FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2040509036> ?p ?o ?g. }

• W2040509036 endingPage "5836" @default.
• W2040509036 startingPage "5829" @default.
• W2040509036 abstract "Bacteriophage terminases are essential molecular motors involved in the encapsidation of viral DNA. They are hetero-multimers whose large subunit encodes both ATPase and endonuclease activities. Although the ATPase domain is well characterized from sequence and functional analysis, the C-terminal region remains poorly defined. We describe sequence-structure comparisons of the endonuclease region of various bacteriophages that revealed new sequence similarities shared by this region and the Holliday junction resolvase RuvC and to a lesser extent the HIV integrase and the ribonuclease H. Extensive sequence comparison and motif refinement led to a common signature of terminases and resolvases with three conserved acidic residues engaged in catalytic activity. Sequence analyses were validated by in vivo and in vitro functional assays showing that the nuclease activity of the endonuclease domain of bacteriophage T5 terminase was abolished by mutation of any of the three predicted catalytic aspartates. Overall, these data suggest that the endonuclease domains of terminases operate autonomously and that they adopt a fold similar to that of resolvases and share the same divalent cation-dependent enzymatic mechanism. Bacteriophage terminases are essential molecular motors involved in the encapsidation of viral DNA. They are hetero-multimers whose large subunit encodes both ATPase and endonuclease activities. Although the ATPase domain is well characterized from sequence and functional analysis, the C-terminal region remains poorly defined. We describe sequence-structure comparisons of the endonuclease region of various bacteriophages that revealed new sequence similarities shared by this region and the Holliday junction resolvase RuvC and to a lesser extent the HIV integrase and the ribonuclease H. Extensive sequence comparison and motif refinement led to a common signature of terminases and resolvases with three conserved acidic residues engaged in catalytic activity. Sequence analyses were validated by in vivo and in vitro functional assays showing that the nuclease activity of the endonuclease domain of bacteriophage T5 terminase was abolished by mutation of any of the three predicted catalytic aspartates. Overall, these data suggest that the endonuclease domains of terminases operate autonomously and that they adopt a fold similar to that of resolvases and share the same divalent cation-dependent enzymatic mechanism. Packaging of the double-stranded DNA of bacteriophages into the viral capsid relies on a molecular machine consisting of several proteins: the portal pore, which is located at one vertex of the capsid, and the terminases. The latter are implicated, first, in bringing the DNA concatemers to the empty prohead through interaction with the portal protein, then in active ATP-driven translocation of the DNA into the capsid, and, finally, in cutting the concatemer to generate the mature viral DNA (1Feiss M. Catalano C. Viral Genome Packaging Machines.in: Catalano C. Kluwer Academic / Plenum Publishers, New York2005: 5-39Google Scholar). Terminases are generally described as hetero-multimeric structures consisting of two subunits. The small regulatory subunit binds to the viral DNA, and the large catalytic subunit carries both an endonuclease activity for DNA cleavage and an ATPase activity required for DNA packaging (2Earnshaw W.C. Casjens S.R. Cell. 1980; 21: 319-331Abstract Full Text PDF PubMed Scopus (339) Google Scholar, 3Catalano C.E. Cue D. Feiss M. Mol. Microbiol. 1995; 16: 1075-1086Crossref PubMed Scopus (169) Google Scholar, 4Black L.W. Annu. Rev. Microbiol. 1989; 43: 267-292Crossref PubMed Scopus (364) Google Scholar, 5Morita M. Tasaka M. Fujisawa H. Virology. 1995; 211: 516-524Crossref PubMed Scopus (14) Google Scholar). Such general organization is well conserved among tailed phages and herpes viruses (6Przech A.J. Yu D. Weller S.K. J. Virol. 2003; 77: 9613-9621Crossref PubMed Scopus (36) Google Scholar). The functional characterization of these proteins and the interplay of the different domains is the object of intensive investigation, specially in the case of phage λ (1Feiss M. Catalano C. Viral Genome Packaging Machines.in: Catalano C. Kluwer Academic / Plenum Publishers, New York2005: 5-39Google Scholar, 7Dhar A. Feiss M. J. Mol. Biol. 2005; 347: 71-80Crossref PubMed Scopus (19) Google Scholar, 8Maluf N.K. Yank Q. Catalano C.E. J. Mol. Biol. 2005; 347: 523-542Crossref PubMed Scopus (51) Google Scholar, 9Catalano C.E. Cell. Mol. Life Sci. 2000; 57: 128-148Crossref PubMed Scopus (146) Google Scholar), T4 (10Kanamaru S. Kondabagil K. Rossmann M.G. Rao V.B. J. Biol. Chem. 2004; 279: 40795-40801Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar, 11Rao V.B. Black L.W. Viral Genome Packaging Machines.in: Catalano C. Kluwer Academic/Plenum Publishers, New York2005: 40-58Google Scholar), and SPP1 (12Droge A. Tavares P. Bacteriophage SPP1 DNA Packaging.in: Catalano C. Kluwer Academic/Plenum Publishers, New York2005: 89-99Google Scholar, 13Gual A. Camacho A.G. Alonso J.C. J. Biol. Chem. 2000; 275: 