Ute MiihS, Charles H. Williams, Jr.S8, and Vincent MasseySn. From the $Department of Biological Chemistry, University of Michigan Medical School, Ann Arbor, ...
THEJOURNAL OF B I O ~ CHEMISTRY I C ~
Vol. 269,No. 11, Issue of March 18,pp. 7989-7993, 1994 Printed in U.S.A.
Lactate Monooxygenase 11. SITE-DIRECTED MUTAGENESIS OF THE POSTULATED ACTIVE SITE BASE HISTIDINE 290* (Received for publication, September 28, 1993, and in revised form, December 2, 1993)
Ute MiihS, Charles H. Williams, Jr.S8, and Vincent MasseySn From the $Department of Biological Chemistry, University of Michigan Medical School, Ann Arbor, Michigan 48109-0606 and the $Department of Veterans Affairs Medical Center, Ann Arbor, Michigan 48105
Lactate monooxygenase catalyzes the oxidation of Llactate with molecular oxygen to acetate, COz, and water. Histidine 290 has been proposedto be the active site base in lactate monooxygenase (Giegel,D. A, Williams, C. H., Jr., and Massey, V. (1990)J. BioZ. Chern. 265,66266632) and was mutated to a glutamine (H290Q). The mutant enzyme showsproperties that support strongly the postulated function of the histidine. The ability of L-lactate toreduce the enzyme flavinis essentially abolished, whereas reoxidation of reduced enzyme with oxygen proceeds at 1.4 x lo4 M - ~s-l, a rate essentially like that found in the wild type enzyme. Thesubstrate, L-lactate, M, and D-lactate, a is bound with a Kd equal to 2.0 x competitive inhibitor with a Kd of 3.1 x lo-’ M. Both values are similar to binding measured in the wild type enzyme. Unlike the situation with wild type enzyme, where thetransitionstate analog oxalate is bound tightly in a two-step reaction involving proton uptake from solution (Ghisla, S., and Massey, V. (1977) J. BioZ. Chem. 252,6729-6735),the mutant enzyme binds oxalate weakly, in a single step reaction, with a Kd in the order of0.1 M. No effect was observed upon varying the pH, indicating that binding does not include a protonation step. Replacing the histidine also has a significant effect on the ability of the enzyme to stabilize the flavin N(5)sulfite adduct. Sulfite is boundat least 1000-foldweaker than it is in the wild type enzyme.
L-lac
LA E
/ B:
E FI,,
2
FI,, ... lac ===
,BH+
E FI,, ... lac-
I E
5
/B:
-;T
E Fired
4
/BH+
- pyr e E FI red
(N5)
- lac
PYr FIG.1. Schematic representation of the reductive half-reaction. Note that Step 5 is not involved in normal catalysis.
The electron pair of the resulting carbanionproceeds to attack the flavin at the N(S)-position to form a covalent bond (Step 3). The ensuing carbinolamine then fragments into the reduced FMN and ketoacid (Step 4 1. In theabsence of oxygen, pyruvate is released, leavingbehind reduced enzyme (Step 5). The evidence that led to thispostulate includeschloride elimination from P-chloro-lactate (Walsh et al., 19731, binding of the transition state analog, oxalate (Ghisla and Massey, 1975, 1977), and reaction with the mechanism-based inhibitor 2-hydroxy-3-butynoic acid (Walsh et al., 1972; Schonbrunn et al., 1976; Ghisla et al., 1976). Proton abstraction requires a base close to the bound substrate. From the results of expericould be a The FMN-dependent enzyme L-lactate monooxygenase (EC ments with oxalate, it was suggested that this base 1.13.12.4) catalyzes theoxidation of L-lactate to acetate,carbon histidine (Ghisla and Massey, 1977). Lactate oxidase binds oxdioxide, and water with theincorporation of molecular oxygen alate in a two-step equilibrium, with proton uptake following binding per se. It is thought that the structure of oxalate is in acetate and water (Hayaishi and Sutton, 1957). In the reductive half-reaction, lactate isoxidized to pyruvate (Fig. 1)to sufficiently close to that of the carbanionic L-lactate, that it give reduced flavin. Catalysis is completed when the reduced raises thepK, of the active site base just as it is raised during a Kdl of 8.3 x M flavin is reoxidized by molecular oxygen yielding H202, which catalysis. Thus, the initial fast step with decarboxylates the bound pyruvate (Lockridge et al.,1972). The leads to the primaryenzyme-oxalate complex with a spectrum overall products of catalysis, acetate, C02, and water are then very similar to that of the uncomplexed enzyme. The second, slow step is proton-dependent with a pK,of 9.8 (Kd2 = 1.9 x released from the active site. MI, compared with an observed pK, of 4.7 for the uncomis initiated by There is general agreement that turnover plexed enzyme. It is accompanied by substantial perturbation proton abstraction from the substrate, and the individual steps of the reductivehalf-reaction are summarized in Fig. 1.Lactate of the absorption spectrum and the uptake of a single proton M at pH 7.0 for the binds to theenzyme and forms a Michaelis complex (Step 1 ). A from the solvent. The overall Kd is 1.6 x base on the enzyme is postulated to initiate catalysis by ab- binding of oxalate. Lactate oxidase is mechanistically relatedto flavocytostracting a proton from the a-carbon of the substrate(Step 2). chrome b2 and glycolate oxidase. Both enzymes show strong * This work was supported by the United States Public Health Ser- homology in the amino acidsequence (Lederer et al., 1985; Volokita and Somerville,1987; Cederlund etal., 1988), and vice, National Institutes of Health Grant GM-11106 (to V. M.) and National Institutes of Health Grant GM 21444 (to C. H. W.) and the their crystal structures have been solved (Xia and Mathews, Health Services and Research Administration of the Department of 1990; Xia et al., 1987; Lindqvist and Branden, 1989). For each Veterans Affairs (to C. H. W.). The costs of publication of this article a histidine residue wasfound close to theisoalloxazine portion were defrayed in part by the payment of page charges. Thisarticle must therefore be hereby marked “aduertisement”inaccordance with 18 of the flavin whichcould function as theactive site base(Fig. 2). The homologous residue in lactate monooxygenase is histidine U.S.C. Section 1734 solely to indicate this fact. 1 To whom correspondenceand reprint requests should be addressed. 290 (Giegel et al., 1990).
7989
7990
Lactate Monooxygenase: Mutagenesis of Histidine 290
FIG.2. Representation of the active site of lactate monooxygenase. The positioningof the functional groups is derived from the crystal structures of glycolate oxidase and flavocytochrome b,; the residues are numberedfor lactate monooxygenase. The first step in catalysisis shown as postulated. Catalysis is initiated by proton abstraction from the lactate a-carbon. Histidine 290 receives this proton and may be stabilized additionally by interaction with aspartate180. Tyrosine 152 may serve to stabilize the ensuing carbanion (adapted from Ghisla andMassey, 1991). Coomassie Blue. All experiments were Histidine 290 was replaced with a glutamine and the result- sis (Laemmli, 1970) stained with ant mutantenzyme referred to as H290Q. Glutamine isa semi- performed in 10 nm ImC1,' pH 7.0, at 25 "C unless noted otherwise. Extinction Coeficient-The extinction coefficient for the mutantform conservative replacement according to thedefinition by Fersht of lactate monooxygenase was determined by denaturing the protein et al. (1987); it has similar spatial requirements to the histidine with SDS as described in the preceding paper (Muh et al., 1994a). and may be able to substitutefor some of the original function Reductive Half-reaction-The enzyme solution was treated with alby its capacity to form hydrogen bonds. However, it lacks the ternate cycles of evacuation and oxygen-free argon in an anaerobic ability to be protonated. H290Q was expressed and purified cuvette. Once the solution was made anaerobic, L-lactate was added at 450 nm with time were measured from Escherichia coli,and theproperties of the mutantenzyme from the side-arm and the changes were studied. The results offer strong support for the hypoth- in the spectrophotometer. Oxidative Half-reaction-Enzyme solution a t aconcentration of esis that histidine 290 is indeed the active site base responsible about 1.6 x M was