DNA Recognition by the Homing Endonuclease PI-SceI Involves a Divalent Metal Ion Cofactor-induced Conformational Change
2004; Elsevier BV; Volume: 279; Issue: 8 Linguagem: Inglês
10.1074/jbc.m311372200
ISSN1083-351X
AutoresAnn-Josée Noël, Wolfgang Wende, Alfred Pingoud,
Tópico(s)Bacterial Genetics and Biotechnology
ResumoPI-SceI, a homing endonuclease of the LAGLIDADG family, consists of two domains involved in DNA cleavage and protein splicing, respectively. Both domains cooperate in binding the recognition sequence. Comparison of the structures of PI-SceI in the absence and presence of substrate reveals major conformational changes in both the protein and DNA. Notably, in the protein-splicing domain the loop comprising residues 53-70 and adopts a “closed” conformation, thus enabling it to interact with the DNA. We have studied the dynamics of DNA binding and subsequent loop movement by fluorescence techniques. Six amino acids in loop53-70 were individually replaced by cysteine and modified by fluorescein. The interaction of the modified PI-SceI variants with the substrate, unlabeled or labeled with tetramethylrhodamine, was analyzed in equilibrium and stopped-flow experiments. A kinetic scheme was established describing the interaction between PI-SceI and DNA. It is noteworthy that the apparent hinge-flap motion of loop53-70 is only observed in the presence of a divalent metal ion cofactor. Substitution of the major Mg2+-binding ligands in PI-SceI, Asp-218 and Asp-326, by Asn or “nicking” PI-SceI with trypsin at Arg-277, which interferes with formation of an active enzyme·substrate complex, both prevent the conformational change of loop53-70. Deletion of the loop inactivates the enzyme. We conclude that loop53-70 is an important structural element that couples DNA recognition by the splicing domain with DNA cleavage by the catalytic domain and as such “communicates” with the Mg2+ binding sites at the catalytic centers. PI-SceI, a homing endonuclease of the LAGLIDADG family, consists of two domains involved in DNA cleavage and protein splicing, respectively. Both domains cooperate in binding the recognition sequence. Comparison of the structures of PI-SceI in the absence and presence of substrate reveals major conformational changes in both the protein and DNA. Notably, in the protein-splicing domain the loop comprising residues 53-70 and adopts a “closed” conformation, thus enabling it to interact with the DNA. We have studied the dynamics of DNA binding and subsequent loop movement by fluorescence techniques. Six amino acids in loop53-70 were individually replaced by cysteine and modified by fluorescein. The interaction of the modified PI-SceI variants with the substrate, unlabeled or labeled with tetramethylrhodamine, was analyzed in equilibrium and stopped-flow experiments. A kinetic scheme was established describing the interaction between PI-SceI and DNA. It is noteworthy that the apparent hinge-flap motion of loop53-70 is only observed in the presence of a divalent metal ion cofactor. Substitution of the major Mg2+-binding ligands in PI-SceI, Asp-218 and Asp-326, by Asn or “nicking” PI-SceI with trypsin at Arg-277, which interferes with formation of an active enzyme·substrate complex, both prevent the conformational change of loop53-70. Deletion of the loop inactivates the enzyme. We conclude that loop53-70 is an important structural element that couples DNA recognition by the splicing domain with DNA cleavage by the catalytic domain and as such “communicates” with the Mg2+ binding sites at the catalytic centers. Homing endonucleases are the products of selfish DNA elements, mostly present in introns (I-homing endonucleases) 1The abbreviations used are: I, intron-encoded homing endonuclease; PI, protein intein homing endonuclease; FRET, fluorescence resonance energy transfer; TMR, tetramethylrhodamine; LC, lower complex; UC, upper complex; Ctr, Candida tropicalis. or encoded in-frame with a precursor protein as an intein (PI-homing endonucleases) (for review see Refs. 1Galburt E.A. Stoddard B.L. Biochemistry. 