Catalytic Properties of Dihydroorotate Dehydrogenase from Saccharomyces cerevisiae: Studies on pH, Alternate Substrates, and Inhibitors

Archives of Biochemistry and Biophysics Vol. 378, No. 1, June 1, pp. 84 –92, 2000 doi:10.1006/abbi.2000.1823, available online at http://www.idealibrary.com on

Catalytic Properties of Dihydroorotate Dehydrogenase from Saccharomyces cerevisiae: Studies on pH, Alternate Substrates, and Inhibitors Douglas B. Jordan,* ,1 John J. Bisaha,† and Michael A. Picollelli* *DuPont Pharmaceutical Company, Experimental Station, Rt. 141 and Henry Clay Road, Wilmington, Delaware 19880-0400; and †DuPont Agricultural Products, Stine-Haskell Research Center, 1094 Elkton Road, P.O. Box 30, Newark, Delaware, 19714

Received October 21, 1999, and in revised form March 17, 2000

Yeast dihydroorotate dehydrogenase (DHOD) was purified 2800-fold to homogeneity from its natural source. Its sequence is 70% identical to that of the Lactococcus lactis DHOD (family IA) and the two active sites are nearly the same. Incubations of the yeast DHOD with dideuterodihydroorotate (deuterated in the positions eliminated in the dehydrogenation) as the donor and [ 14C]orotate as the acceptor revealed that the C5 deuteron exchanged with H 2O solvent at a rate equal to the 14C exchange rate, whereas the C6 deuteron was infrequently exchanged with H 2O solvent, thus indicating that the C6 deuteron of the dihydroorotate is sticky on the flavin cofactor. The pH dependencies of the steady-state parameters (k cat and k cat/K m) are similar, indicating that k cat/K m reports the productive binding of substrate, and the parameters are dependent on the donor–acceptor pair. The lower pK a values for k cat and k cat/K m observed for substrate dihydroorotate (around 6) in comparison to the values determined for dihydrooxonate (around 8) suggest that the C5 pro S hydrogen atom of dihydroorotate (but not the analogous hydrogen of dihydrooxonate), which is removed in the dehydrogenation, assists in lowering the pK a of the active site base (Cys133). The pH dependencies of the kinetic isotope effects on steady-state parameters observed for the dideuterated dihydroorotate are consistent with the dehydrogenation of substrate being rate limiting at low pH values, with a pK a value approximating that assigned to Cys133. Electron acceptors with dihydroorotate as donor were preferred in the following order: ferricyanide (1), DCPIP (0.54), Q o (0.28), fumarate (0.15), and O 2 (0.035). Orotate inhibition profiles versus varied con1 To whom correspondence should be addressed. Fax: (302) 6954083. E-mail: [email protected].

84

centrations of dihydroorotate with ferricyanide or O 2 as acceptors suggest that both orotate and dihydroorotate have significant affinities for the reduced and oxidized forms of the enzyme. © 2000 Academic Press Key Words: dihydroorotate dehydrogenase; substrate specificity; Saccharomyces cerevisiae; deuterium; kinetic isotope effects; flavoprotein dehydrogenase.

Dihydroorotate dehydrogenase (DHOD) 2 follows the equilibrium step catalyzed by dihydroorotase for driving the de novo pyrimidine biosynthetic pathway forward in synthesizing the precursors of RNA and DNA (1). On the basis of its critical physiological role, the enzyme has been targeted for inhibition in antimicrobial (2), anticancer (3), and other medically relevant areas (4, 5). The DHOD catalyst has adapted multiple forms depending on the life form of the organism; so far type I (A and B) and type II enzymes have been identified (6 –12). The DHOD forms are considered to be constant in their reductive half-reaction, with the hydride being transferred from dihydroorotate to a flavin cofactor (Fig. 1). The major differences between the dehydrogenases exist with respect to their electron acceptors (internal and external) and their solubility (membrane bound for the type II enzymes). The simplest form of the enzyme is DHODA (type I), which is soluble and contains one FMN per active site; the yeast 2 Abbreviations used: DHOD, dihydroorotate dehydrogenase; DHO, (S)-dihydroorotate (4 of Fig. 2); D 2-DHO, dideuterated DHO (2 of Fig. 2); DHODA, dihydroorotate dehydrogenase (type IA); DHX, dihydrooxonate (5 of Fig. 2); DCPIP, 2,6-dichlorophenolindophenol; Q o, 2,3-dimethoxy-5-methyl-1,4-benzoquinone; KIE, kinetic isotope effect; dpm, disintegrations per minute; FMN, flavin mononucleotide.

0003-9861/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.

DIHYDROOROTATE DEHYDROGENASE FROM Saccharomyces cerevisiae

85

EXPERIMENTAL PROCEDURES

FIG. 1. Oxidation of dihydroorotate to orotate by the flavoprotein dihydroorotate dehydrogenase.