reduced anaerobically by the electron generating for proton abstraction from the substrate. system, milk xanthine oxidase, and xanthine (Massey, 1991). The reaction proceeded in the presence of 0.8 mM xanthine, M milk xanthine oxidase, 7 PM benzyl viologen, and 5 PM methyl viologen. The reduced enzyme solution was loaded anaerobically into thestopped-flow appaMaterials and instrumentation were as described in the preceding ratus andreacted with 10 mM ImCI, pH 7.0, equilibrated with different paper (Muh et al., 1994a). In addition, rapid reaction studies were oxygen concentrations (obtained by bubbling with Oz/Nzgas mixtures performed at 25 "C in a laboratory built stopped-flow apparatus (Dr. D. purchased from Matheson). P. Ballou and G. Ford). A computer control and analysis system was Photoreduction and Stability of the Semiquinone-Lactate oxidase used, referred t o a s Program A (developed by Dr. D. P. Ballou, Chung- was photoreduced using 5-deazaflavin as catalyst (Massey and HemJen Chiu, Rong Chang, and Joel Dinverno, University of Michigan), merich, 1977, 1978)and as detailed in the preceding paper (Muh et al., which permits the analysis of experimental data by exponential fits 1994a). The thermodynamic stability of the semiquinone species was based on the Marquardt algorithm (Bevington, 1969). The program hastested by photoreducing the mutant lactate monooxygenase tofull been modified t o allow the fixing of as many asfive rate constants and semiquinone, followed by an addition of benzyl viologen (final concensubsequent curve-fitting to evaluate the concentrations of intermediate tration of 6 x M) to act as an electronmediator. species. Mechanism-based Inactiuation-To test the reactivity of the lactate Site-directed Mutagenesis-The vectors usedfor generating mutants monooxygenase mutant with DL-2-hydroxy-3-butynoicacid (Walsh et al., of lactate monooxygenase and the expression plasmid were as described 1972; Ghisla etal., 19761, the enzyme solution was made anaerobic and in thepreceding paper (Muh et al., 1994a). The plasmid pGM07, which the inhibitor was addedfrom the side-arm to a final concentration of 5 is identical t o pGM03 except for a silent mutation of the N-terminal m ~ The . change of absorbance and fluorescence properties were measamino acid serine, was used as the mutagenesis template. The mutation ured with time. reaction of histidine 290 (codon, CAC) was performed with the oligoequiReaction with Sulfite-The enzyme solution was treated in an nucleotide 5'-lTGCTCGAAC~GGCGGGCGCC-3'(the codon for gluof sulfite. Methyl methlibrium titrationwith increasing concentrations tamine is underlined). Successful mutagenesis was screened for the anethiosulfonate (MMTS) was then added to the enzyme-sulfite comremoval of a DsaI site (CKRYGG). The identified clone was sequenced plex at a final concentrationof -30 mM. MMTS reacts rapidly with free from ClaI to MluI to confirm the correct sequence. This 282-base pair sulfite and thereby displaces the equilibrium with enzyme-bound sulfragment was inserted into the ClaI-MluI ofsite pGM07, giving pGM21. fite. Consequently,the return of the absorbance at 450 nm upon treatThe NcoI-KpnI fragment (810 base pairs) of pGM21 was ligated into the ment withMMTS is a direct measurefor k,=(Ghisla andMassey, 1991). appropriate position of pUMO1. Successful insertion was confirmed by The value fork,, was determinedby following the absorbance changes digesting with DsaI. The expression plasmid was designated pUM10. upon reaction with sulfite in the stopped-flow apparatus. Cell Growth and Purification ofH29OQ-Cultures with the genefor H290Q were grownand the mutant enzyme purified as described for the wild type enzyme in the preceding paper (Muh al.,et1994a). The purity The abbreviations used are: ImCl, imidazole/HCl buffer; MMTS, of the protein was determinedby SDS-polyacrylamidegel electrophore- methyl methanethiosulfonate. EXPERIMENTAL PROCEDURES
Lactate Monooxygenase: Mutagenesis of Histidine 290
-p " m
0.015
0.3
453
., .. I
...
300
..
7991
.
.
