Crystal structure of p-hydroxybenzoate hydroxylase complexed with its reaction product 3,4-dihydroxybenzoate

Crystals of the flavin-containing enzyme p-hydroxybenzoate hydroxylase (PHBHase) complexed with its reaction product were investigated in order to obtain insight into the catalytic cycle of this enzyme involving two substrates and two cofactors. PHBHase was crystallized initially with its substrate, p-hydroxybenzoate and the substrate was then converted into the product 3,4-dihydroxybenzoate by allowing the catalytic reaction to proceed in the crystals. In addition, crystals were soaked in mother liquor containing a high concentration of this product. Data up to 2.3 A (1 A = 0.1 nm) were collected by the oscillation method and the structure of the enzyme product complex was refined by alternate restrained least-squares procedures and model building by computer graphics techniques. A total of 273 solvent molecules could be located, four of them being presumably sulfate ions. The R-factor for 14,339 reflections between 6.0 A and 2.3 A is 19.3%. The 3-hydroxyl group of the product introduced by the enzyme is clearly visible in the electron density, showing unambiguously which carbon atom of the substrate is hydroxylated. A clear picture of the hydroxylation site is obtained. The plane of the product is rotated 21 degrees with respect to the plane of the substrate in the current model of enzyme-substrate complex. The 4-hydroxyl group of the product is hydrogen bonded to the hydroxyl group of Tyr201, its carboxyl group is interacting with the side-chains of Tyr222, Arg214 and Ser212, while the newly introduced 3-hydroxyl group makes a hydrogen bond with the backbone carbonyl oxygen of Pro293.     

Oxygen reactivity of p-hydroxybenzoate hydroxylase containing 1-deaza-FAD

The flavin prosthetic group (FAD) of p-hydroxybenzoate hydroxylase (EC 1.14.13.2) was replaced by 1-deaza-FAD (carbon substituted for nitrogen at position 1). An improved method for production of apoenzyme by precipitation with acidic ammonium sulfate was developed. The modified enzyme, in the presence of p-hydroxybenzoate, catalyzed the oxidation of NADPH by oxygen, yielding NADP+ and H2O2, but the ability to hydroxylate p-hydroxybenzoate and other substrates was lost. An analysis of the mechanism of NADPH-oxidase catalysis showed a close analogy between the reaction pathways for native and modified enzymes. In the presence of p-hydroxybenzoate, the rate of NADPH consumption catalyzed by the 1-deaza-FAD form was about 11% that of the native enzyme. Both formed a stabilized flavin-C (4a)-OOH intermediate upon reaction of reduced enzyme with oxygen, but the 1-deaza-FAD enzyme could not utilize this peroxide to hydroxylate substrates, and the peroxide decomposed to oxidized enzyme and H2O2.     

Crystal structure of p-hydroxybenzoate hydroxylase reconstituted with the modified FAD present in alcohol oxidase from methylotrophic yeasts: evidence for an arabinoflavin.

The flavin prosthetic group (FAD) of p-hydroxybenzoate hydroxylase from Pseudomonas fluorescens was replaced by a stereochemical analog, which is spontaneously formed from natural FAD in alcohol oxidases from methylotrophic yeasts. Reconstitution of p-hydroxybenzoate hydroxylase from apoprotein and modified FAD is a rapid process complete within seconds. Crystals of the enzyme-substrate complex of modified FAD-containing p-hydroxybenzoate hydroxylase diffract to 2.1 A resolution. The crystal structure provides direct evidence for the presence of an arabityl sugar chain in the modified form of FAD. The isoalloxazine ring of the arabinoflavin adenine dinucleotide (a-FAD) is located in a cleft outside the active site as recently observed in several other p-hydroxybenzoate hydroxylase complexes. Like the native enzyme, a-FAD-containing p-hydroxybenzoate hydroxylase preferentially binds the phenolate form of the substrate (pKo = 7.2). The substrate acts as an effector highly stimulating the rate of enzyme reduction by NADPH (kred > 500 s-1). The oxidative part of the catalytic cycle of a-FAD-containing p-hydroxybenzoate hydroxylase differs from native enzyme. Partial uncoupling of hydroxylation results in the formation of about 0.3 mol of 3,4-dihydroxybenzoate and 0.7 mol of hydrogen peroxide per mol NADPH oxidized. It is proposed that flavin motion in p-hydroxybenzoate hydroxylase is important for efficient reduction and that the flavin "out" conformation is associated with the oxidase activity.     

