Enzymic Conversion of 3-Hydroxyanthranilic Acid Into Cinnabarinic Acid PARTIAL PURIFICATION and PROPERTIES of RAT-LIVER CINNABARINATE SYNTHASE by P

Biochem. J. (1966) 99, 317 317

Enzymic Conversion of 3-Hydroxyanthranilic Acid into Cinnabarinic Acid PARTIAL PURIFICATION AND PROPERTIES OF RAT-LIVER CINNABARINATE SYNTHASE By P. V. SUBBA RAO AND C. S. VAIDYANATHAN Department of Biochemistry, Indian Institute ofScience, Bangalore, India (Received 20 October 1965)

Rat-liver cinnabarinate synthase (3-hydroxyanthranilic acid-oxygen oxido- reductase) was partially purified. Stoicheiometric studies indicated the consumption of 3 atoms of oxygen/molecule of cinnabarinic acid formed. There was an initial lag in enzyme activity. The reaction had an optimum pH about 7-2 and an optimum temperature of 37°. The enzyme was highly specific for 3-hydroxyanthranilic acid. The system showed an absolute requirement for Mn2+ ions. Several bivalent metal ions and metal-chelating agents inhibited the reaction. Thiol inhibitors had no effect on enzyme activity, but reducing agents such as ascorbic acid were potent inhibitors. There was no requirement for any cofactor other than Mn2+ ions. The probable significance of the reaction in mammals is discussed.

Recent interest in the initiation of bladder tumours in experimental animals (Bryan, Brown & MATERIALS AND METHODS Price, 1964) by certain o-aminophenols such as All reagents were of analytical grade and glass-distilled 3-hydroxyanthranilic acid and 3-hydroxykynure- water was used for the preparation of all solutions. nine, which are normal metabolites, as well as the Chemical&. 3-Hydroxyanthranilic acid was purchased excretion of abnormally large quantities of trypto- from Mann Research Laboratories Inc., New York, N.Y., U.S.A. Horse-heart cytochrome c, horse-radish peroxidase, phan metabolites in the urine of patients bearing ox-liver catalase, NAD, NADP, FMN, FAD and p-chloro- bladder tumours in a normal population (Boyland mercuribenzoate were obtained from Sigma Chemical Co., & Williams, 1956; Brown, Price, Satter & Wear, St Louis, Mo., U.S.A. Sodium arsenite, 2,3-dimercapto- 1960), calls for a fresh search for alternative path- propanol, cysteine and GSH were obtained from British ways for the metabolism of these tryptophan Drug Houses Ltd., Poole, Dorset. Haem and haematin metabolites. The enzymic conversion of 3-hydroxy- were prepared from Sigma haemin. anthranilic acid into cinnabarinic acid (2-amino- Purification ofenzyme. All operations were carried out at 3-oxo-3H-phenoxazine-1,9-dicarboxylic acid) by 0-4o. the nuclear fraction of rat liver has been described Step 1. The nuclear fraction from 25g. of rat liver was isolated according to the procedure of Subba Rao, by Subba Rao, Jegannathan & Vaidyanathan Jegannathan & Vaidyanathan (1965). The nuclear fraction (1964). It was reported by Morgan & Weimorts obtained from 25g. of liver was finally suspended in 65ml. (1964) that acetone-dried powders from livers of of 0.9% NaCl. poikilothermic animals can also catalyse such an Step 2. The nuclear fraction (65ml.) was centrifuged at enzymic oxidation of 3-hydroxyanthranilic acid. 12000g for 5min., the sediment was suspended in 60ml. This enzyme, which has been purified by Morgan, of O-OlM-sodium phosphate buffer, pH7-0, and homo- Weimorts & Aubert (1965), is unspecific and can genized in a VirTis homogenizer for 5-lOmin. The extract oxidize related o-aminophenols also to phenoxazine was centrifuged at 12000g for lOmin. compounds. In view of the carcinogenic activity Step 3. To 54ml. of solubilized extract from step 2 were further studies added 6ml. of 0-lM-MnSO4. The mixture was stirred for ofcinnabarinic acid (Boyland, 1960), 15min. and centrifuged at 12000g for 15min. The precipi- on the enzyme cinnabarinate