DHODA is reported to use fumarate as its physiological electron acceptor in studies on crude enzyme preparations (13). The DHODA isolated from Lactococcus lactis is the most characterized enzyme with respect to structural information and mutant enzyme studies on the involvement of active-site residues in catalysis (8, 10, 11). There are two X-ray structures of the L. lactis DHODA available: one of the resting enzyme (8) and one of the dead-end complex with product orotate (11). The type B DHOD from L. lactis is an ␣ 2␤ 2 heterotetramer, with the second subunit serving to funnel electrons to NAD ⫹ (9). The seminal work on the catalytic properties of DHODA isolated from Crithidia fasciculata suggested that the enzyme-catalyzed dehydrogenation of DHO is a two-step process, with the removal of the C5 proton occurring prior to the hydride transfer (14). In this work we describe our study of the DHODA isolated from Saccharomyces cerevisiae to examine the properties of the catalyzed reaction in the context of previous studies on other DHODA enzymes. The yeast enzyme is approximately 70% identical in sequence to the L. lactis enzyme and so a homology model may be easily constructed. Our model of the yeast DHODA places the C5 pro S hydrogen of DHO 1.6 Å from the sulfur atom of Cys133, which is thought to act as the active-site base. The model of the yeast enzyme places the analogous hydrogen atom of dihyrooxonate (DHX) over 1 Å further away from the sulfur atom. This structural difference is examined through steady-state kinetic analyses of the pH profiles for substrates DHO and DHX. Also, we reexamine the experiments that suggested a two-step mechanism for the dehydration of DHO as reported for the C. fasciculata DHODA (14). Third, electron acceptors for the homogeneous DHODA isolated from S. cerevisiae are studied with DHO as the donor including the reported physiological acceptor fumarate (13). Finally, we report steady-state kinetic and inhibition constants for the yeast DHODA purified from its natural source (blocks of baker’s yeast) as a reference for the heterologously expressed DHODA enzymes.

General. Orotate, (S)-dihydroorotate, oxonate, and other analogs of orotate were obtained from Sigma. [ 14C]Orotate, labeled in the carboxylate, was obtained from NEN. DHX was prepared from oxonate and sodium amalgam as described (15) and the diethylamine salt was generated to ensure its neutrality and solubility in water. Dideuterated (S)-dihydroorotate (2 of Fig. 2) was prepared using Clostridium oroticum DHOD (Sigma) in an incubation with NADH and orotate in D 2O and the material was purified as described (14). 1 H NMR of samples (all in D 2O) were analyzed using a Bruker 360 instrument. Radioactivity in samples was determined by scintillation counting using a LS3801 instrument (Beckman) with an external standard method for correcting for counting efficiency. UV–VIS measurements were obtained using a temperature-controlled HP 8542A spectrophotometer (Hewlett–Packard). The identity of the flavin bound to the yeast DHOD was determined by using a fluorescence method (16) with measurements on an SLM-8000 instrument. SDS–PAGE analyses of proteins were conducted using a PhastSystem (Pharmacia, Uppsala, Sweden). Concentrations of the donor substrates were determined by an end-point method using yeast DHOD with ferricyanide as the acceptor, monitoring the final absorbancy at 420 nm. Kinetic data were fitted to equations using the nonlinear fitting procedures within the computer program RS1 (BBN Research Systems, Cambridge, MA). Hydrodynamic measurements on the protein were conducted using the sucrose-density-gradient centrifugation method of Martin and Ames (17) with bovine serum albumin as the standard 280-nm absorbance reporter and molecular weight indicator for the fractions collected after centrifugation and the catalytic activity of DHOD as its reporter. Homology modeling of the three-dimensional structure of yeast DHOD was through the computer program Biopolymer within the suite of programs of Sybyl (Tripos, St. Louis). Amino acids of the S. cerevisiae DHOD (18) were mutated individually upon the X-ray structure of the L. lactis DHOD (PDB Accession Code 2DOR (10)) followed by a gas-phase minimization, keeping the protein backbone rigid (as an immovable agregate). Orotate was changed to DHO or DHX, which was minimized in the gas phase for positioning within the active site of the yeast enzyme. Purification of S. cerevisiae DHODA. All steps were at 2– 4°C. Locally obtained baker’s yeast (1.0 kg) was suspended in 2.0 l of 20 mM Hepes–NaOH, 0.1 mM EDTA, 1.0 mM dithiothreitol, pH 7.0 (Buffer H) plus 1 mg/l leupetin. The 3.0-L slurry was passed through a microfluidizer instrument (Model 110Y, Microfluidics) at maximum pressure four times with 5-min cooling periods on wet ice between passes. The lysed cells were centrifuged 45 min at 24,000g. The pH of the supernatant (1.8 L) was adjusted to 4.7 by drop-wise addition of 1.7 M acetic acid (with stirring) followed by centrifugation for 30 min at 24,000g. The pH of the supernatant (1.6 L) was adjusted to 7.0 by drop-wise addition of 1.3 M ammonium hydroxide (with stirring). The pH-neutralized sample was loaded onto a column (5 ⫻ 25 cm) of Orange-A resin (Amicon) equilibrated with Buffer H;

FIG. 2. Pyrimidines discussed in the text. Dihydroorotate (4) and orotate (1) are the natural substrate and product, repectively, in the biosynthetic pathway.

