Conformational Mobility in the Active Site of a Heme Peroxidase
2006; Elsevier BV; Volume: 281; Issue: 34 Linguagem: Inglês
10.1074/jbc.m602602200
ISSN1083-351X
AutoresSandip K. Badyal, Michael Joyce, Katherine H. Sharp, Harriet E. Seward, Martin Mewies, Jaswir Basran, Isabel Macdonald, P.C.E. Moody, Emma Lloyd Raven,
Tópico(s)Hemoglobin structure and function
ResumoConformational mobility of the distal histidine residue has been implicated for several different heme peroxidase enzymes, but unambiguous structural evidence is not available. In this work, we present mechanistic, spectroscopic, and structural evidence for peroxide- and ligand-induced conformational mobility of the distal histidine residue (His-42) in a site-directed variant of ascorbate peroxidase (W41A). In this variant, His-42 binds "on" to the heme in the oxidized form, duplicating the active site structure of the cytochromes b but, in contrast to the cytochromes b, is able to swing "off" the iron during catalysis. This conformational flexibility between the on and off forms is fully reversible and is used as a means to overcome the inherently unreactive nature of the on form toward peroxide, so that essentially complete catalytic activity is maintained. Contrary to the widely adopted view of heme enzyme catalysis, these data indicate that strong coordination of the distal histidine to the heme iron does not automatically undermine catalytic activity. The data add a new dimension to our wider appreciation of structure/activity correlations in other heme enzymes. Conformational mobility of the distal histidine residue has been implicated for several different heme peroxidase enzymes, but unambiguous structural evidence is not available. In this work, we present mechanistic, spectroscopic, and structural evidence for peroxide- and ligand-induced conformational mobility of the distal histidine residue (His-42) in a site-directed variant of ascorbate peroxidase (W41A). In this variant, His-42 binds "on" to the heme in the oxidized form, duplicating the active site structure of the cytochromes b but, in contrast to the cytochromes b, is able to swing "off" the iron during catalysis. This conformational flexibility between the on and off forms is fully reversible and is used as a means to overcome the inherently unreactive nature of the on form toward peroxide, so that essentially complete catalytic activity is maintained. Contrary to the widely adopted view of heme enzyme catalysis, these data indicate that strong coordination of the distal histidine to the heme iron does not automatically undermine catalytic activity. The data add a new dimension to our wider appreciation of structure/activity correlations in other heme enzymes. The heme peroxidase enzymes catalyze H2O2-dependent oxidation of a range of substrates through a mechanism that, in all cases, involves formation of an oxidized ferryl intermediate (known as Compound I; see Equation 1) that is subsequently reduced by substrate (Equations 2 and 3) (1Everse J. Everse K.E. Grishham M.B. Peroxidases in Chemistry and Biology. Vol. 1 and 2. CRC Press, Boca Raton, FL1991Google Scholar, 2Dunford H.B. Heme Peroxidases. John Wiley, Chichester, United Kingdom1999Google Scholar). In the majority of cases, reduction of Compound I occurs by two successive single-electron transfer steps, as follows (where P = peroxidase, HS = substrate, S· = 1-electron oxidized form of substrate). P+H2O2k1→Compound I+H2O(Eq. 1) Compound I+HSk2→Compound II+S⋅ (Eq. 2) Compound II+HSk3→P+S⋅+H2O(Eq. 3) Structural information is available for a number of heme peroxidase enzymes, and in all cases the heme iron is poised in a 5- or 6-coordinate environment with the sixth ligand provided by a weakly coordinated water molecule. This