2002; 41: 13851-13860Crossref PubMed Scopus (113) Google Scholar, 2Chevalier B.S. Stoddard B.L. Nucleic Acids Res. 2001; 29: 3757-3774Crossref PubMed Scopus (377) Google Scholar, 3Gimble F.S. FEMS Microbiol. Lett. 2000; 185: 99-107Crossref PubMed Google Scholar, 4Jurica M.S. Stoddard B.L. Cell Mol. Life Sci. 1999; 55: 1304-1326Crossref PubMed Scopus (138) Google Scholar, 5Belfort M. Roberts R.J. Nucleic Acids Res. 1997; 25: 3379-3388Crossref PubMed Scopus (397) Google Scholar). They are found in prokarya, archaea, and eukarya and promote the spread of their genes by catalyzing a double strand break at alleles in which the mobile element is absent. This double strand break is subsequently repaired by homologous recombination thereby integrating the homing endonuclease gene and completing the homing process. A characteristic of homing endonucleases is their extreme specificity, recognizing sequences of up to 40 base pairs in length. In this respect they differ from restriction endonucleases, whose target sequences consist of not more than 8 bp (6Pingoud A. Jeltsch A. Nucleic Acids Res. 2001; 29: 3705-3727Crossref PubMed Scopus (489) Google Scholar). Homing endonucleases are similar to restriction endonucleases in that they require Mg2+ ions for cleavage to generate DNA fragments with a 5′-phosphate and a 3′-hydroxyl, however, they are structurally unrelated. They can be divided into four families, characterized by common sequence motifs: the LAGLIDADG, His-Cys-box, H-N-H, and GIY-YIG. The largest of these is the LAGLIDADG family. Homing endonucleases of this family contain two LAGLIDADG motifs. They either function as homodimers with one LAGLIDADG motif per polypeptide chain, e.g. I-CreI and I-MsoI (7Jurica M.S. Monnat R.J.J. Stoddard B.L. Mol. Cell. 1998; 2: 469-476Abstract Full Text Full Text PDF PubMed Scopus (131) Google Scholar, 8Chevalier B. Turmel M. Lemieux C. Monnat R.J.J. Stoddard B.L. J. Mol. Biol. 2003; 329: 253-269Crossref PubMed Scopus (91) Google Scholar), or as monomers with two motifs per polypeptide chain, e.g. PI-SceI, PI-PfuI, and I-DmoI (9Duan X. Gimble F.S. Quiocho F.A. Cell. 1997; 89: 555-564Abstract Full Text Full Text PDF PubMed Scopus (234) Google Scholar, 10Komori K. Fujita N. Ichiyanagi K. Shinagawa H. Morikawa K. Ishino Y. Nucleic Acids Res. 1999; 27: 4167-4174Crossref PubMed Scopus (29) Google Scholar, 11Silva G.H. Dalgaard J.Z. Belfort M. Van Roey P. J. Mol. Biol. 1999; 286: 1123-1136Crossref PubMed Scopus (86) Google Scholar). Both homodimeric and monomeric LAGLIDADG enzymes have two catalytic centers that cooperate in cleavage of the two strands of their recognition sequence (12Christ F. Schoettler S. Wende W. Steuer S. Pingoud A. Pingoud V. EMBO J. 1999; 18: 6908-6916Crossref PubMed Scopus (41) Google Scholar). The active sites of these enzymes are characterized by a 2-fold (homodimeric enzymes) or pseudo 2-fold (monomeric enzymes) symmetry axis between the two LAGLIDADG helices. They form a two-helix bundle at the interface of the two subunits of the homodimeric LAGLIDADG enzymes or the two subdomains of the monomeric LAGLIDADG enzymes. At the C-terminal end of the LAGLIDADG helix a carboxylate is invariably present whose function is to bind the divalent metal ion cofactor required for catalysis (13Gimble F.S. Stephens B.W. J. Biol. Chem. 1995; 270: 5849-5856Abstract Full Text Full Text PDF PubMed Scopus (97) Google Scholar, 14Schoettler S. Wende W. Pingoud V. Pingoud A. Biochemistry. 