86

JORDAN, BISAHA, AND PICOLLELLI

the column was washed with Buffer H until the 280-nm absorbance was near zero. DHOD was eluted by including 1.0 mM orotic acid in the buffer. Catalytically active fractions were pooled (150 ml) and loaded onto a column (2 ⫻ 25 cm) packed with 2⬘,5⬘-ADP (Pharmacia) equilibrated and developed with Buffer H; this was a reverse-affinity step. DHOD activity was eluted immediately after the void volume and it was collected in 200 ml. The volume was reduced to 40 ml using a pressure cell (Amicon) and the material was loaded onto a 25-ml mono Q column (Pharmacia) equilibrated with 20 mM Hepes– NaOH, 20% glycerol, pH 7.0. The column was developed with a linear gradient included in the equilibration buffer ranging from 0 to 0.5 M NaCl; active fractions were pooled, concentrated to 4.0 ml by using a stirred pressure cell (Amicon), diluted 1:1 by volume with glycerol, and stored at ⫺20°C until use. The final yield was 35% recovery of total catalytic activity from the crude extract and 4.8 mg protein according to a dye-binding assay (19), which was in excellent agreement with the flavin content (0.14 ␮mol) and the calculated molecular weight of the protein (18). Incubations of yeast DHOD with D 2-DHO. The reactions (2.6 ml) included 15.1 ␮mol D 2-DHO (2 of Fig. 2), 1.2 ␮mol [ 14C]orotate (10 ␮Ci), and 130 ␮mol Tris–HCl, pH 7.9. The serum-stoppered reaction mixtures were flushed with argon extensively before the addition of argon-equilibrated yeast DHOD (15 ␮g protein ⫽ 0.43 nmol active sites in 10 ␮l). Incubations (⫾enzyme) were quenched with 0.1 ml of 1 N HCl after 3.5 or 24 h of incubation, neutralized with Tris base, and diluted to 11 ml with water. Ten milliliters of the dilution was loaded onto a 10-ml (by volume) AG1X8 (formate form) column (Bio-Rad) in H 2O; 0.5 ml of the dilution was added to 10 ml of Aquasol scintillation cocktail for determination of radioactivity. The column was washed with water (30 ml) to remove salts. DHO was eluted with 70 mM ammonium formate, pH 3.5, and 5-ml fractions were collected till the end of 14C counts. Orotate was eluted from the column with 100 ml of 1 N HCl. Radioactivity measurements were from aliquots (0.1 ml) of the fractions and subjected to the procedures outlined above. The DHO-containing fractions were further subjected to a AG50 (Bio-Rad) column (5 ml) in H 2O to remove ammonium and other cations. The column was washed with 30 ml of water to elute DHO, which was collected, dried, and prepared for 1H NMR analyses in D 2O. Steady-state kinetic studies. All initial rates were measured by continuous spectrophotometic methods (1 ml, 1-cm path length, 10 s) at 25°C and were initiated with enzyme. All rates reported in this work are for the 2 e ⫺ oxidation of DHO or its analogs and expressed in per second units according to the concentration of the yeast DHOD active sites present in the reactions, as estimated from the protein’s absorbance at 450 nm and assuming an extinction coefficient of 12,200 M ⫺1 cm ⫺1 for the purified flavoprotein. Various pH reactions with DHO, D 2-DHO, or DHX as the varied substrate were conducted in 50 mM Hepes, 50 mM Mes, 100 mM ethanolamine, in the absence or presence of 1.0 mM ferricyanide and with the pH adjusted with HCl or NaOH. Initial velocities were measured continuously on the spectrophotometer at 280 nm for the DHO–ferricyanide and D 2DHO–ferricyanide couples using a delta extinction coefficient (products minus substrates) of ⫺7200 M ⫺1 cm ⫺1 at pH 7.0 with minor corrections at higher pH values (maximum correction ⫽ 3%); for the DHO–O 2 couple, reactions were monitored at 280 nm and the delta extinction coefficient used was ⫺7500 M ⫺1 cm ⫺1 with no corrections for pH; for the DHX–ferricyanide couple the initial rates were monitored at 420 nM using a delta extinction coefficient of 1020 M ⫺1 cm ⫺1. Initial velocities were fitted to the equation k cat A v⫽ , K⫹A

[1]

where v is the observed initial velocity of the reactions, k cat is maximum velocity, A is the varied substrate concentration, and K is the Michaelis constant.

Competitive inhibition constants for orotate at varied pH values were determined for the DHO–ferricyanide couple as above by including varying concentrations of orotate. The initial rate data were fitted to the equation

v⫽

k cat A , A ⫹ K共1 ⫹ I/K i兲

[2]

where I and K i are the inhibitor concentration and the inhibition constant, respectively, and the other definitions are as for Eq. [1]. Initial rate data were also fitted to the equation

v⫽

k cat A , A共1 ⫹ I/K ii兲 ⫹ K共1 ⫹ I/K is兲

[3]

which describes noncompetitive inhibition. K ii and K is are the inhibition constants for the intercept and slope, respectively, and the other definitions are as for Eq. [2]. Kinetic parameters were fitted to equations describing the titration of a single ionizable group causing a decrease on the basic side, the acidic side, and the combination of the two ionizable groups, respectively,

p⫽

P 1 ⫹ K b /关H ⫹ 兴

[4]

p⫽

P 1 ⫹ 关H ⫹ 兴/K a

[5]

p⫽

P , 1 ⫹ K b /关H ⫹ 兴 ⫹ 关H ⫹ 兴/K a

[6]

whereas p represents the measured parameter, P is the pH-independent value for the parameter, [H ⫹ ] is the proton concentration, and K a and K b are the dissociation constants for the groups affecting catalysis. When electron acceptors were being compared, the reactions included 2.0 mM DHO in 50 mM Tris–HCl, pH 8.0. A delta extinction coefficient (reduced minus oxidized) at 600 nm of ⫺22,000 M ⫺1 cm ⫺1 was used for DCPIP. A delta extinction coefficient (reduced minus oxidized) at 420 nm of ⫺1020 M ⫺1 cm ⫺1 was used for ferricyanide. When the parameters for the DHO-fumarate couple were determined, the reactions were in 50 mM Tris–HCl, pH 7.5, at 25°C and they were monitored at 280 nm with a delta extinction coefficient (products minus substrates) of 7240 M ⫺1 cm ⫺1. IC 50 values were determined in reactions that contained 75 mM Tris–HCl, pH 8.0, 10% DMSO, 0.050 mM DHO, and 0.2 mM ferricyanide in the absence or presence of the compounds examined. Reactions were initiated with enzyme and monitored continuously at 280 nm as above. IC 50 values were determined by fitting the initialrate data to the equation

v⫽

Vo , 1 ⫹ I/IC50

[7]

where v is the observed velocity, V o is the uninhibited velocity, I is the inhibitor concentration, and IC 50 is the concentration of inhibitor affording 50% inhibition.