400
500
600
Wavelength (nrn)
700
506 -0.015 400
500 600 700 FIG.3. Spectral properties of HZSOQ. Oxidized enzyme (solid line). The semiquinone (dotted line) was generated by photoreduction. Full reduction(dashed line)could only be achieved in the xanthine/milk Wavelength (nm) xanthine oxidase system. Therefore, there is some contribution of reduced methylviologen a t 400 andat 600 nm. The enzyme concentration FIG.5. Binding of D-lactate. Difference spectra on titration of was1.6 x M as calculated by thedeterminedextinction coefficient H290Q with o-lactate a t 0, 1.6, 6.4, 16, and 61 m~ final concentrations of 12,200 M - ~cm-' in 10 m~ ImC1, pH 7.0, at 460 nm. are shown. All spectra are correctedfor dilution. The sumof the absorbance changes a t 506 and 485 nm are plotted against the D-lactate concentration (inset) in a double-reciprocal fashion and the Kd deter80 minedtobe 3.1 x M . Theenzymeconcentrationwas 1.1x M as I determined by flavin absorbance. The experiment was performed in 10 m~ ImCl a t pH 7.0, 25 "C. presence of 6 x M benzyl viologen, only -5% of the semiquinone had disproportionated to oxidized and reduced flavin. Reductive Half-reaction-The mutated enzyme is essentially inactive. Under anaerobic conditions the enzyme was reduced slowly by L-lactate, but with a half-time of days. From a plot of l/kob, uersus l/[lactatel (resultsnot shown), a limitingrate was estimated at to s-l (wild type enzyme, 230 s-l). As with the K266M mutant form described in the previous paper, 0 20 40 the extremely low rate of reduction of the enzyme flavin by 1 / [L-lactate] ("1) L-lactate makes the enzyme virtually inactive, and precludes FIG.4. Binding of L-lactate. The Kd for L-lactate binding to H290Q the meaningful determination of kcat and K, values. was determined by a n aerobic spectrophotometric titration in 10 m~ Oxidatiue Half-reaction-Reduced H290Q reacts in a ImCl at pH 7.0, 25 "C. The changesat 388 nm were plotted against the L-lactate concentration in a double-reciprocal fashion. The dissociation monophasic fashion with oxygen, i.e. the reaction results in oxidized enzyme without any observable intermediates. The M. Theenzyme constantwascalculated t o beapproximately2 x concentrationwas1.4 x M as determined by flavinextinction. second-order rate constant was determined by a plot of kobs against oxygen concentration to be 1.4 x lo4 M - ~s-l (wild type enzyme, 9.0 x io3 M - ~s-l). RESULTS Binding of~-Lactate-Due to the slow reactivity with L-lacEnzyme Purification-Typically 115 mg of pure enzyme was tate, an aerobic titration monitoring perturbation of the abisolated from 6 liters of bacterial growth, compared with 80 mg sorption spectrum was possible (Fig. 41, giving a value for Kd of of wild type enzyme. 2.0 x lo-' M (wild type enzyme, 5 x lo-' MI. Maximal changes Spectral Properties-Fig. 3 shows the spectral characteriswere measured at 388 nm (A€ -1900 M - ~cm-'1, 462 nm ( A 6 tics of H290Q. The oxidized spectrum has peaks 459 at and 375 -600 M - ~cm-l), 492 nm (A€ -600 M - ~cm-I), and 510 nm (A€ nm. The ratio of A28dA459is 9.5, and the extinction coefficient -400 M - ~cm"), with isosbestic points at 420 and 356 nm. at 459 nm was determined to be 12,200 M - ~cm-l. In contrast to Binding of D-Lactate-Fig. 5 shows the difference spectra wild type enzyme, which is non-fluorescent in the oxidized measured with increasing concentrationsof o-lactate. Maximal state, the mutantenzyme is fluorescent, about 2% that of free differences were found at 453 nm (A€ -1200 M - ~cm-'),485 nm FMN. The excitation spectrum of the oxidized enzyme shows (A6 -600 M - ~cm-'1, and 506 nm ( A € -900 M - ~cm"), with an maxima at 459 nm and at 375 nm, the emission spectrum is isosbestic point at 495 nm. Theinset depicts the determination maximal at 525 nm. Full substratereduction was not achieved, of the value for& as 3.1 x M (wild type enzyme, 1.7 x even by reacting with0.2 M L-lactate for 3 daysunder anaerobic M). The binding characteristics and kinetic properties of this conditions. The reduced enzymecould only be generated by use mutant are compared with those of wild type enzyme in the of an electron generating system such as the reaction with third paper of this series. xanthine and milk xanthine oxidase (Massey, 1991). Upon reBinding of Oxalate-The mutant enzyme binds this transiduction, the enzyme becomes less fluorescent than theoxidized tion state analog weakly. An accurate determination of the form, with an excitation maximum centered around 365 nm dissociation constant was difficult due to the small spectral and an emission maximum at 520 nm. The spectrum of the changes effected by the binding. With the mutant enzyme the semiquinone is typical for a n anionic semiquinone witha maxi- maximal change in absorption due to binding was at 503 nm, mum at 376 and 500 nm and a distinctive shoulder at 400 nm. with an extinction change of only 800 M - ~ cm-'. This is in Stability of the Semiquinone-As in wild type enzyme, the marked contrast towild type enzyme, where binding of oxalate semiquinone is thermodynamically stable. After 36 h in the resultsin extinctionchanges of -2000 cm-l (Ghisla and
Lactate Monooxygenase: Mutagenesis of Histidine 290
7992 0.12
0
50000
l/[Sulfite](M")
1
0
300
500
700
Wavelength (nm) FIG.6. Binding of sulfite. H290Q was titrated with sulfite at the following concentrations: 0,6 m, 16 m, 40 PM, 120 PM,500 and 50 mM. The inset shows a double-reciprocal plotof the absorbance changes at 459 nm against the added sulfite and the determination of =3 x M. The concentration of enzyme was 9.8 x M as determined by the flavin extinction. The experiment was performedin 10 m ImCl at pH 7.0, 25 "C.