Mechanistic studies of p-hydroxybenzoate hydroxylase reconstituted with 2-Thio-FAD

2-Thio-FAD (oxygen substituent at position 2 is replaced by sulfur) was used to reconstitute the apoenzyme of p-hydroxybenzoate hydroxylase. The 2-thio-FAD enzyme differs from native enzyme in several respects. While the native enzyme catalyzes the fully coupled hydroxylation of p-hydroxybenzoate, the 2-thio-FAD enzyme shows no hydroxylation of this substrate, instead reducing molecular oxygen to hydrogen peroxide. The rate of reduction of 2-thio-FAD p-hydroxybenzoate hydroxylase by NADPH in the presence of substrate was 7-fold faster than with the native enzyme. However, the oxygen reactivity of the reduced 2-thio-FAD enzyme was less than 1% that of native enzyme. This slow oxygen reaction results in the very high KmO2 observed in steady state kinetic studies of the modified enzyme. Stopped flow studies of the oxygen reaction of the reduced 2-thio-FAD enzyme in the presence of substrate confirmed the formation of a transient intermediate. The spectrum of this intermediate is very similar to those of the flavin-C(4a) adducts obtained with 2-thio-FMN lactate oxidase. This evidence suggests that reduced 2-thio-FAD p-hydroxybenzoate hydroxylase forms a flavin-C(4a)-hydroperoxide on reaction with oxygen in a reaction analogous to that with native enzyme, but that the resulting peroxyflavin is incompetent as an oxygenating species, breaking down instead to oxidized 2-thio-FAD enzyme and hydrogen peroxide.     

Kinetic studies on the reaction of p-hydroxybenzoate hydroxylase. Agreement of steady state and rapid reaction data.

p-Hydroxybenzoate hydroxylase (EC 1.14.13.2) from Pseudomonas fluorescens is a NADPH-dependent, FAD-containing monooxygenase catalyzing the hydroxylation of p-hydroxybenzoate to form 3,4-dihydroxybenzoate in the presence of NADPH and molecular oxygen. The mechanism of this three-substrate reaction was investigated in detail at pH 6.6, 4 degrees C, by steady state kinetics, stopped flow spectrophotometry, and equilibrium binding experiments. The initial velocity patterns are consistent with a ping-pong type mechanism which involves two ternary complexes between the enzyme and substrates. The first ternary complex is formed by random addition of p-hydroxybenzoate and NADPH to the enzyme, followed by the release of the first product (NADP+). The reduced enzyme . p-hydroxybenzoate complex now reacts with oxygen, the third substrate, to form the second ternary complex. The enzyme-bound p-hydroxybenzoate then reacts with the activated oxygen to give 3,4-dihydroxybenzoate which is released regenerating the oxidized enzyme for the next cycle. The binding of p-hydroxybenzoate to the oxidized enzyme to form a 1:1 complex causes large, characteristic spectral perturbations and fluorescence quenching. The dissociation constant for the enzyme . substrate complex was obtained by titrations in which absorbance and/or fluorescence quenching was measured. The binding constants of NADPH to the enzyme with and without p-hydroxybenzoate were determined kinetically by measuring the rate of reduction of the enzyme at different concentrations of NADPH. The reduction of the enzyme proceeds extremely slowly in the absence of p-hydroxybenzoate. The presence of the substrate causes a dramatic stimulation (140,000-fold) in the rate of enzyme reduction. The anaerobic reduction of the enzyme by NADPH in the presence of p-hydroxybenzoate produces a transient charge-transfer intermediate. On the basis of the proposed mechanism, the dissociation constants for p-hydroxybenzoate and NADPH as well as the Michaelis constants for all the three substrates were calculated from the initial velocity data. The agreement obtained between various kinetic parameters from the initial rate measurements and those calculated from the individual rate constants determined in rapid reactions, strongly supports the proposed mechanism for the p-hydroxybenzoate hydroxylase reaction.     