DIHYDROOROTATE DEHYDROGENASE FROM Saccharomyces cerevisiae

87

FIG. 3. Views of dihydroorotate and dihydrooxonate in the active site of yeast dihydroorotate dehydrogenase. (A) Dihydroorotate. (B) Dihydrooxonate.

RESULTS

DHOD isolated from S. cerevisiae. The yeast DHOD was purified 2800-fold from its natural source with the affinity step of chromatography on Orange A resin providing nearly a 2000-fold enrichment. A single band was observed on SDS–PAGE analysis (with either Coomassie blue or silver staining) and it was located at 35 kDa, in agreement with the value of 34,800 Da calculated from the yeast DHOD protein sequence (18). Yeast DHOD had a sedimentation rate similar to that of bovine serum albumin in sucrosedensity-gradient centrifugation studies, which determined a molecular weight of 72 kDa for the holoenzyme; thus the enzyme is a homodimer. Flavin analysis using a fluorescence method determined that the enzyme contains one (0.88 and 1.03 in two determinations) FMN per subunit. The sequence of yeast DHOD is 70% identical to that of the L. lactis DHODA without there being gaps in the

FIG. 4. pH dependence of the steady-state kinetic parameters for yeast DHOD with DHO as the varied substrate in the DHO–ferricyanide couple. Units for k cat and k cat/K m are s ⫺1 and s ⫺1 mM ⫺1, respectively. Curves are drawn from the fits to Eq. [6].

alignment; the yeast enzyme has a three-amino-acid residue extension at the N terminus in comparison to the L. lactis DHODA. A homology model for the yeast enzyme is easily constructed from the coordinates of L. lactis DHODA. The coordinates of such a homology model of the yeast enzyme are available through Swiss-Model (20). Within a 4-Å radius of the orotate and the FMN molecules in our model of the yeast DHOD, the two active sites differ by two isoleucine for valine substitutions, with the other 25 residues remaining the same. Substitution of orotate with DHO in the yeast active site has the elements of hydrogen that are removed from DHO in the enzyme-catalyzed reaction (the pro S hydrogen from C5 and the hydrogen from C6) in a nearly perfect trans-diaxial orientation for efficient removal through a concerted reaction (Fig. 3A). 3 In our model, the sulfur atom of Cys133 (yeast numbering, Cys130 for the L. lactis DHOD) is in a nearly ideal position to remove the C5 pro S proton and the N5 position of the FMN is in a nearly ideal position to accept the C6 hydride. Substitution of orotate with DHX in the yeast active site has the flat ring of DHX in a good orientation to deliver the hydride to the N5 position of the flavin but in a much poorer orientation (in comparison to DHO) for delivering the proton to the sulfur atom of the cysteine (Fig. 3B). pH studies on the electron donors dihydroorotate and dihydrooxonate. The influence of pH on steady-state kinetic parameters was studied for the substrate pairs DHO–ferricyanide (Fig. 4), DHO–O 2 (Fig. 5), and DHX–ferricyanide (Fig. 6). DHO or DHX concentrations were varied when ferricyanide was the acceptor. In the case of the DHO–O 2 substrate pair, the DHO 3

For a stereo view of a model of the L. lactis enzyme active site complexed with DHO see Fig. 5 of reference 11.

88

JORDAN, BISAHA, AND PICOLLELLI

FIG. 5. pH dependence of the k cat for yeast DHOD with DHO as the varied substrate in the DHO–O 2 couple. Units for k cat are s ⫺1. The curve is drawn from the fits to Eq. [5].

concentration was held at a saturating concentration because the very low K m value for DHO (⬃3 ␮M at pH 8.0) precludes accurate determinations of k cat/K m values. The pH-independent kinetic parameters indicate that the DHO–ferricyanide substrate pair is superior to the other two substrate pairs (Table I). Yeast DHOD is sixfold more specific for DHO over DHX on the basis of the pH-independent k cat/K m values. It should be noted that the equilibrium constant for the DHO– oxonate couple has been determined experimentally and oxonate was found to be a weaker electron acceptor than orotate (22). As well, the pH dependencies are considerably different for the three donor–acceptor pairs: there is a drop in k cat/K m values on the basic side for the DHO–ferricyanide pair but not for the other two substrate pairs; the pK a values are similar and about 6 for the DHO–ferricyanide and DHO–O 2 pairs but about 8 for the DHX–ferricyanide pair. Alternate electron acceptors. At pH 7.5 and in the presence of 2.0 mM DHO, the yeast DHOD prefers electron acceptors in the following relative order: 1.0 mM ferricyanide (1.0); 0.10 mM DCPIP (0.54); 0.50 mM Q 0 (0.28); 2 mM fumarate (0.15); 0.25 mM O 2 (0.035). The steady-state kinetic parameters for the DHO–fumarate substrate pair at pH 7.5 are as follows: with 2 mM fumarate and varied DHO, k cat ⫽ 25 ⫾ 1 s ⫺1, k cat/K m ⫽ 3.0 ⫾ 0.2 s ⫺1 ␮M ⫺1, and K m ⫽ 8.4 ⫾ 0.2 ␮M; with 1 mM DHO and varied fumarate, k cat ⫽ 23 ⫾ 0.6 s ⫺1, k cat/K m ⫽ 0.51 ⫾ 0.03 s ⫺1 ␮M ⫺1, and K m ⫽ 45 ⫾ 3 ␮M. Studies on D 2-DHO. Kinetic isotope effects on the steady-state parameters observed for the D 2-DHO (deuterated in the elements of D 2 removed in the DHOD-catalyzed dehydrogenation) are pH dependent (Fig. 7). As the pH increases, the observed KIE values decrease to 1.0; the pK b value for the KIE is approximatly 7. The KIE determined is a combination of the hydride transfer to the flavin, proton transfer, orotate release, and subsequent oxidation of the flavin by ferricyanide. The kinetic isotope effects determined are made more complicated for interpretation by our observation (see below) that the deuterium at the C6 position of D 2-DHO is retained in an exchange experi-