Upon reaction with MMTS the absorbance at 459 nm returned s-l (wild type in a monophasic fashion at a rate of 6.7 x s-'1. enzyme, 7.2 x H290Q was also reacted with sulfite in thestopped-flow apparatus; the absorbance changes a t 460 nm required fitting with two exponentials, corresponding to approximately 80 and 20%of the total change. The data are consistent with two populations of mutant enzyme in which each population reacts with sulfite. Population 1,-80% of the enzyme, binds sulfite in a two-step fashion with a fast binding equilibrium (Kd = 1.5 x lo-' M) followed bya reaction step that reachesa limiting value equal to 8 s-'. The value of kOE is determined by the MMTS s, and so the overall Kd is calculated to be reaction at 6.7 x 1.3 x M. This is approximately the same as determined in thespectrophotometric titration. Population 2,or 20% of the total enzyme, binds sulfite in a single step equilibrium, with a rate of association equal to 50M-' s-' and a kOR equal to 0.2 s-'. This rate of dissociation is too fast to have been observed in theMMTS reaction. The calculated Kd is therefore 4 x M, which is in the range of Kd(II) measured in the equilibrium titration. DISCUSSION
H290Q shows spectral properties very similar to the wild type enzyme. A notable exception is the fluorescence of the oxidized species. Changing theprotein environment at thepoMassey, 1975). In 10 mM ImCl the value for the Kd was deter- sition of histidine 290 has rendered the oxidized species fluomined as 0.08 2 0.04 M for pH 7.0, and 0.05 * 0.01 M for pH 6.0. rescent, in contrast to the wild type enzyme which is nonis possible thatthe oxidized state.It The fluorescence increased upon binding of oxalate by about fluorescent inthe 5-fold. Like wild type enzyme, the enzyme-oxalate complex was fluorescence of this enzyme form is due to theremoval of the susceptible to photochemistry and a bleaching of the 460 nm histidine residue with its r-electroncloud which was probably absorbance occurred. In wild type enzyme this photoadduct overlapping the r-electron cloud of the flavin in the wild type (flavin-N(5)-carbonate) is a highly fluorescent species (Ghisla structure. By analogy with the crystal structuresof flavocytochrome bz (Xia and Mathews, 1990; Xia et al., 1987) and glyand Massey, 19751, whereas the fluorescence of the mutant enzyme decreased upon irradiation. Presumably the adduct is colate oxidase (Lindqvist and Branden,19891, the side chains of non-fluorescent. However, decay occurred more rapidly than for lysine 266 and histidine 290 are probably close. Interestingly, wild type enzyme, and it was not possible to generatecomplete K266M is also fluorescent in the oxidized state (Muh et al., adduct formation. The rate of decay with wild type enzyme is 1994a). A further reflection of the similarity between K266M of the semiquinone. The sharppeak 5.1 x s-l at 25 "C (Ghisla and Massey, 19751, whereas the and H290Q is the spectrum similar reaction with the mutantenzyme proceeded in a bipha- found in thewild type enzyme at 400 nm is changed to a small s-l at shoulder in K266M and a more resolved shoulder in H290Q. In sic manner with the fast phase approximately 5.4 x contrast, all other investigated mutants exhibit a sharp peak at 25 "C. Reaction with 2-Hydroxy-3-butynoic Acid-H29OQ reacted 400 nm (Muh et al., 199413). The semiquinone of H290Q is, wild type with this mechanism-based inhibitor with a half-time of days, however, thermodynamically stable, justas it is in the comparable to the rate of reduction by L-lactate. Under anaer- enzyme. Most of H290Q binds sulfite with a dissociation constant -3 obic conditions a very slow bleaching of the enzyme was obM, which is substantially weaker than the binding