HOMOGENTISATE OXIDASE OF LIVER*

Transcription

1 HOMOGENTISATE OXIDASE OF LIVER* BY W. EUGENE KNOX AND SALLY W. EDWARDSt (From the Department of Biochemistry and Nutrition, Tufts College Medical Scbol, the Cancer Research Institute, New England Deaconess Hospital, and the Department of Biological Chemistry, Harvard Medical School, Boston, Massachusetts) (Received for publication, December 4, 1954) The liver enzymes which metabolize homogentisate have been separated into an oxidizing fraction and a hydrolyzing fraction by Ravdin and Cranda11 (1). These workers isolated fumarylacetoacetate from the reaction catalyzed by the oxidizing fraction and found that this compound was hydrolyzed by the second fraction to the final products, fumarate and acetoacetate. The present paper describes the preparation and properties of HG oxidase, apparently a single enzyme of a new type, and the reaction which it catalyzed. It oxidized HG not to FAcAc, aa expected, but to another product. The probable identification of this product aa 4-maleylacetoacetate, and its conversion to FAcAc by a specific cis-trans isomerase which uses glutathione as a coenzyme, will be described in later papers (2,3). Enzyme Assays The oxidation of HG wa+~ followed by two methods in parallel, by oxygen uptake at 37 and by a spectrophotometric assay at The homogentisic acid used aa the substrate (white crystals, m.p. 149, corrected) was isolated M the lead salt from alkaptonuric urines (4). The reaction mixtures were identical for both assays. Each contained 1 ml. of 0.1 M 2,3,6-collidine buffer (5)) ph 7.2, enzyme, and 5 to 8 pmoles of homogentisic acid (unneutralized to avoid inhibition by products of its non-enzymatic oxidation) in 3.0 ml. of total volume. Blank reactions without homogentisic acid were always run at the same time. Assay by Oxygen Uptake-The reactions were run in Warburg respirometers at 37, with air as the gas phase. The rate of oxygen uptake was linear until about 90 per cent of the substrate was oxidized, and the rate * This investigation was supported by grants No. A298 and A567 from the National Institute of Arthritis and Metabolic Diseases, United States Public Health Service, and by United States Atomic Energy Commission contract No. AT(30-I)-901 with the New England Deaconess Hospital. t Public Health Service Research Fellow of the National Institutes of Health. 1 In this paper the following abbreviations will be used: HG (homogentisate), FAcAc (fumarylacetoacetate), MAcAc (maleylacetoacetate), and GSH (glutathione). 479

2 480 HOMOGENTISATE OXIDASE was proportional to the concentration of enzyme present. In order to detect the occurrence of any non-enzymic oxidation under the different experimental conditions, amounts of enzyme and substrate were chosen such that the reaction would be complete within an hour, and all the reactions measured were followed to completion. The occurrence of nonenzymic oxidation of HG was shown by a total oxygen uptake of more than the theoretical 1 molecule of oxygen per molecule of HG and by.6- I 2 MIbT TEs4 5 6 FIG. 1. Spectrophotometric assay of homogentisate oxidase at different enzyme concentrations. 0.1 to 0.6 ml. of Fraction II enzyme was assayed as described in the text, without addition of ascorbic acid or glutathione. The dotted lines show the reaction of 0.1 ml. of an ammonium sulfate-precipitated enzyme and its activation with 2.5 pmoles of glutathione. Ascorbic acid (AS) additions had the same effect in this preparation. darkening of the reaction mixture. The results of such experiments were discarded. Spectrophotometric Assay of HG Oxidation-Suitable enzyme preparations were assayed directly in the Beckman spectrophotometer by the change in the optical density at 330 rnp in a 1 cm. cell at The non-absorbing HG was converted by the reaction into a strongly absorbing diketo acid ( e33~mp = 13,500 at ph 7.2 for MAcAc). The optical density changed at a linear rate proportional to enzyme concentration, except at the highest levels of diketo acid accumulation (Fig. 1). The crude soluble liver fractions could not be assayed in t.his way because they accumulated no diketo acid. The removal of GSH by dialysis or alcohol precipitation of the enzyme was the most important factor in making a preparation suitable for