synthase have been tate was discarded and the clear supernatant was dialysed carried out. The enzyme from rat liver has been thoroughly against water and clarified by centrifugation. partially purified and properties of this enzyme, Step 4. The fraction from step 3 (50nml.) was brought to which is specific for 3-hydroxyanthranilic acid, are 30% saturation by the addition of 8-2g. of recrystallized described. (NH4)2SO4, stirred for lBmin. and centrifuged at 12000 g 318 P. V. SUBBA RAO AND C. S. VAIDYANATHAN 1966 Table 1. Purification of cinnabarinate synthase from rat liver The reaction mixture (1 ml.) containing sodium phosphate buffer, pH7*2 (50,umoles), 3-hydroxyanthranilic acid (4,umoles), MnSO4 (0.25/umole) and O lml. of enzyme was incubated at 370 for 30min. (Preparations from steps 1-5 were diluted tenfold and the final preparation was diluted fivefold.) Cinnabarinic acid formed in the reaction mixture was estimated as described in the Materials and Methods section. A unit of activity is defined as the amount of enzyme that catalyses the formation of 1,mole of cinnabarinic acid/min. Specific Total Total activity Volume protein activity (units/mg. Recovery Step (ml.) (mg.) (units) of protein) (%) Purification 1 65 845 112-2 0-13 100 2 60 366 66-0 0-18 58-9 1.4 3 62 149 59.5 0*41 53-1 3*2 4 41 31-2 33-2 1-06 29-6 8-2 5 27 6*75 22-5 3-30 20-0 25-4 6 54 2-38 20-3 8*53 18.1 65-6

for 15min. The precipitate was discarded. The supernatant 0 4 was brought to 60% saturation by the addition of 9 59g. of (NH4)2SO4. After stirring for 15min., the precipitate was collected by centrifugation, suspended in 33ml. of water and dialysed against water. Step 5. The dialysed preparation (step 4) (30ml.) was fractionated at -4° with acetone precooled to -20°. The 0L2 precipitate obtained at 15-40% (v/v) acetone was collected, E dissolved in 20ml. of water and centrifuged to get a clear solution. 01I Step 6. The solution (15ml.) of the precipitate from the acetone fractionation was adjusted to pH6*0 with dilute acetic acid, and tricalcium phosphate gel (45 ml. of a 0 preparation containing 14mg./ml.; Keilin & Hartree, 1938) 340 380 420 460 500 was added. The mixture was stirred for 15min. and the Wavelength (mu) gel was separated by centrifugation and resuspended in 30ml. of 005m-sodium phosphate buffer, pH7-2. After Fig. 1. Visible spectrum of cinnabarinic acid (6.2tLg./ml.) 20min. the mixture was centrifuged and the clear solution in an aqueous solution of 1% (w/v) trichloroacetic acid. was used as enzyme. The results of the purification procedure are given in Table 1. Measurement of enzyme activity. The spectrum of cinna- Oxygen uptake. Warburg manometers contained sodium barinic acid in 1% (w/v) trichloroacetic acid as taken on a phosphate buffer, pH7.2(1001&moles), 3-hydroxyanthranilic Beckman model DB recording spectrophotometer is shown acid (6,umoles), MnSO4 (0.75umole) and 1 ml. of enzyme in Fig. 1. The absorption maxima were at 430mit and in a final volume of 3ml. The reaction was started by the, 450-452m,u. The molar extinction coefficient (e) at 450mi. addition of substrate from the side arm and the oxygen was 17280, and there was linearity between the extinction uptake was measured over a period of 60min. at 300. at 450m,u and concentration of cinnabarinic acid in the Spectrophotometry. The reduction and reoxidation of range 5-40,tg./3 ml. cytochrome c in the presence of 3-hydroxyanthranilic acid The standard reaction mixture consisted of sodium and the enzyme was measured in cuvettes maintained at phosphate buffer, pH7.2 (50,tmoles), 3-hydroxyanthranilic 30°. Difference spectra were recorded on a Beckman model acid (4,umoles), MnSO4 (0 25,umole) and Ol ml. of enzyme DB recording spectrophotometer. The reaction mixture (0.8jg. ofprotein) in