35311-35319Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar). The terminase large subunit is a two-domain protein. Its N terminus is well characterized both from sequence and functional analysis. It consists of a conserved ATPase catalytic center with clear ATP binding motifs (named Walker A and B) (14Koonin E.V. J. Mol. Biol. 1993; 229: 1165-1174Crossref PubMed Scopus (259) Google Scholar). The C-terminal region is less well defined (10Kanamaru S. Kondabagil K. Rossmann M.G. Rao V.B. J. Biol. Chem. 2004; 279: 40795-40801Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar, 13Gual A. Camacho A.G. Alonso J.C. J. Biol. Chem. 2000; 275: 35311-35319Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar, 15Rentas F.J. Rao V.B. J. Mol. Biol. 2003; 334: 37-52Crossref PubMed Scopus (33) Google Scholar, 16Davidson A.R. Gold M. Virology. 1992; 189: 21-30Crossref PubMed Scopus (38) Google Scholar). Mutagenesis studies have shown that the endonuclease activity of phage λ is located within the C-terminal half of the terminase large subunit (gpA; 73 kDa) (17Hwang Y. Catalano C.E. Feiss M. Biochemistry. 1996; 35: 2796-2803Crossref PubMed Scopus (61) Google Scholar). Several mutations affecting this region, among which one was located on Asp-401, were shown to inactivate the endonuclease activity (16Davidson A.R. Gold M. Virology. 1992; 189: 21-30Crossref PubMed Scopus (38) Google Scholar). Phage T4 large subunit terminase (gp17, 70 kDa) exhibits an in vivo endonuclease activity (18Bhattacharyya S.P. Rao V.B. Virology. 1993; 196: 34-44Crossref PubMed Scopus (50) Google Scholar). Extensive site-directed mutagenesis has revealed a cluster of conserved residues (Asp-401, Glu-404, Gly-405, and Asp-409) within the C-terminal half of gp17 that are critical for function (15Rentas F.J. Rao V.B. J. Mol. Biol. 2003; 334: 37-52Crossref PubMed Scopus (33) Google Scholar). Further biochemical analyses revealed only aspartate Asp-401 as an essential residue for T4 phage nuclease activity (15Rentas F.J. Rao V.B. J. Mol. Biol. 2003; 334: 37-52Crossref PubMed Scopus (33) Google Scholar). Expression of the C-terminal half of gp17 was sufficient for nuclease activity (10Kanamaru S. Kondabagil K. Rossmann M.G. Rao V.B. J. Biol. Chem. 2004; 279: 40795-40801Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar). The C-terminal regions of terminases characterized so far share neither a sequence motif nor sequence similarities with any other known protein family. This sequence divergence makes difficult a clear delimitation of the endonuclease domain and a definition of its active site. No relationship with endonucleases already characterized at the structural level was described yet. Indeed, the numerous endonucleases studied so far adopt a great variety of folds ranging from all α to all β, including various α/β and α+β folds, but show no common sequence motifs (19Nishino T. Morikawa K. Oncogene. 2002; 21: 9022-9032Crossref PubMed Scopus (84) Google Scholar). The recent sequencing of the bacteriophage T5 genome (GenBank™ accession numbers AY692264, AY734508, and AY543070) prompted us to search for new sequence structure-function relationships in terminases. The T5 genome is a linear double-stranded DNA of 121,750 bp that contains large direct terminal repeats of ∼10,000 bp (20Rhoades M. Rhoades E.A. J. Mol. Biol. 1972; 69: 187-200Crossref PubMed Scopus (23) Google Scholar, 21Wang J. Jiang Y. Vincent M. Sun Y. Yu H. Bao Q. Kong H. Hu S. Virology. 2005; 332: 45-65Crossref PubMed Scopus (81) Google Scholar). Its replication proceeds via concatemeric intermediates (22Everett R.D. J. Gen. Virol. 1981; 52: 25-38Crossref PubMed Scopus (9) Google Scholar), but the processing pathway of T5 concatemers as well as the mechanism of DNA encapsidation remain unknown. The presence of unusually extensive terminal redundancies and of site-specific single-stranded interruptions (23McCorquodale J.D. Warner H.R. The Viruses.in: Calendar R. 1. Plenum Press, New York1988: 439-476Google Scholar) in the mature as well as in the concatemeric DNA (22Everett R.D. J. Gen. Virol. 1981; 52: 25-38Crossref PubMed Scopus (9) Google Scholar) makes T5 an interesting model of DNA packaging. Here we report an extended bioinformatics analysis of the endonuclease region of double-stranded DNA phage terminases that reveals new sequence similarities shared by this region and the Holliday junction resolvase RuvC and to a lesser extent by the HIV integrase and the ribonuclease H. This analysis is further validated by in vivo and in vitro functional studies of the endonuclease domain of bacteriophage T5 terminase. Bionformatic Procedures—The PSI-BLAST program (24Altschul S.F. Madden T.L. Schaffer A.A. Zhang J. Zhang Z. Miller W. Lipman D.J. Nucleic Acids Res. 1997; 25: 3389-3402Crossref PubMed Scopus (59725) Google Scholar) was applied with standard parameters to search for homologous proteins in the SPTrEMBL and NCBI non-redundant sequence databases. Protein sequences were retrieved from the NCBI Entrez data base (www.ncbi.nlm.nih.gov/entrez). Their accession numbers are given in the legend of Fig. 2A. Fold compatibility for the full-length and truncated sequences of the terminases SPP1, T4, and T5 was searched using 3DPSSM (25Kelley L.A. MacCallum R.M. Sternberg M.J. J. Mol. Biol. 2000; 299: 499-520Crossref PubMed Scopus (1120) Google Scholar), FUGUE (26Shi J. Blundell T.L. Mizuguchi K. J. Mol. Biol. 2001; 310: 243-257Crossref PubMed Scopus (1081) Google Scholar), mGenThreader (27McGuffin L.J. Bryson K. Jones D.T. Bioinformatics. 