Halogenated Protocatechuates As Substrates for Protocatechuate

THE ,JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 258, No. 23, Issue of December 10, pp. 14413-14421,1983 Printed In U.S.A.

Halogenated Protocatechuates As Substrates for Protocatechuate Dioxygenase from Pseudomonas cepacia"

(Received for publication, April 11, 1983)

Terence A. WalshS and David P. Balloug From the Department of Biological Chemistry, The University of Michigan, A n n Arbor, Michigan 48109

Substrates containing electron-withdrawing groups were reacted with protocatechuate 3,4-dioxygenase and oxygen. Haloprotocatechuates (5-fluoro-, 5- chloro-, 5-bromo-, 2-chloro-, and 6-chloroprotocatechuates) are oxygenated by the enzyme at rates 28- to 3000-fold lower than that with the native substrate. These lower rates are due to both deactivation of substrate to O2 attack, and to the formation of abortive enzyme-substrate (ES) complexes. Such ES complexes with haloprotocatechuates are spectrally distinct from the normal ES complex. 6-Chloroprotocatechuate pro- duces changes more like those due to protocatechuate.

The abortive ES complexes, when rapidly mixed with oxygen, decay to free enzyme and product monophasically, and the dependence of the rates on 0 2 concentration shows that a rate-limiting step precedes reaction with 02. Thus these complexes are rather unreactive toward 02, and the rate-limiting step in oxygenation is their conversion to active complexes. In contrast, the reaction of O2 with the enzyme and 6- chloroprotocatechuate is biphasic, the first phase being dependent on O2 concentration (2 X lo4 M-' s-') and the second not (7 s-'). The intermediate formed after the first phase strongly resembles the second intermediate seen in the reaction of enzyme with protocatechuate and O2 (Bull, C., Ballou, D. P., and Otsuka, S., (1981) J. Biol. Chem. 256, 12681-12686), implying that the electron-withdrawing effect of the chlorine slows the O2 addition step considerably while the conversion to the second intermediate is hardly affected. When the enzyme cycles through several turnovers with 6-chloroprotocatechuate, an enzyme species is formed that resembles the unreactive ES complexes seen with the other haloprotocatechuates, indicating that a small amount of the unreactive complex is in equilibrium with the reactive complex and that during successive turnovers the enzyme is slowly converted into the unreactive form. The formation of this form correlates with the observation that in assays the rate of product formation gradually decreases with time.

Catechol dioxygenases from soil bacteria comprise an important class of enzymes catalyzing the ring cleavage of catechol derivatives. Since many of the bacterial catabolic path- ways of complex polycyclic compounds (lignins, alkaloids,

* This research was supported by Grant GM-20877 from the Na- tional Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ Present address, Dow Chemical, New England Lab, Wayland, MA 01778.

Q To whom reprint requests should be sent.

terpenes, etc.) converge on catechol or protocatechuate (Chap- man, 1972; Dagley, 1975), the catechol dioxygenases play a vital role in the degradation of both natural and synthetic phenolic compounds in the soil. Fission of the relatively inert benzene ring of catechols into a ring-opened structure enables bacteria to utilize the product, after some rearrangement, in normal metabolic processes.

The intradiol class of catechol dioxygenases, comprised of enzymes which contain FeC3, catalyzes ring cleavage between the two hydroxyl groups of the substrate to yield the corre- sponding muconic acid, with both atoms of molecular oxygen being incorporated into the product. The extradiol class, also found in soil bacteria, contains Fez+ and catalyzes ring cleavage adjacent, but external, to the catecholic oxygens. This yields an aldehyde group and a hydroxy acid. Protocatechuate 3,4-dioxygenase catalyzes the intradiol cleavage of 3,4-dihy- droxybenzoate (protocatechuate). The enzyme from Pseudo- monm cepacia' (Bull and Ballou, 1981) has a molecular weight of 200,000; it is composed of four a-chains of 23,000 daltons, four &chains of 26,500 daltons, and 4 mol of Fe+'/200,000 daltons, consistent with a 4 ( 4 Fe+3) structure. All of the iron has been shown to be active (Bull and Ballou, 1981).

Resonance Raman experiments indicate that the burgundy- red color of the Fe+3-containing catechol dioxygenases (t4M) - 3000-4000 M" cm") is due to tyrosine to iron charge transfer interactions (Tatsuno et al., 1978; Keyes et al., 1978; Felton et al., 1978; Bull et al., 1979; Que and Heistand, 1979). Anaer- obic addition of substrate (or substrate analogues such as 3,4- dihydroxyphenylpropionate or 3,4-&hydroxyphenylacetate) to a solution of protocatechuate dioxygenase results in a steely gray solution, the spectrum of which has an additional broad band centered at approximately 700 nm (Fujisawa et al., 1972). Resonance Raman data suggest that this band is due to the p-phenolate group of the substrate coordinating to Fe+3 (Que and Heistand, 1979). The ES2 reacts with 0, to form a series of spectrally distinct intermediates (Bull et al., 1981). The first species, ES02, is formed in a second order reaction with 0, ( k = 5 X IO5 M" s-') and is characterized by its featureless spectrum consisting of an almost uniform decrease in absorbance from 350 to 750 nm with a slight shoulder at -420 nm.

This bacterium originally was obtained from D. Ribbons from the University of Miami and was tentatively classified as Pseudomonas fluorescens PHK. Later, it was thought to be Pseudomonas putzda and our publications (Bull and Ballou, 1981; Bull et al., 1981; Bull et al., 1979) used this designation. Recent API 20E tests (Analytab Products, Plainview, NY) and temperature growth tests run by Charles Mountjoy and Debby Loomus of our laboratory with consul- tation from Dr. R. Olsen of the Department of Microbiology cause us to reclassify this phthalate degrading bacterium as a P. cepacia.

The abbreviations used are: ES, enzyme-substrate complex; ES02, first observed intermediate after mixing 0 2 with ES; ES02*, second intermediate observed after mixing 0, with ES; ESO,', the first observed intermediate after mixing 0 2 with enzyme-6-chloroprotocatechuate complex.

