Proton transfer in the oxidative half-reaction of pentaerythritol tetranitrate reductase Structure of the reduced enzyme-progesterone complex and the roles of residues Tyr186, His181 and His184 Huma Khan 1 , Terez Barna 1 , Neil C. Bruce 2 , Andrew W. Munro 1, *, David Leys 1, † and Nigel S. Scrutton 1, † 1 Department of Biochemistry, University of Leicester, UK 2 CNAP, Department of Biology, University of York, UK Pentaerythritol tetranitrate (PETN) reductase was ori- ginally purified from a strain of Enterobacter cloacae (strain PB2) on the basis of its ability to utilize nitrate ester explosives such as PETN and glycerol trinitrate (GTN) as sole nitrogen source. Sequence analysis [1] and structural studies [2] have indicated that PETN reductase is a flavoprotein member of the Old Yellow Enzyme (OYE) family [3]. Other well-defined members include bacterial morphinone reductase (MR) from Pseudomonas putida M10 [4], estrogen binding protein Keywords crystallography; flavoprotein mechanism; kinetics; Old Yellow Enzyme; PETN reductase Correspondence N. S. Scrutton, Faculty of Life Sciences, University of Manchester, Stopford Building, Oxford Road, Manchester, M13 9PT, UK Fax: +44 161 2755586 Tel: +44 161 2755632 E-mail: nigel.scrutton@manchester.ac.uk Present addresses *Manchester Interdisciplinary Biocentre, School of Chemical Engineering and Ana- lytical Science, University of Manchester, The Mill, PO Box 88, Manchester, M60 1QD, UK †Manchester Interdisciplinary Biocentre and Faculty of Life Sciences, Faculty of Life Sciences, University of Manchester, Stopford Building, Oxford Road, Manchester, M13 9PT, UK (Received 13 June 2005, revised 15 July 2005, accepted 21 July 2005) doi:10.1111/j.1742-4658.2005.04875.x The roles of His181, His184 and Tyr186 in PETN reductase have been examined by mutagenesis, spectroscopic and stopped-flow kinetics, and by determination of crystallographic structures for the Y186F PETN reductase and reduced wild-type enzyme—progesterone complex. Residues His181 and His184 are important in the binding of coenzyme, steroids, nitro- aromatic ligands and the substrate 2-cyclohexen-1-one. The H181A and H184A enzymes retain activity in reductive and oxidative half-reactions, and thus do not play an essential role in catalysis. Ligand binding and catalysis is not substantially impaired in Y186F PETN reductase, which contrasts with data for the equivalent mutation (Y196F) in Old Yellow Enzyme. The structure of Y186F PETN reductase is identical to wild-type enzyme, with the obvious exception of the mutation. We show in PETN reductase that Tyr186 is not a key proton donor in the reduction of a ⁄ b unsaturated carbonyl compounds. The structure of two electron-reduced PETN reductase bound to the inhibitor progesterone mimics the catalytic enzyme-steroid substrate complex and is similar to the structure of the oxidized enzyme-inhibitor complex. The reactive C1-C2 unsaturated bond of the steroid is inappropriately orientated with the flavin N5 atom for hydride transfer. With steroid substrates, the productive conformation is achieved by orientating the steroid through flipping by 180°, consistent with known geometries for hydride transfer in flavoenzymes. Our data highlight mechanistic differences between Old Yellow Enzyme and PETN reductase and indicate that catalysis requires a metastable enzyme-steroid complex and not the most stable complex observed in crystallographic studies. Abbreviations EBP, estrogen binding protein; GTN, glycerol trinitrate; MR, morphinone reductase; OYE1, Old Yellow Enzyme 1; PETN, pentaerythritol tetranitrate; TNT, trinitrotoluene. 4660 FEBS Journal 272 (2005) 4660–4671 ª 2005 FEBS (EBP) from Candida albicans [5], glycerol trinitrate reductase from Agrobacterium radiobacter [6], the xenobiotic reductases of Pseudomonas species [7] and 12-oxophytodienoic acid reductase from tomato [8] and Arabidopsis thaliana [9]. These enzymes reduce a variety of cyclic enones, including 2-cyclohexen-1-one and steroids. Some steroids act as substrates, whereas others are potent inhibitors of these enzymes. PETN reductase is also related to the more complex bile-acid inducible flavoenzymes Bai H and Bai C from Eubacte- rium species [10], the bacterial Fe ⁄ S flavoenzymes tri- and dimethylamine dehydrogenases [11,12], the his- tamine dehydrogenase from Nocardiodes simplex [13] and the NADH oxidase of Thermoanaerobium brockii [14]. These latter enzymes utilize diverse substrates, but the catalytic framework has clearly evolved from a common progenitor [15]. PETN reductase is unusual in its ability to degrade major classes of explosive, including nitroaromatic compounds, e.g. trinitrotoluene (TNT), and nitrate esters (GTN and PETN) [16–18]. Degradation of TNT involves reductive hydride addition to the aromatic nucleus [16,19], and key residues involved in this pro- cess have been discerned [20]. The catalytic cycle of PETN reductase comprises two half-reactions. In the reductive half-reaction, enzyme is reduced by NADPH to yield the dihydroquinone form of the enzyme-bound FMN, a reaction known to proceed by quantum mechanical tunneling [21]. In the oxidative half-reac- tion, the flavin is oxidized by nitro-containing explo- sive substrates or, in common with related enzymes, cyclic enone substrates such as 2-cyclohexen-1-one. A detailed kinetic mechanism based on stopped-flow data has been proposed [19]. Studies with OYE have established