3 W. E. KNOX AND S. W. EDWARDS 481 this assay. GSH addition was usually avoided in this assay because of its rale in the further reaction of the diketo acid. The rates of the reaction determined by the two assay methods paralleled each other in all the different enzyme preparations and were of approximately the same absolute magnitude. The molar rate of HG oxidation in the oxygen uptake assay was 2.2, 2.2, and 1.6 times that in the spectrophotometric assay in three different enzyme preparations. This difference could be attributed to the 10 difference in temperature between the two assays. Preparations of Enzyme Fractions Detailed observations were made only on the HG-oxidizing system from rat liver, although similar systems were found in the soluble fraction of rabbit, guinea pig, beef, pork, and horse liver homogenates. It was essential that all manipulations of the enzyme be carried out at 4 or lower, and at ph 7 or above, in order to achieve the two purposes of the preparations, the removal of GSH and the retention of enzymic iron. All preparations were stored at ph 7.5 and -9. Fresh rat livers were homogenized for 2 minutes in twice their volume of 0.9 per cent KCl. The soluble fraction (Fraction I), obtained by 15 minute centrifugation at 12,000 X g, contained all the HG oxidase activity of the original homogenate (approximately 200 ccl. of 02 uptake in 10 minutes per ml.). This fraction was adjusted to ph 7 and cooled to volume of 95 per cent ethyl alcohol was added slowly and the mixture immediately centrifuged. The precipitate was drained well and then resuspended in fresh 32 per cent alcohol at 0 equal in volume to the supernatant fraction removed. After another centrifugation the precipitate was dissolved in water (one-half the volume of the original Fraction I), adjusted to ph 7.5, and dialyzed against water for 2 hours with efficient stirring. This was then frozen at - 9 (Fraction II). When it was thawed at least 24 hours later, a large inactive precipitate was centrifuged off. The fraction contained about one-half the activity, and had about twice the specific activity, of Fraction I. There was a sufficient degree of functional purification to cause the accumulation of MAcAc. The enzyme at this stage was stable at -9 for several months. Part of the activity in Fraction II could be precipitated between 50 and 70 per cent saturation with ammonium sulfate adjusted to ph 7.5 with ammonia. The activity in this preparation did not survive the acidification incidental to dialysis and could be preserved only for several days. The preparation had a somewhat higher specific activity than Fraction II and required activation by reducing agents (see below).

4 482 XOMOGENTISATE OXIDASE Properties of Homogentisate Ox&se HG O&se React&--The reactions catalyzed by the different enzyme preparations are described in Table I. The enzyme preparations all took up approximately 1 molecule of oxygen per molecule of substrate, regardless of which product was formed. The reactions of all the many enzyme prep- TABLE Reaction of Homogentisate in Diferent Liver Fractions The indioated amounts of homogentisate were oxidized to completion as described for the oxygen uptake assay. CO, was not formedjn the reaction, as shown by identical manometric changes with and without NaOH in the center wells. CO2 evolution from bicarbonate-co2 buffer was determined by difference with the 01 uptake measured in collidine buffer. Acetoacetate was determined by aniline citrate decarboxylation in 10 minutes (7). Maleylacetoacetate was determined spectrophotometrically at 330 mf~ (2) Fraction I 11* + supernatantt (Xl+) $04 ppt. In bicarbonate buffer Fraction 115 I I + supernatantj HG added j&m& I On uptake pnolcs co2 evolved Os/HG COr/HG _ MACAC pmoles ACAC Jmolcs % (3.6)/l 0% (4.0)(( 0% * Once precipitated with 32 per cent alcohol. t 0.2 ml. of supernatant fraction from the alcohol precipitation of Fraction II. $ Essentially none. The exact amount was indeterminate in the presence of the slowly decarboxylating diketo acid. 0 Precipitated and washed with 32 per cent alcohol. 11 Determined by aniline citrate decarboxylation in 70 minutes. arations studied here were followed to completion, and none showed evidence of only 1 atom of oxygen taken up per molecule of HG, although the preparations differed in source, age, and previous treatment, and many were assayed in the presence of various inhibitors. In particular, neither the absence of phosphate nor the use of borate buffer resulted in the uptake of 0.5 molecule of oxygen per molecule of substrate, as reported by Suda and Takeda (6), although the enzyme was soon inactivated in borate by acidification of the reaction mixture unless the ph was controlled by other means.

