Metabolism of Resorcinylic Compounds by Bacteria PURIFICATION AND PROPERTIES OF ORCINOL HYDROXYLASE FROM PSEUDOMONAS

Orcinol hydroxylase (EC 1.14.13.6), which catalyzes the first reaction of orcinol catabolism in Pseudomonas putida 01, has been purified to homogeneity, and crystallized. Orcinol hydroxylase catalyzes the hydroxylation of orcinol with equimolar consumption of O2 and NADH (or NADPH) to 2, 3, 5-trihydroxytoluene, which is nonenzymically oxidized to a quinone. The visible absorption spectrum of the enzyme shows maxima at 373 and 454 nm and a shoulder at 480 nm. FAD can be dissociated from the protein. Reconstitution of enzymic activity was achieved with FAD, and to a limited extent by FMN. The enzyme has a molecular weight of 63,000 to 68,000 and contains 1 mol of FAD per mol of protein. K-m values for the three substrates orcinol, NADH, and O2 are 0.03, 0.13, and 0.07mM, RESPECTIVELY. The molecular activity of the crystalline enzyme is 1560 min minus 1. In the absence of orcinol, NADH is only slowly oxidized with formation of H2O2. Several analogs of orcinol also serve as substrates for hydroxylation, namely resorcinol, 4-methylresorcinol, and 4-bromoresorcinol. Other analogs, m-cresol, m-ethylphenol, 4-ethylresorcinol, and phloroglucinol, mimic orcinol as effectors, in that they (a) accelerate electron flow from NADH to the flavin and (b) decrease the apparent K-m for NADH but not to the same extent as the substrates that are hydroxylated. The latter compounds are not hydroxylated. Instead H2O2 accumulates as the only product of O2 reduction. The enzyme therefore behaves either as a hydroxylase or an oxidase. The ratio of hydroxylase to oxidase activities of the enzyme is decreased by an increase in the temperature of incubation; at 60 degrees the reaction with orcinol is almost 50% uncoupled from hydroxylation. The apparent K-m values for the effectors are in good agreement with the D-D values obtained for orcinol, resorcinol, and m-cresol. K-D values were obtained by measurement of the effector-induced perturbations of the visible absorption spectrum of the flavoprotein by difference absorption spectroscopy. The circular dichroism spectrum of orcinol hydroxylase is also altered in the presence of orcinol. The participation of the flavin in the over-all reaction is demonstrated by its rapid reduction under anaerobic conditions by NADH in the presence or orcinol, resorcinol, or m-cresol. Subsequent introduction of oxygen restores the oxidized form and yields H2O2 when m-cresol is the effector, but not when orcinol is the effector. Transfer of reducing equivalents from the reduced flavoprotein to free FAD may also occur. Reduction of orcinol hydroxylase by NADH in the absence of an effector is 10-4-fold slower than in the presence of an effector. The minimal structural requirements for effectors appear to be a 1,3-dihydroxy or 1-alkyl-3-hydorxybenzene, but only the former are substrates for hydroxylation.

agreement with the KD values obtained for orcinol, resorcinol, and m-cresol. KD values were obtained by measurement of the effector-induced perturbations of the visible absorption spectrum of the flavoprotein by difIerence absorption spectroscopy. The circular dichroism spectrum of orcinol hydroxylase is also altered in the presence of orcinol.
The participation of the flavin in the over-all reaction is demonstrated by its rapid reduction under anaerobic conditions by NADH in the presence of orcinol, resorcinol, or m-cresol. Subsequent introduction of oxygen restores the oxidized form and yields H202 when m-cresol is the effector, but not when orcinol is the effector. Transfer of reducing equivalents from the reduced flavoprotein to free FAD may also occur. Reduction of orcinol hydroxylase by NADH in the absence of an effector is 104-fold slower than in the presence of an effector.
The minimal structural requirements for effecters appear to be a 1,3-dihydroxy or 1-alkyl-3-hydroxybenzene, but only the former are substrates for hydroxylation.