a role for Tyr196 in proton donation during the reduction of a ⁄ b unsat- urated carbonyl compounds [22]. This residue is con- served in PETN reductase as Tyr186 [1], and X-ray structural and NMR analyses of PETN reductase in complex with a number of steroid ligands have sugges- ted this residue may likewise function as the key pro- ton donor during the reduction of a ⁄ b unsaturated carbonyl compounds by PETN reductase [2]. This act- ive site tyrosine is not conserved in MR, where it is replaced by cysteine [23,24]. Recent mutagenesis stud- ies have failed to identify the key proton donor in MR, and suggest solvent water is the source of the proton required for reduction of a ⁄ b unsaturated car- bonyl compounds [21,24,25]. Herein, we report solu- tion studies of three mutant forms of PETN reductase. We show that Tyr186 is not the key proton donor in the oxidative half-reaction of PETN reductase, which contrasts with reported studies with the highly homo- logous OYE. We demonstrate that residues His181 and His184 are determinants for substrate ⁄ ligand bind- ing as suggested by the crystal structure of PETN reductase, consistent with a similar role for conserved residues in OYE [26] and MR [25]. We also report the crystal structure of 2-electron reduced PETN reductase in complex with the steroid inhibitor and discuss its implications for the binding and reduction of steroid substrates. Our studies demonstrate a probable role for water in proton donation in PETN reductase and mul- tiple binding modes for steroid ligands in the active site. The work emphasizes the need for (i) detailed evaluation of mechanism, and (ii) caution in inferring mechanistic similarities in structurally highly related enzymes. Results Properties of the H181A and H184A enzymes The structure of PETN reductase solved in complex with prednisone, progesterone, 1,4-androstadiene (Fig. 1A) and 2-cyclohexen-1-one indicate that these ligands bind above the si face of the FMN isoalloxa- zine ring and are held in position by hydrogen bond interactions with His181 and His184 [2]. These residues also form interactions with the hydroxyl group of 2,4- dinitrophenol (an inhibitor) and picric acid (a sub- strate) [19]. In OYE1 the counterpart residues are His191 and Asn194, and are known to have an important role in the binding of phenolic compounds in the active site [3,26]. Likewise, in MR the counter- part residues His186 and Asn189 form key interactions with reducing nicotinamide coenzyme and the oxi- dizing substrate 2-cyclohexen-1-one [25]. Furthermore, NMR and kinetic studies have ruled out a role for His186 in proton donation in the oxidative half-reac- tion of MR [25]. PETN reductase is unusual in having two histidine residues (rather than the His-Asn pair seen in OYE, MR and some of the other member proteins; Fig. 1B) in the active site for ligand and substrate binding. For those members that contain a His-His pair, there has been no report of a systematic analysis of the contribution of each histidine residue to binding and catalysis. Thus, to ascertain the role of each histidine residue, and to identify any differential contribution to binding and catalysis, we isolated the two mutant enzymes H181A and H184A. Both the H181A and H184A enzymes were purified to homogeneity as described and, as for wild-type enzyme [19], UV-visible spectra indicated stoichiometric assembly with the FMN cofactor. Ligand binding titra- tions revealed a substantially increased dissociation H. Khan et al. Proton transfer in PETN reductase FEBS Journal 272 (2005) 4660–4671 ª 2005 FEBS 4661 constant for enzyme-ligand complexes, consistent with a key role for both residues in the binding of substrates and inhibitors (Table 1; Fig. 2). Anaerobic stopped- flow studies of FMN reduction by NADPH in H181A and H184 enzymes indicated reduction of the FMN, but unlike for wild-type enzyme [19] there was no evi- dence at % 560 nm for an oxidized enzyme-NADPH charge-transfer intermediate prior to FMN reduction (Fig. 3A). Plots of observed rate constant vs. NADPH concentration were hyperbolic for the mutant enzymes, and limiting rate constants for FMN reduction were elevated modestly compared with wild-type PETN reductase (Fig. 3A, inset; Table 2). The exchange of His181 and His184 residues by alanine, however, com- promised binding of reducing coenzyme in the active site (Table 2). Stopped-flow studies in which 2-electron reduced enzyme (obtained by titration against dithio- nite) was mixed with 2-cyclohexen-1-one indicated that both the H181A and H184A enzymes are able to trans- fer electrons to 2-cyclohexen-1-one (Fig. 3B). In both cases, single wavelength stopped-flow studies at 450 nm established that observed rate constants for FMN oxi- dation were hyperbolically dependent on 2-cyclohexen- 1-one concentration (Fig. 3B). Kinetic parameters derived from fitting to a standard hyperbolic expression are given in Table 2. The limiting rate constants for oxidation of FMNH 2 by 2-cyclohexen-1-one in the H181A and H184A enzymes are substantially less than that measured for the wild-type enzyme. We suggest that mutation perturbs the binding geometry such that A B Fig. 1. (A) Superposition of steroid com- plexes bound to wild-type oxidized penta- erythritol tetranitrate (PETN) reductase. The bound steroids are prednisone, progester- one and 1,4-androstadiene-3,17-dione. (B) Sequence of PETN reductase and some PETN reductase