differs from the heme coordination geometry in other noncatalytic heme proteins that do not require binding of an exogenous ligand at the metal site. The classic example is the cytochromes, which typically have a strong endogenous protein ligand at the sixth site (usually His or Met) and no vacant coordination site for iron-catalyzed chemistry to occur. The prevailing view that has emerged, therefore, is that the catalytic enzymes (which include the heme peroxidases but also embraces other, more complex heme enzymes such as heme oxygenase, the cytochrome P450s, and cytochrome c oxidase) usually contain 5-coordinate or weakly 6-coordinate heme groups that allow facile reaction with substrate, whereas the electron transfer proteins, for example the cytochromes, have no vacant site at the metal ion for catalysis to occur. In line with the above considerations, there are no known examples of a genuine heme peroxidase with bis-histidine ligation, but there are a few examples in the literature of different heme peroxidases, or site-directed variants thereof, in which coordination of the distal histidine residue has been proposed on the basis of spectroscopic studies (3Cheek J. Mandelman D. Poulos T.L. Dawson J.H. J. Biol. Inorg. Chem. 1999; 4: 64-72Crossref PubMed Scopus (35) Google Scholar, 4Sutherland G.R.J. Zapanta S. Tien M. Aust S.D. Biochemistry. 1997; 36: 3654-3662Crossref PubMed Scopus (69) Google Scholar, 5Smulevich G. Miller M.A. Kraut J. Spiro T.G. Biochemistry. 1991; 30: 9546Crossref PubMed Scopus (130) Google Scholar, 6Vitello L.B. Erman J.E. Miller M.A. Mauro J.M. Kraut J. Biochemistry. 1992; 31: 11524-11535Crossref PubMed Scopus (74) Google Scholar, 7Turano P. Ferrer J.C. Cheesman M.R. Thomson A.J. Banci L. Bertini I. Mauk A.G. Biochemistry. 1995; 34: 13895-13905Crossref PubMed Scopus (31) Google Scholar, 8Youngs H.L. Moenne-Loccoz P. Loehr T.M. Gold M.H. Biochemistry. 2000; 39: 9994-10000Crossref PubMed Scopus (18) Google Scholar). These examples include: the W51A (7Turano P. Ferrer J.C. Cheesman M.R. Thomson A.J. Banci L. Bertini I. Mauk A.G. Biochemistry. 1995; 34: 13895-13905Crossref PubMed Scopus (31) Google Scholar) and D235N (6Vitello L.B. Erman J.E. Miller M.A. Mauro J.M. Kraut J. Biochemistry. 1992; 31: 11524-11535Crossref PubMed Scopus (74) Google Scholar) variants of cytochrome c peroxidase; thermally inactivated manganese peroxidase (4Sutherland G.R.J. Zapanta S. Tien M. Aust S.D. Biochemistry. 1997; 36: 3654-3662Crossref PubMed Scopus (69) Google Scholar); and manganese peroxidase at alkaline pH (8Youngs H.L. Moenne-Loccoz P. Loehr T.M. Gold M.H. Biochemistry. 2000; 39: 9994-10000Crossref PubMed Scopus (18) Google Scholar). Removal of the ligands coordinating to the bound K+-site in ascorbate peroxidase also leads to formation of a low-spin species (3Cheek J. Mandelman D. Poulos T.L. Dawson J.H. J. Biol. Inorg. Chem. 1999; 4: 64-72Crossref PubMed Scopus (35) Google Scholar). In none of these cases has unambiguous structural information been obtained, however. In this work, we present the first crystallographically defined example of a functional peroxidase enzyme with bis-histidine ligation in the W41A variant of ascorbate peroxidase. This variant duplicates the heme coordination geometry of the cytochromes b in the oxidized form but remains fully competent for formation of the catalytic Compound I and Compound II intermediates, as well as for substrate oxidation, by means of a reaction mechanism in which a conformationally mobile ligand (His-42) binds "on" and then swings "off" the iron during catalysis. This switch between the on and off forms is triggered by reaction with hydrogen peroxide (or other ligands) and, under catalytic conditions, is fully reversible, allowing essentially complete