5 W. 1. KNOX AND S. W.. EDWARDS 483 Table Ialso indicates that acetoacetate was a product of the reaction in the crude extract (Fraction I), but that the yield of acetoacetate was decreased after one alcohol precipitation of the enzyme (Fraction II) and eliminated by further treatment of the enzyme. An intermediate which was decarboxylated by aniline citrate (7) more slowly than was acetoacetate replaced it as the reaction product. The addition of the supernatant fraction left from the alcohol precipitation of Fraction II restored the formation of acetoacetate without changing the oxygen consumption of the reaction. This supernatant fraction after dialysis lost the ability to form acetoacetate from the intermediate unless GSH was added, although the same dialyzed fraction without GSH formed acetoacetate from FAcAc. This observation formed the basis for an enzymic differentiation between the new product and FAcAc (3). In the last experiments of Table I, the accumulated precursor of acetoacetate was a dicarboxylic acid, since 1 molecule of CO* was liberated from bicarbonate buber for each molecule of HG oxidized to the intermediate. The conversion of this dicarboxylic acid intermediate to acetoacetate in the presence of the undialyzed supernatant fraction was associated with the liberation of a second molecule of COa. This indicated the formation of an additional acid group which would be expected from the enzymic hydrolysis of FAcAc to fumarate and acetoacetate, the demonstrated products of this reaction (1). However, FAcAc was not accumulated by any of the enzyme preparations. Most preparations of the Fraction II precipitated once with alcohol were in fact able to hydrolyze FAcAc. The new dicarboxylic acid intermediate which was accumulated was a diketo acid distinguished from FAcAc by its lack of any reaction with the dialyzed supernatant fraction from Fraction II which hydrolyzed FAcAc, and by its absorption spectrum, which was used to measure it in the experiments of Table I (2). R8l.e of Iron in HG Ox&fuse-The HG oxidase did not display the usual properties of known iron-containing oxidative enzymes. Prepared as described and kept at a ph of 7 or above, it was not significantly activated by the addition of Fe*. Acidification to ph 5, which was earlier found to inactivate the enzyme (8), resolved it so that activation with Fe++ occurred as has been described (9), but the original activity could be only partially restored and hence such treatment was avoided in the preparation procedures. The HG oxidase reaction did not form hydrogen peroxide detectable by the coupled oxidation of ethyl alcohol with catalase (lo), and the intermediary formation of peroxide did not occur as in the liver tryptophan-peroxidase-oxidase system (11). The latter type of reaction was excluded because no requirement for peroxide was revealed by partial purification of the enzyme, catalase did not cause inhibition, and peroxide generated by glucose oxidase during the reaction did not activate the re-

6 484 HOMOGENTISATE OXIDASE action. CO in the dark did not inhibit the enzyme (Table.11) as it does the iron-porphyrin proteins which react with oxygen. The effects of several other compounds which react with. iron are.given in Table II to define further the role of iron in the HG oxidation reaction. The complete inhibition of oxygen uptake which was caused by.(~,a-dipyridyl was paralleled by complete inhibition of diketo acid or acetoacetate formation in the same experiments. These results provided no evidence for a participation of iron in reactions later than the first oxidation step, TABLE Inhibition of Homogentisate Om dase by Compounds Reacting with Iron These results were obtained in oxygen uptake assays. The compounds tested were in contact with the enzyme 10 minutes before the substrate was added. Similar results with all compounds were obtained in the spectrophotometric assay. II Compound Concentration Inhibition M per cent a+-dipyridyl x lo-3* x 10-s* 80 Naacide x 10-z 83 NaCN x 10-S 27 co-o* yo (Dark) 0 Phosphate... 3 x lo-et Pyrophosphate 3 x IO-St 44 Cysteine... 3 x 10-s 731 Cystine. 4.5 x 10-S w Cyeteic acid. 4.5 x 10-a 15 * The solution was colorless until the substrate was added; then it immediately turned red. t Compared with the collidine-buffered reaction. The inhibition by phosphate was variable. t Not reversed by M GSH added 5 minutes after the inhibitor. as was suggested by Schepartz (9). We observed an excess of oxygen uptake over acetoacetate formation such as he reported only when the diketo acid was accumulated as a result of the lack of GSH or when nonenzymic oxidation of HG occurred. The latter reaction was seen only in certain preparations with low activity, those containing heat-inactivated enzyme, or in the presence of an excess of metal salts or of complex-forming agents. In addition to the inhibition by a!,a-dipyridyl, which reacts with ferrous iron, the reaction was also inhibited by azide and cyanide, which characteristically react with ferric iron. These results suggested a valence change of iron during the reaction. The effects of pyrophosphate and even of phosphate on the reaction emphasized the availability of the iron, already suggested by the easy resolution of the enzymic iron.