related enzymes in the region of His181 and His184. The proteins are PETN (PETN reductase from Entero- bacter cloacae), MR (MR from Pseudomo- nas putida), OYE1 (Old Yellow Enzyme 1 from brewer’s bottom yeast), OYE2, OYE3 (two isoforms of Old Yellow), EBP1 (estro- gen binding protein from Candida albicans ), NER A (glycerol trinitrate reductase from Agrobacterium radiobacter), NEM A (N-ethyl- maleimide reductase from Escherichia coli ), OPDA (12-oxophytodienoate reductase from Arabidopsis thaliana), BAIH (bile acid-indu- cible protein from Eubacterium species), NADH (NADH oxidase from Thermobacillus brockii). The arrows indicate the positions of histidine residues in PETN reductase inferred to be involved in ligand binding and counterpart residues in related enzymes and also the location of the tyrosine residue implicated as proton donor in OYE. Table 1. Dissociation constants for enzyme-ligand complexes calcu- lated from equilibrium titrations. Enzymes (each 10 l M) were titra- ted with picric acid, progesterone and 2,4-dinitrophenol in 50 m M potassium phosphate buffer, pH 7.0, in a 1 mL quartz cuvette. Spectra were recorded after the addition of ligand to the enzyme. From the resultant spectra (examples shown in Fig. 2) the absorp- tion changes at 518 nm were plotted as a function of ligand con- centration and fitted to a hyperbolic or quadratic function from which the dissociation constants were determined. Dissociation constants for wild-type enzyme-ligand complexes are taken from [2,19]. Dissociation constant (K d , lM) Picric acid 2,4-Dinitrophenol Progesterone Wild-type PETN reductase 5.4 ± 1.1 0.95 ± 0.10 0.07 ± 0.03 H181A PETN reductase 92 ± 12 56 ± 7 16 ± 5 H184A PETN reductase 73 ± 16 34 ± 6 15 ± 3 Y186F PETN reductase 11 ± 1 1.9 ± 0.5 0.05 ± 0.09 Proton transfer in PETN reductase H. Khan et al. 4662 FEBS Journal 272 (2005) 4660–4671 ª 2005 FEBS the reducible carbon double bond of the substrate is less optimally positioned with respect to the flavin N5 atom compared with the geometry in the wild-type enzyme. Stopped-flow studies were extended to include the nitroaromatic compound TNT as oxidizing substrate. TNT is a known substrate for the wild-type enzyme, and the kinetics of FMN oxidation with wild-type PETN reductase have been reported [19]. With wild- type enzyme, reduction of TNT follows two parallel pathways; the first pathway involves direct reduction of the nitro group (the nitroreductase pathway), whereas the second pathway involves hydride transfer from FMNH 2 to the aromatic nucleus of the substrate to form a hydride-Meisenheimer product (further details in [19,20]). The hydride-Meisenheimer product is readily detected in stopped-flow studies using a photo- diode array detector as it has a strong absorption band around 560 nm [19]. Photodiode array studies of the oxidative half-reaction of the H181A and H184A enzymes indicated that the hydride-Meisenheimer com- plex is not formed. However, TNT was able to oxidize the flavin resulting in recovery of the oxidized FMN absorption spectrum (Fig. 3C). Single wavelength stud- ies at 453 nm indicated a hyperbolic dependence of the flavin oxidation rate on TNT concentration (Fig. 3D); limiting rate constants and reduced enzyme-TNT disso- ciation constants are presented in Table 2. The reduc- tion potentials of the FMN centres in the H184A and H181A enzymes were determined to be )266 ± 5 mV and )229 ± 5 mV, respectively, which compares with a value of )267 mV for wild-type enzyme [20]. Fig. 2. Analysis of ligand binding by equilibrium titration studies. (A) UV-visible analysis of the titration of H181A PETN reductase with 2,4 dinitrophenol. Inset, detail of the absorbance change around 518 nm used to calculate the dissociation constant for the complex. (B) UV-visi- ble analysis of the titration of H181A PETN reductase with progesterone. (C) Plot of absorbance change as a function of 2,4 dinitrophenol concentration for the data shown in (A). (D) Plot of absorbance change as a function of progesterone concentration for the data shown in (B). Conditions for panels A and B: 50 m M potassium phosphate buffer pH 7.0, 25 °C. Similar plots were generated for other enzyme-ligand combinations. Dissociation constants for the enzyme–ligand complexes are given in Table 1. Arrows indicate direction of absorption change with time. H. Khan et al. Proton transfer in PETN reductase FEBS Journal 272 (2005) 4660–4671 ª 2005 FEBS 4663 Properties of the Y186F PETN reductase Equilibrium binding measurements with picric acid, 2,4-dinitrophenol and progesterone were performed as described for wild-type and the H181A and H184A enzymes. Spectral changes were similar to those repor- ted above and absorption changes at 518 nm were used to determine enzyme-ligand dissociation constants (Table 1). Enzyme-ligand dissociation constants for nitroaromatic ligands are elevated approximately two- fold compared with wild-type enzyme, whereas the binding of progesterone is essentially unaffected by mutation (Table 1). Stopped-flow studies of flavin reduction by NADPH produced mono exponential reaction traces at 464 nm and observed rate constants were independent of coen- zyme concentration (concentration range 200–1300 lm NADPH). The rate constant for flavin reduction (9 s )1 ) is similar to that measured for wild-type enzyme (12 s )1 )at5°C. Absorption changes at 560 nm showed a rapid increase in absorption followed by a slower decay, consistent with the formation and sub- sequent collapse of an oxidized enzyme–NADPH charge-transfer complex (Fig. 4A). The rapid forma- tion of the charge-transfer intermediate prevented accurate analysis of the rate constant for its formation by the stopped-flow method; the rate constant for col- lapse of the charge-transfer species was similar to that measured for hydride transfer at 464 nm, indicating that both processes are kinetically equivalent. Waveleng h tn( )m 00400 5006 0 07 Absorbance 0 50.0 1.0 51.0 2. AB CD 0 [ β NAD PH ] )Mm( 032154 k obs (s -1 ) 0 01 0 2 0 3 0 4 0 5 [ - 2 C y c lho exe nn o e ] ( mM ) 0 2 040 6 0 80 0 01 021 kobs (s -1 ) 0 1. 