a totalvolume of1 ml. After incubation (3.Oml.) consisted of sodium phosphate buffer, pH7.2 for 30min. at 37°, the reaction was stopped by the addition (100l,moles), cytochrome c (0-2,umole), MnSO4 (0 3,umole) of 1 ml. of 3% (w/v) trichloroacetic acid followed by 1 ml. and 3-hydroxyanthranilic acid (0.1 imole). It was scanned of water. After centrifugation, 1 ml. samples were diluted against a blank in which the substrate was omitted. After to 3ml. with 1% (w/v) trichloroacetic acid, and cinnabarinic the addition of 0-5ml. of enzyme to both cuvettes, the acid formed was calculated from the extinction at 450m,u. spectrum was recorded at 5min. intervals. To correct for non-enzymic oxidation of3-hydroxyanthran- flic acid, blanks containing boiled enzyme were set up in RESULTS each experiment. Determination ofprotein. This was done according to the Requirementfor Mn2+ ions. The partially purified method of Lowry, Rosebrough, Farr & Randall (1951). enzyme showed an absolute requirement for Mn2+ Vol. 99 RAT-LIVER CINNABARINATE SYNTHASE 319

Table 2. Effect of Mn2+ ions on cinnabarinic acid 0 5 formation by purified rat-liver cinnabarinate synthase a) Enzyme containing 0-8j,g. of protein was preincubated 0 with the indicated final concentration of MnSO4 for 5min. 0-4 at 370 before the substrate was added. The rest of the procedure was as described in the Materials and Methods 0) section. 0- 3 .0 Concn. of MnSO4 Cinnabarinic acid formed IC (mM) (,umole) 0 o 0 2 - 0 9: 0-01 0-05 .; I 0-025 0-09 Ca0*Ca 0-05 0-12 0-1 _ 0-1 0-15 0 0-25 0-18 0-5 0-12 120 150 1-0 0-10 0 30 60 90 Time (min.) Fig. 2. Time-course of cinnabarinic acid formation in the presence of purified rat-liver cinnabarinate synthase. The ions, the optimum concentration of which was reaction mixture consisted of sodium phosphate buffer 0-25mm in the presence of 4mM-3-hydroxy- (50umoles), 3-hydroxyanthranilic acid (4,umoles), acid MnSO4 anthranilic (Table 2). None of the other (0-25,umole) and enzyme (0-8,g. of protein) in a final bivalent metal ions Cu2+, Fe2+, Ca2+, Mg2+ and volume of lml. The incubation was at 37° for various Zn2+ were effective. The enzyme preparations periods. Cinnabarinic acid formed was estimated as retained activity for at least 10 days without described in the Materials and Methods section. appreciable loss when stored at 4°. Optimum pH. The optimum pH for purified cinnabarinate synthase from rat liver was about 7-2 when determined with sodium phosphate Table 3. Effect of reducing agents on the activity of buffers ranging from pH 6-2 to 7-8. purified rat-liver cinnabarinate synthase Substrate concentration and substrate specificity. The conditions were as described in the Materials and The optimum substrate concentration for the Methods section. reaction was 4nar. The Km calculated from Lineweaver-Burk plots was 1-25 x 10-3M. The Final enzyme showed a remarkably high specificity for concn. Inhibition Addition 3-hydroxyanthranilic acid. The structurally re- (mm) (%) lated compounds 3-hydroxykynurenine and GSH 0-5 21 o-aminophenol were not oxidized when tested 1-0 58 under standard assay conditions. Cysteine 0-5 0 Time-course the reaction. 1-0 68 of As observed with the 2,3-Dimercaptopropanol 0-5 100 nuclear fraction (Subba Rao et al. 1965), the Ascorbic acid 0-1 53 formation of cinnabarinic acid mediated by the 0-5 100 purified enzyme exhibited a lag phase (Fig. 2). Sodium borohydride 2 mg./ml. 100 The lag period could not be abolished by preincu- bating the enzyme with Mn2+ ions for lOmin. In exhibiting an initial lag in enzyme activity, rat- liver cinnabarinate synthase resembles tryptophan Influence of temperature. The enzyme showed pyrrolase (Tokuyama & Knox, 1964). However, maximal activity at 37°. Further increase in tem- preincubation of the purified