2000; 16: 404-405Crossref PubMed Scopus (2721) Google Scholar), PDB-Blast (bioinformatics.burnham-inst.org/pdb_blast/), SAM-T99 (28Karplus K. Barrett C. Hughey R. Bioinformatics. 1998; 14: 846-856Crossref PubMed Scopus (907) Google Scholar) via the metaserver @TOME (29Douguet D. Labesse G. Bioinformatics. 2001; 17: 752-753Crossref PubMed Scopus (80) Google Scholar). Automatic modeling were performed using TITO (30Labesse G. Mornon J. Bioinformatics. 1998; 14: 206-211Crossref PubMed Scopus (63) Google Scholar), SCWRL (31Dunbrack Jr., R.L. Karplus M. J. Mol. Biol. 1993; 230: 543-574Crossref PubMed Scopus (786) Google Scholar), and MODELLER (32Sali A. Blundell T.L. J. Mol. Biol. 1993; 234: 779-815Crossref PubMed Scopus (10505) Google Scholar), and the validity of the resulting models was evaluated with PROSA (33Sippl M.J. Proteins. 1993; 17: 355-362Crossref PubMed Scopus (1767) Google Scholar) and Verify3D (34Eisenberg D. Luthy R. Bowie J.U. Methods Enzymol. 1997; 277: 396-404Crossref PubMed Scopus (1628) Google Scholar). Sequence-structure alignments including various terminases and PDB1HJR (see alignment in Fig. 2A) were manually refined with the help of the program ViTO (35Catherinot V. Labesse G. Bioinformatics. 2004; 20: 3694-3696Crossref PubMed Scopus (50) Google Scholar). Improved three-dimensional models were built for the T5 terminase using MODELLER 6.2 with the loop optimization procedure. Secondary structure assignments were performed using DSSP (36Kabsch W. Sander C. Biopolymers. 1983; 22: 2577-2637Crossref PubMed Scopus (12257) Google Scholar). Cloning Procedure—The gene encoding the terminase large subunit was identified from the complete genome sequence (GenBank™ accession number AY692264). The Endo-Term 3The abbreviations used are: Endo-term, endonuclease domain of phage T5 terminase; Ni-NTA, nickel nitrilotriacetate; WT, wild type; HIV, human immunodeficiency virus. domain, delimited between amino acids 191 and 438, was amplified by PCR using T5 DNA as a template. The oligonucleotides used for PCR were complementary to the coding sequence of the Endo-Term gene fragment over 6 codons at both 5′- and 3′-ends and contained additional bases to create a XhoI and a BamH1 restriction site at the 5′- and 3′-ends of the gene fragment, respectively. The amplified Endo-Term PCR fragment was cloned into the vector pET15b (Novagen), between the XhoI and BamH1 restriction sites of the multicloning site, to generate an N-terminal hexahistidine-tagged Endo-Term domain. Three mutated Endo-Term gene fragments were constructed with the following changes in the sequence: D286N, D342N, and D425N. Site-directed mutagenesis was performed by using the kit “Expand High Fidelity PCR System” from Roche Applied Science according to the method of Ansaldi et al. (37Ansaldi M. Lepelletier M. Mejean V. Anal. Biochem. 1996; 234: 110-111Crossref PubMed Scopus (87) Google Scholar). The sequences of the mutated Endo-Term gene constructs were verified by sequence analysis in the pET15b plasmid. Protein Expression, Purification, and N-terminal Amino Acid Sequencing—Escherichia coli BL21(DE3) pLys-S cells were used for overproduction of the wild type and the mutated endonuclease domains. Cells containing the various pET15b constructs were grown at 28 °C to an A600 of 0.4 in Luria Bertani medium supplemented with ampicillin (100 μg/ml) and chloramphenicol (20 μg/ml). Overexpression was induced by the addition of isopropyl-1-thio-β-d-galactopyranoside (1 mm). Purification of the proteins was conducted at room temperature as follows. The induced cells were harvested by centrifugation at 3000 × g, 10 min. The pellets were resuspended in lysis buffer (20 mm Tris-HCl, pH 8, 100 mm NaCl) and subjected to the French press procedure at 15,000 p.s.i. After centrifugation (30000 × g, 30 min), the supernatant was loaded onto a Ni-NTA Superflow affinity resin (Qiagen). The His-tagged proteins were eluted with 1 m imidazole and loaded onto a HR 10/30 Superose 12 gel filtration column (GE Healthcare) pre-equilibrated with 20 mm Tris-HCl, pH 8, 300 mm NaCl, 5 mm dithiothreitol. The proteins were eluted with the same buffer. They were separated by electrophoresis on SDS-PAGE (12%) and stained with Coomassie blue. N-terminal amino acid sequencing was performed after transfer of the protein bands to a polyvinylidene difluoride membrane. The concentration of the proteins was determined using a value of the molar extinction coefficient of 39070 m-1 cm-1. In Vivo Nuclease Assay—Cells containing the various pET15b constructs were grown to an A600 of 0.4 as described above. 1-ml aliquots of the cultures were withdrawn prior to induction and 60, 120, and 180 min after induction. Each aliquot was centrifuged at 3000 × g for 5 min, and the pellet was used to isolate the DNA by the alkaline lysis procedure. The DNA was further purified using the Quick purification kit Nucleospin; Macherey-Nagel. 