7 W. E. KNOX AND S. W. EDWARDS 485 Cysteine and cystine inhibited HG oxidation and were at the same time rapidly oxidized in the system. Cysteic acid, whose maximal known effect is given in Table II, was considerably less inhibitory. Either chelation of the enzyme iron by cysteine or irreversible oxidation of groups on the en- TABLE Inhibition and Activation of Homogentisate Oxidase As Sulfhydryl Enzyme The substances to be tested were added 10 minutes before the substrate in the oxygen uptake assay and 3 minutes before the substrate in the spectrophotometric assay. All effects were demonstrable by both assays. The activations (+) were all measured in enzymes which had been precipitated with ammonium sulfate at ph 7.5, but were occasionally observed also in Fraction II preparations. HgSOi. Substance tested p-chloromercuribenzoatc.. Hz02 from glucose oxidase.. - III Di5erence oncentration or amount Assay from control activity 3 X lo- M X 10-6M 6 X IO-6 Y 3 x 10-s 0.05 rmole per 10 min. I Benzoquinoneacetic acid. 2.9 x 10-d M Ferriey&ide... 1 x lo-* M Ascybic acid. 8 x x lo--4 Glutathione... 8 X 10-d... 8 x 10-4 Coenzyme A mg. Mercaptoacetic acid... i 8 X 10m4 M Denatured liver proteini. ] 5 mg. per cd * 0-66* -651 Spectrophotometric ; Spectrophotometric +83 I Spectrophotometric Spectrophotometric +3s 02 I +64 * Inhibition completely reversed by glutathione (10-s M) added 5 minutes after the substrate. t Inhibition was prevented by the earlier addition of 1 mg. of crystalline catalase but was not reversible. 1 Precipitated between 60 and 70 per cent alcohol from dialyzed rat liver homogenate and heated at 60 for 5 minutes. Positive nitroprusside test. zyme, perhaps by peroxide, was a more likely cause of the cysteine inhibition than a reaction of cystinc with enzyme thiol groups, since GSH did not reverse the cysteine or cystine inhibitions. Essential Thiol Groups of HG Otidase-The inhibition of the enzyme by Hg+ and by p-chloromercuribenzoate, both reversed by GSH, and the irreversible inhibitions by other reagents given in Table III demonstrated the presence of essent,ial thiol groups on the enzyme. This finding can explain the previously observed inhibitions of this enzyme by benzoquinoneacetic acid, several other quinones, and maleate (9).