0 2.0 3.0 4.0 5.0 [NT T ] ( µ )M 000 2 00 4 00 6 0 08 0 00 1 kobs (s -1 ) 0 5 01 5 1 0 2 Wa v e len g h tn ( m ) 00 3 00 4 00 5 0 0 60 0 7 Absorbance 0 1 . 0 2 . 0 3 .0 Fig. 3. Stopped-flow kinetic analysis of FMN reduction and oxidation in the H181A and H184A PETN reductases. (A) spectral changes accompanying the reduction of H184A PETN reductase (20 l M) by NADPH (200 lM). Four hundred spectra were recorded over a period of one second. One in every 25 of the spectra is displayed. Conditions: 50 m M potassium phosphate buffer, pH 7.0, at 5 °C. Inset: plot of observed rate constant for FMN reduction as a function of NADPH concentration for H181A (filled squares) and H184A (filled circles). (B) dependence of the observed rate for the oxidation of reduced FMN on 2-cyclohexen-1-one concentration for the H181A (filled squares) and H184A (filled circles) PETN reductases. (C) Spectral changes accompanying the oxidation of 2-electron reduced H184A PETN reductase (20 l M) by TNT (200 lM). (D) dependence of the observed rate for the oxidation of reduced FMN on TNT concentration for the H181A ( ) and H184A (d) PETN reductases. Kinetic parameters are given in Table 2. Arrows indicate direction of absorption change with time. Proton transfer in PETN reductase H. Khan et al. 4664 FEBS Journal 272 (2005) 4660–4671 ª 2005 FEBS Stopped-flow studies of the oxidative half-reaction of Y186F PETN reductase (reduced at the 2 electron level with dithionite) with 2-cyclohexen-1-one revealed a hyperbolic dependence of the rate of flavin oxidation on 2-cyclohexen-1-one concentration (Fig. 4D). The limiting rate constant for flavin oxidation is % twofold less than that for wild-type enzyme (Table 2). Stopped- flow studies of the oxidative half-reaction of Y186F PETN reductase with TNT indicated formation and subsequent decay of the hydride Meisenheimer com- plex (Fig. 4B,C), indicating that residue Y186 does not influence strongly the partitioning along the hydride transfer and nitroreductase pathways. The rate of Meisenheimer complex formation measured at 560 nm was shown to be equivalent to that for FMN oxidation at 464 nm, indicating that both processes are kinetic- ally equivalent. Potentiometric titrations indicated that the reduction potential for the concerted 2-electron reduction of the Y186F PETN reduction is )265 ± 5 mV, which is comparable to that measured previously for wild-type enzyme ()267 mV; [20]). Structures of the Y186F PETN reductase, and reduced wild-type enzyme in complex with progesterone The 1.0 A ˚ Y186F structure is essentially identical to the wild-type structure (PDB code 1VYR) with the obvious exception of the mutation. A bound thio- cyanide ion and iso-propanol molecule can be seen occupying the substrate-binding site adjacent to the FMN (Fig. 5). PETN reductase can reduce a ⁄ b unsaturated steroids with a double bond located between C1 and C2; the double bond between C4 and C5 is not susceptible to reduction, and thus progesterone and related steroids (e.g. 4-androstene- 3,17-dione), which lack a double bond between C1 and C2, are inhibitors of PETN reductase (Fig. 6). The overall structure of the 1.05 A ˚ reduced PETN- progesterone complex is virtually identical to the oxidized complex (PDB code 1H60). Two molecules of isopropanol could be resolved and are bound between the progesterone and the protein (Fig. 7A). In comparison to the oxidized structures, the FMN isoalloxazine ring is less planar, with both the N5 atom and the C8-C7 methyl groups moving to signi- ficantly out-of-plane positions. The progesterone sub- strate is bound in a similar manner, albeit shifted by approximately 0.4 A ˚ towards the Thr26 side chain. This places the progesterone C4—C5 double bond in close proximity to the FMN N5 atom (distances 3.47 and 3.61 A ˚ for C4-N5 and C5-N5, respectively). While this distance is ideal for reduction of the dou- ble bond by FMN, the N10-N5-C5 angle value of 92° (Fig. 7B) is distinct form the range of angles (125° to 170°) observed in flavoenzyme-substrate complexes [31]. Any putative motion of the pro- gesterone molecule required to increase the N10-N5- C5 angle to within or close to the 125° to 170° range causes severe steric clashes of the C6 carbon atom with the Thr26 side chain, explaining why progesterone is an inhibitor rather than a substrate for this enzyme. The presence of Cb atom at posi- tion 26 therefore causes the substrate specificity of Table 2. Kinetic parameters for the reductive and oxidative half-reactions of wild-type, H181A, H184A and Y186F PETN reductases. Kinetic data are shown in Figs 3 and 4. Parameters were determined by fitting to a standard hyperbolic expression to obtain values for the limiting rate of flavin reduction or oxidation (k lim ) and the