Mchanism of reaction of 3-hydroxyanthranilic acid with molecular oxygen

Biochimica et Biophysica Acta, 1034 (1990) 207-212

207

Elsevier BBAGEN 23304

Mechanism of reaction of 3-hydroxyanthranilic acid with molecular oxygen Michael K. Manthey, Stephen G. Pyne and Roger J.W. Truscott Department of Chemistry, University of gollongong, Wollongong, N.S.W. (Australia)

(Received9 August 1989)

Key words: Tryptophan metabolism; Oxidation; Oxygen; Catalase; Superoxidedismutase; 3-Hydroxyanthranilicacid

The autoxidation of the tryptophan metabolite, 3-hydroxyanthranilic acid, at pH 7 gives rise to a p-quinone dimer and cinnabarinic acid. A novel dimer formed by radical-radical coupling of 3-hydroxyanthranilic acid is also produced. Labelling studies have shown that the C-2 oxygen in the p-quinone dimer is derived from molecular oxygen. A product versus time study of this reaction has revealed that, in the absence of catalase, cinnabarinic acid is formed but undergoes decomposition by hydrogen peroxide. At pH 7, in the presence of catalase, both the p-quinone dimer and cinnabarinic acid are formed at approximately the same rate and this rate of formation increases with increasing pH. Inclusion of superoxide dismutase was found to increase the rate of formation of cinnaharinic acid, suggesting that superoxide ions may also cause decomposition of cinnaharinic acid. This was confirmed by treating cinnaharinic acid with superoxide. A mechanism involving a common anthranilyi radical intermediate is proposed to account for the formation of the different oxidation products.

Introduction The pathway of tryptophan catabolism is one of the most complex of all the amino acids. Recently, attention has focussed on one of the metabolites, 3-hydroxyanthranilic acid, which has been found to inhibit leucine-stimulated insulin release from rat pancreatic islets [1]. The benzene ring of 3-hydroxyanthranilic acid is normally cleaved by the enzyme 3-hydroxyanthranate oxygenase (EC 1.13.1.6) in a process which requires Fe 2+ ions, molecular oxygen and free sulfhydryl groups [2]. Paradoxically, oxygen also rapidly destroys enzyme activity [2]. If the activity of 3-hydroxyanthranate oxygenase is compromised, the cellular levels of 3-hydroxyanthranilic acid may increase. It is not known whether such a process underlies the increased excretion of 3-hydroxyanthranilic acid by human macrophages in response to treatment with interferon 3' [3]. cad Conceivably the decrease in the intracellular concentration of reduced glutathione which occurs with age [4] could impair the activity of 3-hydroxyanthranate oxygenase and lead to a build up in 3-hydroxy-

anthranilic acid levels. Under these conditions, 3-hydroxyanthranilic acid may undergo autoxidation with the production of H 2 0 2 [5], superoxide radicals [5] and, in the presence of trace amounts of iron, hydroxyl radicals [6]. Oxidised 3-hydroxyanthranilic acid intermediates are also very reactive and are known to bind to proteins [7,8]. We are exploring the possibility that such processes may contribute to the formation of lipofuscin-like pigments with age. There is preliminary evidence in support of such a proposal [9]. We recently reported the formation of the p-quinone dimer 2 upon autoxidation of 3-hydroxyanthranilic acid at pH 11.7 [10]. When the autoxidation of 3-hydroxyanthranilic acid was conducted in the presence of an excess of an aliphatic and aromatic amine or amino acid [11], then p-quinone derivatives could be isolated with good yield. Since both cinnabarinic acid 1 and the p-quinone dimer 2 are produced from the autoxidation of 3-hydroxyanthranilic acid at neutral pH, we were interested to determine the factors which govern the course of oxidation.

Materials and Methods Abbreviations: DMF, dimethylformamide; DMSO, dimethylsulfoxide. Correspondence: R.J.W. Truscott, Department of Chemistry, University of Wollongong, P.O. Box 1144, Wollongong, N.S.W. 2500, Australia.