Chemical modification of tyrosine residues in p-hydroxybenzoate hydroxylase from Pseudomonas fluorescens: assignment in sequence and catalytic involvement

p-Hydroxybenzoate hydroxylase was modified by diethyl pyrocarbonate at pH values greater than 7 and by p-diazobenzoate. Modification of the enzyme by diethyl pyrocarbonate abolishes the affinity of the enzyme for the substrate p-hydroxybenzoate. Modification by p-diazobenzoate has the same effect on the enzyme. The enzyme is protected against these modifications by the effector p-fluorobenzoate. The data indicate that the modification of one tyrosine residue in the active center of the enzyme is responsible for the loss of enzyme activity. This tyrosine residue has been identified by sequence studies using radioactively labeled p-diazobenzoate and was found to be most probably Tyr-222. Diethyl pyrocarbonate reacts with a tyrosine residue in the active center other than Tyr-222; the former could not be identified. Sequence studies further showed that Cys-211 is also partially modified by p-diazobenzoate. In addition, the sequence of residues 343-345 was found to be Ser-Trp-Trp instead of the tentative assignment Ser-Tyr-Trp made earlier. The results are briefly discussed on the basis of the existing three-dimensional model of the enzyme.     

Crystallization of bacteriorhodopsin from bicelle formulations at room temperature

We showed previously that high-quality crystals of bacteriorhodopsin (bR) from Halobacterium salinarum can be obtained from bicelle-forming DMPC/CHAPSO mixtures at 37 degrees C. As many membrane proteins are not sufficiently stable for crystallization at this high temperature, we tested whether the bicelle method could be applied at a lower temperature. Here we show that bR can be crystallized at room temperature using two different bicelle-forming compositions: DMPC/CHAPSO and DTPC/CHAPSO. The DTPC/CHAPSO crystals grown at room temperature are essentially identical to the previous, twinned crystals: space group P21 with unit cell dimensions of a = 44.7 A, b = 108.7 A, c = 55.8 A, beta = 113.6 degrees . The room-temperature DMPC/CHAPSO crystals are untwinned, however, and belong to space group C222(1) with the following unit cell dimensions: a = 44.7 A, b = 102.5 A, c = 128.2 A. The bR protein packs into almost identical layers in the two crystal forms, but the layers stack differently. The new untwinned crystal form yielded clear density for a previously unresolved CHAPSO molecule inserted between protein subunits within the layers. The ability to grow crystals at room temperature significantly expands the applicability of bicelle crystallization.     

Purification, crystallization, and preliminary X-ray data for porcine fumarase

Single crystals of fumarase purified from pig heart have been prepared from solutions containing polyethylene glycol. The crystals give diffraction data corresponding to Bragg spacings of 2.0 A and contain a single subunit of the enzyme in the asymmetric unit of the C222 unit cell. Therefore, the subunits of this tetrameric molecule are arranged with the point symmetry group 222. The present purification scheme and studies of the NH2-terminal amino acid sequences suggest that only a single form of the enzyme is present, and it is thought to be the mitochondrial enzyme.     

Analysis of the active site of the flavoprotein p-hydroxybenzoate hydroxylase and some ideas with respect to its reaction mechanism

The flavoprotein p-hydroxybenzoate hydroxylase has been studied extensively by biochemical techniques by others and in our laboratory by X-ray crystallography. As a result of the latter investigations, well-refined crystal structures are known of the enzyme complexed (i) with its substrate p-hydroxybenzoate and (ii) with its reaction product 3,4-dihydroxybenzoate and (iii) the enzyme with reduced FAD. Knowledge of these structures and the availability of the three-dimensional structure of a model compound for the reactive flavin 4a-hydroperoxide intermediate has allowed a detailed analysis of the reaction with oxygen. In the model of this reaction intermediate, fitted to the active site of p-hydroxybenzoate hydroxylase, all possible positions of the distal oxygen were surveyed by rotating this oxygen about the single bond between the C4a and the proximal oxygen. It was found that the distal oxygen is free to sweep an arc of about 180 degrees in the active site. The flavin 4a-peroxide anion, which is formed after reaction of molecular oxygen with reduced FAD, might accept a proton from an active-site water molecule or from the hydroxyl group of the substrate. The position of the oxygen to be transferred with respect to the substrate appears to be almost ideal for nucleophilic attack of the substrate onto this oxygen. The oxygen is situated above the 3-position of the substrate where the substitution takes place, at an angle of about 60 degrees with the aromatic plane, allowing strong interactions with the pi electrons of the substrate. Polarization of the peroxide oxygen-oxygen bond by the enzyme may enhance the reactivity of flavin 4a-peroxide.     