ment using [ 14C]orotate as the acceptor. That is, the deuterium at the C6 position of DHO affects the hydride transfer from the substrate (DHO) and the electron transfer from the reduced (and deuterated) flavin cofactor. In the exchange experiments that included 5.8 mM D 2-DHO and 0.46 mM [ 14C]orotate at pH 7.85 under argon, it was found that in the absence of enzyme after 24 h, 3.3% of the dpm was in the DHO fraction and 96.7% of the dpm was in the orotate fraction (recovery of total dpm analyzed in comparison to the column load ⫽ 106%); in the presence of yeast DHO (15 ␮g ⫽ 0.43 nmol active sites per 2.6 ml) the 14C composition after 3.5 h of the incubation was 89.2% DHO and 10.8% orotate (recovery of total dpm analyzed in comparison to the column load ⫽ 107%); in the presence of yeast DHO (15 ␮g ⫽ 0.43 nmol active sites) the 14C composition after 24 h of incubation was 87.8% DHO and 12.2% orotate (recovery of total dpm analyzed in comparison to the column load ⫽ 97%). The ratio of DHOto-orotate on the basis of dpm equals 0.033 for the 24-h incubation in the absence of enzyme; the dpm ratio was 8.3 for the 3.5-h incubation in the presence of enzyme; and the ratio was 7.2 for the 24-h incubation in the presence of enzyme. There must have been some O 2 contamination in the enzyme-containing incubations because the calculated 14C equilibrium constant ([ 14C]DHO/[ 14C]orotate from the molar concentrations in the incubations) is 12.6 and the lowering of the ratio with the time of incubation (from 3.5 to 24 h with enzyme) would account for it: O 2 contamination in the reactions would increase the concentration (and dpm determined) of orotate because O 2 is an electron acceptor for the reduced flavoprotein. The DHO-containing fractions of the three incubations were analyzed by 1H NMR. The incubation in the absence of enzyme contained 100% of 2 (Fig. 2). In the presence of yeast DHOD after 3.5 h of incubation, the

FIG. 6. pH dependence of the steady-state kinetic parameters for yeast DHOD with DHO as the varied substrate in the DHX–ferricyanide couple. Units for k cat and k cat/K m are s ⫺1 and s ⫺1 mM ⫺1, respectively. Curves are drawn from the fits to Eq. [5].

89

DIHYDROOROTATE DEHYDROGENASE FROM Saccharomyces cerevisiae TABLE I

Influence of pH on the Steady-State Kinetic Parameters of Yeast DHOD Substrate pair Parameter

DHO–ferricyanide a

DHO–O 2 b

DHX–ferricyanide c

k cat (s ⫺1) d k cat/K m (␮M ⫺1 s ⫺1) d pK a (k cat) pK b (k cat) pK a (k cat/K m) pK b (k cat/K m)

190 ⫾ 20 6.9 ⫾ 0.8 5.9 ⫾ 0.07 8.6 ⫾ 0.07 6.7 ⫾ 0.07 9.3 ⫾ 0.2

6.8 ⫾ 0.1 ND e 6.5 ⫾ 0.02 ND e ND e ND e

39 ⫾ 0.9 1.1 ⫾ 0.09 8.0 ⫾ 0.02 ND e 8.5 ⫾ 0.06 ND e

a

Values from the fitting of data in Fig. 4. Values from the fitting of data in Fig. 5. c Values from the fitting of data in Fig. 6. d pH-independent values. e ND, not determined. b

1

H NMR analysis indicated the presence of 29% of 2, 50% of 3, and 21% of 4. In the presence of yeast DHOD after 24 h of incubation, the 1H NMR analysis indicated the presence of 0% of 2, 31% of 3, and 69% of 4. The ratio of [ 14C]DHO to [ 14C]orotate dpm was 8.3 in the 3.5-h incubation including enzyme and this value indicates that the exchange reaction had reached 66% completion toward isotopic equilibrium between the two species (ratio ⫽ 12.6 at equilibrium) as expected from the concentrations of DHO and orotate in the incubation. When one fits the data to a single exponential with a zero Y intercept, these data yield a half-life of 2.2 h for the exchange of the radiolabel. On the basis of a half-life of 2.2 h, the 24-h incubation should have reached 99% of equilibrium with respect to [ 14C]DHO and [ 14C]orotate. After 3.5 h of incubation, 71% of the deuterium label from the C5 position of 2 had exchanged for hydrogen and this value is considered equivalent to the value of 66% obtained for the radioactive isotope exchange given the errors in the measurements. For the C6 position of 2, only 8 and 69% of the deuterium exchanged for hydrogen in the incuba-