enzyme with haem perature caused rapid loss in activity. The enzyme (0-5,umole) or haematin (0-5,mole) plus ascorbic was inactivated when heated at 600 for 5min. in acid (0-5,umole) or sodium borohydride (0-5,umole) the absence of substrate. did not abolish the lag period. In fact, these Effect of metal ions. These effects, which were reducing agents inhibited the reaction completely determined in the presence of optimum concentra- (Table 3). tion of Mn2+ ions, are summarized in Table 4. Oxygen uptake. In the course of the reaction The ions Cu2+, Hg2+, Ag+, Fe2+, Fe3+, Ni2+ and catalysed by cinnabarinate synthase 1-8S,ug.atoms MoO42- caused appreciable inhibition. Enzyme of oxygen were taken up for the formation of activity was not affected by Ca2+, Mg2+ and Zn2+ 0-63,umole of cinnabarinic acid in lhr. ions. 320 P. V. SUBBA RAG AND C. S. VAIDYANATHAN 1966 Table 4. Effect ofmetal ions on the activity ofpurified Spectrophotometric studies. Examination of ab- rat-liver cinnabarinate syntha8e sorption spectra showed that cytochrome c was immediately reduced by 3-hydroxyanthranilic acid. The reaction mixture containing sodium phosphate buffer, pH7-2 (50,tmoles), MnSO4 (0-25,umole) and O-lml. Nagasawa & Gutmann (1959) reported that cyto- of enzyme (0.8,ug. of protein) was preincubated for 10mi. chrome c is reduced by several o-aminophenols. with the indicated metal ion (final concn. 0-5mM) before When a reaction mixture (3.Oml.) of sodium phos- the addition of 3-hydroxyanthranilic acid (4,umoles). The phate buffer, pH7-2 (100,umoles), cytochrome c rest of the conditions were as described in the Materials (1 ,umole) and 3-hydroxyanthranilic acid (0.1 umole) and Methods section. was incubated for 30min. at 370, a yellow product Inhibition was formed which was extractable into ether at Addition pH 2-5. This was identified as cinnabarinic acid by (%) comparing its absorption spectrum in ethanol and AgNO3 95 CdCl2 5 Rp values on paper chromatograms with those of NiSO4 64 an authentic sample (Subba Rao et al. 1965). CaCl2 0 However, under our experimental conditions, the MgSO4 5 cytochrome c reduced by 3-hydroxyanthranilic acid CuSO4 68 was not reoxidized by rat-liver cinnabarinate FeSO4 36 synthase. Further, cytochrome c prepared by FeCl3 27 reduction with sodium borohydride was not reoxi- ZnSO4 9 dized by the enzyme preparation in the absence of Na2MoO4 100 3-hydroxyanthranilic acid or any other reducing CoCl2 18 HgCl2 91 agent. Effect of substrate analogUes. Anthranilic acid, 3-hydroxykynurenine, o-aminophenol and o-, m- and p-hydroxybenzoic acid had no effect on the reaction. Catechol and p-aminobenzoic acid in- Table 5. Effect of metal-binding agents on the activity hibited the reaction by about 20% at a final ofpurified rat-liver cinnabarinate syntha8e concentration of 0-5mM. Lack of activation by The conditions were as described in Table 4. catechol of cinnabarinic acid formation indicates that catechol oxidase (or polyphenol oxidase) is Final concn. Inhibition not responsible for the formation of cinnabarinic Addition (mM) (%) acid from 3-hydroxyanthranilic acid. Sodium azide 0-5 68 Influence of and catalase. 1-0 77 hydrogen peroxide Diethyldithiocarbamate 0-5 59 Hydrogen peroxide (final concn. 0.1%) neither 1-0 64 removed the initial lag phase (Fig. 2) nor activated o-Phenanthroline 1-0 - the reaction. Catalase (1mg.) had no effect on oax'-Bipyridyl 1.0 - enzyme activity. These results suggest that EDTA 0-5 9 hydrogen peroxide is not formed during the 1-0 86 oxidation of 3-hydroxyanthranilic acid. Potassium cyanide 0-5 73 DISCUSSION The results presented in this paper show that Effect of other s8ubstances on the reaction. The the enzyme from rat-liver nuclei which catalyses metal-binding agents EDTA, diethyldithiocar- the oxidation of 3-hydroxyanthranilic acid to bamate, cyanide and sodium azide inhibited the cinnabarinic acid (2-amino-3-oxo-3H-phenoxazine- reaction to a considerable extent (Table 5). 