14413

by guest on February 12, 2018http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

14414 Halogenated Substrates for Protocatechuate Dioxygenase

ESO, rapidly converts to a second intermediate, ES02*, ( k = 450 s-') that is spectrally distinguishable by virtue of a small shoulder a t 500-550 nm with - 2000 M" cm". Neither intermediate has any Iong wavelength absorbance analogous to ES. At pH 8.5, ES02* apparently breaks down directly into free enzyme and product.

No redox change of the ferric iron has been detected in the various enzyme complexes so far reported, although the tran- sient nature of the ESO, complex observed in the rapid kinetic experiments (Bull et al., 1981) has prevented characterization of this species beyond the optical spectrum. Based on ideas initially developed by Hamilton (1974), Que et al. (1977) have proposed a mechanism in which the iron remains as Fe+3 throughout the reaction. It involves activation of the substrate by Fe+3 (acting as a Lewis acid) for attack by 02. At present, it is not possible to clearly ascribe any of the spectrally known intermediates to any of those in the proposed mechanism.

We have investigated the reaction of protocatechuate dioxygenase with a number of halogenated protocatechuates, involving single chlorine substitution a t different positions (2- chloro-, 5-chloro-, and 6-chloroprotocatechuate) or a series of halogen substitutions at a single position (5-fluoro-, 5- chloro-, and 5-bromoprotocatechuate). This study has enabled us to probe some of the requirements of protocatechuate dioxygenase for efficient substrate binding and oxygenation, both in terms of steric factors and in terms of electron distribution around the ring system. The effectiveness with which protocatechuate dioxygenase deals with halogenated protocatechuates also has environmental relevance since halogenated phenolic compounds are potentially long-lived and harmful pollutants in the soil (Wood, 1982).

EXPERIMENTAL METHODS3

RESULTS

Properties of Halogenated Substrates-The electron-withdrawing effects of various ring substituents on protocatechuate were estimated by comparing the pK, values of the first hydroxyl ionizations for the substrates employed in this study. Changes in the UV spectrum of the appropriate halogenated protocatechuate occurring upon titration with acid or alkali were plotted versus pH and fitted to a theoretical pH titration curve to yield the pK, value. These experiments were conducted anaerobically, as halogenated protocatechuates readily autoxidize at alkaline pH. The results are summarized in Table I, together with the extinction coefficients at pH 6.5 and the relative rates of oxygenation by protocatechuate dioxygenase. The values for the parent compound, protocatechuate, are also given for comparison. As expected, halogen substitution lowers the pK, value of the first hydroxyl ionization from that found for protocatechuate, ranging from -0.54 for 6-chloroprotocatechuate to -1.9 pH units for 5- bromoprotocatechuate. We have assumed the first hydroxyl ionization to be the 4-hydroxy group of the protocatechuate, but with 2-chloroprotocatechuate, in particular, this may not be the case. The inductive effect of the chloro substitution adjacent to the 3-hydroxy position may be sufficient for this group to ionize first. ~~~~~ ~~

The "Experimental Methods" are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard

of Biological Chemistry, 9650 Rockville Pike, Bethesda, MD 20814. magnifying glass. Full size photocopies are available from the Journal

Request Document No. 83M-971, cite the authors, and include a check or money order for $1.60 per set of photocopies. Full size photocopies are also included in the microfilm edition of the Journal that is available from Waverly Press.

TABLE I Properties of halogenated protocatechuates

The pK. values were determined by UV spectral titration. The wavelength and extinction values are those of the absorbance peaks at pH 6.5 which were used to determine the concentrations of solutions of the compounds. They were calculated from samples dried over Pz06. Relative rates were determined by adding enzyme to 200 p M solutions of the appropriate compounds in air-saturated buffer and measuring the rate of oxygenation spectrophotometrically.

Compound lic PK. extinction values Relative rate First Wavelenth and

value

nm, M' cm" P C P 8.90 290,3850 100 6-Chloro-PCA 8.36 288,3000 4.3 2-Chloro-PCA 8.00 283, 2700 0.2 5-Fluoro-PCA 7.60 250,7900 2.1 5-Chloro-PCA 7.32 253,7990 0.1 5-Bromo-PCA 7.02 252,7220 0.03 PCA, protocatechuate.

Wavelength, nm Wovelenglh. n m Wovelength. nm

FIG. 1. Spectra of haloprotocatechuates and the products of oxygenation by protocatechuate dioxygenase. The sample cell contained 1 ml of 50 mhf potassium phosphate buffer, pH 6.5, and (a) 92 p~ 5-fluoroprotocatechuate, ( b ) 109 p~ 5-chloroprotocatechuate, (c) 164 p~ 5-bromoprotocatechuate, (d ) 247 pM 6-chloroprotocatechuate, (e) 104 p~ 2-chloroprotocatechuate, and ( f ) 200 pM protocatechuate (solid lines) before addition of 1-2 p~ protocatechuate dioxygenase to both sample and reference cells (doshed l ims).

All of the halogenated protocatechuates are oxygenated by protocatechuate dioxygenase at significantly lower rates than the native substrate. The UV spectra of the substrates before and after oxygenation at pH 6.5 are shown in Fig. 1. The spectra of the products are typical of those of @-carboxy- cis,cis-muconates, although the product of 6-chloroprotocatechuate oxygenation has no defined peak, only a shoulder a t 260 nm.

