[CANCER RESEARCH 39, 3070-3073, August 1979] 0008-5472/79/0039-0000502.00 Synthesis of the Glucuronic Acid Conjugate of Methylazoxymethanol'

Hiromu Matsumoto,2 Ronald H. Takata,3 and David V. Komeiji

Department of Agricultural Biochemistry, University of Hawaii, Honolulu, Hawaii 96822

ABSTRACT enters the duodenum. The glucuronide could migrate to the lower intestinal tract where it is hydrolyzed by bacterial @8- The glucuronic acid conjugate of methylazoxymethanol was glucuronidase to release free MAM (17). Such a scheme could synthesized by oxidizing the primary of the explain the formation of both duodenal and colonic tumors moiety of cycasin (methylazoxymethanol-fl-D-glycopyranoside) induced by MAM administered parentenally or by DMH admin to a . The oxidation was carried out by bubbling isteredviaanyroute. oxygen gas through a cycasin solution in the presence of a There is no evidence thus far to indicate that MAM-GIcUA is platinum-on-carbon catalyst. formed in the liver. If MAM-GIcUA were available, it could be A band at 1 71 5 cm1, not present in the cycasin infrared used to develop sensitive analytical procedures for the quan spectrum, appeared in the spectrum of the oxidized cycasin titative determination of the compound in the bile and , product, establishing the presence of a canboxylic acid group. and these procedures could be used to unequivocally demon The oxidation product is methylazoxymethanol-fJ-D-glucosidu strate the presence or absence of MAM-GIcUA in the bile or ronic acid because, when hydrolyzed with @8- urine of animals given injections of DMH or MAM. The synthesis glucuronidase, it produced methylazoxymethanol and glucu of MAM-GIcUA, its physical and biochemical properties, and nonic acid and also indicated retention of the fl-linkage of its acute toxicity and mutagenicity are described in this paper. cycasin. Varying quantities of the synthesized methylazoxymethanol glucosiduronic acid, injected into Wistanrats of both sexes and MATERIALS AND METHODS of varying weights, were not acutely toxic. The compound was Cycasin. Cycasin was isolated from Cycas circinalis and . mutagenic to Salmonella typhimurium when preincubated with purified by the method of Nishida et al. (16). E.coli,6-glucuronidase,butnotwhenpreincubatedwithbovine MAM-GIcUA. MAM-GIcUA was prepared by catalyticoxida liven glucuronidase. tion of cycasin. The procedure is described in “Results.― Enzymes. fl-Glucuronidase (/1-D-glucuronideglucunonohy INTRODUCTION drolase; EC 3.2.1 .31) from Escherichia coli and bovine liven were purchased from Sigma Chemical Co., St. Louis, Mo. The MAM,4 the aglycone of cycasin (methylazoxymethanol-f1-D- prepared enzyme solutions were filtered through a Swinnex-25 glycopyranoside), has been shown to be toxic and cancino bacterial filter unit (Millipone Corp., Bedford, Mass.) before use. genic (6, 14). Carcinomas of the duodenum were found in The activity of $-glucuronidase was determined by the method significant numbers, in addition to tumors of the liver, kidney, of Fishman (4) with phenolphthalein glucunonide as substrate; and colon, in rats which had received multiple i.p. injections of MAM was measured by liquid chromatography when MAM MAM. The predominant localization of the tumors in the middle GIcUA was used as substrate. third of the duodenum and the extensive changes in the intna Chromatography. Thin-layer chromatography was carried hepatic biliary system suggested that MAM on its metabolite out on silica gel using a developing solvent mixture of n-butyl may have been excreted with the bile (8). The observation of alcohol:acetic acid:waten (2: 1:1, v/v/v). The plate was sprayed tumors in the gallbladder of hamsters given i.p. injections of with sulfuric acid: (1:1) and charred at 110°for visual MAM (9) also suggests that MAM on its conjugate may be ization. A Hewlett-Packard Model 1084A high-pressure liquid excreted with the bile. The symmetrical DMH has been shown chromatograph with a 3.2- x 250-mm Vydak C18column (Altex, by Fiala (3) to be transformed in the liven by a series of Berkeley, Calif.), using a 1% methanol:waten eluent at a flow oxidations through azomethane and azoxymethane to the prox rate of 1.0 mI/mm, was utilized to analyze MAM and MAM imate carcinogen MAM, as postulated by Druckney (2). lt has GIcUA. The eluate was monitored at 215 nm. A Bendix 2500 been suggested that MAM might be conjugated with glucuronic gas chromatograph with a flame ionization detector and a I 80- acid in the livenand that MAM-GIcUA is excreted with the bile. cm SE-30 column (inside diameter, 0.2 cm; 1.5% on 60 to 80 Free MAM might affect the bile duct, the gallbladder, or the mesh Chromosorb W) at 215°was used to obtain gas chro upper intestinal tract, especially where the concentrated bile matograms. Peak areas were determined by an on-line Autolab