Some of the data have been presented earlier, in preliminary form (3,11

Puri$cation and Crystallization
of Orcinol Hydroxylase-A summary of the purification procedure outlined under "Experimental Procedure" is given in Table  I and elution profiles are shown in Fig. 1. The loss of enzymic activity from the DEAE-cellulose chromatography may appear large but this is due to the complete separation of the ring cleavage enzyme from orcinol hydroxylase. The specific activities recorded for crude and protamine sulfatetreated extracts are optimistic by a factor of 2, because the polarographic assay measures the two sequential reactions of the catabolic pathway catalyzed by orcinol hydroxylase and 2,3,5trihydroxytoluene oxygenase, and the activity of the latter enzyme is in excess. Each of these enzymes catalyzes the fixation of 1 mol of oxygen per mol of substrate (Scheme 1).
Orcinol hydroxylase readily crystallizes from (NH&S04 solutions as yellow plates.
Stability of Orcinol Hydroxylase and pH Optimum-Orcinol hydroxylase activity is stable in crude extracts for days at 4", but thiol reagents, such as mercaptoethanol, are required to stabilize purified preparations.
Orcinol, EDTA, and FAD singly or in combination were not as effective as stabilizers.
The enzyme was most stable at pH 7. This pH optimum for stability is broadened by the presence of 0.3% mercaptoethanol.
The pH optimum for activity is difficult to evaluate because the values do not take into account the nonenzymic oxidation rate of the product of reaction, which increases lo-fold between pH values of 6 and 8. At pH 6.8 in the presence of 0.3% mercaptoethanol suspensions of crystals and solutions at 5 to 10 mg of protein per ml of orcinol hydroxylase lose about FIG. 3. Plot of log c (c = fringe displacement) against the square of the distance from the axis of rotation (r2) for orcinol hydroxylase in 20 mM KH2POa-NaOH buffer, pH 6.8, containingO.l% 2-mercaptoethanol. Fringe patterns were obtained 20 hours after attaining a speed of 26,522 rpm. The initial protein concentration was 0.6 mg per ml. Temperature, 10".
50% activity in 8 to 10 weeks, at 4". Comparable measurements in the absence of mercaptoethanol were not made because large losses of enzymic activity occurred during purification without this supplementation.
Homogeneity-Orcinol hydroxylase obtained from hydroxylapatite columns yields two diffuse bands after polyacrylamide disc gel electrophoresis at pH 8.6 in Tris-HCl buffer. In sodium dodecyl sulfate gels a single band appears. The enzyme appears homogeneous upon ultracentrifugation giving a single symmetrical schlieren pattern (Fig. 2) and linear relationships between log C versus r* from the sedimentation equilibrium data (Fig. 3). The latter measurements were made between 0.3 and 1.2 mg of protein per ml and gave a molecular weight value of approximately 65,000. Analysis of the NHz-terminal amino acid of this preparation showed that isoleucine was the only NH*-terminal amino acid present in the sample. From the purification data it can be calculated that the orcinol hydroxylase content of Pseudomonas puticla accounts for about 0.67" of the total protein of cells grown on orcinol as the sole carbon source, bearing in mind that the sensitivity of the assay is reduced 2-fold when the ring cleavage enzyme has been removed. Molecular Weight Determination-The molecular weight of orcinol hydroxylase was estimated as 68,000 by dodecyl sulfate disc gel electrophoresis and as 63,000 by gel filtration chromatography, respectively, values in reasonable agreement with those obtained from sedimentation equilibrium measurements.
Flavin Content of Orcinol Hydroxylase-The visible absorption spectrum of orcinol hydroxylase is shown in Fig. 4 and compared with that of FAD. The yellow color of the enzyme is abolished by dithionite but reappeared on aeration. The identification of the enzyme-bound flavin as FAD was achieved by a combination of chromatography and absorption spectroscopy of the chromophore dissociated from the enzyme. The cofactor co-chromatographed with FAD in Solvents 1 and 3 of Ref. 6, although faint spots were usually observed with Rp values similar to FMN; other minor components were also observed; the absorption maxima and minima of the protein-free cofactor coincided with those of FAD. A molar extinction at 450 nm (assuming a molecular weight of 65,000) of orcinol hydroxylase was estimated as 10,500, which indicated that 1 mol of FAD is bound per mol of orcinol hydroxylase (a value of 0.94 is obtained using the molar extinction of 11,300 for FAD). Reconstitution of enzymic activity by the addition of flavin nucleotides to the apoenzyme is shown in Table II. Under these conditions most of the activity of the apoenzyme was recon- stituted by the addition of FAD; FMN was a poor substitute for FAD. Stoichiometry-Initial attempts to determine the stoichiometric relationships of the hydroxylation reaction were thwarted by the rapid nonenzymic oxidation of the product of orcinol hydroxylation, 2,3,5-trihydroxytoluene, to a quinone (3, 11). This occurred too rapidly even at pH 6.8, as shown in Fig. 5, which compares the time course of the reaction at pH 6.8 and at pH 8.0. The amount of oxygen consumed is in excess of that required for a monooxygenase (mixed function oxidase) reaction, due to the subsequent nonenzymic oxidation of the product. Formation of the quinone is shown by the increase in absorbance at 490 nm.