Complexation of Halogenated Protocatechuutes by Protocat- echuate Dioxygenase-Formation of ES with haloprotocatechuates is accompanied by characteristic changes in the visi- ble spectrum of the enzyme (Fig. 2). All of the halogenated compounds tested induce a red shift of varying degree in the 460 nm peak of the enzyme, in contrast to the blue shift noted for halogenatedp-hydroxybenzoate inhibitor complexes (May et al., 1978). Anaerobically, the enzyme in complex with 5- fluoroprotocatechuate has a peak at 505 nm with an increased extinction ("4500 M" cm") whereas 5-bromo- or 2-chloroprotocatechuate gives rise to a peak at 540 nm when added


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Halogenated Substrates for Protocatechuate Dioxygenase 14415

350 450 550 650 750 350 450 550 650 750 0 " " " " "

Wavelength, nrn Wavelength, nrn

FIG. 2. Spectra of protocatechuate dioxygenase in complex with haloprotocatechuates. a, resting enzyme (-), enzyme in complex with 5-fluoroprotocatechuate (- - -), and 5-chloroprotocatechuate (- . - .); b, resting enzyme (-), enzyme in complex with 6-chloroprotocatechuate (- - -), and protocatechuate (- . - . ). Concentrations of substrates were -4-fold in excess of enzyme concentration. Buffer in all cases was 50 mM potassium phosphate, pH 6.5. The spectra of the enzyme in complex with 5-bromo- and 2- chloroprotocatechuate are almost identical to that with 5-chloroprotocatechuate.

anaerobically to protocatechuate dioxygenase, with a concom- itant decrease in extinction to -2900 M-' cm". The spectra of these complexes are somewhat different from that of the enzyme in complex with protocatechuate; the latter is characterized by relatively small changes in the 460 nm region and a large increase in absorbance around 700 nm which extends to wavelengths greater than 1 pm. This is due to the formation of a new Fe+3-catecholate charge transfer band (Que and Heistand, 1979). This band is either absent in the complexes of the enzyme with haloprotocatechuates or is blue- shifted to coincide with the Fe+3-tyrosinate absorbance. It appears that the 460 nm peak of the uncomplexed enzyme has merely undergone a red shift, the resulting peak being relatively symmetrical in shape. The difference spectra between protocatechuate dioxygenase and the haloprotocatechuate complexes illustrate this as they are smooth derivative shapes (not shown).

The spectra of the ES obtained with &bromo-, 5-chloro-, and 2-chloroprotocatechuate resemble those seen with dicarboxylate inhibitor complexes of protocatechuate dioxygenase (e.g. terephthalate or glutarate, Que and Epstein, 1981). In contrast to the spectra of the dicarboxylate inhibitor complexes which exhibit X,,, values of 532 nm at pH 6.0 and 495 nm at pH 9.0, the spectra of the haloprotocatechuate complexes are not pH-dependent.

Although we have stressed the difference between the normal ES seen with protocatechuate and those obtained with the haloprotocatechuates, one compound tested, 6-chloroprotocatechuate, did give rise to spectral changes somewhat more like that of the normal ES. The spectrum of the enzyme in complex with 6-chloroprotocatechuate has a peak at 500 nm, less red-shifted than the other chloroprotocatechuates. It is perhaps significant that, of all the compounds investigated, the pK, value of the 6-chloro derivative is closest to that of protocatechuate, suggesting that the stronger acidity of the p- phenolate group of the other haloprotocatechuates has re- sulted in modified binding to the enzyme.

Anaerobic titration of protocatechuate dioxygenase with substoichiometric amounts of each of the haloprotocatechuates (Fig. 3) indicated that binding was relatively tight at pH 6.5 (Kd < 10 p ~ ) , comparable to the value obtained with protocatechuate (Kd - 4 pM). This is in contrast to the dicarboxylate inhibitors which have Kd values on the order of 1-20 mM. Also it is apparent that binding of the substituted

protocatechuates is less tight at higher pH, in common with halogenated p-hydroxybenzoate inhibitors (May et al., 1978). This reflects the preferential binding of the protonated form of the substrate.

An unusual, and repeatable, feature of titrations of the enzyme with 6-chloro- and 5-chloroprotocatechuate was the presence of a lag phase in the change in absorption during the first few additions of the haloprotocatechuate (Fig. 3b). This phenomenon was verified in a number of ways: (a) titrations of the same stock protocatechuate dioxygenase samples with protocatechuate yielded normal changes, ( b ) strict anaerobiosis was verified indicating that the lag was not due to a small amount of o2 contamination, ( c ) a 5-fold increase in protocatechuate dioxygenase concentration showed the same titration characteristics, and (d) the changes were re- producible. Isosbestic points were maintained during the titrations. The lag may be due to a small proportion of tight nonspecific binding of the haloprotocatechuates, perhaps due to the increased acidity of the p-hydroxy group. Unfortu- nately, the presence of this lag phase made precise determi- nation of the Kd difficult. Nevertheless, it is apparent that the haloprotocatechuates all bind to the enzyme relatively tightly.

Stopped Flow Studies of ES formation-Stopped flow ex-

+ 0

0 1 2 3

[2-chloro-PCA] / [Enzyme]

Q) 100 I I

I- o

n - 0 1 2 3

[6-chloro-PCA] / [Enzyme]

FIG. 3. Anaerobic titrations of protocatechuate dioxygenase with haloprotocatechuates. a, titration of the enzyme with 2- chloroprotocatechuate in 50 mM potassium phosphate buffer, pH 6.5 (0) and in 50 mM Tris-sulfate, pH 8.5 (A). Enzyme concentration was 45 p ~ . b, as a except the enzyme was titrated with B-chloroprotocatechuate.


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


periments in which protocatechuate dioxygenase solutions were mixed anaerobically with varying concentrations of the 5-haloprotocatechuates produced biphasic spectral changes at all wavelengths examined (Fig. 4). At 0.5 mM substrate, the majority of the reaction with all of the 5-haloprotocatechuates was complete within the dead time of the stopped flow apparatus (-2.5 ms). A slower phase occurred over the next 200 ms with all three 5-haloprotocatechuates. The proportion of this slower phase varied with wavelength, occupying -10% of the total change a t wavelengths below 480 nm and 10-25% at longer wavelengths. The spectrum of the species formed at the end of the first phase was similar to the final ES spectrum having virtually the same isobestic points, although the A,,, was slightly blue-shifted by "5 nm. The rates for the slow phase were 16, 14, and 12 s" for 5-flUOrO-, 5-chloro-, and 5- bromoprotocatechuate, respectively.

At lower concentrations of the 5-haloprotocatechuates (125 p ~ ) , a larger proportion of the first fast phase could be resolved, particularly with 5-chloro- and 5-bromoprotocatechuate. Thus, the following estimates for the second order rate constant for the first reaction could be made: 5-fluoroprotocatechuate, >4 X lo6 M" s-'; 5-chloroprotocatechuate, -3.2 X lo6 M" s-'; 5-bromoprotocatechuate, 0.9 X lo6 M" s-'. The rate of the smaller, slower phase was unaltered by changing substrate concentration. Therefore, it appears that initial complexation proceeds rapidly (as with the native substrate, protocatechuate), although the size of the halogen at the 5-position may attenuate this rate as the kinetic constants for this reaction vary inversely according to substituent size. The substrate concentration-independent phase may represent a conformational change of the protein to accommodate the halogen in the active site.