8 436 HOMOGENTI8A!4YE OXIDASE Certain enzyme preparations (notably the ammonium sulfate fractions) required addition of reducing agents for full activity. This requirement was at first overlooked because of the inhibition caused by cysteine. The activation of HG oxidase by ascorbic acid and several thiols is shown in Table III. This activation was more readily demonstrated in the oxygen uptake assay than under the milder conditions and shorter reaction period of the spectrophotometric assay. Examples from both assays are given, including activation by GSH in the spectrophotometric assay in which the activation could be demonstrated only in a preparation nearly free of the enzymes which removed the diketo acid (see also Fig. 1). Activation by a thiol-containing protein emphasized the lack of specificity of this effect. A small activation of the oxidation of HG to acetoacetate was reported by Williams and Sreenivasan, who suggested that it was a specific effect of ascorbic acid (12). However, they did not test GSH or the other effective reducing agents in this reaction. Ascorbic acid also activated resolved HG oxidase by reducing ferric to ferrous iron (6). Ferrous iron did not activate the preparations used in the present experiments. The activations observed were most probably the result of protection of thiol groups of the enzyme from oxidation during the reaction. DISCUSSION The most important factors in obtaining active and stable HG oxidase preparations which accumulated the new product of the reaction were the removal of GSH, avoidance of resolution of the enzymic iron, and the addition of reducing agents such as ascorbic acid when these were required. The reaction catalyzed by these preparations of HG oxidase showed the stoichiometry expected for the formation of a compound similar to FAcAc, which has been isolated from similar preparations. 1 molecule of oxygen was taken up and 1 molecule of Cot was displaced from bicarbonate, per molecule of HG oxidized. The accumulated compound was not FAcAc, but was converted to acetoacetate (and presumably fumarate) by the fraction of liver separated from the oxidase. The oxidation of HG with 1 molecule of oxygen to the new diketo acid appeared to be the result of the action of a single enzyme, to which we restrict the name HG oxidase. The enzyme can be partially defined in terms of its reaction, the nature of its contained iron, and its essential thiol groups. These properties distinguish it from other types of known enzymes with the possible exception of two bacterial enzymes, pyrocatecase and proto- catechuic acid oxidase, whose similarities with the present enzyme have already been recorded by Starrier and Ingraham (13). It may be that HG oxidase represents a new type of non-porphyrin, iron-containing oxidative enzyme. An association of the iron with the protein thiol groups in this enzyme has been suggested by recent experiments of Crandall(14).

9 W. E. KNOX AND 8. W. EDWARDS 487 SUMMARY 1. Homogentisate oxidase, prepared from liver by alcohol precipitation, formed a new diketodicarboxylic acid of the same oxidation level as fumarylacetoacetate. The accumulation of this initial product of the oxi- dation was the basis of a spectrophotometric assay of the enzyme reaction. 2. The participation of iron in this reaction was confirmed. There was evidence that the iron behaved unlike an iron porphyrin and changed valence during the reaction. 3. The enzyme possessed essential sulfhydryl groups, and certain prep- arations were activated non-specifically by ascorbic acid, glutathione, and other thiol compounds except cysteine. Dr. Charles H. DuToit, Massachusetts General Hospital, and Dr. James Patterson, New England Center Hospital, kindly made available to us urine collections from their alkaptonuric patients as a source of homogentisic acid. The technical assistance of Miss Anita Aspen is gratefully acknowledged. BIBLIOGRAPHY 1. Ravdin, R. G., and Crandall, D. I., J. Biol. Chem., 189, 137 (1951). 2. Knox, W. E., and Edwards, S. W., J. Biol. Chcm., 216,489 (1955). 3. Knox, W. E., and Edwards, S. W., Federation Proc., 13,242 (1954). 4. Neuberger, A., Rimington, C., and Wilson, J. M. G., Biochem. J., 41,438 (1947). 5. Gomori, G., Proc. Sot. Exp. Biol. and Med., (1946). 6. Suds, M., and Takeda, Y., J. Biochem., Japan, 37,381 (1950). 7. Edson, N. L., Biochem. J., 29, 2032 (1935). 8. Knox, W. E., and LeMay-Knox, M., Biocbem. J., 49, 086 (1951). 9. Schepartz, B., J. Biol. Chem., 906, 185 (1953). 10. Keilin, D., and Hartree, E. F., Biochem. J., 39, 293 (1945). 11. Knox, W. E., B&him. et biophye. acta, 14, 117 (1954). 12. Williams, J. N., Jr., and Sreenivasan, A., J. BioZ. Chem., (1953). 13. Stanier, R. Y., and Ingraham, J. L., J. BioZ. Chem., 210, 799 (1964). 14. Crandall, D. I., in McElroy, W. D., and Glass, H. B., A symposium on amino acid metabolism, Baltimore, 867 (1955).

10 HOMOGENTISATE OXIDASE OF LIVER W. Eugene Knox and Sally W. Edwards J. Biol. Chem. 1955, 216: Access the most updated version of this article at Alerts: When this article is cited When a correction for this article is posted Click here to choose from all of JBC's alerts This article cites 0 references, 0 of which can be accessed free at tml#ref-list-1