activity to be maintained. These data indicate that strong coordination of the distal histidine residue to the heme iron does not automatically undermine peroxidase activity. Materials—l-Ascorbic acid (Aldrich), guaiacol (Sigma), and the chemicals used for buffers (Fisher) were of the highest analytical grade (more than 99% pure) and were used without further purification. Hydrogen peroxide solutions were freshly prepared by dilution of a 30% (v/v) solution (BDH Chemicals); exact concentrations were determined using the published absorption coefficient (ϵ240 = 39.4 m1- cm-1 (9Nelson D.P. Kiesow L.A. Anal. Biochem. 1972; 49: 474-478Crossref PubMed Scopus (826) Google Scholar)). Aqueous solutions were prepared using water purified through an Elgastat Option 2 water purifier, which itself was fed with deionized water. All pH measurements were made using a Russell pH-electrode attached to a digital pH-meter (Radiometer Copenhagen, model PHM 93). Mutagenesis and Protein Purification—Site-directed mutagenesis on recombinant soybean cytosolic APX (rsAPX) 2The abbreviations used are: APX, ascorbate peroxidase; rsAPX, recombinant soybean cytosolic ascorbate peroxidase; sh, shoulder; MALDI, matrix-assisted laser desorption ionization; TOF, time-of-flight. was performed according to the QuikChange™ protocol (Stratagene Ltd., Cambridge, UK). Two oligonucleotides encoding the desired mutation were synthesized and purified (Invitrogen). For W41A, the primers were: 5′-GCTCCGTTTGGCAGCGCACTCTGCTGGAACC-3′ (forward primer) and 3′-GGTTCCAGCAGAGTGCGCTGCCAAACGGAGC-5′ (reverse primer). DNA sequencing of the entire coding region using an Applied Biosystems 3730 DNA analyzer was used to confirm the desired mutation and the absence of spurious mutations. Bacterial fermentation of cells and purification of rsAPX and W41A was carried out according to published procedures (10Metcalfe C.L. Ott M. Patel N. Singh K. Mistry S.C. Goff H.M. Raven E.L. J. Am. Chem. Soc. 2004; 126: 16242-16248Crossref PubMed Scopus (40) Google Scholar, 11Lad L. Mewies M. Raven E.L. Biochemistry. 2002; 41: 13774-13781Crossref PubMed Scopus (72) Google Scholar). Enzyme purity was assessed by examination of the Asoret/A280 value; in all cases an Asoret/A280 value of >1.7 for rsAPX and W41A was considered pure. Enzyme purity was also assessed using SDS-PAGE, and the preparations were judged to be homogeneous by the observation of a single band on a Coomassie Blue-stained reducing SDS-polyacrylamide gel. Enzyme concentrations for the W41A mutant were determined using the pyridine hemochromagen method (12Antonini M. Brunori E. Hemoglobin and Myoglobin and Their Reactions with Ligands. North Holland Publishers, Amsterdam1971Google Scholar); an absorption coefficient of ϵ405 = 125 mm-1 cm-1 was determined for W41A. Enzyme concentrations of rsAPX were determined using the ϵ407 = 107 mm-1 cm-1 (13Jones D.K. Dalton D.A. Rosell F.I. Lloyd Raven E. Arch. Biochem. Biophys. 1998; 360: 173-178Crossref PubMed Scopus (40) Google Scholar). Mass Spectrometry—The integrity of the tryptophan mutant was confirmed by MALDI-TOF mass spectrometry. A 10 pmol/μl stock of the W41A mutant was made up in water and 0.5 μl of the 1:1 protein/matrix mixture (sinapinic acid (5 mg/ml) in 1:1 acetonitrile/water, 0.1% trifluoroacetic acid) was spotted onto a MALDI target plate using the drying droplet method. The MALDI-TOF mass spectrometer (Applied Biosystems) was calibrated in the range of 5000 to 35,000 Da with a protein mass standard kit (Sequazyme, Applied Biosystems). Spectra were accumulated in the same mass range using an average of at least 250 laser shots. The spectra were analyzed using Data Explorer software (Applied Biosystems). The MALDI-TOF mass spectrum of W41A gives a mass of 28,209.36 ± 0.05% Da (calculated mass for W41A = 28,203.74 Da), indicating that no posttranslational modification has occurred. Steady-state Kinetics—Steady-state measurements (sodium phosphate, pH 7.0, μ = 0.1 m, [enzyme] = 25 nm, 25 °C) were carried out according to published protocols (14Mittler R. Zilinskas B.A. Plant Physiol. 