The rates of dissociation of the complexes of protocatechuate dioxygenase with haloprotocatechuates were determined by anaerobic displacement with protocatechuate. The enzyme was titrated stoichiometrically with the appropriate haloprotocatechuate and then mixed with 2 or 5 mM protocatechuate in the stopped flow apparatus. Thus, the observed rate-limiting step for enzyme-protocatechuate complex formation was the dissociation rate of the enzyme-haloprotocatechuate complex. (A control experiment in which uncomplexed protocatechuate dioxygenase was mixed anaerobically

with protocatechuate showed that ES formation was complete within the dead time of the apparatus.) Dissociation rates for the enzyme complex (as observed in the 450 or 550 nm regions) with 5-fluoro-, 5-chloro-, and 5-bromoprotocatechuate were 6, 2, and 0.2 s-', respectively. I t is difficult to determine whether the decreasing magnitude of these rates reflects steric hindrance to leaving due to the size of the halogen or the strength of the Fe+3 coordination to the substrate.

In contrast to the results obtained with 5-haloprotocatechuates, complex formation between the enzyme and 2- chloro- and 6-chloroprotocatechuates was monophasic with rates that were dependent on substrate concentration. A plot of the observed rates versus the concentration of 2-chloro- or 6-chloroprotocatechuate was linear in both cases (Fig. 5) yielding second order rate constants of 7 X lo3 and 8.8 X lo3 termined by the displacement method described above were 0.025 s" for 2-chloroprotocatechuate and 0.12 s" for 6- chloroprotocatechuate, significantly lower than the dissociation rates observed with the 5-substituted compounds. These kinetic values yield Kd values of 3.4 and 13.6 PM, as compared to the values of 2.5 and -5 p~ determined by titration. It is noteworthy that the chlorine substitution in the 2- and 6- positions has decreased the ES formation rate by "1000-fold from that found for the native substrate and the 5-haloprotocatechuates; it is the slow rates of dissociation of the 2- and 6-chloroprotocatechuates that account for the Kd values of the compounds being similar. The smaller ES formation and dissociation rates with the 2- and 6-substituted compounds may indicate some steric problems in binding substrate with groups in these positions.

Reaction of Oxygen with the Enzyme in Complex with Hal- oprotocatechuates-The relatively tight binding of the haloprotocatechuates to protocatechuate dioxygenase enabled us to titrate the enzyme with substoichiometric amounts of substrate so that only minimal quantities of uncomplexed substrate remained in solution. This avoided the complication of multiple turnovers in rapid mixing experiments with 02. To observe multiple turnover events, further substrate could easily be titrated into the enzyme solution in the tonometer on the stopped flow apparatus.

"1 s-l , respectively. The rate constants for dissociation de-

a, 0 S 0 Ll L 0 v) Ll Q

p ) --(a) ,

0 100 200 300 400

n VI

0 n V Y

VI

n v)

0 Y

n V

'I-

Time, ms [CHLORO-PCA] , m M I/ [oxygen] , m ~ "

FIG. 4 (left). Stopped flow kinetic time courses of the anaerobic reaction of protocatechuate dioxygenase with 5-chloroprotocatechuate, followed at 680 nm. Concentration of enzyme was 24 PM and concentration of 5-chloroprotocatechuate was (a) 0.09 mM and ( b ) 0.5 mM. The traces are offset for clarity and the absorbance of uncomplexed enzyme before mixing is indicated. FIG. 5 (center). Dependence of the observed rate of reaction for the anaerobic formation of ES on 2-

chloroprotocatechuate (A) and 6-chloroprotocatechuate (0) concentration. Enzyme concentrations and the rate constants derived from the plots were: A, 27 PM, 7 X lo3 M-' s-'; 0 , 2 4 p ~ , 8.8 X lo3 M" s-'. FIG. 6 (right). Double reciprocal plot of the observed rate of the reaction of protocatechuate dioxy-

genase in complex with 2-chloroprotocatechuate with various oxygen concentrations. A solution of 96 p~ enzyme, titrated with 76 p~ 2-chloroprotocatechuate was mixed in the stopped flow apparatus with varying concentrations of oxygen, and the monophasic reaction was followed a t 460 or 700 nm. The extrapolated maximum rate is 0.17 s-'.


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


Substrates Halogenated at the 5-Position-In marked contrast to the multiphasic reaction of 0, with the enzyme in complex with protocatechuate, the analogous reaction using 5-chloroprotocatechuate was monophasic at all wavelengths, with a rate constant of 0.005 s-l at 0.98 mM 02. More surpris- ingly, the rate of the reaction was independent of the concentration of 02. The disappearance of ES was slow enough for multiple spectral scans to be taken in the stopped flow apparatus, showing that the decay was isosbestic at 510 and 370 nm (the same as those between the spectra of resting enzyme and the enzyme in complex with 5-chloroprotocatechuate). Multiple turnover experiments demonstrated that the spectrum of the steady state complex was identical to that of ES, and the decay rate constant for this complex converting to free enzyme when substrate was depleted was also 0.005 s-’. Steady state kinetic experiments using 5-chloroprotocatechuate concentrations of between 12.5 and 125 p M (and monitoring the oxygenation of the substrate at 270 nm catalyzed by 2 p~ protocatechuate dioxygenase and with an O2 concentration of 0.98 mM) gave rates of between 0.0055 and 0.0067 s-’. This shows the insensitivity of the rates to substrate concentrations in this range and agrees well with the rate constant for the reaction observed in the rapid mixing experiments.

