258 Biochimica et Biophysica Acta 953 (1988) 258-262 Elsevier

BBA 33096

O x i d a t i o n - r e d u c t i o n potential studies on p-hydroxybenzoate hydroxylase

from Pseudomonas fluorescens

Gary Will iamson a,., Dale E. Edmondson a, and Franz Mi~ller b , . .

a Department of Biochemistry, Emory University School of Medicine, Atlanta, GA (U.S.A.) and b Department of Biochemistry, Agricultural University, Wageningen (The Netherlands)

(Received 9 December 1987)

Key words: Oxidation-reduction potential; p-Hydroxybenzoate hydroxylase; 4-Hydroxybenzoate 3-monooxygenase; Flavoprotein hydroxylase; ( P. fluorescens )

The oxidation-reduction potential of p-hydroxybenzoate hydrolylase (4-hydroxybenzoate, NADPH: oxygen oxidoreductase (3-hydroxylating), EC 1.14.13.2) from Pseudomonas fluorescens has been measured in the presence and absence of p-hydroxybenzoate using spectrocoulometry. The native enzyme demonstrated a two-electron midpoint potential of -129 mV during the initial reductive titration. The midpoint potential observed during subsequent oxidative and reductive titrations was -152 inV. This marked hysteresis is proposed to arise from the oxidation and reduction of the known air-sensitive thiol group on the enzyme (Van Berkei, W.J.H. and Miiller, F. (1987) Eur. J. Biochem. 167, 35-46). Redox titrations of the enzyme in the presence of substrate showed a two-electron midpoint potential of -177 mV. No spectral or electro- chemical evidence for the thermodynamic stabilization of any flavin semiquinone was observed in the titrations performed. These data show that the affinity of the apoenzyme for the hydroquinone form of FAD is 150-fold greater than for the oxidized flavin and that the substrate is bound to the reduced enzyme with a 3-fold lower affinity than to the oxidized enzyme. These data are consistent with the view that the stimulatory effect of substrate binding on the rate of enzyme reduction by NADPH is due to the respective geometries of the bound FAD and NADPH rather than to a large perturbation of the oxidation-reduction potential of the bound flavin coenzyme.

Introduction

Kinetic studies of p-hydroxybenzoate hydrox- ylase (4-hydroxybenzoate, NADPH : oxygen oxidoreductase (3-hydroxylating), EC 1.14.13.2)

(pHBH) from Pseudomonas fluorescens [1] have demonstrated the following reaction scheme for its catalytic mechanism (Scheme I). This mechanistic scheme is referred to as a bi uni uni uni ping pong mechanism [2] and is also followed by the similar

* Present address: ARC Food Research Institute, Norwich, U.K.

* * Present address: Sandoz AG, Agro Division, Department of Toxicology, Basel, Switzerland.

Abbreviations: pHBH, p-hydroxybenzoate hydroxylase (EC 1.14.13.2); DCPIP, dichlorophenol indophenol.

Correspondence: D.E. Edmondson, Department of Biochem- istry, Emory University School of Medicine, Atlanta, GA 30322, U.S.A.

p OHB NADPH ~ NADP* 02 DOHB,H20

~ O H B ' N A D P H EH 2-p-OHB EH2-P-OHB'O 2 E

l l EH2,P-OHB NADP" E'DOHB" H20 NADPH p-OHB

Scheme I. Reaction pathway for pHBH catalysis. The abbrevi- ations used are: p-OHB, p-hydroxybenzoate; E, oxidized form of the enzyme; EH 2, reduced form of the enzyme; and DOHB,

dihydroxybenzoate. (Taken from Ref. 1.)

0167-4838/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)

flavoprotein hydroxylase, salicylate hydroxylase [3]. The stimulatory effect of substrate or effector binding on the rate of NADPH reduction of the bound FAD in flavoprotein hydroxylases is well- known [1,4]. A stimulation of reduction rate of 1.4-105 is observed on p-hydroxybenzoate bind- ing to pHBH as compared to the free enzyme [1]. Recent X-ray crystallographic studies [5] show that, in the absence of substrate, the isoalloxazine ring is moved into the vacant substrate-binding site. Model-building studies suggest that the orien- tation of the N-5 position of the flavin ring with respect to the 4 position of the bound NADPH makes hydride transfer geometrically more im- probable than when the substrate is bound [5].