1 This investigation was supported in part by Grant CA 1 8631 , awarded by Vidar 6300 digital integrator. The MAM-GIcUA, cycasin, and the National Cancer Institute, Department of Health, Education, and Welfare. glucuronic acid were treated with an excess of Tni-Sil tn Published as Journal Paper 2330, Hawaii Agricultural Experiment Station. methylsilylation reagent (Pierce Chemical Co., Rockford, II.), 2 To whom requests for reprints should be addressed, at Department of Agricultural Biochemistry, University of Hawaii, 1800 East-West Road, Honolulu, and after 1 hr each mixture was injected into the chromato Hawaii 96822. graph. 3 Present address: Pathology Associates, Medical Laboratories, 4400 Kalan ianaole Highway, Honolulu, Hawaii 96821. Spectrometry. Mass spectra were obtained with a Finnigan 4 The abbreviations used are: MAM, methylazoxymethanol; DMH, 1 ,2-dimeth Series 3000 mass spectrometer coupled to a VanianAerograph ylhydrazine; MAM-GIcUA, methylazoxymethyl-$-D-glucopyranosiduronic acid Series 1400 gas chromatograph with a 180-cm SE-30 column (methylazoxymethanol-glucosiduronic acid, methylazoxymethanol glucuronide); MAM-OAc, methylazoxymethyl acetate. (inside diameter, 0.2 cm; 2% on 60 to 80 mesh Chromosorb Received November 6, 1978; accepted May 8, 1979. W). IR spectra of cycasin and its oxidation product, MAM

3070 CANCERRESEARCHVOL. 39

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1979 American Association for Cancer Research. Synthesis of MAM-GIcUA