Qualitatively similar results were obtained with two other 5-substituted protocatechuate derivatives, 5-bromoprotocatechuate and 5-fluoroprotocatechuate. The reaction of 0 2

with the enzyme in complex with 5-bromoprotocatechuate was even slower (0.0018 s-’) than that using 5-chloroprotocatechuate. This value was again independent of O2 concentration. The reaction of enzyme-5-fluoroprotocatechuate complex with O2 was faster than that with the other 5-substituted compounds and had a weak dependence on O2 concentration; the rate constants for the reaction were 0.22, 0.15, and 0.13 s” for 0.98, 0.3, and 0.125 mM 02, respectively.

2-Chloroprotocatechuate-The spectral changes on mixing O2 with protocatechuate dioxygenase in complex with 2-chloroprotocatechuate in the stopped flow apparatus were also monophasic, but in this case there was an appreciable nonlin- ear dependence of the observed rate constant on O2 concentration. A double reciprocal plot of these data is shown in Fig. 6, from which an extrapolated limiting rate of 0.17 s-’ can be derived. A turnover number of 0.058 s-’ was obtained from steady state kinetic experiments monitoring the oxygenation of 2-chloroprotocatechuate (8-125 p ~ ) catalyzed by 2 p~ protocatechuate dioxygenase at an O2 concentration of 0.98 mM. This agrees well with the rapid kinetics data ( k = 0.053 s-’) at the same O2 concentration.

6-Chloroprotocatechuate-Fig. 7 shows the reaction of 0.98 mM O2 with protocatechuate dioxygenase in complex with 6- chloroprotocatechuate. The kinetic traces were clearly biphasic at nearly all the wavelengths inspected (360-750 nm), both phases being complete within 0.8 s of mixing. These two processes were followed by a much slower reaction occurring over an 80 s time period which was most prominent in the 600 nm region. Even there it only consisted of “10% of the total spectral change. The significance of this slower phase will become more apparent in the description of multiple turnover experiments.

The fastest phase of the reaction was dependent on the O2 concentration while the subsequent two processes were not. The dependence of the first phase could be accurately measured by observing the reaction at 620 nm where the con- tribution of the second phase was negligible. The plot of observed rate constant uersus O2 concentration was linear (Fig. 8) with a second order rate constant for O2 addition of 2 X

io4 M” s-l (as compared to 5 x IO5 M-’ s” for the native substrate). No evidence for the reversibility of the reaction is apparent as the intercept passes through zero.

It was possible to derive a spectrum of the first intermediate that occurred in the oxygenation of 6-chloroprotocatechuate by the enzyme from the kinetic traces taken over the wavelength region 350-750 nm using the method described under “Experimental Methods.” This is shown in Fig. 9. It is apparent that the spectrum of this species and that of the second intermediate (ES02*) in protocatechuate oxygenation (Bull et al., 1981) are very similar, both having a low shoulder centered around 520 nm. The species observed with 6-chloroprotocatechuate, ES02’ , does have more long wavelength absorbance. The fact that only the second intermediate is seen suggests that the initial oxygen addition step has been slowed significantly by the halogen substitution at position 6, whereas the second step is affected very little.

1

I

ll 0.0 0.2 0.4 0.6 0.8

Time, s FIG. 7. Stopped flow kinetic traces (at various wavelengths)

of the reaction of oxygen with protocatechuate dioxygenase in complex with 6-chloroprotocatechuate. A solution of 110 p M enzyme, titrated with 80 p~ 6-chloroprotocatechuate, was mixed In the stopped flow apparatus with 1.96 mM oxygen. The lowest trace (- - -) was recorded over 80 s (to show the slow ( k = 0.08 s-’), third phase) and is shown at half the absorbance scale of the other traces.

v x 5 0 ::K 0 0.0 0.2 0.4 0.6 0.8 1.0

[OXYGEN], mM FIG. 8. Dependence on oxygen concentration of the rate of

the first phase of the reaction of oxygen with protocatechuate dioxygenase in complex with 6-chloroprotocatechuate. Con- ditions were as in Fig. 7 except that the oxygen concentration was varied and the reaction was followed at 620 nm.


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


r' ._ c 0 .- t X

W

Wavelength, n m FIG. 9. Spectra of enzyme species in the reaction of proto-

catechuate dioxygenase with 6-chloroprotocatechuate and oxygen. Resting enzyme (-), the first intermediate seen in the reaction of oxygen with the enzyme in complex with 6-chloroprotocatechuate (M), the species that is formed during several turnovers of substrate (- - -), and the enzyme in complex with product (-. -. ). The turnover species was generated by adding a concentrated solution of 6-chloroprotocatechuate, to give a final concentration of 0.95 mM, to 1 ml of 20 b M enzyme solution equilibrated with oxygen a t 2 "C, and recording the spectrum immediately after mixing. The enzyme-product complex is the species formed after all spectral changes had finished.

When sufficient 6-chloroprotocatechuate was added to the protocatechuate dioxygenase solution to give "5 turnovers of substrate, and this was rapidly mixed with O2 in the stopped flow apparatus, the first two phases were again observed within 0.8 s of mixing. This was followed by a period of steady state which lasted until substrate was depleted. However, during this steady state, the enzyme spectrum gradually changed. Spectra recorded on the stopped flow apparatus during this period revealed that the species being formed was none of the species previously observed in single turnover. Rather, it appeared more like the spectra of the ES obtained with the 5-substituted compounds. After substrate was depleted, the spectrum of the steady state species returned to that of free enzyme with a rate constant of 0.08 s-'. To better characterize this new intermediate, the following experiment was performed using the Cary 219 spectrophotometer. A solution of 21 p~ enzyme was equilibrated with 100% O2 gas a t 2 "C (to give almost 2 mM O2 in solution) and then an aliquot of concentrated 6-chloroprotocatechuate was added to give a final substrate concentration of 0.94 mM. Immediately after mixing, repeated spectral scans were recorded between 340 and 740 nm (at 10 nm/s). The species that arose during steady state is shown in Fig. 9. This spectrum closely resembles that obtained by anaerobic titration of the enzyme with 2- and 5- substituted protocatechuates (Fig. Za). Thus, the species aris- ing during steady state appears to be an ES that is not seen in appreciable amounts in anaerobic titrations of the enzyme. When substrate was depleted, this species converted ( k = 0.08 SKI) to a stable species with a spectrum consistent with an enzyme-product complex.