2000; 39: 15895-15900Crossref PubMed Scopus (20) Google Scholar). Within the LAGLIDADG family both, intron-encoded and intein homing endonucleases occur. The latter are more complex than the former, because they have an additional protein splicing activity. Different from the I-homing endonucleases for which structural information is available (I-CreI, its homolog I-MsoI, and I-DmoI), the PI-homing endonucleases (PI-SceI and PI-PfuI) have an extra domain that harbors the catalytic center for protein splicing (9Duan X. Gimble F.S. Quiocho F.A. Cell. 1997; 89: 555-564Abstract Full Text Full Text PDF PubMed Scopus (234) Google Scholar, 10Komori K. Fujita N. Ichiyanagi K. Shinagawa H. Morikawa K. Ishino Y. Nucleic Acids Res. 1999; 27: 4167-4174Crossref PubMed Scopus (29) Google Scholar). This explains the different sizes of the I- and PI-homing endonucleases. In addition, this domain is involved in DNA recognition. For PI-SceI it was shown that the isolated splicing domain (domain I) is capable of strong and specific binding to the recognition sequence as well as to a fragment corresponding to the downstream product obtained upon cleavage of the recognition sequence (15Grindl W. Wende W. Pingoud V. Pingoud A. Nucleic Acids Res. 1998; 26: 1857-1862Crossref PubMed Scopus (34) Google Scholar). In contrast, the isolated endonucleolytic domain (domain II) binds only weakly and non-specifically to DNA and is catalytically inactive. Cleavage activity is not restored by addition of domain I to domain II. Furthermore, neither domain I or II, by itself or in combination, is capable of strong DNA distortion observed for intact PI-SceI in electrophoretic mobility shift assays upon specific complex formation, both in the absence and presence of divalent metal ions. DNA bending occurs in two steps with apparent bending angles of 45° and 75°, respectively (16Wende W. Grindl W. Christ F. Pingoud A. Pingoud V. Nucleic Acids Res. 1996; 24: 4123-4132Crossref PubMed Scopus (49) Google Scholar, 17Gimble F.S. Wang J. J. Mol. Biol. 1996; 263: 163-180Crossref PubMed Scopus (81) Google Scholar). Bending serves two purposes: (i) to accommodate a recognition sequence of ∼35 bp on the surface of a globular protein and (ii) to position the scissile phosphodiester bonds vis-à-vis the catalytic centers. The recently determined structure of a PI-SceI·DNA complex (18Moure C.M. Gimble F.S. Quiocho F.A. Nat. Struct. Biol. 2002; 9: 764-770Crossref PubMed Scopus (84) Google Scholar) confirms (and of course extends with detailed information) models for the interaction between PI-SceI and its DNA substrate (19Christ F. Steuer S. Thole H. Wende W. Pingoud A. Pingoud V. J. Mol. Biol. 2000; 300: 867-875Crossref PubMed Google Scholar, 20Hu 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), proposed on the basis of footprinting, interference, and cross-linking experiments as well as mutagenesis studies for both the protein and the DNA (17Gimble F.S. Wang J. J. Mol. Biol. 1996; 263: 163-180Crossref PubMed Scopus (81) Google Scholar, 19Christ F. Steuer S. Thole H. Wende W. Pingoud A. Pingoud V. J. Mol. Biol. 2000; 300: 867-875Crossref PubMed Google Scholar, 20Hu 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, 21Pingoud V. Thole H. Christ F. Grindl W. Wende W. Pingoud A. J. Biol. Chem. 1999; 274: 10235-10243Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar, 22Hu D. Crist M. Duan X. Gimble F.S. Biochemistry. 1999; 38: 12621-12628Crossref PubMed Scopus (21) Google Scholar, 23Posey K.L. Gimble F.S. Biochemistry. 