(I)20.5 (7.5)aNumbers in parentheses are for the resolution range 2.07-2.00 Å.R merge0.046 (0.262)aNumbers in parentheses are for the resolution range 2.07-2.00 Å.Mean redundancy11.6Number of protein molecules per asymmetric unit1Number of protein atoms1,107Number of water molecules52R work0.210R freebR free is calculated for the 10% subset of the unique reflections.0.263Root mean square deviations from idealitiesBond lengths (Å)0.013Bond angles (degree)2.0Average B factors (Å2)Protein30.2Waters33.4a Numbers in parentheses are for the resolution range 2.07-2.00 Å.b R free is calculated for the 10% subset of the unique reflections. Open table in a new tab A model of the wild type DnaB mini-intein precursor was generated by replacing Ala-1 and Ala-154 in the determined intein structure with Cys and Asn, respectively. The two splice junctions are located at the center of the horseshoe structure at the ends of two adjacent antiparallel β-strands. Intriguingly, the modeled structure precursor reveals the presence of two catalytic sites that are in close proximity to each other but which do not share catalytic components (Figs. 2 and 3). Several residues in the Ssp DnaB intein N-terminal region appear to play major roles in aligning and activating the reactive groups at the N-terminal splice junction (Fig. 2D). The important residues include Gly-1 (the native extein residue preceding the scissile bond) and Cys-1, flanking the N-terminal splice site, Thr-70, Asn-72, and His-73 of block B, and Thr-51. Contributions from the highly conserved block B residues have been observed in the x-ray structures of the Sce VMA and Mxe GyrA inteins (22Poland B.W. Xu M.Q. Quiocho F.A. J. Biol. Chem. 2000; 275: 16408-16413Abstract Full Text Full Text PDF PubMed Scopus (125) Google Scholar, 23Mizutani R. Nogami S. Kawasaki M. Ohya Y. Anraku Y. Satow Y. J. Mol. Biol. 2002; 316: 919-929Crossref PubMed Scopus (65) Google Scholar, 26Klabunde T. Sharma S. Telenti A. Jacobs Jr., W.R. Sacchettini J.C. Nat. Struct. Biol. 1998; 5: 31-36Crossref PubMed Scopus (203) Google Scholar). The S-H bond of the modeled Cys-1 is polarized by hydrogen bonding with the hydroxyl group of Thr-51 and the carbonyl oxygen of Gly-1. The conformation of Thr-51 appears to be maintained in the appropriate orientation for hydrogen bonding by interactions with Lys-54 and Val-134. The arrangement of this active site is capable of facilitating the initial step in the splicing pathway, a nucleophilic attack by the thiol of Cys-1 on the carbonyl carbon of the peptide bond between Cys-1 and Gly-1 to form a thioester intermediate (Fig. 4A).Fig. 4A chemical mechanism proposed for splicing of the Ssp DnaB intein. The red arrows indicate the routes of nucleophilic attacks in the splicing pathway. Dashed lines indicate hydrogen bonds. The tetrahedral intermediate formed by an N-S acyl rearrangement at Cys-1 is not shown.View Large Image Figure ViewerDownload Hi-res image Download (PPT) The second catalytic site is formed by the conserved residues in block F and G in the intein C-terminal subdomain and surrounds the Asn-154 and Ser-155 that are involved in two crucial steps of the splicing reaction (Figs. 2E, 3, and 4). This site may be responsible for enhancing the nucleophilicity of the Asn-154 side chain and stabilizing the resulting tetrahedral intermediate. In addition, the scissile bonds at the N- and C-terminal splice junctions are in the ordinary trans conformation. The backbone angle N-Cα-C (τ) at the C-terminal scissile bond (Ser-155) measures 122°, deviating from the ideal angle of 110° by 12°. The implications of these findings are discussed in the next sections. Structure and Function of the C-terminal Catalytic Site—This study provides the first structural evidence that several highly conserved block F and G residues, located in

Referência(s)