measserved at 460 nm, with a concurrent increase in absorbance at x postulated 320 nm, indicatingflavin-adduct formation. The adduct isfluo- ured in wild type enzyme or the mutations that are rescent with an emission maximum at 520 nm, unlike the ad- to hydrogen bond with the substrate (Muhet al., 1994b). This duct with wild type enzyme, which emits maximally at 502 nm again points to the similaritywith K266M, which shows bind(Ghisla etal., 1976).Afurther difference from wild type enzyme ing that is stillweaker. The mutation wasdesigned to test thepossible role of histiis the stability of the adduct. In wild type enzyme the flavinadduct is reasonably unstable, losing its characteristic absor- dine 290 as theactive site base, initiating catalysis by abstracbance and fluorescence properties with a half-life of several tion of a proton from the a-carbon atom of the substrate (see hours, at pH 7.0,25 "C. In H290Q, however, the development of Ghisla andMassey, 1991, for a review). In agreementwith this the 320-nm peak due to the adduct, and its characteristic fluo- postulate, the mutantenzyme is seven to eight ordersof magrescence, continued to increase progressively over a period of at nitude less active than thewild type enzyme, underscoring the importance of the histidine residuefor catalysis. On the other least 5 days. Reaction with Sulfite"H290Q bound sulfite giving two dis- hand, the capacity to bind L-lactate and D-lactate is not much sociation constants (Fig.6). wasdeterminedas approxi- different from that determinedfor the wild type enzyme. This M, whereasKd(II)can only be estimated to be in result supports thedesign of the mutation asa semiconservamately 3 x tive substitution; presumably no major conformational change therange of to M (wild type enzyme, one dissociation constant, equal t o 5.6 x lo-* M). Note that the spectrum in the has occurred. The rateof reoxidation by oxygen was alsomeasured to be in the same order of magnitude as it is found in the presence of 40 V M sulfite in Fig. 6 is nearly equivalent to&(I), yet shows about 40% ofthe totalabsorbance changesmeasured wild type enzyme, thus indicating that the mutant enzyme is upon complete bleaching of the absorbance at 459 nm. This impaired mainly in its ability to oxidize lactate. Since it is possible to reduce H290Q with substrate, albeit 80%of the enzyme binds sulfite withKd(1). indicates that about
Lactate Monooxygenase: Mutagenesis
of
Histidine 290
7993
oxalate,in with a half-life of several days, the possibility of catalysis by the medium on binding thetransitionstateanalog imidazole buffer was tested. Both the anaerobic reduction with sharp distinction to the uptakeof a single proton on binding of substrateandthe anaerobicreactionwith 2-hydroxy-3-bu- oxalate to wild type enzyme (Ghisla and Massey, 1977). These tynoic acid were performed in the presence of 10 mM Bis-Tris, results provide strong supportto the concept that histidine290 pH 7.0. In each case the reaction was found to be identical to plays an essential role in catalysis, by abstraction of a proton that in the presence of imidazole. The very small residual ac- from the a-carbon of the substrate, the resulting carbanion tivity may therefore be due to conformational strain on the transferring its electrons very rapidly to the flavin, probably via a transient N(51-substrate adduct (see Ghisla and Massey, substrate upon binding at the active site. The reaction of H290Q with 2-hydroxy-3-butynoic acid is 1991, for a review). similar to theflavin reduction with substrate since it proceeds REFERENCES with a half-life of several days.Again it islikely that this is due to a conformational strain on the inhibitor upon binding. No Berington, P. R. (1969) Data Reduction and Error Analysis for the Physical Sciences, pp. 235-242, McGraw-Hill Book Co., New York further conclusion can be drawn, however, concerning the Cederlund, E., Lindqvist, Y., Soderlund, G . , Branden, C. I., and Jomwall, H. (1988) mechanism by which the covalent adduct is formed. Eur: J . Biochem. 