We have obtained indirect evidence that orcinol hydroxylase catalyzes a reaction in which equimolar consumption of orcinol, 02, and NADH occurs. In the presence of limiting quantities of NADH, excess orcinol, and large quantities of enzyme, approximately 1.5 pmol of 02 were consumed per pmol of NADH supplied and the reaction mixtures typically turned brick-red (X,,, 485 nm) indicating that the quinone accumulated. By incorporat- ing large quantities of 2,3,5trihydroxytoluene 1,2-oxygenase, purified to the DEAE-cellulose stage (3)) into these reaction mixtures, quinone formation was not detected by its visible absorption spectrum; the amount of 02 consumed however increased to approximately 2 mol of O2 per mol of NADH supplied (Table  III), a result expected for the two sequential reactions of the orcinol pathway (Scheme 1). As expected, approximately 2 mol of 02 are similarly consumed per mol of orcinol supplied (excess NADH) when the ring cleavage enzyme is present.
Aromatic Substrate-Effector Specificity of Orcinol Hydroxylase-The difficulties encountered in directly determining the stoichiometry of orcinol hydroxylation led to an examination of the reactions with analogs of orcinol. It was known that resorcinol and m-cresol aIso stimulate NADH oxidation by orcinol hydroxylase and that one of the presumed analogous products, 3-methyleatechol (from m-cresol) was more stable to nonenzymic oxidation by oxygen than 2,3,5-trihydroxytoluene (3). An attempt to establish the stoichiometry with m-cresol revealed that 02 and NADH were consumed in equimolar quantities, but in excess of that required for a simple hydroxylation.
An equimolar amount of hydrogen peroxide was shown to be formed (11). 3-Methylcatechol, the expected product from m-cresol hydroxylation, was not detected chromatographically, nor by the dioxygenase assay, for which it is known to be a substrate (1,3) and hydrogen peroxide was formed in amounts equivalent to the NADH and oxygen consumed (Table IV). Hydroxyquinol was detected as a product of the reaction with resorcinol as effector by (a) chromatography, (b) formation of its quinone, and (c) oxidation by the ring cleavage enzyme, 2,3,5-trihydroxytoluene oxygenase (and hydroxyquinol 1,2-oxygenase purified from Ps. putida ORC) .I Fig. 6 shows the course of reaction when resorcinol is the aromatic substrate; during the initial stages of the reaction both 02 concentration and AsdO nm rapidly decreased, but after 4 min the Asdon,,, values began to rise again, probably due to the nonenzymic formation of hydroxybenzoquinone from hydroxyquinol, the product of resorcinol oxidation. Sequential additions of catalase and 2,3,5-trihydroxytoluene 1,2-oxygenase increased   (Fig. 6). The glucinol completely uncouple electron flow from hydroxylation combined measurements of Hz02 and hydroxyquinol formed how-with consequent formation of hydrogen peroxide. Fig. 7   peroxide as a product of oxygen reduction. Only orcinol, 4-methylresorcinol, 4-bromoresorcinol, and to a lesser extent resorcinol have been found as effecters that also undergo hydroxylation (Table IV). Compounds which did not stimulate oxygen reduction by NADH included: orsellinic acid, pyrogallol, 3,5-dihydroxybenzoate, 2,3-dihydroxybenzoate, 2,4-dihydroxybenzoate, and those compounds listed in Table VII. E$ector and Nucleotide SpeciJicity, and their Kinetic Constants-Apparent K, values for the aromatic substrates (effecters), electron donors, and oxygen, for orcinol hydroxylase are given in Tables V and VI. Although these values were not obtained in the presence of saturating concentrations of the other substrates, other experiments have shown that the apparent K, values for orcinol are not appreciably affected between NADH concentrations of 50 and 830 PM; similarly, the apparent K, for NADH is not altered between orcinol concentrations of 200 and 500 PM. V mall values obtained are not recorded because they have only comparative value within a particular group of experiments. The most notable feature of these results is that in the presence of orcinol, 4-methylresorcinol, and 4-bromoresorcinol, i.e. the substrates that do not show extensive uncoupling of electron flow from hydroxylation, the apparent K, values for NADH are relatively low. With other aromatic effecters, apparent Km values for NADH are much greater. The KD values (see later, Figs. 14 and 15) for the enzyme-effector complexes compare favorably with these kinetic values obtained for orcinol, resorcinol, and m-cresol (Table V). The apparent K, value for orcinol is not significantly altered by the use of different electron donors (Table  VI)  Inhibitors of Orcinol Hydroxylase-A preliminary survey of possible inhibitors of orcinol hydroxylase is given in Table VII. These results show that several analogs of orcinol which are not effecters significantly inhibit the reaction, as do high concentrations of orcinol and phloroglucinol.