In an effort to further investigate the factors responsible for the stimulatory effect of substrate binding on reduction rate, we established a col- laborative effort to study the oxidation-reduction potentials of the bound FAD both in the presence and in the absence of substrate. Studies of the effect of competitive inhibitor binding on the re- dox potentials of D-amino-acid oxidase [6] have demonstrated quite large effects of ligand binding on the FAD potentials. Thus, the possibility that the stimulation in NADPH reduction of the flavin of pHBH on substrate binding could be due, at least in part, to an alteration in redox potential deserved to be investigated. Previous binding ex- periments of p-hydroxybenzoate to the oxidized and reduced forms of the enzyme have shown the respective K d values to differ by only a factor of 2 (oxidized from K d = 41 /tM [1]; reduced form K d = 21 /~M [7]). Calculation of the difference in redox potentials from thermodynamic considera- tions using these data suggests that the redox potential of pHBH is unaffected on binding of the substrate.

The results presented here demonstrate that substrate binding to pHBH shifts the two-electron potential by only 25 mV more negative than that of the free enzyme. In addition, it is suggested that the measured potential of the enzyme is in- fluenced by the auto-oxidation of a catalytically nonessential thiol group which serves to demon- strate the requirement for showing complete re- versibility in redox potential titrations of enzymes.

259

Experimental procedures

The isolation and purification of pHBH was as described earlier [8]. Enzyme samples were stored as a precipitate in saturated ammonium sulfate solutions and dialyzed against 50 mM potassium phosphate buffer (pH 7.0) before the titration experiments. Oxidation-reduction potential mea- surements were performed using spectrocoulome- try in a system similar to that published by Stankovich [9]. The specific instrumentation and procedures used have been described earlier [10]. All measurements were performed at 10°C in 50 mM potassium phosphate buffer (pH 7.0). Methyl viologen (100 #M) served as mediator reductant and DCPIP (20 /tM) was used as mediator oxi- dant. Of various coupling dyes tested (which were required to facilitate equilibration of the enzyme with the electrodes), a mixture of Indigo di- sulfonate (2 /~m) and 2-hydroxy-l,4-naphtho- quinone (2 /~m) gave the best results and were used in the experiments reported here.

Results and Discussion

Initial spectrocoulometric experiments on pHBH in the presence of 2/~M riboflavin and 2 /~M pyocyanin as coupling dyes and 100 /tM methyl viologen as the reductive mediator titrant showed a 94.5% current efficiency for enzyme reduction; however, equilibration with the measur- ing electrodes was quite slow (over 1 h to achieve a stable potantial for each data point). The re- oxidation of the enzyme using 20 #M DCPIP as mediator oxidant also demonstrated a high cur- rent efficiency, but a marked hysteresis in the apparent midpoint potential by approx. -35 mV as compared to the reductive phase of the titra- tion.

To determine whether this apparent hysteresis was due to problems associated with electrode equilibration or was due to an inherent property of the enzyme, the experiment was repeated using a fresh sample of enzyme and using Indigo di- sulfonate (2 /~M) and 2-hydroxy-l,4-naphtho- quinone (2 /~M) as coupling dyes. A much more rapid equilibration of the enzyme with the measur- ing electrodes was observed using this system. The potential for the initial reductive phase of the

260

~oo 90[ 80 r 7o k

so t ,~o~

eL - 200 480 460 -140 420 -100

E h (mY}

Fig. 1. Oxidation-reduction potential titration of native pHBH. o, initial reduction of enzyme; zx, oxidative titration after standing overnight at 10 ° C in the presence of reduced methyl viologen; and o, reductive titration immediately after electro- chemical reoxidation. The solid lines are theoretical plots for

two-electron potentials at -129 mV and at - 1 5 2 inV.

titration was observed to be approx. 20 mV higher than those observed on subsequent oxidation and re-reduction. The data are shown in Fig. 1. The initial reduction of pHBH (as judged by the bleaching of absorbance due to the bound FAD) occurred with an apparent midpoint potential of - 1 2 9 mV (n = 2). The reduced enzyme solution was allowed to equilibrate overnight at 10 o C. 20 /LM DCPIP added, and subsequent reoxidation yielded a midpoint potential of - 152 mV (n = 2). Re-reduction of the enzyme (Fig. 1) yielded data identical with those observed in the oxidative phase. Thus, a reversible E m value for the two- electron oxidation-reduction potential of pHBH of - 152 mV is observed. No spectral evidence for the intermediate formation of any flavin semi- quinone was observed during the titration. The enzyme retained greater than 95% of its catalytic activity when assayed at the end of the experiment which demonstrates no irreversible inhibition from contact with the dyes used in the experiment.