Ultraviolet visible spectra were recorded on a Shimadzu UV-160 spectrophotometer. N M R spectra were recorded on a JEOL FX 90 N M R spectrophotometer. Microanalyses were performed by the Microana-

0304-4165/90/$03.50 © 1990 ElsevierSciencePublishers B.V, (BiomedicalDivision)

208 lytical Division, Australian National University, Canberra. Melting points are uncorrected. Solvents were of analytical grade. Column chromatography was performed using silica gel (0.063-0.2 mm (Merck)). Catalase, superoxide dismutase and xanthine oxidase were obtained as suspensions from Sigma Chemical Company. Enzymatic assays were carried out as described by the manufacturer with the following results: catalase, 113 000 units/ml; xanthine oxidase, 14 000 units/ml; superoxide dismutase, 320 units/ml. 1802 was obtained from Novachem Pty. Ltd., Australia in 97.5% purity.

3-hydroxyanthranilic acid. The solid was washed with acetone and recrystallized from DMSO to give 3a as a yellow powder. Yield, 8 mg (1.5%); m.p., 236-238°C; molecular weight = 314 g / m o l (FAB-MS). Analysis for C16HlnN205. Calcd.: C, 61.14; H, 4.49; N, 8.91. Found: C, 60.88; H, 4.72; N, 8.66. ~kmax ( D M S O / C H 3 O H (2:3)) 255.0 nm (log e 4.27), 328.5 nm (log e 3.82), 378.5 nm (log e 4.09). I H - N M R (d6-DMSO) 3.84 (3H,s), 3.93 (3H,s), 6.73 (2H,brs), 7.04 (2H, brs), 7.296 (1H, d J = 8.8 Hz), 7.328 (1H, d J = 8 . 6 Hz), 7.432 (1H, d J = 8 . 6 Hz), 7.635 (1H, d J = 8.8 Hz). 13C-NMR (d6-DMSO) 51.66, 56.00, 101.41, 107.98, 108.00, 109,25, 115.41, 121.81, 125.41, 125.97, 137.28, 140.08, 142.67, 146.52, 161.54, 167.35.

Cinnabarinic acid l To a solution of 4-hydroxyanthranilic acid (700 mg) in ethanol (400 ml) at 40 ° C, was added over 1 h freshly prepared p-benzoquinone (800 mg) in ethanol (70 ml). After stirring for 2 h at 40 ° C, then at room temperature overnight, the resulting orange solid was collected by filtration, washed with ethanol and recrystallized from pyridine to give 1 m.p. > 300°C ~max (0.2 M sodium phosphate, pH 3.0/DMF(20%)) 253.5 nm (log e 4.28).

180 labelling 1802 labelling was carried out using 3-hydroxyanthranilic acid (10 mg) in 4 ml of 0.2 M Na3PO 4 buffer adjusted to 11.7. After 5 h of oxidation, the red solution was acidified to pH 3.0 with HC1, and extracted with ethyl acetate (10 ml). FAB-MS with pure unlabelled -2 as reference showed an incorporation of 180 which was found to average > 90%.

Synthesis of 6-amino-3-[(2-carboxy-6-hydroxyphenyl)amino]-2, 5-dioxo-1, 3-cyclohexadiene-1 -carboxylic acid 2__ This chemical was synthesized as previously described [10]. Amax (0.2 M sodium phosphate, pH 3.0/DMF(20%)), 333.0 nm (log e 4.11); 517 nm (log 3.21).

Product versus time study For each set of experiments, 3-hydroxyanthranilic acid solutions were made up in 0.2 M sodium phosphate buffer (approx. 1 m g / m l ) and divided into 40 ml aliquots. Enzymes were added as described under Figs. 2-6. For reactions involving the use of xanthine, stock solutions of 3-hydroxyanthranilic acid were stirred under N 2 with an excess of xanthine for 1 h. The excess xanthine was removed by filtration and the 3-hydroxyanthranilic acid/xanthine solutions used as described. The 40 ml aliquots were oxygenated and at pre-determined intervals 1 ml aliquots were taken, D M F (0.2 ml) and 1 M HC1 (40 ffl) were added and 1 and 2 were separated by chromatography on LH-Sephadex (18 cm × 1.2 cm) using 0.2 M phosphate buffer, pH 3 . 0 / D M F (20%) as the eluent. 1 eluted first followed by 3-hydroxyanthranilic acid and -2. Fractions were collected