FIG. 7. pH dependence of the kinetic isotope effects determined for the steady-state kinetic parameters for yeast DHOD with DHO or D 2-DHO as the varied substrate in the DHO–ferricyanide couple. k cat, solid symbols; k cat/K m, open symbols.

tion with enzyme after 3.5 and 24 h, respectively. From the 3.5- and 24-h data the half-life for the exchange of the C6 deuterium for hydrogen is 14 h, approximately sixfold slower than the rate of radioisotope exchange in the reaction. Inhibition of the yeast DHOD by orotate and its analogs. Orotate analogs were studied as inhibitors against the yeast DHOD at pH 8.0 using DHO (50 ␮M) as the donor and ferricyanide as the oxidant. Under these conditions, product orotate is a modest inhibitor having an IC 50 value of 75 ⫾ 12 ␮M. Orotate analogs had the following IC 50 values: 5-nitroorotate, 76 ⫾ 2.6 ␮M; 2-thioorotate, 18 ⫾ 1.1 ␮M; oxonic acid, 9.9 ⫾ 0.4 ␮M; 5-bromoorotate, 16 ⫾ 0.6 ␮M; 5-fluoroorotate, 55 ⫾ 8.6 ␮M; orotate methyl ester, 710 ⫾ 160 ␮M; 5-aminoorotate, 1600 ⫾ 270 ␮M; and uracil, 2500 ⫾ 490 ␮M. Further, it was found that orotate was a competitive inhibitor with respect to substrate DHO (Fig. 8A) and its K i value was pH dependent (Fig. 9). When the data of Fig. 8A are fitted to Eq. [3], which describes noncompetitive inhibition, the determined K ii value is very large and insignificant in comparison to the K is value (K ii ⫽ 47 ⫾ 28 M; K is ⫽ 10.4 ⫾ 1.1 ␮M). The pH-independent value was not obtained because of the limits of the pH values examined, but at the highest pH studied (pH 9.05) the K i value determined was 5.7 ⫾ 0.9 ␮M. If one assumes that the other inhibitors are competitive with respect to substrate, then their respective IC 50 values may be divided by 4 to generate K i values because the K m for DHO, under these conditions, is 16 ⫾ 1.6 ␮M at pH 8.0. The K i value determined for orotate at pH 8.0, under similar conditions, is 10.2 ⫾ 0.9 ␮M (Fig. 9), which compares favorably with the K D value of 13 ␮M for the oxidized L. lactis DHODA determined by monitoring spectral perturbations of the enzyme-bound flavin (10).

90

JORDAN, BISAHA, AND PICOLLELLI

FIG. 8. Inhibition of yeast DHOD by orotate at varying DHO concentrations and pH 8.0. Units for k cat are s ⫺1. Units for [DHO] are mM. (A) The DHO–ferricyanide couple. Concentrations of orotate are zero (solid circles), 10 ␮M (open circles), 40 ␮M (solid squares), and 100 ␮M (open squares). Curves are drawn from the fits of the data to Eq. [2]. (B) The DHO–O 2 couple. Concentrations of orotate are zero (solid circles), 5.0 ␮M (open circles), 20 ␮M (solid squares), and 50 ␮M (open squares). Curves are drawn to connect the data points as a visual aid.

It was further found that orotate produced curvilinear double reciprocal plots when it was an inhibitor of the yeast DHOD with DHO as the electron donor and the weak electron acceptor O 2 at pH 8.0 (Fig. 8B). At low concentrations of DHO, the plots approximate a noncompetitive pattern. High concentrations of DHO displace orotate from its inhibitory position. DISCUSSION

pH Studies. The pH profiles for the DHO–ferricyanide couple, the DHO–O 2 couple, and the DHX–ferricyanide couple provide kinetic evidence for the involvement of the C5 pro S hydrogen atom of DHO, but not the analogous atom of DHX, in assisting the deprotonation of the SH group of Cys133. A deprotonated Cys133 is thought to serve as a general base for removing the C5 pro S hydrogen atom from DHO in the dehydrogenation reaction. The observed pH dependence of k cat in all three substrate pairs reflects a deprotonation step in catalysis. Since the pH profiles of k cat/K m are similar to those of k cat, the pK a of the former parameter very likely reports the formation of a productive enzyme–substrate complex rather than the deprotonation of the free enzyme or substrate to allow binding between the two (24). Indeed, the pK a values