The marked hysteresis observed on comparing the initial reduction of the enzyme with subse- quent oxidation and re-reduction suggests a reduc- tion of another component of the enzyme which does not reoxidize readily during the oxidative phase of the spectrocoulometric titration. Since there are no other known redox groups on the enzyme, the most obvious candidate is the air-sen- sitive thiol group (cysteine-116) in the enzyme [11,12]. Oxidation of this thiol is known not to

100

8O

~6o

f 2C-

-220 -200 180 160 440 -120 E h (mY)

Fig. 2. Oxidation-reduction potential titration of pHBH in the presence of 0.3 mM p-hydroxybenzoate. The enzyme sample had been previously reduced, air-reoxidized, and dialyzed aerobically overnight before the titration. O, initial reductive titration; A, average of two oxidative titrations; and m, second reductive titration. The solid line is a theoretical plot for a

two-electron potential of - 1 7 7 mV.

influence the catalytic activity of the enzyme, and samples used in this work were not resolved from air-oxidized materials although the samples were stored under conditions known to minimize this oxidation. Thus, incubation of the reduced en- zyme overnight in the presence of excess reduced methyl viologen probably reduces the oxidized thiol group and subsequent anaerobic oxidative and reductive titrations show good reversibility.

Previous binding data have shown that p-hy- droxybenzoate binds to the oxidized enzyme with about one-half the affinity as to the reduced en- zyme [1,7] (Ko× = 41 ~M; Kre d = 21 /~M). From these values, calculations suggest that substrate binding to pHBH should have tittle or no in- fluence on the measured oxidation-reduction potential of the enzyme. To determine whether this expectation is indeed valid, spectrocoulomet- ric titrations were performed on samples of en- zyme (that had been pre-reduced and incubated under reducing conditions prior to potential mea- surements) in the presence of 0.3 mM p-hydroxy- benzoate. As expected, no hysteresis was observed on comparing the initial reductive titrations with subsequent oxidative and re-reductive titration data (Fig. 2). The midpoint potential of the en- zyme is lowered by 25 mV on substrate binding relative to the free enzyme (Em = - 1 7 7 mV; n = 2). Current efficiencies of 93% were found for both oxidative and reductive phases of the titra- tion.

261

E * FAD

K d : 4 5 n M K d : 41 }.IM

-'~ E FAD S :- E" FAD "- S 4

AG l = - 9 . 9 k c a l / m o l e A G 2 : - 5 . 9 k c a l / m o l e

[l G3:.9.4kc°,/m°,eEm 2osmv II 152mv 11 GS .8.°kc°'/m°'eEm 177mv

K d : 0.3 nM Kd = 110 JJM

b i.- E FADH 2 "S E + FADH 2 q E 'FADH 2 ,, S

A G 6 = -12.3 k c a l / m o l e A G 7 : - 4.9 k c a l / m o l e

Fig. 3. Born cycle for the binding of oxidized and reduce dFAD to the apoenzyme and for the binding of substrate ( p-hydroxybenzo- ate) to the oxidized and reduced forms of the holoenzyme.

These data permit the calculation of the bind- ing constants for the reduced ravin to the apoen- zyme and for the substrate to the reduced enzyme using the Born cycle shown in Fig. 3. The binding constants for oxidized FAD association to the apoenzyme (K 0 = 45 nM) [11] and for substrate binding to the oxidized holoenzyme (K d = 41 #M) [1] have been determined previously. These calcu- lations show that the apoenzyme has a greater affinity (150-fold) for the hydroquinone form than for the oxidized form of FAD. X-ray crystallo- graphic data [5] show the reduced ravin in the enzyme-binding site has a planar conformation similar to that of the oxidized form. The increase in binding energy (2.4 kcal/mol) probably arises from the favorable interaction of the anionic N-1 position on the reduced isoalloxazine ring with the positive end of a helix dipole oriented towards the N-1 position.