Synthesis of 4, 7-diamino-8-methoxy-6H-dibenzo[b,d]pyran-6-one-3-carboxylic acid methylester 3_a_a 3-Hydroxyanthranilic acid (250 mg) in 0.1 M sodium phosphate buffer pH 7.0 (5 m g / m l ) was saturated with 02 for 3.5 days, after which time the pH was adjusted to 6.0 with 1 M HCI and extracted with ethyl acetate (4 × 100 ml). The organic fractions were combined, dried with NaaSO 4, filtered and the solvent was removed to give a light red solid which was methylated with diazomethane. Column chromatography of the solid after removal of the solvents (ethyl acetate/hexane (1 : 1), ethyl acetate and ethyl a c e t a t e / C H 3 O H (20%) as the eluents] gave 3__a_awhich eluted after fully methylated

OR

COOH COOH

OH

COOH

N

~

o ~ O

O ~

I

NH2

NH2 COOH

COOR

~_ R=H 32 R = CH3 Structures of l, 2 and _3.

209 into volumetric flasks, made up to volume with the eluent and the absorption at Xmaxwas determined (454.5 nm 1; 517 nm, 2). Results are expressed in terms of percentage conversion of 3-hydroxyanthranilic acid to ! or 2.

Decomposition of cinnabarinic acid by superoxide

30

t-

.9

20

0

Cinnabarinic acid (4.08 mg) was dissolved in 0.2 M phosphate buffer (pH 7.0, 250 ml) and the solution was saturated with xanthine as previously described. The solution was divided into 25 ml aliquots and treated as described in the text (Fig. 6). The decrease in the characteristic absorbance of 1 was recorded and the results were expressed in terms of the percentage of 1_ remaining.

0

(D

C

• o

"10

(3-

I

Results

J

I

2000

=

I

3000

Time(minutes)

180 labelling experiment To confirm our previous suggestion [10] that the C-2 oxygen in 2 was derived from molecular oxygen, a labelling experiment was undertaken using 1801. Autoxidation of 3-hydroxyanthranilic acid at pH 11.7 in a 97.5% 1802 atmosphere, followed by isolation of 2 resuited in 90% 180 incorporation by mass spectra analysis.

Product versus time study The progress of the autoxidation of 3-hydroxyanthranilic acid at pH 7 was monitored at specific time intervals by removing aliquots of the reaction mixture and separating 3-hydroxyanthranilic acid, ! and 2 by chromatography on LH-Sephadex. The amounts of 1 and 2, determined by quantitative ultraviolet analysis, were expressed as percentage conversion based on the initial amount of 3-hydroxyanthranilic acid. The results of this study are shown on Fig. 1. Initially, ! is ap-

Fig. 2. Percentage conversion of 3-hydroxyanthranilic acid into 1 and 2 at pH 7.0 in the presence of excess catalase. Catalase (5600 units/40 rnl reaction volume) was added every 2 h. [3, 1; A, 2.

parently the sole autoxidation product as determined by TLC. However, after a lag phase of approx. 0.5 h, a significant amount of 2 can be detected. After approx. 8 h, the amount of 1 slowly declines, while the formation of 2 shows a typical autoxidation product versus time profile [12] (Fig. 1). We suspected that 1 was being decomposed, to as yet unidentified compounds, by reaction with the hydrogen peroxide produced during the course of the autoxidation [5]. This proposal was readily tested by performing the autoxidation of 3-hydroxyanthranilic acid at pH 7 in the presence of excess catalase to scavenge hydrogen peroxide (Fig. 2). Under these conditions, both 1 and 2 were formed at essentially the same rate.

Dependence of the autoxidation of 3-hydroxyanthranilic acid on pH

30

The autoxidation of 3-hydroxyanthranilic acid was carried out in 0.1 M phosphate buffer at pH 6.0, 7.0, 8.0 and 11.7. In each case (except pH 11.7), catalase was added at regular intervals to quench H202 produced. The product versus time profile for the formation of 1 and 2 is presented in Fig. 3a, and b, respectively. The rate of formation of both 1 and 2 from autoxidation of 3-hydroxyanthranilic acid in the presence of catalase was enhanced at higher pH values, (Figs. 3, a and b). At pH 6, the formation of 2 was barely detectable.

2O > t-0

0

g

c

P

I

1000

10

,

OI

0

I

1000

,

I

2000

,

I

3000

Time (minutes)

Fig. 1. Percentage conversion of 3-hydroxyanthranilic acid into cinnabarinic acid (1) and the p-quinone dimer (2) at p H 7.0. [3, 1_; I , _2.