Properties of p-hydroxybenzoate hydroxylase when stabilized in its open conformation

p-Hydroxybenzoate hydroxylase is extensively studied as a model for single-component flavoprotein monooxygenases. It catalyzes a reaction in two parts: (1) reduction of the FAD in the enzyme by NADPH in response to binding of p-hydroxybenzoate to the enzyme and (2) oxidation of reduced FAD with oxygen in an environment free from solvent to form a hydroperoxide, which then reacts with p-hydroxybenzoate to form an oxygenated product. These different reactions are coordinated through conformational rearrangements of the protein and the isoalloxazine ring during catalysis. Until recently, it has not been clear how p-hydroxybenzoate gains access to the buried active site. In 2002, a structure of a mutant form of the enzyme without substrate was published that showed an open conformation with solvent access to the active site [Wang, J., Ortiz-Maldonado, M., Entsch, B., Massey, V., Ballou, D., and Gatti, D. L. (2002) Proc. Natl. Acad. Sci. U.S.A. 99, 608-613]. The wild-type enzyme does not form high-resolution crystals without substrate. We hypothesized that the wild-type enzyme without substrate also forms an open conformation for binding p-hydroxybenzoate, but only transiently. To test this idea, we have studied the properties of two different mutant forms of the enzyme that are stabilized in the open conformation. These mutant enzymes bind p-hydroxybenzoate very fast, but with very low affinity, as expected from the open structure. The mutant enzymes are extremely inactive, but are capable of slowly forming small amounts of product by the normal catalytic pathway. The lack of activity results from the failure of the mutants to readily form the out conformation required for flavin reduction by NADPH. The mutants form a large fraction of an abnormal conformation of the reduced enzyme with p-hydroxybenzoate bound. This conformation of the enzyme is unreactive with oxygen. We conclude that transient formation of this open conformation is the mechanism for sequestering p-hydroxybenzoate to initiate catalysis. This overall study emphasizes the role that protein dynamics can play in enzymatic catalysis.     

Mechanistic insights into p-hydroxybenzoate hydroxylase from studies of the mutant Ser212Ala.

In the crystal structure of native p-hydroxybenzoate hydroxylase, Ser212 is within hydrogen bonding distance (2.7 A) of one of the carboxylic oxygens of p-hydroxybenzoate. In this study, we have mutated residue 212 to alanine to study the importance of the serine hydrogen bond to enzyme function. Comparisons between mutant and wild type (WT) enzymes with the natural substrate p-hydroxybenzoate showed that this residue contributes to substrate binding. The dissociation constant for this substrate is 1 order of magnitude higher than that of WT, but the catalytic process is otherwise unchanged. When the alternate substrate, 2,4-dihydroxybenzoate, is used, two products are formed (2,3,4-trihydroxybenzoate and 2,4, 5-trihydroxybenzoate), which demonstrates that this substrate can be bound in two orientations. Kinetic studies provide evidence that the intermediate with a high extinction coefficient previously observed in the oxidative half-reaction of the WT enzyme with this substrate is composed of contributions from both the dienone form of the product and the C4a-hydroxyflavin. During the reduction of the enzyme-2,4-dihydroxybenzoate complex by NADPH with 2, 4-dihydroxybenzoate, a rapid transient increase in flavin absorbance is observed prior to hydride transfer from NADPH to FAD. This is direct evidence for movement of the flavin before reduction occurs.     

Purification and crystallization of recombinant Escherichia coli malate dehydrogenase

Malate dehydrogenase from Escherichia coli has been crystallized with polyethylene glycol and citrate buffer at pH 5.7. The enzyme was obtained from an E. coli strain in which the chromosomal malate dehydrogenase gene was contained on a pBR322 vector. Two types of crystals have been observed; a monoclinic C2 form and an orthorhombic C222(1) form, which is found infrequently. Monoclinic crystals were used as seeds in several rounds of crystallization until large crystals suitable for diffraction analysis were available. A complete X-ray data set to 2.0 A has been collected.     

Preliminary crystallographic investigations of glycinamide ribonucleotide transformylase

Crystals of glycinamide ribonucleotide transformylase have been grown from 0.4 to 1 M ammonium sulfate, 0.6 to 1 M sodium-potassium phosphate, or 0.65 to 1 M citrate in the pH range 4.5-7.0. The single crystals display variable morphology with varying pH. The crystals belong to the orthorhombic space group C222 with cell dimensions a = 141.4 A, b = 98.2 A, c = 103.5 A. Co-crystals have also been obtained in the presence of the inhibitor 5,8-dideazafolate (KI = 18 microM) under similar crystallization conditions. Crystals of a chemically modified enzyme, iodinated at Cys-21, were grown under similar conditions within the pH range 6.5-7.0. These crystals are isomorphous with the unmodified enzyme. Crystals suitable for high resolution (less than 2.5 A) x-ray diffraction studies have been obtained for each of the above.     