2002; 41: 2184-2190Crossref PubMed Scopus (18) Google Scholar). According to the structure analysis (18Moure C.M. Gimble F.S. Quiocho F.A. Nat. Struct. Biol. 2002; 9: 764-770Crossref PubMed Scopus (84) Google Scholar), the splicing domain contacts ∼17 bp and the endonuclease domain ∼15 bp. Two discrete DNA distortions are observed for the DNA region bound to the protein-splicing domain (22°) and for the stretch of DNA bound to the endonuclease domain (50°). In addition to many phosphate backbone contacts, base-specific contacts are made to 11 of the 31 bp of the minimal recognition sequence. For these contacts to be formed, substantial conformational changes of both enzyme and substrate are required. A comparison of the structures of the apoenzyme and the complex shows that three major conformational changes are evident for the induced fit of enzyme and substrate (Fig. 1): (i) the loop comprising residues 53-70 of domain I, which is located at the interface between the two domains, moves toward the DNA upon complex formation; (ii) the β-hairpin loop comprising residues 369-375 of domain II undergoes a hinge-flap motion relative to the ligand-free structure, and (iii) the 269-284 loop located at the extreme of the endonuclease domain (domain II) becomes ordered in the ligand-bound structure. Incidentally, cleavage of this loop by trypsin at position 277 inactivates the enzyme (24Pingoud V. Grindl W. Wende W. Thole H. Pingoud A. Biochemistry. 1998; 37: 8233-8243Crossref PubMed Scopus (13) Google Scholar). These conformational changes of the protein are required for binding, bending, and recognition of the DNA as well as coupling of recognition and catalysis. This is of particular importance for a protein like PI-SceI, which utilizes for recognition a domain remote from the catalytic center. In this work, we sought to study the dynamics of the interaction between PI-SceI and its DNA substrate with emphasis on the conformational transition from an open to a closed conformation of the loop comprising residues 53-70. This loop was shown to be essential for DNA cleavage by PI-SceI (25Wende W. Schottler S. Grindl W. Christ F. Steuer S. Noel A.J. Pingoud V. Pingoud A. Mol. Biol. (Moscow). 2000; 34: 1054-1064Crossref PubMed Scopus (1) Google Scholar, 26Steuer S. Pingoud V. Pingoud A. Wende W. Chembiochem. 2003; (in press)Google Scholar), presumably because it couples recognition by domain I and cleavage by domain II. For this purpose we have introduced a single fluorescein group at six different positions in the tip of the loop and a single tetramethylrhodamine group at three different positions of the target DNA. Changes in fluorescence as well as fluorescence resonance energy transfer (FRET) that occur during DNA binding by PI-SceI were used to determine the equilibrium and the on/off rate constants. We also determined rate constants for a conformational change of the loop, which we interpret as a movement that leads to the insertion of the loop into the minor groove of the DNA. Interestingly, although the on and off rates of DNA binding by PI-SceI are similar in the absence or presence of divalent metal ions, the conformational change of the loop requires the presence of divalent metal ions bound at the active site and is only seen when PI-SceI is able to form an active enzyme·substrate complex. Thus, our results suggest a straightforward mechanism for communication between domains I and II, which is mediated by divalent metal ion binding to the active site in domain II and loop53-70 in domain I closing onto the DNA. Mutagenesis and Purification of the PI-SceI Variants—Site-directed mutagenesis to create the variant 5C/S (C1S, C17S, C249S, C398S, and C416S) was performed essentially as described by Kirsch and Joly (27Kirsch R.D. Joly E. Nucleic Acids Res. 1998; 26: 1848-1850Crossref PubMed Scopus (205) Google