1991; 97: 962-968Crossref PubMed Scopus (189) Google Scholar). Oxidation of ascorbate was monitored at 290 nm (ϵ290 = 2.8 mm-1 cm-1) (15Asada K. Methods Enzymol. 1984; 105: 422-427Crossref Scopus (285) Google Scholar) and initial rates were multiplied by a factor of 2 to account for the fast disproportionation (k ≈ 106m-1 s-1) of the mono-dehydroascorbate radical to ascorbate and dehydroascorbate (16Kelly G.J. Latzko E. Naturewissenschaften. 1979; 66: 617-618PubMed Google Scholar). For oxidation of guaiacol, stock solutions (100 mm) were prepared in sodium phosphate buffer containing 30% ethanol (17Santimone M. Can. J. Biochem. 1975; 53: 649-657Crossref PubMed Scopus (43) Google Scholar), and oxidation to tetraguaiacol was monitored at 470 nm (ϵ470 = 22.6 mm-1 cm-1 (17Santimone M. Can. J. Biochem. 1975; 53: 649-657Crossref PubMed Scopus (43) Google Scholar)). Values for kcat were calculated by dividing the maximum rate of activity (μm-1 s-1) by the micromolar concentration of the enzyme. Values for Km were determined by a fit of the data to the Michaelis-Menten equation using a nonlinear regression analysis program (Kalieda-Graph, version 3.09, Synergy Software). All reported values are the mean of three independent assays. Transient-state Kinetics—Transient-state measurements were performed using an SX.18MV microvolume stopped-flow spectrophotometer (Applied Photophysics) fitted with a Neslab RTE-200 circulating water bath (±0.1 °C). Reported values of kobs are an average of at least five measurements. Individual traces were monophasic in all cases. All kinetic data were analyzed using nonlinear least squares regression analysis on an Archimedes 410-1 microcomputer (Applied Photophysics) using Spectrakinetics software. All curve fitting was performed using the Grafit 5 software package (Grafit version 5.0.3, Erithacus Software Ltd.). Pseudo-first-order rate constants for the formation of Compound I (k1,obs) were monitored by a decrease in absorbance (corresponding to the formation of Compound I) at 407 nm for rsAPX and 405 nm for W41A in single mixing mode by mixing enzyme (1 μm) with various concentrations of H2O2. Multiple wavelength absorption studies were carried out using a photodiode array detector and X-SCAN software (Applied Photophysics). Spectral deconvolution was performed by global analysis and numerical integration methods using PROKIN software (Applied Photophysics). EPR Spectroscopy—EPR spectra ([W41A] = 300 μm) were recorded in sodium phosphate buffer (pH 7.0, μ = 0.10 m) and glycerol (50% v/v). X-Band EPR spectra were recorded on a Bruker ER-300D series electromagnet and microwave source interfaced to a Bruker EMX control unit and fitted with an ESR-9 liquid helium flow cryostat from Oxford Instruments and a dual mode microwave cavity from Bruker (ER-4116DM). Crystal Growth and Structure Determination—Crystals of rsAPX and the W41A mutant were obtained using the previously published procedures (18Sharp K.H. Mewies M. Moody P.C.E. Raven E.L. Nat. Struct. Biol. 2003; 10: 303-307Crossref PubMed Scopus (157) Google Scholar). Crystals of the cyanide-bound forms of rsAPX and the W41A mutant were obtained by dissolving a few crystals of potassium cyanide in 10 ml of mother liquor (0.1 m Hepes, pH 8.3, and 2.25 m lithium sulfate). Crystals of the NO-bound heme complexes were obtained by soaking crystals in sodium dithionite in the mother liquor followed by the addition of crystals of potassium