enzyme-substrate dissociation constants (K d ). All reactions were performed in 50 mM potas- sium phosphate buffer, pH 7.0. ND, not determined. Owing to the very rapid formation of the charge-transfer intermediate and the small absorption changes it was not possible to evaluate the dissociation constant for the oxidized enzyme-NADPH charge-transfer species from analysis of its rate of formation as a function of NADPH concentration. Reductive half-reaction k lim (s )1 ) K d (lM) Wild-type PETN reductase a 11.6 ± 0.2 33.4 ± 8.5 H181A PETN reductase 31.2 ± 0.3 113 ± 6 H184A PETN reductase 46.6 ± 0.6 973 ± 28 Y186F PETN reductase 9.0 ± 0.2 ND Oxidative half-reaction 2-Cyclohexen-1-one TNT k lim (s )1 ) K d (mM) k lim (s )1 ) K d (lM) Wild-type PETN reductase a 35 ± 2 9.5 ± 1.6 4.5 ± 0.1 78.4 ± 11.7 H181A PETN reductase 0.34 ± 0.01 19.2 ± 1.5 9.3 ± 0.4 194 ± 27 H184A PETN reductase 0.49 ± 0.02 44 ± 4 15.8 ± 0.3 134 ± 11 Y186F PETN reductase 14 ± 1 2.1 ± 0.2 6.9 ± 0.3 164 ± 24 a Data taken from [19]. H. Khan et al. Proton transfer in PETN reductase FEBS Journal 272 (2005) 4660–4671 ª 2005 FEBS 4665 the enzyme to be limited to enones with a primary or secondary b-carbon such as steroids with a double bond located between C1 and C2. However, crystal structures of the oxidized enzyme in complex with C1-C2 unsaturated steroids reveal a mode of binding similar to that observed for progesterone, placing the C4-C5 bond rather than the reactive C1-C2 bond in close proximity of the N5 atom. A putative flipping motion of the progesterone molecule along an axis parallel to the FMN plane aligns the C1-C2 with the plane of the isoalloxazine ring and in close proximity to the N5, with possibility of adopting a conforma- tion with an appropriate N10-N5-C1 angle for cata- lysis. Previous NMR data and modelling suggested that the preferred orientation of steroid substrates (with a C1-C2 double bond) might be different dependent on the oxidation state of the protein [2]. Similar behaviour was also proposed recently for Escherichia coli nitroreductase in complex with the antibiotic nitrofurazone [32]. Our present data Time (s) 10 3 245 Absorbance 0 50. 0 1.0 51.0 2. AB CD 0 i T m e( s ) 00 .005 . 00 1 . 0 0 15 0. 02 Absorbance .0015 . 0 0 2 .00 2 5 . 0 0 3 . 0 035 . 0 0 4 Wavelength (nm) 030 400 050 600 070 Absorbance 0 . 01 .02 .03 .04 .05 .06 Wavelength (nm) 03 0 40 0 0 5 0 60 0 07 0 Absorbance 0 .01 .02 .03 .04 .05 .06 [2-Cyclohexenone] (mM) 05 0151025230 k obs (s -1 ) 0 2 4 6 8 0 1 21 41 61 Fig. 4. (A) Typical stopped-flow transient obtained for the reductive half-reaction of Y186F PETN reductase at 464 nm. The absorption trace was measured by mixing NADPH (200 l M) with Y186F PETN reductase (20 lM)in50mM potassium phosphate buffer, pH 7.0, at 5 °C. Inset: the small absorption change observed at 560 nm indicating formation of an oxidized enzyme-NADPH charge-transfer species over 0.02 s. Measurements over longer time periods indicate the absorption decreases with a rate constant similar to that observed at 464 nm, which indicates reduction of the FMN. Conditions: 50 m M potassium phosphate buffer, pH 7.0, at 5 °C. (B) Time-dependent spectral chan- ges following the reaction of dithionite-reduced Y186F PETN reductase and TNT, using stopped-flow spectroscopy. Dithionite-reduced Y186F PETN reductase (40 l M) was mixed with (400 lM) TNT at 25 °C, in 1% acetone, potassium phosphate buffer, pH 7.0, under anaer- obic conditions. Spectral changes are shown over 2 s, and indicate re-oxidation of the flavin and the formation of the hydride-Meisenheimer complex of TNT. (C) As for panel B except the spectral changes are recorded over an extended time period and show degradation of the hydride-Meisenheimer complex which generates the oxidized form of the enzyme. (D) The concentration-dependence of the rate of hydride transfer from dithionite-reduced Y186F PETN reductase to 2-cyclohexen-1-one. Anaerobic single wavelength spectroscopy at 464 nm was performed with 20 l M enzyme and a range of 2-cyclohexen-1-one concentrations, in 50 mM potassium phosphate buffer, pH 7.0 at 25 °C. The solid line shows the fit of the data to a hyperbolic function. Kinetic parameters are given in Table 2. Arrows indicate direction of absorp- tion change with time. Proton transfer in PETN reductase H. Khan et al. 4666 FEBS Journal 272 (2005) 4660–4671 ª 2005 FEBS strongly indicate that the preferred steroid binding mode is independent of enzyme redox state and is in a nonproductive conformation for C1-C2 unsaturated steroids. The slow observed turnover values with these steroids are therefore a likely consequence of the low concentration of enzyme complexes in the productive conformation with the C1-C2 bond optimally aligned for hydride transfer. There are two models for productive binding of steroid substrates: in the first, a (small) sub population of enzyme binds directly the steroid in the reactive conformation, with the remainder of the enzyme binding the unreactive conformation. Rescue of the unreactive conformation involves release of steroid and re-binding in the reactive conformation. In the second model, the ster- oid flips its conformation from the unreactive to reactive binding mode whilst resident in the enzyme active site. At this stage, we cannot categorically rule out either model. However, flipping within the active site would require a change in steroid conformation A B Fig. 7. Structure of the reduced