Effect of superoxide It has been reported that the autoxidation of 3-hydroxyanthranilic acid leads to the production of superoxide ions in addition to H202 [5]. In order to test whether superoxide decomposes 1, the autoxidation of

210 20

20

a

¢.0

tO 0

10

(D')

"E

og

lo

g-

s

O)

2

(1.) 13...

0

,

200

0

400 Time

600

I

800

(minutes) 0

6O

"

'

100

b

'

'

200

300

'

400

'

500

I

600

700

Time (minutes) 50

Fig. 4. Percentage conversion of 3-hydroxyanthranilic acid into 1 and 2 in the presence of catalase and superoxide dismutase. 1.27 mg/ml 3-hydroxyanthranilic acid in 0.2 M phosphate buffer (pH 7.0) with 2 hourly additions of catalase (5600 units/40ml) and superoxide dismutase (4.8 units/40 ml). zx, 1 (without superoxide dismutase additions); *, 1_;D, _2(without superoxide dismutase additions); B, 2.

>~ 4o 8 o

30

g-

lo

o 0

200

400

600

800

Time (minutes)

Fig. 3. pH dependence of the conversion of 3-hydroxyanthranilic acid into (a) 1 and (b) 2. 3.11 mg/ml 3-hydroxyanthranilic acid in 0.2 M phosphate buffer (50 ml) with hourly additions of catalase (as for Fig. 2). (a) 13, pH 6.0; A, pH 7.0; I, pH 8.0. (b) zx, pH 6.0; rn, pH 7.0; *, pH 8.0; II, pH 11.7.

3-hydroxyanthranilic acid was repeated in the presence of superoxide dismutase as well as catalase. The inclusion of superoxide dismutase markedly increased the net rate of formation of 1 without affecting the overall rate of synthesis of 2 (Fig. 4). A similar effect of superoxide dismutase on cinnabarinic acid formation has been described by Dykens et al. [13], Additional evidence in support of the proposal that superoxide ions decompose cinnabarinic acid was provided by studies in which x a n t h i n e / x a n t h i n e oxidase was included in the autoxidation mixture containing both 3-hydroxyanthranilic acid and catalase. U n d e r these conditions, the rate of formation of 2 remained essentially unaffected whereas the formation of 1 was significantly reduced (Fig. 5). Since other factors could conceivably play a role in the observed alteration in the rate of cinnabarinic acid formation from 3-hydroxyanthranilic acid, cinnabarinic acid was synthesized and then treated with xanthine and xanthine oxidase. A rapid decomposition was observed (Fig. 6) which was blocked by the inclusion of

superoxide dismutase. Catalase was added together with the superoxide dismutase in this experiment to eliminate the H 2 0 2 decomposition of 1, since it is k n o w n that H 2 0 2 is formed b y disproportionation of 02-" [14]. A similar decomposition of cinnabarinic acid was observed following addition of potassium superoxide and catalase (data not shown). Addition of catalase alone caused only a small decrease in the rate of cinnabarinic acid b r e a k d o w n (Fig. 6). Inclusion of mannitol had no

t,,..

o

> C 0

o

c~

[]

8

0 0

100

200

300

400

500

Time (minutes) Fig. 5. Percentage conversion of 3-hydroxyanthranilic acid into 1 and _2 in the presence of catalase and xanthine/xanthine oxidase. 1.06 mg/ml 3-hydroxyanthranilic acid in 0.2 M phosphate buffer, (pH 7.0, saturated with xanthine) with hourly additions of catalase (11200 units/40 ml) and xanthine oxidase (630 units/40ml). [3, 1 (without xanthine oxidase additions); II, 1_; zx, 2 (without xanthine oxidase additions); A, _2.

211 110

r-

"2

100,

E

~

-&

8o 70

,

0

I 2

,

I 4

,

I 6

,

=,

"&

I

I

I

8

10

12

Time (minutes) Fig. 6. Effect of catalase on the decomposition of 1 by superoxide. 1.63-10 -2 mg/ml 1 in 0.2 M phosphate buffer (pH 7.0, 25 ml) saturated with xanthine. After addition of catalase, xanthine oxidase (350 units) was added and the decrease in hm~x recorded, n, Without catalase addition; - , with catalase addition. (2800, 5600 or 11200 units). A reading of 100% was obtained if superoxide dismutase (128 units) was added before xanthine oxidase.

effect on the rate of decomposition of 1_suggesting that hydroxyl radicals are not involved in this reaction.