for the ring nitrogens of DHO, DHX, and orotate are above pH 10 and, therefore, removed from consideration in the assignment of the pK a. The pK a values for the kinetic parameters determined for DHO with O 2 or ferricyanide as the acceptor are around 6, whereas the pK a values for the kinetic parameters determined for DHX with ferricyanide as the acceptor are around 8. The differences in pK a values of the kinetic parameters between substrates suggest that DHO (but not DHX) lowers the pK a of Cys133. DHO has a puckered ring configuration, which orients the C5 pro S hydrogen toward the sulfur atom of Cys133 with a distance of 1.6 Å between the two atoms, whereas the ring of DHX is flat and its analogous hydrogen atom is not displayed toward the sulfur of Cys133 so the distance between the two atoms is 2.7 Å (Fig. 3). The proximity differences between the substrates and the sulfur of Cys133 could account for the differences in pK a values. Assignment of the pK a values to Cys133 is supported by the pH dependence of the KIE values determined for D 2DHO (Fig. 7). KIE values decline with increasing pH, which likely reports the reductive half-reaction becoming less rate determining with increased ionization of Cys133. The pK b values of around 9 for the kinetic parameters in the DHO–ferricyanide couple could report the neutralization of either Lys167 or Lys46 side-chain amines. Lys167 is positioned near the flavin’s N1 and the adjacent carbonyl oxygen, and it has a potential role of stabilizing the reduced flavin. Lys46 is near the carboxyl group of DHO (or orotate) and is thought to contribute to productive binding (10). An alternative explanation is that the reduction of ferricyanide becomes rate limiting at elevated pH values and the pK b values of around 9 report the lowered redox potential for ferricyanide in driving the oxidative half-reaction. The reason for the absence of drops on the basic side for the kinetic parameters determined for the DHO–O 2 and DHX–ferricyanide couples is uncertain, but could result from the limited pH range studied and from lower catalytic rates for the donor–acceptor pairs. The DHX–ferricyanide couple could mask a rate drop on the basic side because the pK a values (assigned to the

FIG. 9. pH dependence of the K i value of orotate for yeast DHOD acting on the DHO–ferricyanide couple. Units of K i values are ␮M. A curve is not drawn for the fitted data because of the poor fit to Eq. [4] (there is less than one proton in the drop of K i values).

DIHYDROOROTATE DEHYDROGENASE FROM Saccharomyces cerevisiae

SH of Cys133) are high (8.0 for k cat and 8.5 for k cat/K m) and close to the highest pH studied. A sticky deuterium. The seminal work on the catalytic properties of the C. fasciculata DHOD (whose sequence is unavailable, yet this enzyme must belong to family IA of the DHOD tree because of its properties that resemble those of the L. lactis DHODA and the S. cerevisiae DHOD) reported by Pascal and Walsh suggested a two-step mechanism in the reductive halfreaction with proton removal from the C5 of dihydroorotate constituting the initial step (14). They found that upon incubation of the enzyme with 2 under anaerobic conditions in H 2O that the C6 deuteron exchanged more slowly with solvent than the pro S deuteron at C5 (14). The latter result suggested that the C5 pro S proton is removed prior to the hydride transfer from C6 in the overall transfer of H 2. In the same work it was pointed out that the differential exchange rates of deuterons for protons between C5 and C6 of 2 could also be explained by a sticky deuterium atom as transferred from the C6 position of 2 to the N5 position of FMN and then retained as a deuterium in the back transfer to an orotate molecule. The sticky deuterium idea seemed unlikely because of the method by which 2 was prepared for their work (and this work) that depended on solvent exchange (reaction in D 2O) of the hydrogen atom for deuterium on the Zymobacterium oroticum DHOD so a deuteride could be transferred to the C6 position of orotate to produce a deuterium at C6 of 2. In this work we demonstrate that the deuterium atom from the substrate used to reduce the oxidized FMN of the yeast DHOD is relatively slow to exchange with bulk water in comparison to the rate of its capture by an orotate molecule. Previously we reported experiments that used 5-fluoroorotate as the acceptor of the deuterium instead of [ 14C]orotate as described in this work (23). The two methodologies used for detecting the sticky deuterium atom on the yeast DHOD catalyst complement one another in that the 5-fluoorotate acceptor study had a 15-fold greater concentration of yeast DHOD in the incubations in comparison to the incubations described in this work. Also, the concentration of 5-fluoorotate was 13-fold greater than the concentration of [ 14C]orotate. Comparisons of the rates of exchange with the H 2O solvent with respect to the deuterium atom transferred from the C6 position of D 2-DHO to the two carbon atom acceptors (5-fluoroorotate and [ 14C]orotate) indicate that the solvent exchange rate is inversely dependent on the concentration of the electron acceptor. The exchange rate of the C5 deuterium of 2 with H 2O was found to equal the rate of oxidation of 2 (this work). Any steady-state analysis of kinetic isotope effects on the transfer of the hydride from C6 to an acceptor must consider the slow exchange of the hydrogen because the measured ki-

91

netic isotope effect will likely reflect two processes: the transfer of a deuteride to the flavin and the oxidation of the reduced and deuterated flavin by the acceptor. Inhibition by orotate. Competitive inhibition by orotate versus DHO with ferricyanide as the electron acceptor (Fig. 8A) violates the rule for a classical pingpong– bi-bi kinetic mechanism where the product (orotate) does not bind to the same enzyme form as the reactant (DHO) and noncompetitive inhibition is seen at unsaturating concentrations of the second substrate (21). In the incubations, the strong electron acceptor, ferricyanide, was at saturating concentrations where the concentration of the oxidized enzyme dominates and no inhibition by orotate is expected in a classical ping-pong– bi-bi kinetic mechanism. Competitive inhibition by orotate with respect to DHO suggests that the oxidized enzyme has a significant affinity for orotate. Certainly, there is room for orotate in the active site of the resting enzyme (oxidized) as proven by the X-ray structure of the L. lactis DHODA– orotate binary complex (11). Shifts in the oxidized-enzyme flavin absorption spectrum of the L. lactis DHODA upon titration with orotate determined K d values for orotate that are similar to the K i values reported in this work (10). Noncompetitive inhibition by orotate versus DHO was observed with the L. lactis DHODA when the second substrate was unsaturating and there was a weaker electron acceptor than ferricyanide (10). The weaker electron acceptor favors population of the reduced enzyme in the steady-state allowing for orotate to bind to a form of the enzyme which has low affinity for DHO as in the classical case. There were curvilinear double reciprocal plots with orotate as the inhibitor, DHO as the varied substrate, and O 2 as the electron acceptor (Fig. 8B). Dioxygen is a weak electron acceptor for yeast DHOD and the reduced form of the enzyme dominates in the steady state. The observed inhibition pattern is consistent with orotate binding to the reduced form of the enzyme. At low concentrations of DHO the inhibition pattern of Fig. 8B appears noncompetitive, which is the expected pattern for product inhibition in a classical ping-pong– bi-bi kinetic mechanism (21). It is not until very high concentrations of DHO are reached that inhibition by orotate is relieved. It is likely that substrate DHO binds to both the oxidized and reduced forms of the enzyme; DHO would prefer the oxidized form (just as orotate would prefer the reduced enzyme) and the different affinities for DHO by the reduced and oxidized enzyme forms would account for the nonlinear inhibition patterns. In this scenario, DHO (high affinity to the oxidized enzyme) binds and forms a reactive complex with the oxidized enzyme, and DHO (low affinity to the reduced enzyme) displaces orotate from the reduced enzyme (and its reversal of the reaction) so that