10 μl of DNA were run on a 1% (w/v) agarose gel. In Vitro Nuclease Assays—All assays were performed soon after elution of the purified proteins from the gel filtration column to avoid their proteolysis. Purified proteins (36 μl; 0.2 mg/ml) were incubated at 37 °C with purified phage T5 DNA (4 μl; 0.32 mg/ml) in the presence of MgCl2 (5 mm). 10-μl aliquots of the reaction mixture were withdrawn 30, 60, 120, and 180 min after the beginning of the reaction. Nuclease activity was stopped by addition of EDTA (10 mm). DNA cleavage was followed on a 1% (w/v) agarose gel. The nuclease activity of the purified proteins was also assessed from fluorescence measurements using the DNA intercalant YO-PRO-1 (38Boulanger P. le Maire M. Bonhivers M. Dubois S. Desmadril M. Letellier L. Biochemistry. 1996; 35: 14216-14224Crossref PubMed Scopus (89) Google Scholar) (1 mm in dimethyl sulfoxide; Molecular Probes). The dye was 1000-fold diluted in a 1-ml cuvette containing 20 mm Tris-HCl, pH 8, 300 mm NaCl, 5 mm MgCl2, and 20 μg of T5 DNA. The purified proteins (200 μg) were then added. The fluorescence intensity was monitored at 30 °C as a function of time on a Varian fluorimeter with excitation and emission wavelengths set at 491 and 509 nm, respectively. Bioinformatic Analysis of Terminase Sequences—The T5 genome encodes a putative 49-kDa large subunit terminase, and the latter was used as a representative member of the terminase family together with those of bacteriophages SPP1 and T4 (10Kanamaru S. Kondabagil K. Rossmann M.G. Rao V.B. J. Biol. Chem. 2004; 279: 40795-40801Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar, 13Gual A. Camacho A.G. Alonso J.C. J. Biol. Chem. 2000; 275: 35311-35319Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar). A sequence similarity search using the program PSI-BLAST (24Altschul S.F. Madden T.L. Schaffer A.A. Zhang J. Zhang Z. Miller W. Lipman D.J. Nucleic Acids Res. 1997; 25: 3389-3402Crossref PubMed Scopus (59725) Google Scholar) and the large subunit of the T5 terminase as a query only hit several terminase sequences at convergence. This search also confirmed a higher level of homology at the N terminus (residues 1-165) and a higher divergence at the C-terminal part (residues 170-438). We further searched for structural similarities using the full-length terminases of bacteriophages SPP1, T4, and T5 as queries. Among the 5-fold recognition servers used, four highlighted strong similarities (>95% certainty) to ATP-dependent helicases and related ATPases. Highly significant scores were computed by mGenThreader and FUGUE for phages T4 and T5 terminases (see supplementary materials at www.infobiosud.cnrs.fr/bioserver/TER/suppl.html). However, the detected similarities covered only roughly one half of the terminase sequences. No significant hit corresponded to the C-terminal half. However, analysis of the full-size protein sequence could hide other structural and functional relationships that the C-terminal part might share with other proteins. Delimitation of the C- and N-terminal regions was therefore undertaken starting from the above similarity searches and from additional fold-compatibility searches. To further confirm and refine the proposed structural alignments, sequence analyses were performed with the N terminus and the C-terminal sequences, separately. According to sequence comparisons, the N-terminal region corresponds to an ATPase domain. It possesses characteristic Walker A (Motif, gXXXGKs where “X” stands for any amino acid and small and capital letters represent mostly present and strictly conserved amino acids, respectively) and Walker B (Motif, ////DE where “/” stand for a hydrophobic residue) Mg-ATP binding motifs and an ATPase-coupling motif (motif III). These motifs are located at positions 62-68, 150-155, and 184-186 in the T5 terminase (Fig. 1). This signature is present in most terminases (11Rao V.B. Black L.W. Viral Genome Packaging Machines.in: Catalano C. Kluwer Academic/Plenum Publishers, New York2005: 40-58Google Scholar, 39Mitchell M.S. Matsuzaki S. Imai S. Rao V.B. Nucleic Acids Res. 2002; 30: 4009-4021Crossref PubMed Scopus (107) Google Scholar, 40Mitchell M.S. Rao V.B. Virology. 2004; 321: 217-221Crossref PubMed Scopus (35) Google Scholar) and is related to the ATP binding signature of helicases (41Parris W. Rubinchik S. Yang Y.C. Gold M. J. Biol. Chem. 1994; 269: 13564-13574Abstract Full Text PDF PubMed Google Scholar) (notably the particular Walker B motif////DEad). The minimal ATPase region should encompass these motifs, and sequence-structure comparisons using the full-length terminases suggested that the N-terminal domain is roughly 160 residues in length. Fold-compatibility searches using the derived N-terminal region revealed the same trends: highly significant similarities (e-values below e-4 corresponding to >95% certainties) were observed with various helicases despite a rather low level of sequence identity (∼15%). Molecular modeling using MODELLER (32Sali A. Blundell T.L. J. Mol. Biol. 1993; 234: 779-815Crossref PubMed Scopus (10505) Google Scholar) was performed to analyze the proposed structural alignment and to precisely delimit the helicase-like ATP-binding domain of the terminases (data not shown). It suggested that the ∼160 residues of the N-terminal domain could form a stable and functional ATP-binding domain autonomously. Fold-recognition searches were then performed using