Scholar). A single cysteine residue was introduced at positions Arg-57, Ala-58, His-59, Lys-60, Ser-61, or Ser-64, respectively, using the 5C/S construct as a template to generate the variants 5C/S+R57C, 5C/S+A58C, etc. For control measurements, the Cys-17 residue was retained in the variant 5C/S-C17. Finally, the variants 5C/S+K60C+D218N and 5C/S+K60C+D218N+D326N were created to study the effects of metal ion binding. All PI-SceI variants were expressed and purified as described by Wende et al. (16Wende W. Grindl W. Christ F. Pingoud A. Pingoud V. Nucleic Acids Res. 1996; 24: 4123-4132Crossref PubMed Scopus (49) Google Scholar) from JM109 cells containing the pREP4 plasmid. After elution from the Ni-nitrilotriacetic acid affinity column, the fractions containing PI-SceI protein were pooled and extensively dialyzed at 4 °C against storage buffer (10 mm Hepes-KOH, pH 7.0, 200 mm KCl, and 50% (v/v) glycerol). Oligodeoxynucleotides—Oligodeoxynucleotides were purchased from MWG Biotech (Ebersberg, Germany) and Thermo Electron (Ulm, Germany). Specific Labeling of PI-SceI with Fluorescein Maleimide—In preparation for the labeling reactions, a 100 μm solution of the PI-SceI variant in storage buffer was incubated with 10 mm dithiothreitol at 25 °C for 15 min. After overnight dialysis at 4 °C in a thoroughly degassed labeling buffer (20 mm Hepes-KOH, pH 7.0, 150 mm KCl, and 10% (v/v) glycerol), the protein concentration was adjusted to 50 μm and 0.1 mm TCEP-HCl (Pierce, Rockford) was added to keep the cysteine residue in the reduced state. The labeling reaction to modify the single cysteine in the DNA binding loop was initiated by the addition of 150 μm fluorescein-5-maleimide (Molecular Probes, Eugene, OR) or 300 μm for the variant Cys-17. The reaction was carried out in the dark for 2 h at ambient temperature. A control for the specificity of labeling was carried out in parallel with the variant 5C/S (which does not have a solvent-exposed Cys residue) under the same conditions. Removal of free dye was achieved by gel filtration using a Sephadex G-25 column (Amersham Biosciences). After addition of 50% (v/v) glycerol, the final concentrations of the purified labeled variants varied from 2 to 12 μm. Five μg of the labeled proteins was electrophoresed on a 15% SDS-PAGE and subsequently visualized by both fluorescence using UV transillumination (365 nm) and Coomassie Blue staining. Determination of the Efficiency of Labeling of PI-SceI—The total fluorescein (covalently or non-covalently bound) concentration in the protein preparation was evaluated following digestion of 100 nm of the labeled proteins with 0.1 mg/ml proteinase K (Merck, Darmstadt, Germany) in (20 mm Hepes-KOH, pH 8.5, 150 mm KCl, 10% (v/v) glycerol) at 50 °C for 1 h. A standard curve of the relative fluorescence as a function of the fluorescein concentration was determined under the same conditions. The total fluorescein concentration was calculated from the emission spectrum recorded with a F-4500 Hitachi Fluorescence Spectrophotometer at an excitation wavelength of 497 nm. The non-covalently bound fluorescein concentration was assessed by precipitating 10 μl of the purified proteins with 10% trichloroacetic acid in a final volume of 150 μl. After centrifugation, the supernatant was adjusted to pH 9 using a 2 m Tris solution, and the emission spectra were recorded as described above. A standard curve was obtained under the same conditions, and the percentage of labeling was calculated by subtracting the free fluorescein concentration from the total fluorescein concentration of each sample relative to the protein concentration. Determination of DNA Binding and Cleavage by PI-SceI—Cleavage assays, using the supercoiled