Effecters with only two substituents in the benzene ring do not show "substrate" inhibition at high concentrations.
FIG. 8. Anaerobic reduction of orcinol hydroxylase by NADH in the absence of substrate. Orcinol hydroxylase (1 ml containing 3 mg of protein in 20 mM KH2POa-NaOH buffer, pH 6.8) was added to the cuvette and flushed with N, (Spectrum 0). Then, 25 mM NADH (lOk1) was added and the spectrum was retraced after 2,5, 10,15,20,30,and 40 min. Temperature, 26". The inset shows the time course of FAD reduction at 454 nm.
Reduction of Orcinol Hydroxylase by NADH-The preparation of large amounts of pure orcinol hydroxylase has allowed us to examine the reduction of flavin in the enzyme by NADH. In the absence of an effector and oxygen, the rate of reduction of the FAD of orcinol hydroxylase by NADH is low (Fig. 8) and appears biphasic; the final spectrum obtained is similar to that for the enzyme reduced by dithionite. The molecular activity calculated for the initial rate of reduction by NADH is approximately 0.05 min-i as compared with the value of 1560 mini for the over-all hydroxylation reaction when orcinol is the effector. Addition of oxygen to the reaction mixtures restores the oxidized flavin spectrum. Similar anaerobic experiments in the presence of orcinol, resorcinol, m-cresol, or m-ethylphenol gave instantaneous bleaching of the flavin spectrum. Fig. 9 shows experiments where the flavin is titrated to its reduced form by NADH in the presence of (a) orcinol, (b) resorcinol, and (c) m-cresol.
Reoxidation of Reduced Orcinol Hydroxylase- Fig.  10 shows the spectral changes that occur when aerated buffer is introduced to orcinol hydroxylase that had been reduced previously by NADH, in the presence of orcinol or m-cresol as described in Fig. 9. With m-cresol, addition of successive increments of 02 to the reduced enzyme produced increases of A 451 of only one-half the magnitude (Curves 1, 2, and S) produced by similar quantities of 02 with orcinol as substrate. After the addition of catalase the AM increased (Curves .9 to 6) by approximately those values observed when orcinol was present. Therefore, in confirmation of the data bi RtSOkClNOL' /,, (c) m-Cresol, Curves 1, 2, 5, 4, and 5 were obtained after sequential additions of 3, 2, 2, 3, and 3 ~1 of 7 mM NADH.
of Table  IV, it appears that Hz02 is formed by addition of 02 to reduced orcinol hydroxylase in the presence of m:cresol, but not orcinol. The increment seen upon the addition of catalase supports this notion (Curues S and 4, Fig. 10, right).

Reoxidation of Reduced Orcinol Hydroxylase by Other Electron
Acceptors-Reduced orcinol hydroxylase is able to transfer reducing equivalents to a variety of electron acceptors, other than oxygen. These include free FAD, ferricyanide, cytochrome c, acetylpyridine-NAD, and tetrazolium salts. Fig. 11 shows the rapid reduction of free FAD by NADH in the presence of orcinol hydroxylase and orcinol, which also compares the slow nonenzymic rate of FAD reduction by NADH. Perturbation of Flavin Spectrum of Orcinol Hydroxylase by Effectors-The absorption spectra of flavoproteins are often perturbed in the presence of their natural substrates or analogs of them. Figs. 12 and 13 show the perturbations of the flavin spectrum of orcinol hydroxylase by orcinol, and by resorcinol and m-cresol. The difference spectra (enzyme-effector complex minus enzyme) are qualitatively similar to each other with the exception of a broad band and shift in X,,, at lower wavelengths for the m-cresol-enzyme complex (Fig. 13). The extent of the spectral changes as a function of effector concentrations are shown in Figs. 14 and 15. The insets for Figs. 14 and 15 are double reciprocal plots for the absorption changes induced by the effecters versus the effector concentration and allow estimates for the KD of the enzyme-effector complexes to be made. They are 0.026, 0.3, and 0.28 mM for orcinol, resorcinol (not shown), and m-cresol respectively. These values are in reasonable agreement with the apparent K, values obtained earlier, i.e. 0.03, 0.19, and 0.2 mM, respectively (Table V).