the C-terminal region from distinct terminases. No significant score for any known structure was observed in the various runs performed using PSI-BLAST, in agreement with previous studies. On the contrary, fold-recognition using the metaserver @TOME suggested distant relationships with various structures, among which were some endonucleases (see supplementary materials at www.infobiosud.cnrs.fr/bioserver/TER/suppl.html). However, the compatibility scores were rather weak, and most structural alignments seemed to match only to the very C terminus (residues 270-438). The searches were resumed with shorter sequences according to the primary matches. In addition, the 14 top-scoring sequence-structure alignments were systematically tested using the automatic procedure available on the @TOME server. Promising three-dimensional models were observed for the C-terminal domain of the T5 and T4 phage terminases with scores ranging between -0.45 and -0.65 in PROSA and 0.25 and 0.35 in Verify3D. However, the overall sequence identity was low (∼10% over 150 residues). In most cases the models were built using three structurally related templates: a resolvase RuvC (PDB1HJR), a ribonuclease H (PDB1IO2), and two retroviral integrases (PDB1A5V and PDB1B9D). These four proteins possess a similar catalytic domain sharing a common topology (α/β) and the same enzymatic mechanism (42Yang W. Steitz T.A. Structure. 1995; 3: 131-134Abstract Full Text Full Text PDF PubMed Scopus (215) Google Scholar). Among these enzymes, resolvase RuvC appeared to be the most closely related to the terminases (data not shown). Resolvases were used for further alignment refinement and validation of the structural comparison. The first sequence-structure alignments including various terminases and the resolvase RuvC (PDB: 1HJR, see alignment in Fig. 2A) were manually refined and used to build new three-dimensional models for the T5 terminase. The resulting model (Fig. 2B) appeared significantly better and more satisfactory, taking into account the low level of sequence identity (PROSA score below -1.0 versus -1.2 for the template PDB 1HJR; verify3D score above 0.32 and 0.35, respectively). These results supported the proposed structural alignment of the catalytic domain of resolvases with the C-terminal endonuclease domain of the terminases. Furthermore, two sequence motifs (/g/D/g and dX/DXXXXX/) corresponding to the active site of RuvC resolvases (residues 5-10 and 139-148, respectively) are conserved in the terminases (in T5, residues 283-288 and 422-431, respectively). This suggested a potential functional homology and also enabled us to delimit more precisely the C-terminal domain (residues 281-438). To further confirm this new relationship, the sequence motifs containing two putatively catalytic aspartates were refined using the program PATTINPROT (43Combet C. Blanchet C. Geourjon C. Deleage G. Trends Biochem. Sci. 2000; 25: 147-150Abstract Full Text Full Text PDF PubMed Scopus (1431) Google Scholar). Extensive sequence comparison and motif refinement led us to a specific sequence signature (see alignment in Fig. 2A) of the resolvase-terminase subfamily with a highly significant e-value of 4.4e-11. This signature contains three well conserved acidic residues (aspartates 286, 342, and 425 according to the T5 terminase numbering). They are equivalent to the residues of the Holliday junction resolvase (PDB 1HJR) that form a catalytic triad requiring divalent cations (Mg2+ or Mn2+) for activity (44Rice P. Craigie R. Davies D.R. Curr. Opin. Struct. Biol. 1996; 6: 76-83Crossref PubMed Scopus (180) Google Scholar, 45Nowotny M. Gaidamakov S.A. Crouch R.J. Yang W. Cell. 2005; 121: 1005-1016Abstract Full Text Full Text PDF PubMed Scopus (499) Google Scholar). This local sequence and overall fold conservation suggested that the terminases possess a similar endonuclease domain sharing a common enzymatic mechanism with resolvases. To validate this proposal, we expressed the wild type T5 endonuclease domain and three recombinant proteins in which each of the predicted catalytic aspartates was substituted by an asparagine and measured their endonuclease activity. Expression of Endo-term, the Endonuclease Domain of Phage T5 Terminase: Mutations of the Three Predicted Catalytic Aspartates Abolish in Vivo Endonuclease Activity—The constructs used (respectively, named WT, D286N, D342N, D425N Endo-term) were delimited between amino acids 191 and 438 and contained an additional N-terminal hexahistidine sequence. The recombinant proteins were expressed in E. coli BL21(DE3) pLysS to lower their basal expression and subsequent toxicity. Isopropyl-1-thio-β-d-galactopyranoside induction resulted in an immediate arrest of growth of the bacteria carrying the WT Endo-term plasmid but only slightly affected the growth of those carrying the D286N, D342N, D425N mutations. Aliquots of the cells were taken at different times after induction and tested for protein expression. Although the level of expression of WT Endo-term appeared significantly lower than that of the mutated proteins (data not shown), its nuclease activity was higher. Indeed, WT Endo-term expression resulted in cleavage of the corresponding pET15b-derived plasmid and of the genomic DNA (see below), generating a