nitrite, which reacts with excess dithionite to produce nitric oxide (19Leys D. Backers K. Meyer T.E. Hagen W.R. Cusanovich M.A. Van Beeumen J.J. J. Biol. Chem. 2000; 275: 16050-16056Abstract Full Text Full Text PDF PubMed Scopus (32) Google Scholar). For H2O2 soaks of W41A, crystals were soaked in 0.1 m H2O2, left for 5 min, and then frozen in liquid nitrogen. The heme coordination geometry in cytochrome c peroxidase is known to be sensitive to the presence of noncoordinating anions (e.g. phosphate, nitrate; see for example, Refs. 6Vitello L.B. Erman J.E. Miller M.A. Mauro J.M. Kraut J. Biochemistry. 1992; 31: 11524-11535Crossref PubMed Scopus (74) Google Scholar, 7Turano P. Ferrer J.C. Cheesman M.R. Thomson A.J. Banci L. Bertini I. Mauk A.G. Biochemistry. 1995; 34: 13895-13905Crossref PubMed Scopus (31) Google Scholar, and 20Ferrer J.C. Turano P. Banci L. Bertinin I. Morris I.K. Smith K.M. Smith M. Mauk A.G. Biochemistry. 1994; 33: 7819-7829Crossref PubMed Scopus (56) Google Scholar, 21Vitello L.B. Huang M. Erman J.E. Biochemistry. 1990; 29: 4283-4288Crossref PubMed Scopus (75) Google Scholar, 22Vitello L.B. Erman J.E. Miller M.A. Wang J. Kraut J. Biochemistry. 1993; 32: 9807-9818Crossref PubMed Scopus (143) Google Scholar, 23Erman J.E. Vitello L.B. Miller M.A. Shaw A. Brown K.A. Kraut J. Biochemistry. 1993; 32: 9798-9806Crossref PubMed Scopus (186) Google Scholar). These anions were not used in the crystallization experiments reported here. Glycerol also was not used in the crystallization studies. Diffraction data for W41A and the NO- and cyanide-bound forms were collected at beamline ID23-EH1 using an ADSC Quantum-315 detector, whereas data for the W41A-H2O2 soak was collected at beamline ID14-EH4 using an ADSC Quantum-4 detector all at the ESRF, Grenoble. Data for the rsAPX-NO and rsAPX-CN complexes were collected at DESY (Hamburg) at beamline X-11 using a MAR 165 mm CCD detector. All synchrotron data were collected at 100 K. Data collected at the ESRF was indexed and scaled using MOSFLM (24Leslie A.G.W. Joint CCP4 + ESF-EAMCB Newsletter on Protein Crystallography, No. 26. 1992; Google Scholar) and SCALA (25Collaborative Computational ProjectActa Crystallogr. Sect. D Biol. Crystallogr. 1994; 50: 760-763Crossref PubMed Scopus (19762) Google Scholar), and data collected at DESY was indexed and scaled using the HKL programs DENZO and SCALEPACK (26Otwinoski Z. Minor W. Methods Enzymol. 1997; 227: 366-396Google Scholar). Data collection and processing statistics are shown in Table 1; 5% of the data were flagged for the calculation of Rfree and excluded from subsequent refinement. The structures were refined from a model derived from the 1.45-Å rsAPX-ascorbate complex (Protein Data Bank accession code 1OAF) by the removal of bound ligand and water molecules. Several cycles of refinement using REFMAC5 (27Murshudov G.N. Vagin A.A. Dodson E.J. Acta Crystallogr. Sect. D Biol. Crystallogr. 1997; 53: 240-255Crossref PubMed Scopus (13865) Google Scholar) from the CCP4 suite (25Collaborative Computational ProjectActa Crystallogr. Sect. D Biol. Crystallogr. 1994; 50: 760-763Crossref PubMed Scopus (19762) Google Scholar) and manual rebuilding of the protein model using COOT (28Emsley P. Cowtan K. Acta Crystallogr. Sect. D Biol. Crystallogr. 