Circular Dichroism of Orcinol Hydroxylase-The circular dichroism spectrum of orcinol hydroxylase is shown in Fig. 16. In the presence of the substrate orcinol the circular dichroism spectrum is altered by a slight shift and decrease in positive values to longer wavelengths (368 -371 nm), whereas the negative band became less intense at 455 nm (Fig. 16). (1 ml, 2.7 mg of protein in 20 mM KH2P04-NaOH buffer, pH 6.8) Curve 1. Then 25 mM orcinol (10 ~1) and 25 mM FAD (2~1) were added after flushing the enzyme with N2 (Curve 2). Four sequential additions of 25 rnM NADH (2 pl) were added and gave instantaneous bleaching of the flavin spectra (Curves S, 4, 6, and 6). The reduction of total flavin present is less than that anticipated from the amount of NADH added. This may be due to a time-dependent transfer of reducing equivalents from the enzyme to free FAD, or to incomplete removal of O2 from the cuvette.
Effect of Temperature on Hydroxylase and Oxidase Activities- Fig. 17 shows the effect of temperature on the catalytic activities of orcinol hydroxylase. Clearly at elevated temperatures orcinol hydroxylase loses some of its regulatory properties for effectordependent oxidation of NADH. NADH oxidase activity becomes a substantial contributor for the reduction of oxygen at higher temperatures, with concomitantly less hydroxylation of orcinol. Similar observations were made for the hydroxylase and oxidase activities when resorcinol was the effector, i.e. elevated temperatures of incubation reduce hydroxylase in favor of oxidase activity. The addition of urea (0.1 to 2.0 M) to reaction mixtures did not give similar results, but showed a progressive loss of activity as the urea concentration was raised.

DISCUSSION
Orcinol hydroxylase, the first enzyme in the pathway that enables Pseudomonas putida to catabolize orcinol to acetate and pyruvate (Scheme 1), catalyzes a typical monooxygenase (mixed function oxidase) reaction to yield 2,3,5-trihydroxytoluene which is rapidly and nonenzymically oxidized to a quinone (1, 11). The enzyme has been purified to homogeneity and crystallized as yellow plates. The specific activity of the crystalline preparations, which varied between 18 and 24 pmol of 02 consumed, min-1 mg of protein-i, is higher than that previously reported (3, 11) because previous determinations had used limiting concentrations of NADH in the assay mixtures. This represents a maximum molecular activity of 1560 min-i. 14. Spectrophotometric determination of the binding constant for orcinol hydroxylase and orcinol. The data given were provided from the difference spectra shown in Fig. 12 (with additional data of the experiment). Inset, reciprocal plots of orcinol concentration and A(Aa~A385).
orcinol hydroxylase are summarized in Table VIII. It appears to consist of a single poiypeptide chain and to be monomeric at a variety of concentrations, containing 1 mol of FAD. The activity of the apoenzyme was mostly reconstituted by FAD and only slightly by FMN (Table II). With these, and its catalytic properties, orcinol hydroxylase resembles three other flavoprotein monooxygenases, salicylate hydroxylase (4, 5, 24). p-hydroxybenzoate hydroxylase (6)(7)(8)25), and melilotate hydroxylase (9, 10) which were also obtained from several strains of Pseudomonas, and shown to contain FAD. Thus, salicylate hydroxylase from Ps. putida studied by Katagiri and co-workers (4, 24) and the p-hydroxybenzoate hydroxylases from Ps. putida A3-12 (6), Ps. putida M-6 (25), Pseudomonas fluorescens (8) (5) is a dimer of identical subunits each possessing 1 mol of FAD per molecular weight subunit of about 45,000 and melilotate hydroxylase (10) has recently been shown to be a tetramer of similar subunits of molecular weight 65,000, each containing 1 mol of FAD. Neujahr and Gaal (27) have also described a flavoprotein hydroxylase from the yeast, Trichosporon cufaneum.