complex mixture of DNA fragments that migrated as a characteristic smear on agarose gels (Fig. 3, lanes 2 and 3). The DNA was totally cut at t = 180 min (lane 4). On the contrary, the plasmids harboring each of the conservative mutations D286N, D342N, or D425N showed no sign of degradation over 180 min (lanes 6-8, 10-12, 14-16). These results suggested that mutation of the conserved aspartate residues strongly affected the in vivo endonuclease activity of the proteins. To validate these results and to determine how the mutations affected the nuclease activity, we purified the four proteins and determined their in vitro endonuclease activity. Purification and in Vitro Endonuclease Activity of the WT, D286N, D342N, and D425N Endo-term Proteins—All four proteins, expressed as described above, were found in the supernatant of cells centrifuged after a French pressure treatment. They were purified to near homogeneity by successive passage on an affinity column and on a gel filtration column as described. WT Endo-term migrated at an apparent molecular mass of 31 kDa, corresponding to that expected for the cloned DNA fragment (Fig. 4, lanes 3 and 5). The protein underwent limited proteolysis with time. A single product with molecular mass of 25 kDa was obtained after 72 h (Fig. 4, lane 7). Sequencing indicated that the protein was missing its N-terminal hexahistidine tag. Because proteolysis resulted in a shift of the molecular mass from 31 to 25 kDa this indicated that the cleavage had essentially occurred in the C-terminal region. The release of these 6 kDa results in a loss of the critical aspartate 425 and is also likely to affect the protein fold, explaining why the proteolyzed protein showed no nuclease activity when assayed by agarose gel electrophoresis (data not shown). The purified D286N, D342N, and D425N Endo-term proteins showed a pattern of time-dependent proteolysis similar to that of WT Endo-term (data not shown). Further functional assays were therefore only performed on freshly prepared proteins that were retained on the Ni-NTA column. Each of the four proteins was mixed with phage T5 DNA in a one:one protein:nucleotide molar ratio. Aliquots were taken at given times, and DNA cleavage was followed on agarose gel. Fig. 5A, lanes 5-8, shows that WT Endo-term efficiently cut T5 DNA in less than 60 min. DNA cleavage by the three mutants was slower: D286N and D425N Endo-term showed the same activity and cut the DNA in less than 180 min (lanes 9-12 and 17-20). D342N Endo-term showed the lowest nuclease activity, and the DNA was only slightly cut after 180 min (lanes 13-16). In all cases was the cut nonspecific, confirming that the smear observed in the in vivo assay corresponded to small pieces of E. coli genomic DNA. Endonuclease activity had an absolute requirement for divalent cations (MgCl2 or MnCl2) because it was not observed in the absence of these cations or arrested by addition of an excess of EDTA (data not shown). The decreased activity of the mutated proteins compared with the wild type protein was not due to a misfolding of the proteins because all four proteins retained the same content of secondary structure as judged by their circular dichroism spectrum (data not shown). Endonuclease activity was further quantified by measuring the dependence with time of the DNA intercalant YO-PRO1 fluorescence, which is enhanced upon binding to DNA and decreased when the DNA is cut by nucleases (38Boulanger P. le Maire M. Bonhivers M. Dubois S. Desmadril M. Letellier L. Biochemistry. 1996; 35: 14216-14224Crossref PubMed Scopus (89) Google Scholar). Fig. 5B highlights the difference in nuclease activity of the four proteins. The t1/2 of endonuclease activity was 14 and 50 min for the WT and D342N Endo-term proteins, respectively, and 30 min for both the D286N and D425N Endo-term proteins. This unambiguously demonstrated that the aspartate mutations affected nuclease activity and that all three substitutions were not equivalent. In this study we have used a combination of bioinformatics and biochemical analysis to precisely define the catalytic residues involved in the endonuclease activity of the large subunit of tailed phage terminases. Extensive sequence comparisons led to the conclusion that, despite primary sequence divergence, terminases including those of phage T7, P1, λ, P22, SPP1, T1, P2, T4, and T5 contain three well conserved aspartate residues that are equivalent to the residues of the Holliday junction resolvases and, to a lesser extent, the HIV integrase and ribonuclease H. The role of these residues in the endonuclease activity of the terminase large subunits was attested by functional assays performed on phage T5. From sequence comparisons and from the known location of the functional domains of phage T4 (10Kanamaru S. Kondabagil K. Rossmann M.G. Rao V.B. J. Biol. Chem. 2004; 279: 40795-40801Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar) and λ terminases (16Davidson A.R. Gold M. Virology. 