plasmid pBSVDEX or the 47-bp oligonucleotides as substrates, were performed in the presence of 2.5 mm MgCl2 or MnCl2 as described by Wende et al. (16Wende W. Grindl W. Christ F. Pingoud A. Pingoud V. Nucleic Acids Res. 1996; 24: 4123-4132Crossref PubMed Scopus (49) Google Scholar). The gel-shift assays were performed at equimolar concentrations (15 nm) of PI-SceI and a 311-bp PCR fragment, in the presence of either 2.5 mm CaCl2 or 1 mm EDTA, also as described (16Wende W. Grindl W. Christ F. Pingoud A. Pingoud V. Nucleic Acids Res. 1996; 24: 4123-4132Crossref PubMed Scopus (49) Google Scholar). However, instead of nonfat dry milk, acetylated bovine serum albumin (Sigma) at 0.1 mg/ml was used in the binding buffer and loading buffer was not added. Fluorescence Equilibrium Measurements—All fluorescence measurements, unless otherwise noted, were carried out in 20 mm Hepes-KOH, pH 8.5, 150 mm KCl, 10% (v/v) glycerol, and 0.1 mg/ml acetylated bovine serum albumin, in the presence or absence of 2.5 mm Ca2+ using the F-4500 Hitachi fluorescence spectrophotometer. The excitation wavelength was set at 497 nm (5.0-nm slit width) and the emission (2.5-nm slit width) was recorded from 475 to 625 or 650 nm, depending on whether fluorescein-labeled proteins and unlabeled DNA were used or FRET experiments with fluorescein-labeled proteins and tetramethylrhodamine (TMR)-labeled DNA were performed, respectively. In all cases, the differences in fluorescein emission intensity of 100 nm of labeled variants were compared before and after addition of an excess (500 nm) of the different oligonucleotide substrates to ensure that all the protein was bound to substrate. The change in fluorescein emission from 510 to 625 nm of the labeled variants was calculated by comparing the emission spectra of the protein before and after addition of unlabeled 47-bp substrate (Sub 0 (for definition see Fig. 2)) either in the presence or absence of 2.5 mm CaCl2. Results are expressed in percent emission change relative to the emission of labeled protein alone; error bars represent the standard deviation for three individual data sets. Similar measurements were performed using both unlabeled nonspecific substrate and with domain I ligand (for definition see Fig. 2). A similar procedure was used for the FRET experiments with Sub 2 (for definition see Fig. 2), except that the percent quenching of fluorescein was calculated for emission in the range of 510-540 nm, i.e. well separated from the emission spectrum of TMR. The distance between the fluorescent donor/acceptor (fluorescein/TMR) in the protein, and the internally labeled DNA substrate, respectively, was approximated using Förster's equation, R=R0(1/E-1)1/8(Eq. 1) where R0 is the distance for 50% transfer efficiency E. The R0 was taken as being 46 Å, which is an average value for this Förster pair (28Wu P. Brand L. Anal. Biochem. 1994; 218: 1-13Crossref PubMed Scopus (1148) Google Scholar). The efficiency of fluorescence energy transfer (E) was calculated as the fractional decrease of the PI-SceI donor fluorescence due to binding of the TMR-labeled acceptor, E=1-FDA/FD(Eq. 2) where FDA and FD are the relative fluorescence yield of the donor in the presence and absence of the acceptor, respectively. The equilibrium constant for Ca2+ binding in the PI-SceI·DNA complex was determined by titration experiments using the labeled variant 5C/S+K60C (enzymatically active) or 5C/S+K60C+D218N+D326N (inactive) in complex with Sub 0. The change in fluorescein emission was recorded as described above. Fluorescence Titrations with the Labeled 5C/S+K60C Variant Subjected to Limited Tryptic Digestion—The tryptic digestion of PI-SceI was carried out similarly as described previously (24Pingoud V. Grindl W. Wende W. Thole H. Pingoud A. Biochemistry. 