2004; 60: 2126-2132Crossref PubMed Scopus (23349) Google Scholar) followed by the addition of water molecules were carried out until the Rfree and Rfactor values converged. In total, six crystal structures and diffraction data have been deposited with the following identifiers: Protein Data Bank codes 2GGN (ferric W41A), 2GHC (W41A-NO complex), 2GHD (W41A-CN complex), 2GHE (W41A-H2O2 soak), 2GHH (rsAPX-NO complex), and 2GHK (rsAPX-CN complex). The final refinement statistics of all structures are presented in Table 1.TABLE 1Data collection and refinement statistics r.m.s.d., root-mean-square deviation; PDB, Protein Data Bank. Values for outer shells are given in parentheses.StatisticsW41AW41A-CNrsAPX-CNW41A-NOrsAPX-NOW41A-H2O2Data collection Resolution (Å)28.88-1.3536.81-1.458-2.0045.88-1.2518.86-2.0128.99-1.75 Outer shell(1.42-1.35)(1.436-1.4)(2.07-2.0)(1.32-1.25)(2.064-2.01)(1.84-1.75) Total observations599,082248,23258,820299,27059,132360,725 Unique observations56,19352,94015,79869,77316,45526,381 I/σI17.8 (4.0)13.3 (3.1)15.9 (2.98)12.8 (3.9)15.8 (3.12)30.6 (11.6) Rmerge0.0970.0880.0520.1070.0510.059 Completeness99.4 (99.9)92.8 (96.6)95.22 (95.7)98.6 (99.9)97.32 (95.9)99.8 (100) Multiplicity10.74.73.74.33.613.7Refinement Rfactor0.1900.2060.1870.1890.1830.179 Rfree0.2050.2290.2350.2070.2580.220 r.m.s.d. angle (°)1.0861.1101.2901.0491.4571.188 r.m.s.d. bonds (Å)0.0070.0080.0120.0060.0150.011 PDB accession code2GGN2GHD2GHK2GHC2GHH2GHE Open table in a new tab Heme Coordination Geometry—The electronic spectrum of the ferric derivative of W41A is shown in Fig. 1. The spectrum (λmax/nm (ϵ/mm-1 cm-1 = 405 (125), 525, 564, 630) differs from that of the wild type enzyme (λmax/nm (ϵ/mm-1 = 407 (107), 525, ≈630 (11Lad L. Mewies M. Raven E.L. Biochemistry. 2002; 41: 13774-13781Crossref PubMed Scopus (72) Google Scholar); Fig. 1) and shows a peak in the visible region (564) that is consistent with the presence of low-spin heme. (The corresponding W41A variant in pea cytosolic APX was also examined (data not shown); in this variant similar wavelength maxima are observed but the Soret band is shifted to 412 nm.) The high-spin peak at ≈630 nm in the wild type enzyme is still visible in the variant (Fig. 1). No evidence for the formation of (low-spin) hydroxide-bound heme was observed at alkaline pH (data not shown); hence, the low-spin species was tentatively assigned as arising from coordination of an (internal) protein ligand. Reaction of ferric W41A with various noncatalytic ligands is informative because it allows us to assess whether the proposed protein ligand is reversibly or irreversibly bound to the metal. Addition of potassium cyanide to the ferric derivative of W41A leads to a spectrum in which complete formation of low-spin heme is now observed (λmax/nm = 418, 540, 561sh; Fig. 1). This suggests that addition of a strong exogenous ligand leads to displacement of the existing (internal) ligand. As shown below, the crystal structure of the cyanide-bound derivative of W41A confirms this observation. Additionally, we have observed that NO binds to the reduced heme (λmax = 417, 541, 574 nm); the crystal structure of the NO-bound derivative also confirms that His-42 is displaced (see below). EPR spectroscopy provides further evidence in support of a low-spin heme species (Fig. 2). The EPR spectrum of ferric W41A is dominated by a rhombic low-spin species with observed g-values of 3.22 and 2.05 (the third feature is too broad to be observed). A second, minor component with g-values of 5.67 and 1.99 is consistent with an axial, high-spin species. The low-spin species is the majority species at 10 K and is consistent with bis-histidine ligation; the g-values are consistent with an orientation in which the imidazole planes are not parallel to each other. Neither species resembles those observed in the EPR spectrum of the wild type enzyme, which has a rhombic high-spin species (g = 6.04, 5.27, and 1.98) and a small amount of rhombic low-spin heme (g = 2.69, 2.21, and 1.79) (13Jones D.K. Dalton D.A. Rosell F.I. Lloyd Raven E. Arch. Biochem. Biophys. 