These results can be accounted for if it is assumed there are two modes of binding substrate: one analogous to that seen with the native substrate, characterized by a band at -700 nm, the other analogous to the binding of 5-substituted protocatechuate derivatives to the enzyme. The species with substrate in the binding mode comparable to protocatechuate (ES-I) can react readily with 02, but the other species (ES- 11) reacts with O2 either relatively slowly or not at all (Scheme 1). At equilibrium, with no O2 present, the Kd values for the two species are such that only 10-15% of the total ES is

present as ES-11. Thus in a single turnover experiment, 85- 90% of the enzyme rapidly cycles through ES02' and back to resting enzyme, while the remaining ES-I1 slowly converts with a rate constant of 0.08 s-', giving rise to the small, slow phase observed at the end of the single turnover experiment. When multiple turnovers occur, at each cycle a proportion of the enzyme (10-15%, depending on the relative on rates to each ES form) will be removed from the fast O2 reaction cycle and be sequestered as ES-11. With subsequent turnovers, more ES-I1 is formed, as observed experimentally. When substrate is depleted, ES-I1 slowly and monophasically converts to a stable enzyme-product complex.

Steady state kinetic time courses following the disappearance of substrate (ie. normal assay procedures) reflect this self-inhibition as no true linear disappearance of substrate is observed at concentrations of 6-chloroprotocatechuate up to 275 p ~ ; rather, the time courses are markedly curved from the point of mixing, even when the reactions are followed in the stopped flow apparatus. Another factor involved in the curvature of these traces is the occurrence of product inhibition. I t is apparent from the following experiment that product inhibition does take place. The oxygenation of a solution of 275 p~ 6-chloroprotocatechuate catalyzed by 1.1 p~ protocatechuate dioxygenase was followed to completion. The solution was re-oxygenated and a further aliquot of 6-chloroprotocatechuate was added to again give 275 p~ substrate. The initial rate of oxygenation was 14-fold slower in the presence of product. In a similar experiment, it was also shown that the rate of native substrate oxygenation was also markedly reduced in the presence of the halogenated product.

Fig. 9 shows that the product of 6-chloroprotocatechuate oxygenation does produce a small but significant perturbation in the spectrum of protocatechuate dioxygenase and suggests a Kd value below 1 mM. A 5-pl aliquot of this enzyme sample was fully active in the normal assay system (Bull and Ballou, 1981), indicating that the inhibition was reversible. The small quantities of 6-chloroprotocatechuate we possess has as yet precluded a formal analysis of this product inhibition. It is known that the native product, P-carboxy-cis,&-muconate, binds to the enzyme (Bull and Ballou, 1981), the affinity being higher below neutral pH. This is also the case with the terephthalate and glutarate complexes that presumably mimic the product (Que and Epstein, 1981).

EP (ESO 2)

SCHEME 1. Kinetic scheme to account for the observations of the reaction of protocatechuate dioxygenase with haloprotocatechuates and oxygen. See text for description.


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


DISCUSSION

The substitution of halogens onto the benzene ring of protocatechuate has a profound effect on how protocatechuate dioxygenase reacts with and oxygenates these substrate analogues, as compared to its reaction with the native substrate. The simple prediction that halogen substituents would lower the rate of turnover due to their inductive electron-withdrawing effects was substantiated, but the rapid mixing experiments show that the reasons for the lower rates are not straightforward. The enzyme forms very different types of complexes with the halogenated substrates than with protocatechuate (in four out of the five halogenated compounds investigated). Only with 6-chloroprotocatechuate could intermediates be resolved spectrally in rapid mixing experiments with 0,. This leads to insights into some of the catalytic requirements that protocatechuate dioxygenase has for substrates.

In catalytic systems, care needs to be used in distinguishing an intrinsic deactivation of substrate towards 0, as opposed to a slow overall turnover rate as measured by steady state assays. The slow turnover rate may be due to other factors such as slow release of product. The relative steady state turnover rates of substrate analogues are therefore not always useful for comparing the rate of initial 0, attack on different substrates. It is essential to use rapid kinetic techniques to monitor the individual steps of the reaction to accurately gauge the effect of a substrate analogue on a given step.

The only compound tested with which we were able to observe the initial 0, reaction clearly and unequivocally was 6-chloroprotocatechuate. In this case, 0, addition was 28-fold slower than with protocatechuate as substrate, and the first intermediate seen (ESO,') during oxygenation of B-chloroprotocatechuate corresponds closely to the second intermediate (ESO,*) observed in protocatechuate oxygenation. This suggests that 0, attack has been sufficiently slowed by the effect of the chlorine at C-6 to effectively render the conversion of ESO, to ESO,* kinetically invisible i e . , ESO, is converted to ESO,' as fast as it is formed. If this is so, we can conclude that, with 6-chloroprotocatechuate, the rate analogous to the decay of ESO, to ES02* must be in excess of 200 s" because the maximum observed rate of oxygenation was 19 s-'. If the rate of the next step had been any slower than 200 s-', our data would have permitted kinetic resolution at appropriate wavelengths. Kinetic traces at all wavelengths were rigorously examined for such evidence, but none was found.

These data fit current ideas on the mechanism of the intradiol dioxygenases. It has been proposed (Que et al., 1977) that the iron coordinates thep-hydroxyl group of the substrate and promotes ketonization at C-3 to form a carbanion a t C- 4. Then electrophilic attack by 0, yields an a-peroxyketone species which can rearrange (either via a Criegee-type rearrangement or a dioxetane species) to give a seven-membered anhydride. The anhydride then decays to product. This mechanism adheres to the concept that the iron undergoes no formal change in oxidation state during catalysis although the data presented here do not rule out the proposal by Bull et al. (1981) that the first intermediate formed may have ferrous iron character, but this is not necessarily a prerequisite.

Clearly, in the mechanism of Que et al. (1977), 0, attack will be slowed by electron-withdrawing groups such as halogens. However, rearrangement of the peroxy intermediate may either be accelerated or slowed by such a substitution according to the mechanism of its decomposition (Sawaki and Ogata, 1975, 1978). The effect of the 6-chloro substituent is to lower the rate of formation of ESO, from 5 X IO5 to 1.9 X IO4 M"

s-l, as predicted, while the following conversion is not as greatly affected the rate is 450 s" with protocatechuate and at least 200 s-' with 6-chloroprotocatechuate. (The actual rate could in fact be much greater.) 6-Chloroprotocatechuate is therefore the first substrate for the intradiol oxygenases in which the true rate of O2 addition has been shown to be significantly decreased. A more complete discussion of these effects can be found in the following paper (Walsh et al., 1983).