Other 3-hydroxyanthranilic acid oxidation products After close examination of the reaction mixture obtained by exposure of 3-hydroxyanthranilic acid to molecular oxygen, a third but minor autoxidation product, the lactone _3, was isolated in low yield (1-2%). _3 was characterized as the dimethyl derivative (3a) by treatment with diazomethane, since this prevented aerial oxidation of _3 and gave an easily recrystallizable product. The structure of methylated _3 was consistent with the ]H- and ]3C-NMR spectral data. Discussion

3-Hydroxyanthranilic acid, a normal metabolite of the amino acid tryptophan, is readily autoxidised. In a number of diseases, elevated levels of urinary tryptophan metabolites have been reported [15,16]. Patients with cancer of the bladder, for example, have been found to excrete increased amounts of 3-hydroxyanthranilic acid and 3-hydroxykynurenine [16,17]. The possible involvement of these tryptophan metabolites in the etiology of this disease is supported by the demonstration that 3-hydroxyanthranilic acid and cinnabarinic acid are carcinogens linked with the induction of both breast and bladder carcinomas [18,20]. It is known that bladder cancer in experimental animals can be induced by implantation of 3-hydroxyanthranilic acid pellets [17,19]. Evidence that oxidation may be associated with the carcinogenicity of 3-hydroxyanthranilic acid has come from studies which have shown a marked protective effect of simultaneously administered vitamin C [21].

3-Hydroxyanthranilic acid is normally metabolised in the body to 2-amino-3-carboxymuconic-6-semialdehyde in a reaction catalysed by the enzyme 3-hydroxyanthranilate oxygenase (EC 1.13.1.6). This enzyme has a requirement for 02 , Fe 2+ ions and free SH groups. However, oxygen also destroys enzyme activity [2]. If the activity of 3-hydroxyanthranate oxygenase becomes limiting, for example as a result of a change from a reducing to an oxidising environment, then 3-hydroxyanthranilic acid may be autoxidised with the possible production of superoxide [5], hydroxide radicals [6], H202 [5] and dimeric 3-hydroxyanthranilic acid oxidation products [10]. In this paper, we have demonstrated that the two major oxidation products of 3-hydroxyanthranilic acid 1 and 2 are formed at similar rates at physiological pH. However, cinnabarinic acid is readily decomposed by reaction with either H202 or superoxide ions, both of which are produced during the course of the autoxidation. The red p-quinone dimer is resistant to further oxidation by these reagents. This is the first report of cinnabarinic acid degradation by superoxide, although several authors have described the H202-induced decomposition of cinnabarinic acid [5,13]. We have isolated the yellow H202-oxidation product of cinnabarinic acid and will describe the structural elucidation of this compound in a future publication. This compound, along with several others, is also formed by exposure of cinnabarinic acid to superoxide. The relative rates of formation of the p-quinone dimer 2 and cinnabarinic acid 1 were found to be markedly pH dependent (Fig. 3). Under our initially reported conditions of oxidation used for the preparation of 2 [10] no 1 could be detected. We suggest that 1 is probably formed during autoxidation of 1 at pH 11.7, but is rapidly decomposed by hydrogen peroxide [5] and superoxide. The stability of O~- is known to be enhanced at higher pH values [14]. Other workers [6,13,7] who have observed that inclusion of superoxide dismutase in solutions containing 3-hydroxyanthranilic acid increases the rate of formation of cinnabarinic acid, have suggested that this occurs by alteration of a proposedly reversible electron transfer reaction: 3-hydroxyanthranilic acid + 02 ~ 3-OHanthranilyl radical + O~

Our results suggests than an alternative explanation for the phenomenon is that superoxide dismutase is removing 0 2 ions which would otherwise decompose cinnabarinic acid. In experiments where the concentration of 0 2 was increased by the inclusion of xanthine/ xanthine oxidase, the rate of formation of 2 was unaffected (Fig. 5). Furthermore, when O ; levels were decreased by addition of superoxide dismutase, the rate of formation of 2 also remained unchanged (Fig. 4).

212

COOH

COO pH 7.0

STEP I

02-.