1998; 360: 173-178Crossref PubMed Scopus (40) Google Scholar, 29Gadsby P.M. Thomson A.J. J. Am. Chem. Soc. 1990; 112: 5003-5011Crossref Scopus (144) Google Scholar). A minor low-spin rhombic species (gz = 2.95, as a shoulder on the g = 3.22 feature) and the positive lobe of gy feature (gy = 2.30) in W41A are likely to arise from a histidine/histidine-ligated heme in which the imidazole planes are parallel to each other (note that there is also evidence for multiple conformations of the proximal histidine in the crystal structure of the cyanidebound form of W41A; see above). X-ray Crystallography of W41A and Its Cyanide- and NO-bound Derivatives—The crystal structure of the cyanide-bound derivative of W41A (Fig. 3 and Table 1) confirms the observations made in solution. In this structure (Fig. 3B), which closely maps onto that for the cyanide-bound form of rsAPX (Fig. 3A), the iron is ligated by the cyanide ligand and the nitrogen of the bound ligand is hydrogen-bonded (2.8 Å) to Nϵ of His-42 (2.5 Å in rsAPX). For the W41A-CN complex, the electron density observed for the proximal His-163 residue is consistent with two orientations of this side chain (Fig. 3B). One orientation has His-163 hydrogen-bonded to Asp-208 (3.3 Å) as for rsAPX (3.1 Å); the other orientation has His-163 hydrogen-bonded (2.8 Å) to the backbone carbonyl of Ser-160 (3.8 Å in rsAPX). In the rsAPX-CN structure a single conformation of His-163 is observed; however, the data are to lower resolution (2.0 Å compared with 1.4 Å), and a second orientation of the proximal histidine may not be observed because of the large effect of the heme electron density. The crystal structure of the NO-bound derivative of W41A also confirm that His-42 is displaced (Fig. 3D). In this case, and in contrast to rsAPX-NO (Fig. 3C), removal of the hydrogen bond to Trp-41 means that the NO ligand now adopts two conformations. To clarify the nature of the low-spin heme in ferric W41A, diffraction data to 1.75 Å were obtained (Fig. 4 and Table 1). W41A is very similar in its overall structure to the wild type protein; the root-mean-square deviation between Cα positions (residues 2-249) for this structure and ferric rsAPX is 0.260 Å (determined using LSQKAB (30Kabsch W. Acta Crystallogr. Sect. A. 1976; 32: 922-923Crossref Scopus (2351) Google Scholar)). The structure of W41A shows that space previously occupied by the Trp side chain is now filled by two water molecules (labeled 1 and 2 in Fig. 4A). Although the overall structure and most of the active site structure is similar to rsAPX, there are local changes in protein conformation around His-42 (Fig. 4B). Hence, the main chain of His-42 moves toward the heme in W41A such that Nϵ of His-42 is now within bonding distance (2.3 Å, compared with 5.5 Å in rsAPX) of the iron. We refer to this as the "on" form. A further water molecule (labeled 3 in Fig. 4A) is located in the region that was previously occupied by His-42. The Fe-Nϵ(His-163) distance is essentially identical in both W41A and rsAPX (2.1 and 2.0 Å, respectively). This new histidine ligand replaces a water molecule that is bonded to the iron in the wild type protein (2.1 Å). This movement of the main chain of His-42 toward the heme in the on form and the subsequent alteration in heme geometry are the consequences of the removal of the bulky Trp-41 residue, which allows the His-42 side chain to ligate to the iron. Examination of the electron density for W41A around His-42 indicates that there is positive Fo - Fc density above His-42 and also close to the main chain of His-42 (carbonyl oxygen; Fig. 4). This density overlays with the orientation of His-42 in rsAPX and is consistent with the presence of a minority (presumed high-spin) heme species in which His-42 is not ligated. There are two explanations of this observation. First, the spectroscopic data presented above are also consistent with some high-spin heme in the ferric form; in this case, the residual electron density would arise from a mixed population of high- and low-spin iron in the crystal. Second, although we used ferric enzyme in our experiments, we recognize that partial reduction of heme is possible during data collection (31Berglund G.I. Carlsson G.H. Smith A.T. Szoke H. Henriksen A. Hajdu J. Nature. 