It might be expected that the effect of halogen substitution closer to the 0,-sensitive C-3-C-4 bond would be to deacti- vate the system to 0, attack even further. Although this may be the case, it appears that substitution in these positions modifies substrate binding such that the attack of 0, on the ES can no longer be directly observed. Thus, the reaction rate is independent, or only weakly dependent, on 0, concentration. Certainly, the spectra of the ES with 5-fluoro-, 5- chloro-, 5-bromo-, and 2-chloroprotocatechuate (Fig. 2) are different from that seen with the native substrate, suggesting a different mode of binding for the haloprotocatechuates. If this different type of ES does not react with 02, then the rate of conversion to an active complex will determine the rate of oxygenation, provided the conversion is slow relative to the oxygenation step. The observed rate of oxygenation will therefore be independent of 0, concentration, as is seen with the 5-chloro- and 5-bromoprotocatechuate complexes. Alterna- tively, if the rate of conversion to an active ES is comparable to or greater than that for 0, attack on ES-I, then the observed rate will show a curved dependence on 0, concentration with a limiting rate equal to the rate of conversion from inactive to active substrate complex. This situation appears to occur when 2-chloroprotocatechuate and 5-fluoroprotocatechuate are used as substrates. Using 2-chloroprotocatechuate, the observed rate of 0.05 s-' at the maximum 0, concentration we could attain (0.96 mM) does not approach the saturated rate of 0.17 s-' very closely. This saturated rate was extrapolated from a double reciprocal plot of the observed rates against the 0, concentration (Fig. 6). With 5-fluoroprotocatechuate, the observed rates are close to maximal, therefore only showing a weak dependence on 0, concentration.

We therefore propose that there are two distinct modes of binding substrate which we have termed ES-I and ES-11. ES- I is typified by the enzyme in complex with protocatechuate and is reactive with O,, whereas ES-I1 is typified by the enzyme in complex with 5-chloroprotocatechuate and is quite unreactive towards 0,. 6-Chloroprotocatechuate exhibits both types of binding, the ES-I type predominating (85-90%). We have constructed a kinetic scheme that encompasses all of our observations on the reaction of haloprotocatechuates with protocatechuate dioxygenase (Scheme 1). Note that two forms of ES-I1 (ES-I1 and ES-IIa) have been proposed in order to accommodate the observation of biphasic kinetics seen in the anaerobic reaction of the enzyme with 5-haloprotocatechuates. The two forms are spectrally very similar. It is not yet clear in the case of 6-chloroprotocatechuate whether ES- I1 converts to ES-I via free enzyme through kv3 and k-, to undergo oxygenation or whether ES-I1 converts to ES-I without prior dissociation of substrate i.e. through k,. The conversion of ES-I1 to a reactive species with the other haloprotocatechuates does not involve dissociation of the substrate (via k--5 and k-,) and reassociation to form an active species (via kl) since computer simulations of this portion of the scheme using the rates of dissociation of the ES-I1 complexes as measured by the protocatechuate displacement experiments are not large enough to be consistent with the rate limits for oxygenation. Thus, the rate-limiting step in their oxygenation


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/


is k4 in the scheme. It is assumed that in all complexes, except that with 6-chloroprotocatechuate, the equilibria favor the ES-I1 forms.

What distinguishes ES-I and ES-I1 in terms of Fe+3 coordination? This is difficult to answer at present as the precise nature of ES-I i.e. the native substrate complex, remains unclear. In model complexes resonance Raman data (Que and Heistand, 1979) suggest that protocatechuate chelates the Fe+3 center via both phenolate groups, rather than by monodentate coordination via the p-phenolate which would leave a free m-phenol. For the enzyme, however, evidence suggests that substrate binds as the dianion (Que, 1980). No model complex with both a monodentate p-phenolate coordination to Fe+3 and a free rn-phenolate has been investigated. I t is clear that straightforward coordination of the p-hydroxyl group, such as with the p-hydroxybenzoate inhibitor complexes, does not give rise to typical ES-I spectra, indicating some important role for the m-hydroxy group. I t should be noted that the present mechanistic proposals specify p-phenolate coordination to the Fe+3, with the m-hydroxy group free to undergo ketonization. ES-I and ES-I1 may differ in that ES-I1 type complexes may chelate the Fe+3 via both phenolic groups such that the conversion of the chelated complex to one involving monodentate coordination (ES-I) could be the rate-limiting step before O2 attack.

We have already noted the similarity of the spectra of ES- I1 and of the carboxylate inhibitor complexes. This raises the possibility of inverted binding of substrate analogues where it has been suggested (May et al., 1978) that the p-phenolate group binds at the carboxylate binding site (possibly to an arginyl residue, Carlson et al., 1980) because of its increased acidity. This has been invoked for the complex of enzyme with 3,4-dihydroxynitrobenzene, a potent inhibitor of the enzyme (Tyson, 1975), where a more electron-withdrawing nitro group has replaced the carboxyl group of protocatechuate. All of the substrates that form ES-I1 type complexes do have more acidic p-hydroxyl groups than that of protocatechuate. However, inverted binding seems unlikely for a number of reasons. (i) 3-Fluoro-4-hydroxybenzoate is a potent inhibitor of protocatechuate dioxygenase with a Kd = 0.56 p~ and a phenolic pK, value of 7.8, close to the pK, values of 8.0 for 2-chloroprotocatechuate and 7.6 for 5-fluoroprotocatechuate. Resonance Raman studies (Felton et al., 1978) have conclusively shown the p-hydroxy group of this compound to be directly coordinated to the iron. (ii) Conversion of a re- verse-complexed substrate molecule to one with correct coordination enabling oxygenation to take place is unlikely to occur without prior dissociation of the substrate. (iii) I t is unlikely that inverted binding involving carboxylate-FeC3 coordination would have a Kd < 10 p~ as carboxylate inhibitors typically have Kd values of >0.5 mM. We intend to undertake resonance Raman studies to show conclusively whether the phenolic groups of the compounds giving ES-I1 complexes are coordinated to the Fe+3, and how the coordination differs between that found with the native substrate, the p-hydroxybenzoate, and the carboxylate inhibitor complexes.