1992; 189: 21-30Crossref PubMed Scopus (38) Google Scholar, 46Hwang Y. Feiss M. J. Mol. Biol. 1996; 261: 524-535Crossref PubMed Scopus (36) Google Scholar), we could delimit the endonuclease domain (Endo-Term) of the T5 terminase. Expression of the plasmid-encoded corresponding protein resulted in in vivo cutting of both plasmid and genomic DNA. Such in vivo nuclease activity was considerably reduced by mutating each of the critical conserved aspartate residues (respectively D286N, D342N, and D425N). Further purification of the corresponding expressed proteins confirmed the role played by the aspartate residues and allowed determination of the relative nuclease activity of the proteins: WT Endo-term showed the highest nuclease activity, and the activity of D286N and D425 was about half that of WT Endo-term. Finally, D342 showed the lowest nuclease activity. The defect in DNA cleavage activity was not due to misfolding, because all four proteins retained the same content of secondary structure. The endonuclease activity of the proteins had an absolute requirement for divalent cations (mainly Mg2+ or Mn2+) in agreement with that found for the nuclease domain of phage λ terminase (1Feiss M. Catalano C. Viral Genome Packaging Machines.in: Catalano C. Kluwer Academic / Plenum Publishers, New York2005: 5-39Google Scholar) and for Holliday junction resolvases, among which is the E. coli protein RuvC (47Saito A. Iwasaki H. Ariyoshi M. Morikawa K. Shinagawa H. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 7470-7474Crossref PubMed Scopus (83) Google Scholar). A divalent metal cation is also a cofactor required for the catalytic activity of the proteins belonging to the integrase/ribonuclease-H superfamily (45Nowotny M. Gaidamakov S.A. Crouch R.J. Yang W. Cell. 2005; 121: 1005-1016Abstract Full Text Full Text PDF PubMed Scopus (499) Google Scholar). All these proteins have similar active sites containing three to four catalytic carboxylate-bearing residues (44Rice P. Craigie R. Davies D.R. Curr. Opin. Struct. Biol. 1996; 6: 76-83Crossref PubMed Scopus (180) Google Scholar). It is therefore likely that phage endonucleases and these proteins share a common catalytic mechanism in which the acidic residues are involved in the coordination of the metal required for the cleavage of the phosphodiester bond. Does such a mechanism explain the differences observed in the nuclease activity of the three mutated proteins? According to the model shown in Fig. 2B, Asp-286 and D425N are located in the vicinity of the aspartate 291 residue so that chelation of the metal might occur via this third aspartate if any of the two others is mutated. As a result, some of the nuclease activity would be restored. On the other hand, the nuclease activity would not be restored by mutation of Asp-342 because this residue lies too far away from any other aspartate to permit chelation. Exhaustive mutagenesis experiments have allowed critical amino acids involved in the endonuclease activity of phage T4 and λ terminases to be defined (15Rentas F.J. Rao V.B. J. Mol. Biol. 2003; 334: 37-52Crossref PubMed Scopus (33) Google Scholar, 16Davidson A.R. Gold M. Virology. 1992; 189: 21-30Crossref PubMed Scopus (38) Google Scholar, 17Hwang Y. Catalano C.E. Feiss M. Biochemistry. 1996; 35: 2796-2803Crossref PubMed Scopus (61) Google Scholar). In particular, it was shown that residue Asp-401 (which corresponds to Asp-286 in T5) from phage T4 and λ terminases was absolutely required for in vivo and in vitro DNA cutting activity, but not for in vitro packaging of the phage genome. On the basis of the T5 data and given the high conservation of the three-aspartate motif (among which is Asp-401 from λ and T4) (see Fig. 2A), we propose that the aspartate residues are required, and sufficient, to promote the endonucleotidic activity of phage T4 and λ terminases. Therefore phage endonucleases share the same catalytic activity independently of whether DNA cleavage is sequence specific as in λ or nonspecific as in T4. Altogether, our results suggest that bacteriophage-encoded terminases share a conserved endonuclease domain that is distantly related to the resolvase/integrase/ribonuclease-H superfamily and that behaves as a properly folded and independent domain. The interplay between the nuclease and packaging domains starts to be understood in the case of phage T4 (10Kanamaru S. Kondabagil K. Rossmann M.G. Rao V.B. J. Biol. Chem. 2004; 279: 40795-40801Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar) and λ (7Dhar A. Feiss M. J. Mol. Biol. 2005; 347: 71-80Crossref PubMed Scopus (19) Google Scholar, 8Maluf N.K. Yank Q. Catalano C.E. J. Mol. Biol. 2005; 347: 523-542Crossref PubMed Scopus (51) Google Scholar). The recent sequencing of the T5 genome has allowed identification of the genes encoding the different partners of the encapsidation machinery. The unusual characteristics of the T5 genome (large terminal repeats, presence of genetically defined single-stranded interruptions) and of its mechanism of transfer into E. coli cells (recently reviewed in Ref. 48Ponchon L. Mangenot S. Boulanger P. Letellier L. Biochim. Biophys. Acta. 2005; 1724: 255-261Crossref PubMed Scopus (9) Google Scholar) makes it a new and interesting challenge to define the rules, and exceptions, that govern phage DNA packaging. We thank M. Santamaria for assistance in the terminase constructs, P. Decottignies for protein sequencing, Chantal Janmot for technical assistance, and G. Craescu for access to the circular dichroism apparatus." @default.
• W2040509036 created "2016-06-24" @default.