PETN reductase-progesterone complex. (A) Active site of the reduced protein–progesterone com- plex. Progesterone binds to reduced enzyme in the same con- formation as observed for the oxidized enzyme, i.e. with the nonreducible C4-C5 double of the steroid positioned close to the flavin N5. The reduced protein-progesterone atoms are displayed in coloured sticks, with the oxidized protein-progesterone overlayed for comparison and displayed in white sticks. The sigmaA weighted F 0 -F c density for the progesterone molecule is displayed in blue. (B) schematic displaying the steric clashes between enones with tertiary Cb atoms and the Cb atom at position 26 for N10-N5-Cb angles in the range of 125° to 170°. R A C B O O O O O O O 1 2 3 4 5 6 7 8 9 01 1 1 2 1 3 1 4 1 51 6 1 7 1 go r Penoretse ANPD + H H + P DAN + i dats or dna- 4,1id-7 1, 3 en ee n o i d -71 , 3-e n etsor d na- 4 eno Fig. 6. Steroid nomenclature and chemical structures. (A) Nomen- clature for atom labelling in 3-oxo steroids; (B) chemical structure of the inhibitor progesterone; (C) reaction catalysed with the sub- strate 1,4-androstadiene-3,17-dione. Fig. 5. Superposition of the thiocyanide complex structures for Y186F PETN reductase and wild-type PETN reductase. The struc- ture of Y186F PETN reductase (shown in atom coloured sticks) is similar to the wild-type enzyme (shown in white), confirming that the mutation of Tyr186 to Phe186 does not grossly perturb the overall framework of the enzyme or the active site. The active sites of the enzymes are shown in stick format with the sigmaA weigh- ted 2F 0 -F c density for the Y186F mutant displayed in blue. H. Khan et al. Proton transfer in PETN reductase FEBS Journal 272 (2005) 4660–4671 ª 2005 FEBS 4667 and is hindered by a number of unfavourable steric interactions with the protein. We suggest therefore that flipping does not occur within the active site and that the productive conformation is formed directly by the binding of free steroid from solution. Discussion A number of crystal structures are now available for members of the Old Yellow Enzyme family [2,3,24,33– 35]. Each of these indicates that the counterpart resi- dues of His181 and His184 of PETN reductase are implicated in ligand binding. Solution studies with MR [25] and OYE1 [26], which contain a His-Asn pair at this position support this inferred role, and the data reported in this paper indicate that the His181 and His184 pair are likewise determinants for the binding of coenzyme, nitroaromatic ligands, steroids and 2-cyclohexen-1-one. In this regard, the structural simi- larity seen in the Old Yellow Enzyme family members is consistent also with similar functional roles. An unexpected finding, however, is that mutation of Tyr186 to phenylalanine in PETN reductase does not prevent reduction of 2-cyclohexen-1-one. In OYE1, the counterpart residue (Tyr196) has been shown to be the key proton donor required for reduction of 2-cyclo- hexen-1-one and related substrates, and its mutation leads to inactivation of the oxidative half-reaction [22]. Clearly, in PETN reductase proton transfer is not from Tyr186 despite (i) the very similar structural architec- ture of OYE1 and PETN reductase, and (ii) geometry for the binding of steroid compounds and 2-cyclohe- xen-1-one [2,19]. The structure of the Y186F enzyme is essentially identical to that of wild-type PETN reduc- tase, which therefore rules out major structural change as a result of mutation that might have been respon- sible for the recruitment of a surrogate proton donor in the oxidative half-reaction. The overall structural similarities of OYE1 and PETN reductase previously led us to propose that Tyr186 (PETN reductase) might function in a role analogous to that of Tyr196 (OYE1) [2]. We now conclude, however, that as with MR [25,36], proton transfer is most likely from water. This therefore highlights subtle mechanistic differences between family members despite their overall structural similarity. No significant differences in position of Y186 ⁄ 196 and nearby residues can be observed for PETN reductase and available OYE structures (PDB codes 1OYA, 1OYB). Detailed comparison of water molecules in or near the active site is precluded due to either lack of corresponding complexes with identical molecules bound in the active site or sufficiently high resolution data for the OYE complexes. The structure of the 2-electron reduced form of PETN reductase bound to progesterone is very similar to that of the oxidized enzyme-progesterone complex. Moreover, in these inhibitor complexes the steroid is bound in the active site in a similar manner to the bind- ing of steroid substrates in oxidized PETN reductase. We noted previously from structures of oxidized PETN reductase in complex with steroid substrates that the C1-C2 reactive bond of the steroid substrate is inappro- priately aligned with the flavin N5 atom to enable hydride transfer from the flavin in the reduced form of the enzyme [2]. This led us to propose that steroid sub- strates would need to flip by 180°, such that the react- ive bond was optimally aligned with the flavin N5 atom in the catalytically relevant 2-electron reduced form of the enzyme-steroid complex. We conjectured that the reduced form of the enzyme might direct binding of the steroid in the flipped conformation, thus facilitating catalysis. Our structure of the 2-electron reduced form of PETN reductase in complex