2002; 417: 463-468Crossref PubMed Scopus (743) Google Scholar). The electronic spectrum of ferrous W41A shows maxima (λmax/nm = 428, 556, 581) that are similar to those for rsAPX (λmax/nm = 431, 555, 585) and are not consistent with bis-histidine ligation, indicating that dissociation of His-42 probably occurs on reduction. In this case, the residual electron density would arise from partial reduction of the heme during data collection. We were not able to obtain crystallographic data for the ferrous W41A derivative. Whatever the origin of the residual density, the important feature is that the data clearly indicate that His-42 is flexible and can adopt more than one conformation. Reactivity toward H2O2—The data presented above indicated that, although the heme is largely 6-coordinate and low-spin in ferric W41A (with two strong histidine ligands), His-42 is not irreversibly bound to the heme and may dissociate under certain conditions. We refer to this dissociated form as the "off" form. If this is the case, then reaction with H2O2 and turnover of substrate may still be possible. Because there is, to our knowledge, no unambiguous example of a crystallographically defined bis-histidine-ligated peroxidase in the literature, it was of critical interest to establish whether strong axial ligation, analogous to that observed for example in the cytochromes b, would preclude reaction with H2O2. Rate constants for Compound I formation in W41A were determined under pseudo-first-order conditions ([H2O2] = 10-125 × [W41A]). The W41A variant was shown to be competent for formation of Compound I under these conditions, albeit at a lower rate, as evidenced by a decrease in absorbance at 405 nm. These changes in absorbance duplicate those observed for rsAPX (11Lad L. Mewies M. Raven E.L. Biochemistry. 2002; 41: 13774-13781Crossref PubMed Scopus (72) Google Scholar). Observed rate constants for this process, k1,obs, showed a clearly nonlinear dependence on the concentration of hydrogen peroxide (Fig. 5A). This is in contrast to the data for rsAPX, in which a linear dependence on [H2O2] is observed in the experimentally accessible concentration range (k1 = (3.3 ± 0.1) × 107m-1 s-1 (11Lad L. Mewies M. Raven E.L. Biochemistry. 2002; 41: 13774-13781Crossref PubMed Scopus (72) Google Scholar)). The nonlinear dependence is consistent with a mechanism that requires a conformational change of the protein, proposed to be conversion between the on and off forms, prior to reaction with H2O2, as shown in Equations 4 and 5 (CI = Compound I). W41Aonk1⇌k−1W41Aoff(Eq. 4) W41Aoff+H2O2k2→Cl+H2O(Eq. 5) In the presence of excess H2O2, the observed rate constant, k1,obs, can be expressed as follows (Equation 6). k1,obs=k1k2[H2O2]k−1+k2[H2O2](Eq. 6) A fit of these data for W41A to Equation 6 (Fig. 5A) yields values for the limiting first-order rate constant, k1, of 2370 s-1 and the composite second order rate constant, k1k2/k-1, of 6.6 × 106m-1 s-1. In separate experiments, the reaction of W41A with H2O2 was studied using photodiode array detection (Fig. 5B). Data collected over a period of 500 ms were best-fitted to a two-step model (A → B → C, as shown previously for rsAPX (11Lad L. Mewies M. Raven E.L. Biochemistry. 2002; 41: 13774-13781Crossref PubMed Scopus (72) Google Scholar)), where A is ferric W41A, B is Compound I, and C is Compound II. Spectra for the Compound I (λmax/nm = 410, 530, 569sh, 640) and Comp
Referência(s)