Although we have used the relative acidity of the p-hydroxy group of the substrates as a measure of the effect of a substituent in terms of electron density at the oxygen-sensitive C-3-C-4 bond, this single consideration is not the only factor involved in how well a substrate is used by protocatechuate dioxygenase. For example, the p-phenolic pK, for 3,4-dihydroxybenzaldehyde is 7.05,4 close to the value of 6.96 for that

' T. Walsh and D. P. Ballou, unpublished results.

of 5-bromoprotocatechuate, yet 5-bromoprotocatechuate is cleaved by the enzyme, albeit relatively slowly, whereas 3,4- dihydroxybenzaldehyde is not, even over long periods of in- cubation with enzyme (greater than 18 h). We have investigated other analogues in which the carboxyl group of the substrate is replaced by a resonance-stabilizing group (3,4- dihydroxybenzonitrile, 3,4-dihydroxynitrobenzene, and 34- dihydroxyacetephenone) and have found that none of these compounds is cleaved by protocatechuate dioxygenase, although all bind to the enzyme. It therefore seems that substituent groups that increase the acidity of the p-phenolic group through resonance effects, rather than inductive ones, are not oxygenated by protocatechuate dioxygenase. It may be that the lack of reactivity is due to the resonance stabilization of the p-quinonoid form of these compounds when bound to the enzyme. Current mechanistic proposals specify that the m- phenolic group undergoes ketonization, which would be pro- hibited by the formation of the p-quinonoid form of the substrate in the ES. This may also indicate why no bands attributable to iron to phenolate coordination are seen in the resonance Raman spectrum of protocatechuate dioxygenase in complex with 3,4-dihydroxynitrobenzene (Que, 1980), although the compound binds very tightly to the enzyme.

Acknowledgments-We would like to thank Dr. V. Massey and C. P. Mountjoy for critically reading the manuscript, Dr. L. M. Schopfer for valuable discussions on the kinetic aspects of this work, and Dr. S. Dagley for help in expediting this work.

REFERENCES Beaty, N. B., and Ballou, D. P. (1981) J. Biol. Chem. 256,4611-4618 Bull, C., and Ballou D. P. (1981) J. Biol. Chem. 256 , 12673-12680 Bull, C., Ballou, D. P., and Otsuka, S. (1981) J. Biol. Chem. 256,

Bull, C., Ballou, D. P., and Salmeen, I. (1979) Biochem. Biophys. Res. Commun. 87,836-841

Carlson, R. E., Wood, J. M., and Howard, J. B. (1980) Fed. Proc. 31 , 2090

Chapman, P. J. (1972) in Degradation of Synthetic Organic Molecules in the Biosphere, National Academy of Sciences Report, Washing- ton, D. C.

12681-12686

Dagley, S. (1975) Am. Sci. 6 3 , 681-689 Entsch, B., Ballou, D. P., and Massey, V. (1976) J. Biol. Chem. 251 ,

Felton, R. H., Cheung, L. D., Phillips, R. S., and May, S. W. (1978)

Fujisawa, H., Hiromi, K., Uyeda, M., Okuno, S., Nozaki, M., and

Hamilton, G. A. (1974) in Molecular Mechanisms of Oxygen Activation

Keyes, W. E., Loehr, T. M., and Taylor, M. L. (1978) Biochem.

May, S. W., Phillips, R. S., and Oldham, C. D. (1978) Biochemistry

Moore, J. W., and Pearson, R. G. (1981) Kinetics and Mechanism,

Que, L., Jr. (1980) S t r u t . Bonding 40, 39-72 Que, L., Jr., and Epstein, R. M. (1981) Biochemistry 20,2545-2549 Que, L., Jr., and Heistand, R. H. (1979) J. Am. Chem. SOC. 101,

Que, L., Jr., Lipscomb, J. D., Munck, E., and Wood, J. M. (1977)

Sawaki, Y., and Ogata, Y. (1975) J. Am. Chem. SOC. 9 7 , 6983-6989 Sawaki, Y., and Ogata, Y. (1978) J. Am. Chem. SOC. 100,856-860 Tatsuno, Y., Saeki, Y., Iwaki, M., Yagi, T., Nozaki, M., Kitagawa, T.,

Tyson, C. A. (1975) J. Biol. Chem. 2 5 0 , 1765-1770 Walsh, T., Ballou, D. P., Mayer, R., and Que, L., Jr. (1983) J. Biol.

Chem. 258, 14422-14427 Williams, C . H., Arscott, L. D., Matthews, R. G., Thorpe, C., and

Wilkinson, K. D. (1979) Methods Enzymol. 6 2 , 185-198 Wood, J. M. (1982) Enuiron. Sci. Technol. 16 , 291A-297A

2550-2563

Biochem. Biophys. Res. Commum. 85,844-850

Hayaishi, 0. (1972) J. Biol. Chem. 247 , 4422-4428

(Hayaishi, O., ed) pp. 443-445, Academic Press, New York

Biophys. Res. Commun. 83,941-945

17,1853-1860

pp. 290-291, John Wiley and Sons, New York

2219-2221

Bwchim. Bwphys. Acta 485,60-74

and Otsuka, S. (1978) J. Am. Chem. SOC. 100,4614-4615


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/



ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

T A Walsh and D P BallouPseudomonas cepacia.

Halogenated protocatechuates as substrates for protocatechuate dioxygenase from

1983, 258:14413-14421.J. Biol. Chem.

http://www.jbc.org/content/258/23/14413Access the most updated version of this article at

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here

http://www.jbc.org/content/258/23/14413.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/content/258/23/14413

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;258/23/14413&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/258/23/14413

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=258/23/14413&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/258/23/14413

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/258/23/14413.full.html#ref-list-1

http://www.jbc.org/

Documents

Halogenated Protocatechuates As Substrates for Protocatechuate