with progesterone, how- ever, indicates this is not the case, and that the reduc- tion state of the flavin does not ‘signal’ a change in the binding geometry. Shifts in substrate position following reduction of a cofactor have been postulated in other enzyme systems, including cytochrome P450 BM3 [37] and nitroreductase [32], in an attempt to reconcile the apparent conflict between the postulated structure of the catalytically active reduced enzyme-substrate com- plex and the observed crystallographic structures of the oxidized enzyme-substrate complex. However, in some cases, the substrates studied are not closely related to the physiological substrates and one cannot assume such compounds are good mimics of the physiological substrate. It is probable that in both redox states the majority of substrate-enzyme complexes adopt the crystallographically determined structure and that the catalytically active conformation is populated to only a small extent and therefore represents a less sta- ble form of the reduced enzyme-substrate complex. The preferred binding mode that we observed in our crystal- lographic studies is catalytically incompetent, and ster- oid must either be released from the active site to allow binding in the active configuration, or flipping of the steroid must occur whilst resident in the active site. An in situ flipping mechanism seems unlikely given the extensive steric clashes and change in steroid conforma- tion that must occur to allow rotation through 180°, and we therefore favour a ‘release-and-rebinding’ mechanism. The low occupancy of the catalytically competent form of the enzyme-substrate complex no doubt contributes to the very low turnover numbers for the PETN reductase-catalysed reduction of steroid substrates [2]. Proton transfer in PETN reductase H. Khan et al. 4668 FEBS Journal 272 (2005) 4660–4671 ª 2005 FEBS Experimental procedures Chemicals All chemicals were of analytical grade where possible. Com- plex bacteriological media were from Unipath Ltd (Basing- stoke, UK), and all media were prepared as described in Sambrook et al. [27]. Mimetic Orange 2 affinity chromato- graphy resin was from Affinity Chromatography Ltd (Cam- bridge, UK). Q-Sepharose resin was from Amersham Biosciences (Piscataway, NJ, USA). NADPH, glucose 6-phosphate dehydrogenase, glucose 6-phosphate, benzyl viologen, methyl viologen, 2-hydroxy-1,4-naphthaquinone, phenazine methosulfate and 2,4 dinitrophenol were from Sigma (St Louis, MO, USA). 2-cyclohexen-1-one was from Acros Organics (Geel, Belgium). S. Nicklin (UK Defence and Evaluation Research Agency) supplied TNT. The following extinction coefficients were used to calculate the concentra- tion of substrates and enzyme: NADPH (e 340 ¼ 6.22 · 10 3 m )1 cm )1 ); PETN reductase (e 464 ¼ 11.3 · 10 3 m )1 cm )1 ). Stock solutions of TNT (600 mm) were made up in acetone. Dilutions were then made into 50 mm potassium phos- phate buffer, pH 7.0, and the acetone concentration was maintained at 1% (v ⁄ v). The presence of acetone in buffers at 1% (v ⁄ v) was shown not to affect enzyme activity. Mutagenesis and purification of enzymes Site-directed mutagenesis was achieved using the Quik- change mutagenesis method (Stratagene, La Jolla, CA, USA) and the following oligonucleotides: 5¢-TTCACTCTG CGCACGGTTTTCTGCTGCATCAGTTC-3¢ (Y186F for- ward primer), 5¢-GAACTGATGCAGCAGAAAACCGTG CGCAGAGTGAA-3¢ (Y186F, reverse primer), 5¢-CTTCG ACCTGGTTGAGCTTGCGTCTGCGCACGGTTACCTG- 3¢ (H181A forward primer), 5¢-CAGGTAACCGTGCGCG ACGCAAGCTCAACCAGGTCGAAG-3¢ (H181A, reverse primer), 5¢-GTTGAGCTTCACTCTGCGGCGGGTTACC TGCTGCATCAG-3¢ (H184, forward primer) and 5¢-CTG ATGCAGCAGGTAACCCGCCGCAGAGTGAAGCTCA AC-3¢ (H184, reverse primer). Plasmid pONR1 [1] was used as template for mutagenesis reactions. All mutant genes were completely sequenced to ensure that spurious changes had not arisen during the mutagenesis reaction. The expres- sion and purification of the wild-type and mutant PETN reductase enzymes was as described previously for wild-type enzyme [1]. Owing to poor retention by the Mimetic Orange 2 affinity chromatography resin used for purifica- tion of wild-type PETN reductase, the H181A and H184A enzymes were purified by Q-Sepharose ion exchange chro- matography (pre-equilibrated with 50 mm Tris ⁄ HCl buffer, pH 7.5; buffer A). The enzymes were subsequently eluted using a salt gradient (0–200 mm NaCl contained in 500 mL buffer A). Following an initial purification using Q-Seph- arose resin a second chromatography step using the same resin was required (gradient of 15–150 mm NaCl contained in 1 L buffer A) to obtain pure protein. Redox potentiometry, ligand binding and stopped-flow kinetic analyses Redox titrations and the determination of redox potential of the enzyme-bound FMN were performed as described previously for wild- type PETN reductase [19]. Ligand binding studies were also as performed previously with wild-type PETN reductase [19], relying on the perturbation of the flavin electronic absorption spectrum on binding lig- and in the active site of PETN reductase. Data were collec- ted in the UV-visible region (250–600 nm), and the absorption at 518 nm plotted as a function of ligand con- centration. Data for the wild-type enzyme were analysed by fitting to the quadratic function (Eqn 1). DA ¼ DA max 2E T ðL T þ E T þ K d ÞÀ ðL T þ E T þ K d Þ 2 Àð4L T E T Þ ÀÁ 0:5 hi ð1Þ where DA max is the maximum absorption change at 518 