1,9-dicarboxylic acid) can be extensively purified. o-Phenanthroline and aa'-bipyridyl had no in- The optimum pH of 7-2 for rat-liver cinnabarinate hibitory effect. synthase is similar to that reported for liver The reaction was not inhibited by the thiol preparations from poikilothermic animals by inhibitors p-chloromercuribenzoate, N-ethylmalei- Morgan et al. (1965). However, it is different from mide, sodium arsenite and iodoacetic acid at Streptomyces (Katz & Weissbach, 1962) and spinach concentrations varying from 0-1 to 1 mm. (Nair & Vining, 1964) enzymes, which showed The electron acceptors NAD, NADP, FMN, maximal activity at pH5-1 and 9-0 respectively. FAD and menadione had no effect on the enzymic The system apparently showed no requirement formation of cinnabarinic acid, but 0-01mM- for any cofactor other than Mn2+ ions. The cytochrome c inhibited the reaction by 20%. inability ofthe nucleax cinnabarinate synthase from Atabrine (1 mM) had no effect on enzyme activity. rat liver to reoxidize reduced cytochrome c in the Vol. 99 RAT-LIVER CINNABARINATE SYNTHASE 321 C02H CO2H C02H - NH O2H H C02H ~NNH2 w NH 0 N .H 2

OH 0 N N0 0 (I) (II) (III) CO2H XXNH2 OR I C02H HHC02H CO2H CO2H H H XN>N< NH2 X Nz HNH2

0- -OH OH 0 H (V) (IV)

CO2H C02H C02H CO2H Nz:z NH2 S N\> N, NH2

0 0 HI H (VII) (VI) Scheme. 1. Suggested mechanisms for the biosynthesis of cinnabarinic acid from 3-hydroxyanthranilic acid. presence of 3-hydroxyanthranilic acid rules out the It is noteworthy that Ehrensvard, Liljekvist & possibility of a system analogous to the mito- Heath (1960) observed that compounds (IV) and chondrial cytochrome c-cytochrome oxidase system (VII) (Scheme 1) are formed during the oxidation (Nagasawa & Gutmann, 1959; Nagasawa, Gutmann of 3-hydroxyanthranilic acid by human serum, & Morgan, 1959). the process being activated by Cu2+ or Mn2+ The overall reaction catalysed by cinnabarinate ions. They suggested that this enzymic formation synthase involves the conversion of 2 molecules of of cinnabarinic acid (VII), which was cyanide- 3-hydroxyanthranilic acid into 1 molecule of cinna- and azide-sensitive, accounts for 15% of the 3- barinic acid (Subba Rao et al. 1965) with the hydroxyanthranilic acid (I) that is oxidized. They consumption of 3 atoms of oxygen. Such an assume that the rest of the 3-hydroxyanthranilic oxidative dimerization may involve an inter- acid underwent a non-specific oxidation to com- mediary formation of the o-quinoneimine of 3- pound (IV) catalysed by metal-protein complexes. hydroxyanthranilic acid. As suggested by Morgan In our experiments also the formation of cinna- et al. (1965),thehighlyreactive transient o-quinone- barinic acid by rat-liver cinnabarinate synthase is imine (II) can condense in two ways, namely strongly inhibited by cyanide and azide. either with another molecule of o-quinoneimine (II) 3-Hydroxyanthranilic acid and 3-hydroxy- or with a molecule of 3-hydroxyanthranilic acid kynurenine are known to induce neoplasms under (I), as depicted in Scheme 1. Cinnabarinic acid certain conditions. Allen, Boyland, Dukes, Horning (VII) is the final product of both pathways. The & Watson (1957), and more recently Bryan et al. initial lag in the formation of cinnabarinic acid, as (1964), have unequivocally established that these well as the extreme sensitivity of the enzyme to two normal metabolites can induce neoplasms reducing agents, suggests that an intermediate locally when implanted into the bladder of mice. quinoneimine is involved in the reaction. It has been repeatedly emphasized that the binding 11 Bioch. 