It, like the enzymes from pseudomonads possesses 1 mol of FAD, although it has a much higher molecular weight (146,000). Additionally, it shares with orcinol hydroxylase the ability to hydroxylate resorcinol, one of several substrates the yeast uses for growth. All five enzymes, and also the flavoproteins 3-hydroxybenzoate 6-hydroxylase (28) and 3-hydroxybenzoate 4-hydroxylase (29) catalyze the introduction of a hydroxyl group into the benzenoid ring, o-or p-to an existing hydroxyl substituent. This occurs with a concomitant decarboxylation for salicylate hydroxylase. Direct determination of the stoichiometry of the orcinol hydroxylase reaction has been particularly t.roublesome due to the rapid nonenzymic oxidation of the product, 2,3,5-trihydroxytoluene by oxygen, and reduction of the quinone so formed by NADH. In addition, 2,3,5-trihydroxytoluene may serve as an effector of NADH oxidation by the enzyme in a manner observed for p-hydroxybenzoate hydroxylase where the product, protocatechuate is an effector of electron flow, as is gentisate for 3-hydroxybenzoate-6-hydroxylase (28). Indirect evidence of the expected quantitative relationships of oxygen and NADH consumption is, however, available. Thus, when reactions are limited by NADH, the molar ratio of substrate consumption is oreinol-NADH-02 (1:1:1.5). When 2,3,5-trihydroxytoluene 1,2-oxygenase is present such that the ring fission rate is in excess of the nonenzymic rate of oxidation of 2,3,5-trihydroxytoluene, these ratios change to I : 1:2; the 2nd mol of 02 being consumed by the dioxygenative reaction, and the quinone is not formed.
The catalytic properties of orcinol hydroxylase closely resemble those of salicylat,e, p-hydroxybenzoate, and melilotate hydroxylases. The rate of enzyme-catalyzed oxidation of NAD(P)H by molecular oxygen (or other electron acceptors) is elevated by at least 4 orders of magnitude in the presence of orcinol or some of its structural analogs. The enzyme probably catalyzes the quantitative hydroxylation of orcinol in vivo, although small and variable amounts of hydrogen peroxide are formed in vitro, possibly due to the product accumulating in reaction mixtures and acting as an effector. Orcinol hydroxylase does not hydroxylate m-cresol, phloroglucinol, m-ethylphenol, or 4-ethylresorcinol. Instead the only detected product of oxygen reduction for the last four effectors is hydrogen peroxide. 4-Methylresorcinol and 4-bromoresorcinol are good substrates for the hydroxylation reaction but resorcinol is intermediate in its role as a substrate or as an effector, in that substantial quantities of both hydroxyquinol and hy-drogen peroxide are formed as products of oxygen reduction. These latter three substrates, and orcinol, have dual roles of acting as (a) substrates for hydroxylation, and (b) as effecters for the facilitated oxidation of reduced nicotinamide nucleotides by the FAD in the enzyme. In addition, orcinol also regulates the synthesis of this inducible enzyme, a property not shown for the other effectors.2 It is not possible from these studies to determine if the elevated rate of flavin reduction observed is due only to a change in the apparent K,,, for pyridine nucleotide or also to a change in the kred (of flavin reduction), or both. However, more detailed studies of the reaction mechanisms of p-hydroxybenzoate hydroxylase (6)(7)(8)25,30,31)) salicylate hydroxylase (4, 5, 32,33), and melilotate hydroxylase (9, 10, 34), by steady state kinetic analyses, and stopped flow techniques indicate that the kred is an altered parameter. The apparent K,,, values for the nucleotide in the presence of different analogs are also considerably different, as we have observed for orcinol hydroxylase (Table V) The absorption spectrum of orcinol hydroxylase is perturbed by the presence of orcinol (Fig. 12), resorcinol, and m-cresol (Fig.  13) ; the effector-induced changes in the flavin spectrum are remarkably similar to those observed for p-hydroxybenzoate (8,25,30) and salicylate hydroxylases (5) and provide estimates of KD values which are in good agreement with the apparent K, values (Table V and Figs. 14 and 15). Orcinol has also been shown to substantially change the circular dichroism spectrum of orcinol hydroxylase (Fig. 16). The fluorescence spectrum of orcinol hydroxylase is also quenched in the presence of orcinol.3 The flavoprotein hydroxylases possess a built-in regulatory property, that prevents indiscriminate oxidation of NAD(P)H and transfer of reducing equivalents to the flavin, a property shown also by pteridine (35), P-450 (36-38), and dioxygenase hydroxylases (39). Once the flavin in the enzyme has been reduced, then alternative routes exist for the reduction of molecular oxygen to either hydroxylated product and water, or hydrogen peroxide, and these are presumably determined by the orientation and reactivity of the reduced enzyme-effector-complex, when reoxidation of the flavin occurs. The reduced species of the flavin hydroxylases all appear to be readily oxidized by molecular oxygen, irrespective of the method used to reduce the enzyme, e.g. by reduced nucleotides, EDTA, and light or dithionite.