1998; 37: 8233-8243Crossref PubMed Scopus (13) Google Scholar). Labeled 5C/S+K60C (9.6 μm) in stock buffer was digested using a substrate:protease ratio of 150:1 (w/w) at 25 °C. Aliquots of the trypsinization reaction mixture were withdrawn after the time intervals indicated, and the reaction was terminated with 50 μg/ml trypsin inhibitor from bovine lung (Serva Biochemicals, Heidelberg, Germany). 1 μg of the aliquots was analyzed on 15% (w/v) SDS-PAGE; after silver staining of the gel, the relative amount of the uncleaved PI-SceI toward the trypsin product (N-Ter domain and C-Ter domain) was evaluated using the BioDoc-Analyze 1.0 software from Biometra. The fluorescein emission increase upon Ca2+ addition to the complex consisting of trypsinized 5C/S+K60C, and Sub 0 complex was measured as described above using concentrations of 40 and 200 nm of labeled 5C/S+K60C and Sub 0, respectively. Fluorescence Stopped-flow Measurements—Time-resolved fluorescence measurements (for an overview of the experiments see Table II) were carried out using a Bio-Logic SFM-300 stopped-flow instrument (Claix, France) operated at 25 °C. Changes in fluorescein emission, using a 436-nm excitation wavelength, were monitored at wavelengths above 515 nm using a cut-off filter. The change in FRET efficiency was measured by the change in the emission of TMR and was recorded at wavelengths above 570 nm utilizing a cut-off filter. The fluorescence change of TMR was recorded with an excitation wavelength of 546 nm using the same filter. Using two 20-ml syringes filled to 3.2 ml, equal volumes of the reactants (100 μl) were rapidly combined in the mixing chamber with a dead time of 3.1 ms. All concentrations specified are for the final combined reactants in the mixing chamber, unless otherwise stated. A minimum of 15-30 single curves per experiment were averaged using the Biokine 2.07 software, and all the curve fits were computed using the Solver accessory furnished with Microsoft Excel®.Table IIOverview of the fluorescence stopped-flow experiments with PI-SceI 5C/S variantsExp.Syringe 1Syringe 2Signal observedRate constantIS61C-F1 + Ca2+47-mer + Ca2+Fluorescein ↑konII-C17-F1 + Ca2+Sub 2 + Ca2+FRET (TMR ↑)konIIIS61C + EDTASub 3 + EDTATMR ↓konIVS61C-F1 + 47-mer + Ca2+S61C + Ca2+Fluorescein ↓koff, kopen+Ca2+VS61C + Sub 3 + EDTA47-mer + EDTATMR ↑koffVIS61C-F1 + Sub 1Ca2+FRET (TMR ↑)kclose, kconf.chVIIS61C-F1 + 47-mer + Ca2+EGTAFluorescein ↓kopen-Ca2+ Open table in a new tab The association rate constant for DNA binding by PI-SceI in the presence of 5 mm CaCl2 was determined for the labeled 5C/S+S61C variant by monitoring the change in fluorescein emission obtained using Sub 0. Three data sets of increasing equimolar concentrations of substrate and protein were analyzed: 17.5, 35, and 50 nm. In a similar manner, the association rate constant for DNA binding by the labeled control variant 5C/S-C17 was measured with Sub 2 by recording the change in TMR emission using FRET obtained after mixing equimolar concentrations (12.5, 37.5, and 100 nm) of substrate and protein. DNA binding of PI-SceI in the absence of CaCl2 was measured by monitoring the change in TMR emission of Sub 3 in equimolar concentration (35 and 70 nm) with unlabeled 5C/S+S61C in the presence of 1 mm EDTA. Global fitting of the curves obtained at different concentrations used for association experiments was computed using the equation, dF/dt=d[ES]/dt=kon[E][S] where F is fluorescence signal, ES is enzyme·DNA complex, E is enzyme, S is substrate, and t is time. Analysis of the dissociation of the protein·DNA complex consisting of labeled 5C/S+S61C variant and Sub 0 in the presence of