Substrate binding affinity to the reduced en- zyme is reduced by a factor of 3 (approx. 1 kcal/mol) relative to oxidized pHBH. The calcu- lated K d of 110 /~M from the redox potential experiments differs from that reported by Entsch et al. [7] by a factor of 5. If these values were used in the calculation of the redox potential of the enzyme in the presence of substrate; no alterations in midpoint-potential would be observed. Perhaps the discrepancy between the two methods of de-

termination could arise from the presence of oxidized thiol groups on the enzyme that appear to be the source of the hysteresis shown in Fig. 1. Entsch et al. [7] used enzyme reduced by the light-EDTA method. Whether this reductive method would result in the reduction of the oxidized thiol on the enzyme is currently un- known.

The results presented here on pHBH are in accord with those redox potential data published on the related enzyme, salicylate hydroxylase [13]. Stankovich and co-workers [13] have found that salicylate binding to salicylate hydroxylase lowers the redox potential of the bound FAD by only 20 mV which shows the binding of substrate to the reduced enzyme is decreased in affinity 5-fold less than to the oxidized enzyme. Thus, for both flavoprotein hydroxylases examined, the stimula- tory effect of substrate binding on the rate of enzyme reduction by NADPH is due to alter- ations in the respective geometries of the bound FAD with respect to the bound pyridine nucleo- tide rather than to a large perturbation of the redox potential of the bound FAD.

Acknowledgements

This work was supported by grants from the National Institutes of Health (GM-29433 to

262

D .E .E . ) and f r o m the N e t h e r l a n d s O r g a n i z a t i o n

for the A d v a n c m e n t o f Pure R e s e a r c h ( Z W O / S O N

to F .M.) . T h e c o l l a b o r a t i v e e f fo r t was m a d e poss i -

ble by a N A T O t rave l g r an t ( 0 2 0 0 / 8 7 to F .M. a n d

D .E .E . T h e au tho r s also wish to t hank Mr. W . J . H .

and Berkel for the p r e p a r a t i o n o f the enzyme .

References

1 Husain, M. and Massey, V. (1979) J. Biol. Chem. 254, 6657-6666.

2 Cleland, W.W. (1970) in The Enzymes (Boyer, P., ed.), 3rd Edn., Vol. 2, p. 8, Academic Press, New York.

3 Wang, L.H. and Tu, S.C. (1984) J. Biol. Chem. 247, 10682-10688.

4 White-Stevens, R.H. and Kamin, H. (1972) J. Biol. Chem. 247, 2358-2370.

5 Schreuder, H.A., Van der Laan, J.M., Thunnissen, M.M.G.M., Kalk, K.H., Swarte, M.B.A., Hol, W.G.J. and

Drenth, J. (1988) in Flavins and Flavoproteins (Edmond- son, D.E. and McCormick, D.B., eds.), pp. 527-538, de Gruyter, Berlin, in press.

6 Van den Berghe-Snorek, S. and Stankovich, M.T. (1985) J. Biol. Chem. 260, 3373-3379.

7 Entsch, B., Ballou, D.P. and Massey, V. (1976) J. Biol. Chem. 251, 2550-2563.

8 Miiller, F., Voordouw, G., Van Berkel, W.J.H., Stennis, P.J., Visser, S. and Van Rooijen, P.J. (1979) Eur. J. Bio- chem. 101,235-244. Stankovich, M.T. (1980) Anal. Biochem. 109, 295-308. WiUiamson, G. and Edmondson, D.E. (1985) Biochemistry 24, 7790-7797. Miiller, F. and Van Berkel, W.J.H. (1982) Eur. J. Biochem. 128, 21-27. Van Berkel, W.J.H. and Miiller, F. (1987) Eur. J. Biochem. 167, 35-46. Einarsdottir, G.H., Stankovich, M.T., Wang, L.H. and Tu, S.C. (1987) in Flavins and Flavoproteins (Edmondson, D.E. and McCormick, D.B., eds.), pp. 569-572, de Gruyter, Berlin, in press.

9 10

11

12

13