Many of the flavoprotein hydroxylases are particularly versatile in the catalytic activities they possess, and those "isofunctional" enzymes isolated from different strains of related bacteria possess quantitatively different but often overlapping specificities for both the substrates (effecters) and the products of reaction (hydroxylation or hydrogen peroxide). Thus p-hydroxybenzoate hydroxylases isolated from the fluorescent pseudomonads, P.S. juorescens (8) and Ps. putida (25) and the nonfluorescent species Ps. de.smoZyfica (acidovoruns) (7, 30) exhibit common capabilities for the transformation of p-hydroxybenzoate, 2,4-dihydroxybenzoate to their 3-hydroxylated products, but this is not shared by them for the transformation of benzoate and 2,3,4-trihydroxybenzoate.
Similar differences in the substrate or effector specificities have been shown for the orcinol hydroxylase that another strain of Ps. pufidu (not that used in this study) elaborates during growth on orcinol.3 Further versatility of substrate hydroxylation is shown by the 3-hydroxybenzoate 6-and 4-hydroxylases from Pseudomonas aeruginosu and Pseudomonas leslosteroni respectively (28, 29). These enzymes hydroxylate several 3-hydroxybenzoates substituted in the 2,4,5-and 6-posi-2 Unpublished observations. 3 Unpublished data.
tions of the benzene nucleus, albeit with different efficiencies. Likewise the phenol hydroxylase from Trichosporon cutuneum catalyzes the sequential hydroxylation of phenol, and the product, catechol, as well as the three cresols, and all of the isomeric fluoro-and chlorophenols (27). Re sorcinol, a partial substrate of orcinol hydroxylase is equally as good a substrate for phenol hydroxylase as phenol. Phenol hydroxylase however is restricted to the use of NADPH as reductant (27). This enzyme then is possibly used during growth of the yeast on resorcinol, as is the orcinol hydroxylase of the mutant of Ps. putidu 01 (strain OlOC), being recruited by mutation to constitutivity.2 Substrate specificity and analog inhibitor studies with orcinol hydroxylase indicate that a necessity for competent binding to orcinol hydroxylase is a 1,3-substitution of either (a) two hydroxyl groups, or (b) an alkyl and a hydroxyl group. However, for hydroxylation to occur, case (a) is a prerequisite, and limited substitution in the 4-and 5-positions is allowed, e.g. 4-or 5-methylresorcinols, but not 4-ethylresorcinol are hydroxylated, the latter is an effector only. Those compounds that possess 1,3,5-substitution patterns, e.g. orcinol and phloroglucinol, also show "substrate" inhibition at high concentrations.
The effect of elevated temperatures on the catalytic activity of orcinol hydroxylase is to reduce its efficiency for hydroxylation of substrates, such that reoxidation of reduced enzyme-effector complexes by oxygen is altered in that different ratios of products are formed (Fig. 17). The increase in oxidase activity over hydroxylase activity is paralleled by increases in the apparent K, values for NADH (0.13 mM at 30" to 1.7 mM at 50"), and reduction of tritide transfer from 4R-4-[aH]NADH (Ref. 11 and footnote 3), but the significance of these observations is not clear, and the phenomena may be unrelated. However, the hydroxylase and oxidase activities are related in the proposals made by Palmer and Massey (40) for the alternative routes of flavoprotein oxidation. Thus, the evolutionary origin of flavoprotein hydroxylases from the oxidases is suggested, because physical (temperature) or chemical (41) modification of these proteins results in different products of catalysis.