nm, L T is the total ligand concentration and E T the total enzyme concentration. Data for mutant enzymes were calculated by fitting to the standard hyperbolic expression (Eqn 2). DA ¼ DA max L T K d þ L T ð2Þ Rapid reaction kinetic experiments using single wavelength absorption and photodiode array detection were performed Table 3. Crystallographic data and refinement statistics. Property Y186F Red enzyme– progesterone complex Crystal properties Spacegroup P2 1 2 1 2 1 P2 1 2 1 2 1 Cell dimensions (A ˚ )a¼ 56.6 b ¼ 68.6 c ¼ 88.6 a ¼ 58.2 b ¼ 68.6 c ¼ 88.4 Data collection Completeness (%) 98.2 (84) 99.1 (98.4) Redundancy 4.2 (3.1) 3.8 (3.5) I ⁄ rI 11.1 (1.9) 12.2 (2.2) R sym 0.087 (0.489) 0.067 (0.438) Resolution 15–1.0 (1.03–1.0) 15–1.05 (1.08–1.05) Refinement R factor 0.127 (0.345) 0.117 (0.28) R free 0.147 (0.389) 0.143 (0.296) RMSD bond lenghts (A ˚ ) 0.017 0.016 RMSD bond angles (°) 0.003 0.003 H. Khan et al. Proton transfer in PETN reductase FEBS Journal 272 (2005) 4660–4671 ª 2005 FEBS 4669 [...]... & Clausen T (2005) The 1.3 A crystal structure of the flavoprotein YqjM reveals a novel class of Old Yellow Enzymes J Biol Chem 280, 27904–27913 Messiha H, Bruce N, Sattelle B, Sutcliffe M, Munro A & Scrutton N (2005) Reaction of morphinone reductase with 2-cyclohexen-1-one and 1-nitrocyclohexene: proton donation, ligand binding, and the role of residues Histidine 186 and Asparagine 189 J Biol Chem... Scrutton NS & Moody PC (2002) Crystal structure of bacterial morphinone reductase and properties of the C191A mutant enzyme J Biol Chem 277, 30976–30983 Messiha HL, Munro AW, Bruce NC, Barsukov I & Scrutton NS (2005) Reaction of morphinone reductase with 2-cyclohexen-1-one and 1-nitrocyclohexene: proton donation, ligand binding and the role of residues His186 and Asn-189 J Biol Chem 280, 10695–10709... [29] and refmac [30] The atomic resolution structure (PDB code 1VYR) was used as the starting model for refinement of both structures Coordinates and structure factors have been deposited with the PDB, access codes 2ABA (reduced enzyme in complex with progesterone) and 2ABB (Y186F PETN reductase) Acknowledgements The work was funded by the UK Biotechnology and Biological Sciences Research Council, The. .. H-transfers of the catalytic cycle of morphinone reductase and in the reductive half-reaction of the homologous pentaerythritol tetranitrate reductase J Biol Chem 278, 43973– 43982 Kohli RM & Massey V (1998) The oxidative half-reaction of Old Yellow Enzyme The role of tyrosine 196 J Biol Chem 273, 32763–32770 French CE & Bruce NC (1995) Bacterial morphinone reductase is related to Old Yellow Enzyme Biochem... Crystal structure of pentaerythritol tetranitrate reductase: ‘flipped’ binding geometries for steroid substrates in different redox states of the enzyme J Mol Biol 310, 433–447 ˚ 3 Fox KM & Karplus PA (1994) Old yellow enzyme at 2 A resolution: overall structure, ligand binding, and comparison with related flavoproteins Structure 2, 1089–1105 4 French CE & Bruce NC (1994) Purification and characterization of. .. & Fox BG (2005) X-ray structure of Arabidopsis At2g06050, 12-oxophytodienoate reductase isoform 3 Proteins 58, 243–245 Breithaupt C, Strassner J, Breitinger U, Huber R, Macheroux P, Schaller A & Clausen T (2001) X-ray structure of 12-oxophytodienoate reductase 1 provides structural insight into substrate binding and specificity within the family of OYE Structure 9, 419–429 Kitzing K, Fitzpatrick TB,... resolution structures and solution behavior of enzymesubstrate complexes of Enterobacter cloacae PB2 pentaerythritol tetranitrate reductase Multiple conformational states and implications for the mechanism of nitroaromatic explosive degradation J Biol Chem 279, 30563–30572 Basran J, Harris RJ, Sutcliffe MJ & Scrutton NS (2003) Hydrogen tunneling in the multiple H-transfers of the catalytic cycle of morphinone... sequencing and expression of the gene encoding NADH oxidase from the extreme anaerobic thermophile Thermoanaerobium brockii Biochim Biophys Acta 1174, 187–190 15 Scrutton NS (1994) alpha ⁄ beta barrel evolution and the modular assembly of enzymes: emerging trends in the flavin oxidase ⁄ dehydrogenase family Bioessays 16, 115–122 16 French CE, Nicklin S & Bruce NC (1998) Aerobic degradation of 2,4,6-trinitrotoluene... progesterone The reduced progesterone complex crystals were grown and flash-cooled to 100 K under anaerobic conditions The redox state of the PETN crystals could easily by assessed and verified by bright yellow colour of the oxidized crystals vs the colourless reduced form All data were processed and scaled using the denzo ⁄ scalepack package [28] (Table 3) Model building and refinement were carried out using... acid-inducible NADH: flavin oxidoreductase from Eubacterium sp strain VPI 12708 J Bacteriol 175, 3002–3012 11 Boyd G, Mathews FS, Packman LC & Scrutton NS (1992) Trimethylamine dehydrogenase of bacterium W3A1 Molecular cloning, sequence determination and over-expression of the gene FEBS Lett 308, 271– 276 12 Yang CC, Packman LC & Scrutton NS (1995) The primary structure of Hyphomicrobium X dimethylamine . Proton transfer in the oxidative half-reaction of pentaerythritol tetranitrate reductase Structure of the reduced enzyme-progesterone complex and the roles. from Thermobacillus brockii). The arrows indicate the positions of histidine residues in PETN reductase inferred to be involved in ligand binding and counterpart