92

JORDAN, BISAHA, AND PICOLLELLI

the forward reaction may proceed at its regular rate. An example of a similar situation that produces nonlinear inhibition patterns has been examined and the steady-state velocity equation contains a squared term for the substrate concentration (21, p. 836). Conclusions. The yeast DHOD is a simple flavin dehydrogenase that has significant affinities for DHO and orotate in both its oxidized and reduced forms, thus accounting for the nonclassical inhibition patterns presented. The enzyme organizes the C5 pro S hydrogen of its natural substrate, DHO, to assist in its deprotonation by the sulfur atom of Cys133 in the reductive half-reaction. At physiological pH values and in vivo, the rates of catalysis are likely limited by the oxidation of the enzyme rather than the reduction of the enzyme because fumarate and Q o are rather poor electron acceptors. The more rapid exchange of the C5 deuterium of 2 than the deuterium of the C6 position is attributed to a sticky deuteride as transferred from C6 rather than taken as evidence for a carbanion mechanism in the reductive half reaction. ACKNOWLEDGMENTS We thank Ya-Jun Zheng and Alan Rendina for helpful discussions.

REFERENCES 1. Jones, M. E. (1980) Annu. Rev. Biochem. 49, 253–279. 2. Howie, C., Suckling, C. J., and Wood, H. C. S. (1990) J. Chem. Soc., Perkin Trans. 1, 3129 –3135. 3. Chen, S. F., Perrella, F. W., Behrens, D. L., and Papp, L. M. (1992) Cancer Res. 52, 3521–3527. 4. Knecht, W., and Loffler, M. (1998) Biochem. Pharmacol. 56, 1259 –1264.

5. Batt, D. B. (1999) Exp. Opin. Ther. Patents 9, 41–54. 6. Nielsen, F. S., Andersen, P. S., and Jensen, K. F. (1996) J. Biol. Chem. 271, 29359 –29365. 7. Nielsen, F. S., Rowland, P., Larsen, S., and Jensen, K. F. (1996) Protein Sci. 5, 852– 856. 8. Rowland, P., Nielsen, F., Jensen, K. F., and Larsen, S. (1997) Structure 5, 239 –252. 9. Rowland, P., Nielsen, F. S., Jensen, K. F., and Larsen, S. (1997) Acta Crystallogr. D53, 802– 804. 10. Bjornberg, O., Rowland, P., Larsen, S., and Jensen, K. F. (1997) Biochemistry 36, 16197–16205. 11. Rowland, P., Bjornberg, O., Nielsen, F. S., Jensen, K. F., and Larsen, S. (1998) Protein Sci. 7, 1269 –1279. 12. Bjornberg, O., Gruner, A.-C., Roepstorff, P., and Jensen, K. F. (1999) Biochemistry 38, 2899 –2908. 13. Nagy, M., Lacroute, F., and Thomas, D. (1992) Proc. Natl. Acad. Sci. USA 89, 8966 – 8970. 14. Pascal, R. A., Jr., and Walsh, C. T. (1984) Biochemistry 23, 2745–2752. 15. Blitz, H., and Giesler, E. (1913) Chem. Ber. 46, 3410 –3425. 16. Faeder, E. J., and Siegel, C. M. (1973) Anal. Biochem. 53, 332– 336. 17. Martin, R. G., and Ames, B. N. (1961) J. Biol. Chem. 236, 1372–1379. 18. Roy, A. (1992) Gene 118, 149 –150. 19. Bradford, M. M. (1976) Anal. Biochem. 72, 248 –254. 20. Guex, N., and Peitsch, M. C. (1997) Electrophoresis 18, 2714 – 2723. 21. Segel, I. H. (1975) Enzyme Kinetics, Wiley, New York. 22. Thompson, J. E., and Jordan, D. B. (1998) Anal. Biochem. 256, 7–13. 23. Jordan, D. B., Bisaha, J. J., and Picollelli, M. A. (1999) in Flavins and Flavoproteins 1999 (Ghisla, S., Kroneck, P., Macheroux, P., and Sund, H., Eds.), pp. 611– 614, Agency for Scientific Publ., Berlin. 24. Cleland, W. W. (1982) Methods Enzymol. 87, 390 – 405.