Reaction of 2-thio-FAD-reconstituted p-hydroxybenzoate hydroxylase with hydrogen peroxide. Formation of a covalent flavin-protein linkage

Hydrogen peroxide reacts with 2-thio-FAD-reconstituted p-hydroxybenzoate hydroxylase to yield a long wavelength intermediate (lambda max = 360, 620 nm) which can be isolated in stable form on removal of excess H2O2. The blue flavin derivative slowly decays in a second peroxide-dependent reaction to yield a new flavin product lacking long wavelength absorbance (lambda max = 408, 472 nm). This final peroxide-modified enzyme binds p-hydroxybenzoate with a 10-fold lower affinity than does the native enzyme; furthermore, substrate binding leads to the inhibition of enzyme reduction by NADPH. Trichloroacetic acid treatment of the final peroxide-modified enzyme results in the quantitative conversion of the bound flavin to free FAD. However, gel filtration of the modified enzyme in guanidine hydrochloride at neutral pH leads to the co-elution of protein and modified flavin. The nondenatured peroxide product reacts rapidly with hydroxylamine to yield 2-NHOH-substituted FAD. These observations indicate that the secondary reaction of peroxide with the blue intermediate from 2-thio-FAD p-hydroxybenzoate hydroxylase results in the formation of an acid-labile covalent flavin-protein linkage within the enzyme active site, involving the flavin C-2 position.     

Cloning, expression, and regulation of the Pseudomonas cepacia protocatechuate 3,4-dioxygenase genes.

The genes for the alpha and beta subunits of the enzyme protocatechuate 3,4-dioxygenase (EC 1.13.11.3) were cloned from the Pseudomonas cepacia DBO1 chromosome on a 9.5-kilobase-pair PstI fragment into the broad-host-range cloning vector pRO2317. The resultant clone was able to complement protocatechuate 3,4-dioxugenase mutations in P. cepacia, Pseudomonas aeruginosa, and Pseudomonas putida. Expression studies showed that the genes were constitutively expressed and subject to catabolite repression in the heterologous host. Since the cloned genes exhibited normal induction patterns when present in P. cepacia DBO1, it was concluded that induction was subject to negative control. Regulatory studies with P. cepacia wild-type and mutant strains showed that protocatechuate 3,4-dioxygenase is induced either by protocatechuate or by beta-carboxymuconate. Further studies of P. cepacia DBO1 showed that p-hydroxybenzoate hydroxylase (EC 1.14.13.2), the preceding enzyme in the pathway, is induced by p-hydroxybenzoate and that beta-carboxymuconate lactonizing enzyme, which catalyzes the reaction following protocatechuate 3,4-dioxygenase, is induced by both p-hydroxybenzoate and beta-ketoadipate.     

Metabolism of cyclohexane carboxylic acid by Alcaligenes strain W1.

Thirty-three microorganisms capable of growth with cyclohexane carboxylate as the sole source of carbon were isolated from mud, water, and soil samples from the Aberystwyth area. Preliminary screening and whole-cell oxidation studies suggested that, with one exception, all of the strains metabolized the growth substrate by beta-oxidation of the coenzyme A ester. This single distinctive strain, able to oxidize rapidly trans-4-hydroxycyclohexane carboxylate, 4-ketocyclohexane carboxylate, p-hydroxybenzoate, and protocatechuate when grown with cyclohexane carboxylate, was classified as a strain of Alcaligenes and given the number W1. Enzymes capable of converting cyclohexane carboxylate to p-hydroxybenzoate were induced by growth with the alicyclic acid and included the first unambiguous specimen of a cyclohexane carboxylate hydroxylase. Because it is a very fragile protein, attempts to stabilize the cyclohexane carboxylate hydroxylase so that a purification procedure could be developed have consistently failed. In limited studies with crude cell extracts, we found that hydroxylation occurred at the 4 position, probably yielding the trans isomer of 4-hydroxycyclohexane carboxylate. Simultaneous measurement of oxygen consumption and reduced nicotinamide adenine dinucleotide oxidation, coupled with an assessment of reactant stoichiometry, showed the enzyme to be a mixed-function oxygenase. Mass spectral analysis enabled the conversion of cyclohexane carboxylate to p-hydroxybenzoate by cell extracts to be established unequivocally, and all of our data were consistent with the pathway: cyclohexane carboxylate --> trans-4-hydroxycyclohexane carboxylate --> 4-ketocyclohexane carboxylate --> p-hydroxybenzoate. The further metabolism of p-hydroxybenzoate proceeded by meta fission and by the oxidative branch of the 2-hydroxy-4-carboxymuconic semialde-hyde-cleaving pathway.     