• W2040509036 creator A5010833952 @default.
• W2040509036 creator A5047238292 @default.
• W2040509036 creator A5066658693 @default.
• W2040509036 creator A5070275531 @default.
• W2040509036 date "2006-03-01" @default.
• W2040509036 modified "2023-10-10" @default.
• W2040509036 title "The Endonuclease Domain of Bacteriophage Terminases Belongs to the Resolvase/Integrase/Ribonuclease H Superfamily" @default.
• W2040509036 cites W1494276408 @default.
• W2040509036 cites W1965104216 @default.
• W2040509036 cites W1966204777 @default.
• W2040509036 cites W1974719543 @default.
• W2040509036 cites W1983419528 @default.
• W2040509036 cites W1985718554 @default.
• W2040509036 cites W1986118467 @default.
• W2040509036 cites W1989228404 @default.
• W2040509036 cites W1997325885 @default.
• W2040509036 cites W1998777238 @default.
• W2040509036 cites W2003003250 @default.
• W2040509036 cites W2004228538 @default.
• W2040509036 cites W2006212265 @default.
• W2040509036 cites W2008708467 @default.
• W2040509036 cites W2011983782 @default.
• W2040509036 cites W2011989198 @default.
• W2040509036 cites W2013257494 @default.
• W2040509036 cites W2015876589 @default.
• W2040509036 cites W2027189612 @default.
• W2040509036 cites W2029887349 @default.
• W2040509036 cites W2034256074 @default.
• W2040509036 cites W2052036458 @default.
• W2040509036 cites W2053389329 @default.
• W2040509036 cites W2062614190 @default.
• W2040509036 cites W2065283382 @default.
• W2040509036 cites W2067871083 @default.
• W2040509036 cites W2069151831 @default.
• W2040509036 cites W2077487491 @default.
• W2040509036 cites W2077511423 @default.
• W2040509036 cites W2080057721 @default.
• W2040509036 cites W2080558450 @default.
• W2040509036 cites W2084466582 @default.
• W2040509036 cites W2088092681 @default.
• W2040509036 cites W2101232776 @default.
• W2040509036 cites W2103471942 @default.
• W2040509036 cites W2106154817 @default.
• W2040509036 cites W2120967544 @default.
• W2040509036 cites W2134082301 @default.
• W2040509036 cites W2134955152 @default.
• W2040509036 cites W2140244239 @default.
• W2040509036 cites W2142529984 @default.
• W2040509036 cites W2143456881 @default.
• W2040509036 cites W2150503843 @default.
• W2040509036 cites W2158714788 @default.
• W2040509036 doi "https://doi.org/10.1074/jbc.m511817200" @default.
• W2040509036 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/16377618" @default.
• W2040509036 hasPublicationYear "2006" @default.
• W2040509036 type Work @default.
• W2040509036 sameAs 2040509036 @default.
• W2040509036 citedByCount "32" @default.
• W2040509036 countsByYear W20405090362012 @default.
• W2040509036 countsByYear W20405090362014 @default.
• W2040509036 countsByYear W20405090362015 @default.
• W2040509036 countsByYear W20405090362017 @default.
• W2040509036 countsByYear W20405090362018 @default.
• W2040509036 countsByYear W20405090362019 @default.
• W2040509036 countsByYear W20405090362020 @default.
• W2040509036 crossrefType "journal-article" @default.
• W2040509036 hasAuthorship W2040509036A5010833952 @default.
• W2040509036 hasAuthorship W2040509036A5047238292 @default.
• W2040509036 hasAuthorship W2040509036A5066658693 @default.
• W2040509036 hasAuthorship W2040509036A5070275531 @default.
• W2040509036 hasBestOaLocation W20405090361 @default.
• W2040509036 hasConcept C104317684 @default.
• W2040509036 hasConcept C105793898 @default.
• W2040509036 hasConcept C10879258 @default.
• W2040509036 hasConcept C141231307 @default.
• W2040509036 hasConcept C14430832 @default.
• W2040509036 hasConcept C185592680 @default.
• W2040509036 hasConcept C2776441376 @default.
• W2040509036 hasConcept C2777028655 @default.
• W2040509036 hasConcept C2778976237 @default.
• W2040509036 hasConcept C2909909245 @default.
• W2040509036 hasConcept C4918238 @default.
• W2040509036 hasConcept C54355233 @default.
• W2040509036 hasConcept C547475151 @default.
• W2040509036 hasConcept C552990157 @default.
• W2040509036 hasConcept C67705224 @default.
• W2040509036 hasConcept C69766542 @default.
• W2040509036 hasConcept C86803240 @default.
• W2040509036 hasConcept C92028520 @default.
• W2040509036 hasConceptScore W2040509036C104317684 @default.
• W2040509036 hasConceptScore W2040509036C105793898 @default.
• W2040509036 hasConceptScore W2040509036C10879258 @default.
• W2040509036 hasConceptScore W2040509036C141231307 @default.
• W2040509036 hasConceptScore W2040509036C14430832 @default.
• W2040509036 hasConceptScore W2040509036C185592680 @default.
• W2040509036 hasConceptScore W2040509036C2776441376 @default.
• W2040509036 hasConceptScore W2040509036C2777028655 @default.

• next