1966, 99 322 P. V. SUBBA RAO AND C. S. VAIDYANATHAN 1966 of carcinogens to proteins is a prerequisite for their P. S. Sarma for his keen interest. Our thanks are also due carcinogenic activity. Gutmann, Peters & Burtle to Dr G. S. Krishna Rao for his valuable suggestions on the (1956) suggested that a carcinogen can be covalently probable mechanism of the enzyme reaction. bound to protein if the former is converted into a highly reactive intermediate such as a quinoneimine. Support for the formation of such inter- REFERENCES mediates with quinonoid structure was provided Allen, M. J., Boyland, E., Dukes, C. E., Horning, E. S. & by Nagasawa & Gutmann (1959), who demonstrated Watson, J. G. (1957). Brit. J. Cancer, 11, 212. that the mammalian cytochrome c-cytochrome Boyland, E. (1960). Acta Un. int. Cancr. 16, 273. oxidase system can catalyse the oxidation of car- Boyland, E. & Williams, D. C. (1956). Biochem. J. 64, cinogenic o-aminophenols to reactive quinone- 578. imines. However, in contrast with the cytochrome Brown, R. R., Price, J. M., Satter, E. J. & Wear, J. B. c-cytochrome oxidase system and with the liver (1960). Acta Un. int. Cancr. 16, 299. Bryan, G. T., Brown, R. R. & Price, J. M. (1964). Cancer enzyme from poikilothermic animals (Morgan et al. Res. 24, 596. 1965), rat-liver cinnabarinate synthase is highly EhrensvArd, G., Liljekvist, J. & Heath, R. G. (1960). Acta specific for 3-hydroxyanthranilic acid. This may chem. 8cand. 14, 2081. mean that this enzyme is significant in the aetiology Gutmann, H. R., Peters, J. H. & Burtle, J. G. (1956). of tumours since it is involved in the metabolism J. biol. Chem. 222, 373. of a known carcinogen, namely 3-hydroxyanthran- Katz, E. & Weissbach, H. (1962). J. biol. Chem. 237, ilic acid. 882. Schultz (1962) postulated that mammalian Keilin, D. & Hartree, E. F. (1938). Proc. Roy. Soc. B, cancer may be a manifestation of deranged trypto- 124, 397. phan metabolism that may lead to the production Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951). J. biol. Chem. 193, 265. of an endogenous carcinogen via the condensation Morgan, L. R., jun. & Weimorts, D. M. (1964). Biochim. of 3-hydroxykynurenine and o-aminophenol. Since biophys. Acta, 82, 645. it is known that tryptophan metabolites such as Morgan, L. R., jun., Weimorts, D. M. & Aubert, C. C. 3-hydroxyanthranilic acid and 3-hydroxykynure- (1965). Biochim. biophys. Acta, 100, 393. nine are excreted in abnormally high proportions Nagasawa, H. T. & Gutmann, H. R. (1959). J. biol. Chem. by patients bearing bladder tumours (Boyland & 234, 1593. Williams, 1956; Brown et al. 1960), it seems Nagasawa, H. T., Gutmann, H. R. & Morgan, M. A. (1959). reasonable to assume that deranged tryptophan J. biol. Chem. 234, 1600. leadtothemaintenance ofanabnor- Nair, P. M. & Vining, L. C. (1964). Proc. 6th int. Congr. metabolismmay Biochem., New York, IV-1 18. mally high concentration of 3-hydroxyanthranilic Schultz, R. D. (1962). Hypothesis for Chemical Induction acid in the system. This in turn may lead to the and Chemotherapy of Cancer, Parts 1, 2 and 3, Rep. SID formation of cinnabarinic acid by cinnabarinate 62-559, SID 62-1296, SID 62-1297, North American synthase. The finding (Boyland, 1960) that the Aviation Inc., Space and Information Systems Division, phenoxazine dimer of 3-hydroxyanthranilic acid Donwey, Calif. produces significant incidence of bladder tumours Subba Rao, P. V., Jegannathan, N. S. & Vaidyanathan, in experimental animals lends further support to C. S. (1964). Biochem. biophys. Res. Commun. 16, 145. this view. Subba Rao, P. V., Jegannathan, N. S. & Vaidyanathan, C. S. (1965). Biochem. J. 95, 628. The authors are grateful to Professor A. Butenandt for Tokuyama, K. & Knox, W. E. (1964). Biochim. biophys. his generous gift of cinnabarinic acid, and to Professor Acta, 81, 201.