The temperature and pH dependence of some properties of p-hydroxybenzoate hydroxylase from Pseudomonas fluorescens

The free and complexed flavoprotein, p-hydroxybenzoate hydroxylase, was studied by light-absorption, circular-dichroism and fluorescence techniques as a function of the pH. The following compounds served as ligands for the enzyme: p-hydroxybenzoate, p-fluorobenzoate, benzoate, p-aminobenzoate and tetrafluoro-p-hydroxybenzoate. Depending on the technique used, the various ligands exhibit pH-dependent physical properties and dissociation constants. The data can be fitted with pKa values in the range 7.7-7.9. It is suggested that this pKa value belongs to a tyrosine residue in the active center of the enzyme. This assignment is supported by published data and additional experiments.     

Crystal structure of the reduced form of p-hydroxybenzoate hydroxylase refined at 2.3 A resolution.

The crystal structure of the reduced form of the enzyme p-hydroxybenzoate hydroxylase from Pseudomonas fluorescens, complexed with its substrate p-hydroxybenzoate, has been obtained by protein X-ray crystallography. Crystals of the reduced form were prepared by soaking crystals of the oxidized enzyme-substrate complex in deaerated mother liquor containing 300-400 mM NADPH. A rapid bleaching of the crystals indicated the reduction of the enzyme-bound FAD by NADPH. This was confirmed by single crystal spectroscopy. X-ray data to 2.3 A were collected on oscillation films using a rotating anode generator as an X-ray source. After data processing and reduction, restrained least squares refinement using the 1.9 A structure of the oxidized enzyme-substrate complex as a starting model, yielded a crystallographic R-factor of 14.8% for 11,394 reflections. The final model of the reduced complex contains 3,098 protein atoms, the FAD molecule, the substrate p-hydroxybenzoate and 322 solvent molecules. The structures of the oxidized and reduced forms of the enzyme-substrate complex were found to be very similar. The root-mean-square discrepancy for all atoms between both structures was 0.38 A. The flavin ring is almost completely planar in the final model, although it was allowed to bend or twist during refinement. The observed angle between the benzene and the pyrimidine ring is 2 degrees. This value should be compared with observed values of 10 degrees for the oxidized enzyme-substrate complex and 19 degrees for the enzyme-product complex. The position of the substrate is virtually unaltered with respect to its position in the oxidized enzyme. No trace of a bound NADP+ or NADPH molecule was found.     

A study of p-hydroxybenzoate hydroxylase from Pseudomonas fluorescens. Improved purification, relative molecular mass, and amino acid composition.

The purification procedure for p-hydroxybenzoate hydroxylase has been modified by replacement of the DEAE-cellulose (DE-32) column in the original procedure by a Sephadex--Cibacron-blue affinity column. In this way the yield of enzyme could be improved from 16% to about 40--50%. Preparative gel chromatography indicated that the enzyme does not exist as a monomeric species as earlier believed but mainly as a dimer. Sodium dodecyl sulfate gel electrophoresis of purified enzyme revealed a minimum relative molecular mass (Mr) of 43000--45000. Analytical gel chromatography, sedimentation equilibrium and sedimentation velocity experiments showed that the enzyme exists in solution mainly as a dimer but also in higher-order quaternary structures (presumably tetramer and hexamer). Temperature dependence of the distribution of the oligomers suggests that the association is of hydrophobic nature. The amino acid composition of the enzyme is also presented. The enzyme contains no disulfide but five sulfhydryl groups. In the native state of the enzyme only one sulfhydryl group is accessible to N-ethylmaleimide or 5,5'-dithiobis(2-nitrobenzoic acid). The iso-electric point of the enzyme was found to be 5.8.