COO

COO-

COOH

i n d u c t i o n of b l a d d e r cancer b y 3 - h y d r o x y a n t h r a n i l i c acid. T h e a u t o x i d a t i o n of 3 - h y d r o x y a n t h r a n i l i c a c i d is therefore complex. E x p e r i m e n t s are u n d e r w a y in our l a b o r a t o r y to c h a r a c t e r i s e the a d d u c t s f o r m e d b y reaction of p r o t e i n s with oxidised solutions of 3-hydroxya n t h r a n i l i c acid, since this will h o p e f u l l y reveal the n a t u r e of s o m e of the reactive species involved in the modification of biomacromolecules by 3-hydroxya n t h r a n i l i c acid. References

(B)

(A)

-H20

1l02 COO-

COO "O-OH'~H2

Ac•HA , 02

02

1_

2_

3OHA

COOH-O~"~-~__~ NH2

O2 Scheme I. These results suggest t h a t Step 1 (Scheme I), which leads to the f o r m a t i o n of the 3 - h y d r o x y a n t h r a n i l y l radical, m a y b e irreversible. It is of interest to n o t e that c i n n a b a r i n i c a c i d was also d e c o m p o s e d following the a d d i t i o n of a c e t a l d e h y d e d u r i n g p r e l i m i n a r y e x p e r i m e n t s involving the generation of s u p e r o x i d e radicals using x a n t h i n e o x i d a s e / acetaldehyde. T h e i d e n t i f i c a t i o n of __3, a p r o d u c t f o r m e d b y orthopara coupling o f 3 - h y d r o x y a n t h r a n i l y l radicals, indicates that these r a d i c a l species, together with s u p e r o x i d e [5] a n d h y d r o x y [6] radicals, could also p l a y a role in the

1 Rogers, K.S. and Evangelista, S.J. (1985) Proc. Soc. Exp. Biol. Med. 178, 275-278. 2 Greenberg, D. (1969) in Methods in Enzymology (Greenberg, D., ed.), Vol. 111, p. 163, Academic Press, New York. 3 Werner, E.R., Fuchs, D., Hausen, A., Lutz, H., Reibnegger, G. and Wachter, H. (1985) Biochem. Cfin. Aspects Pteridines 4, 473-484. 4 Harding, J.J. (1970) Biochem. J. 117, 957-960. 50gawa, H., Nagamura, Y. and Ishigura, I. (1983) Hoppe-Seyler's Z. Physiol. Chem. 364, 1507-1518. 6 Iwahashi, H., Ishii, T., Sugata, R. and Kido, R. (1988) Biochem. J. 251, 893-899. 7 King, C.M. and Kriek, E. (1965) Biochim. Biophys. Acta 111, 147-153. 8 Truscott, R.J.W. and Martin, F. (1989) Exp. Eye Res. 49, 927-935. 9 Hegeduz, Z.L., Frank, H.A., Altschule, M.D. and Nayak, U. (1986) Arch. Int. Physiol. Biochem. 94, 339-348. 10 Manthey, M.K., Pyne, S.G. and Truscott, R.J.W. (1988) J. Org. Chem. 53, 1486-1489. 11 Manthey, M.K., Pyne, S.G. and Truscott, R.J.W. (1989) Tetrahedron 45, 2803-2810. 12 Walling, C. (1957) Free Radicals in Solution, pp. 397-466, Wiley & Sons, New York. 13 Dykens, J.A., Sullivan, S.G. and Stern, A. (1986) in Superoxide and Superoxide Dismutase in Chemistry, Biology and Medicine (G. Rotilia, ed.), pp. 92-94, Elsevier, Amsterdam. 14 Lee-Ruff, E. (1977) Chem. Rev. 6, 195-214. 15 Costa, C., Angi, M.R., De Cavli, M., Vanzan, S. and Allegri, G. (1982) Bull. Soc. Ital. Biol. Sper. 58, 333-339. 16 Boyland, E. and Williams, D.C. (1956) Biochem. J. 64, 578-582. 17 Abul-Fadl, M.A.M. and Khalafallah, A.S. (1961) Br. J. Cancer 15, 479-484. 18 Bryan, G.T., Brown, R.R. and Price, J.M. (1964) Cancer Res. 24, 596-602. 19 Bryan, G.T., Brown, R.R. and Price, J.M. (1964) Cancer Res. 24, 582-590. 20 Allen, M.J., Boyland, E., Dukes, C.E., Horning, E.S. and Watson, J.G. (1957) Br. J. Cancer 11,212-228. 21 Schlegel, J.U., Pipkin, G.E., Nishimura, R. and Shultz, G.N. (1970) J. Urol. 103, 155-159.