173,523-530 In the absence of an active site base, the binding of oxalate Fersht, A.R., Leatherbarrow, R. J., and Wells, T. N. C. (1987) Biochemistry 26, 60304038 should no longer be proton-dependent. This assumption was Ghisla, S., and Massey, V. (1975) J. Biol. Chem. 250, 577-584 confirmed, since the dissociation constant for oxalatewas Ghisla, S., and Massey, V. (1977) J . B i d . Chem. 252, 67294735 within experimental error the same when the pH was changed Ghisla, S., and Massey, V. 11991) in Chemistry and Biochemistry ofFlauwnzymes (Muller, F., ed) Vol. 11, pp. 243-289, CRC Press, Boca Raton by 1unit. The tightness of binding isonly slightly weaker than Ghisla, S., Ogata, H., Massey, V., Schonbrunn, A,,Abeles, R. H., and Walsh, C. T. the first step of binding found in thewild type enzyme (Kd = 8.3 11976) Biochemistry 15, 1791-1797 Giegel, D. A., Williams, C. H. Jr., and Massey, V. (1990) J. Biol. Chem. 265, x M). This indicates that oxalate is still being bound at the 66264632 active site in a manner similar to monocarboxylic acids, but Hayaishi, O., and Sutton, W. B. (1957) J . Am. Chem. Soc. 79,4809 that there is no second pH-dependent step, as there iswith wild Laemmli, U. K. (1970) Nature 227, 6 8 0 4 8 5 Lederer, F., Cortial, S., Becam, A. H., Haumont, P. Y., and Perez, L.(1985)Eur: J. type enzyme (Ghisla andMassey, 1977). Interestingly, the phoBiochem. 1 5 2 , 4 1 9 4 2 9 toadduct of the oxalate-enzyme complex is not fluorescent as it Lindqvist, Y., and Branden, C. I. (1989)J. B i d . Chem. 264.36263628 is with wild type enzyme. As pointed out for the oxidized spe- Lockridge, O., Massey, V., and Sullivan, P. A. (1972)J. Bioi. Chem. 247,809743106 V. (1991) inFlauins and Flavoproteins (Curti, B., Ronchi, S., and Zanetti, cies, changing the protein environment around the isoallox- Massey, G., eds) pp. 5 9 4 6 Walter de Gruyter, Berlin azine moiety has a significant effect on the fluorescent proper- Massey, V., and Hemmerich, P. (1977) J. Bid. Chem. 262, 5612-5614 Massey, V., and Hemmerich, P. (1978) Biochemistry 17, 9-17 ties of the flavin. H., Jr. (1994a)J.Bid. Chem. 269,7982-7988 Miih, U., Massey, V., and Williams, C. In conclusion, the properties measured for H290Q are con- Miih, U., Williams, C. H., Jr., andMassey, V. (1994b)J . Biol. Chem. 269,79948000 sistent with its postulated function as the active site base. It Schonbrunn, A., Abeles, R. H., Walsh, C. T., Ghisla, S., Ogata, H., and Massey, V. (1976) Biochemistry 15, 1798-1807 appears that the replacement of histidine 290 with a glutamine Volokita, M., and Somerville, C. R. (1987) J. Biol. Chem. 262, 15825-15828 has not resulted in a conformationaldisruption, since substrate Walsh, C., Schonbrunn, A., Lockridge, 0.. Massey, V., and Abeles, R. H. (1972) J . Bid. Chem. 247,60044006 and inhibitorsbind to the mutant enzyme in a similar fashion Walsh, C., Lockridge, O., Massey, V., and Abeles, R. (1973) J. B i d . Chem. 248, to their binding to wild type enzyme. However, the ability of the 7049-7054 mutant enzyme to be reduced by L-lactate, or to form a covalent Xia, 2. X., and Mathews, S. F. 11990) J . Mol. Biol. 212, 837 Xia, Z. X.,Shamala, N., Bethge, P. H., Lim, L. W.. Bellamy, H. D., Xuong, N. H., adduct with a-hydroxy-butynoate, is decreased dramatically. Lederer, F., and Mathews, S. E (1987) Proc. Natl. Acad. Sci. U. S . A. 84, 2629This is correlated with the absence of proton uptake from the 2633