Ca2+, was carried out by mixing preformed specific complex at 17.5, 25, or 35 nm, respectively, with 350 nm of the unlabeled 5C/S+S61C variant, and recording the change in fluorescein emission. The best global curve fit was obtained using the biexponential equation, dF/dt=Aoffexp[-kofft]+Aopenexp[-kopen+Ca2+t] where Aoff and Aopen are the amplitudes for two processes and koff and kopen+Ca2+ are the rate constants for dissociation and a conformational change, respectively (see “Results”). The Kd was calculated using Equation 5. Kd=koff/kon(Eq. 5) To analyze the protein·DNA complex dissociation in the absence of Ca2+, either 100 or 50 nm of the preformed complex consisting of unlabeled 5C/S+S61C variant and Sub 3 was mixed with 500 nm of Sub 0, and the change in TMR emission was recorded. On average, 10 curves were globally fit with a single first order constant (Equation 6). dF/dt=Aoffexp[-kofft](Eq. 6) The divalent metal ion cofactor-induced conformational changes were measured by mixing the specific complex formed by 35 nm of the labeled 5C/S+S61C variant and a 2-fold excess of Sub 1 (Fig. 2) in the first syringe with varying CaCl2 concentrations in the second syringe (0.5, 1, and 10 mm, respectively). At 0.5 mm, the final concentration is thus 0.25 mm and corresponds to the Kd value determined in titration experiments with CaCl2. The transients obtained by recording the FRET between fluorescein and TMR, were globally fit using Equation 7. dF/dt=Acloseexp[-kcloset]+A2exp[-kconf.cht](Eq. 7) The open conformation of the loop, which we define as the movement of the loop away from the DNA, was induced in the absence of Ca2+ by mixing the complex composed of the labeled 5C/S+S61C variant and Sub 0 (17.5, 12.5, and 8.75 nm, respectively) formed in the presence of 5 mm Ca2+ with 20 mm EGTA. Changes in fluorescein emission were recorded and a global curve fitting was carried out using Equation 8. dF/dt=Aoffexp[-kopen-Ca2+t](Eq. 8) A control experiment was performed under the same conditions using the nonspecific substrate. Generation and Characterization of Fluorescently Labeled PI-SceI Variants and Substrates—To study the dynamics of the interaction of PI-SceI with its DNA substrate, fluorescence techniques were used. For this purpose, PI-SceI was labeled with fluorescein and DNA was labeled with tetramethylrhodamine (TMR) (Fig. 2). To produce singly and specifically labeled proteins with fluorescein maleimide, site-directed mutagenesis was used to exchange the five exposed cysteine residues (Cys-1, Cys-17, Cys-249, Cys-398, and Cys-416) of PI-SceI with serine resulting in the variant 5C/S. A single buried cysteine (Cys-75) that previously had been demonstrated to be unreactive with thiol-specific reagents (20Hu 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) was not changed. Using 5C/S as a base construct, a single cysteine residue was introduced in the DNA binding loop (comprising residues 53-70, cyan and yellow in Fig. 2A), at positions where protein/DNA interactions should be monitored: Arg-57, Ala-58, His-59, Lys-60, Ser-61, and Ser-64 (yellow in Fig. 2A). For control experiments, the 5C/S to S17C reversion was generated, because this residue is located at the surface of the protein without being part of a flexible structural element and is distal to the DNA binding site (yellow ball-and- stick representation in Fig. 2A). The protein labeling efficiency with fluorescein maleimide varied from 70 to 100% with a free fluorophore contamination of about 3%. As FRET partners for the fluorescein labeled variants, a 47-bp oligonucleotide was singly labeled at three different positions with TMR (Fig. 2B). The positions of the TMR labels in Sub 1 and Sub 3 (Fig. 2B) we
Referência(s)