GIcUA, were obtained with a Beckman lR-8 IR spectrometer. ethanol:water (3:1) gave fine white needles which were used A Vanian XL 100 NMR spectrometer was used to determine for the characterization of the compound. The product was the protein magnetic resonance spectra of MAM-GIcUA methyl homogeneous by thin-layer chromatography (RF 0.46 versus ester in CDCI3and CDCI3plus D2O. RF 0.52 for cycasin). High-pressure liquid chromatography Acute Toxicity Tests. MAM-GIcUAwas injectedi.p. intorats showed a main peak (99 + %) at 1.1 mm, followed by a small to test its acute toxicity. The compound was neutralized with cycasin peak at 2.2 mm. The UV absorption maximum was an equivalent amount of sodium bicarbonate, made .up to determined to be 215 nm by varying the wavelength of the volume, and in appropriate volumes injected as the sodium detector ovenseveral injections. Gas chnomatographic analysis salt. Male Wistar rats weighing 100 ±5(S.D.) g each in groups of the tnimethylsilylated derivative of the oxidation product gave of 4 were given i.p. injections with 4 dosage levels, 125, 250, a single peak at 8.3 mm, distinct from the 7.0-mm retention of 500, and 1000 mg MAM-GIcUA per kg body weight. The the tnimethylsilylated derivative of cycasin. number of deaths within 48 hr in each group was observed. Physical Characterization of MAM-GIcUA. Melting point: Larger male rats weighing 200, 300, and 400 ±10 g each 109-111°. were given injections with 268 mg (1 mmol) MAM-GIcUA pen kg body weight. Three female rats weighing 250 g each were C8H14N2O8.H2O given 300 mg MAM-GIcUA pen kg body weight. Calculated:C33.80, H 5.63, N 9.86, 0 50.70 Mutagenicity Test. MAM-GIcUA was tested for mutagenic Found: C 33.99, H 5.63, N 9.18, 0 50.57 activity with Salmonella typhimurium by a modification of the IR (KBr) spectrum of the cycasin oxidation product was Ames test (1). The modification has been described pnevi significantly different from that of cycasin. The oxidation prod ously.5 The S. typhimurium strain used was His G46, a histidine uct displayed the following prominent bands: 3480 (OH), 2920, auxotroph which detects base pain change mutation. The S. 1715 (COOH), 1540 (azoxyl), 1420, 1340, 1290, 1160, 1080, typhimurium culture was provided by Professor Bruce N. Ames and 950 cm'. The 1715 cm1 band established the presence (University of California, Berkeley, Calif.). MAM-GIcUA was of a canboxylic acid group, while the 1540 cm1 absorption preincubated at 37°with 2000 units each of E. coli and bovine indicated that the MAM moiety remained intact. The IR spectra livenf1-glucuronidase. of cycasin and its oxidation product, MAM-GIcUA, are shown in Chart 1. RESULTS The methyl ester was obtained by treatment with excess ethereal diazomethane. The purified product gave the following Synthesis of MAM-GIcUA. Cycasin was oxidized to MAM physical data. The IR spectrum displayed a band at 1740 cm ‘, GIcUA with gaseous oxygen in the presence of a platinum-on characteristic of a methyl ester. The proton magnetic neso carbon catalyst prepared by the method of Mehltnetter et aI. nance spectra in CDCI3and CDCI3plus D2Oare listed in Table (15). Cycasin (10.1 g, 40 mmol), 7.5 gof the catalyst, and 250 1. The signals at 84 and 85 ppm are attributable to the MAM ml of water were placed in a 500-mI 3-necked flask. A gas moiety and agree with the observations on cycasin (7) and bubbler and a pH electrode were placed in the side necks, and MAM-acetate (13). The 3 exchangeable protons indicate the a water condenser was placed in the middle one. The flask was presence of 3 hydroxyl groups, the correct number for a placed in a water bath, and the mixture was magnetically glucosidunonic acid ester. The methylated cycasin oxidation stirred. The mixture was neutralized with sodium bicarbonate, product was tnimethylsilylated and analyzed by gas chromatog and oxygen was passed through it at a rate of approximately raphy. A peak with retention time of 6.4 mm was obtained. 85 liters/hr. The temperature of the water bath was raised and Mass spectrometry on this peak gave high-mass ions at m/e maintained at 55°.Isobutyl alcohol was added intermittently to 481 (M—CH3),449,423, and 409. The mass spectral evidence reduce foaming and overoxidation (18). Sodium bicarbonate is compatible with a tnimethylsilylated methyl ester, wherein the was added, as the oxidation proceeded, to keep the pH in a m/e 409 fragment is generated by the loss of a methyl group 6.8 to 7.7 range. At the end of 4 hn,the mixture was cooled to and a (CH3)2Si=CH2 molecule from the molecular ion. room temperature and filtered. The catalyst was washed thor Biochemical Properties. The cycasin oxidation product was oughly with hot water and finally with 50% ethanol. The liquid hydrolyzed with $-glucuronidase (E. coil), and a portion of the was passed through a Dowex 50 (H@)column to convert the hydnolysate was analyzed by high-pressure liquid chromatog compound to the free acid. The eluate was concentrated under raphy. A peak with the same retention time as authentic MAM reduced pressure in a rotary evaporator to a syrupy consist was observed. The remaining hydrolysate was lyophilized, and ency; then 3 volumes of ethanol were added, and the product the solid residue was tnimethylsilylated and analyzed by gas was crystallized in a refrigerator. The crystals were filtered, chromatography. A peak with the same retention time as glu and a second crop was obtained from the filtrate. The crude cunonic acid was observed. The results indicate that the agly yield of MAM-GIcUA for 2 oxidations averaged 6.8 g or 60% cone is MAM, which is attached to glucuronic acid. The original per oxidation. The combined solid was dissolved in a small fl-glycosidic linkage of cycasin is retained. The compound is volume of water and recrystallized with the addition of 3 vol MAM-fl-D-glucosidunonic acid. umes of ethanol. The process was repeated until the compound Acute Toxicity of MAM-GIcUA. The acute toxicitytest of i.p. showed no cycasin on thin-layer chromatography. injected MAM-GIcUA was set up to determine the 50% lethal Purity of the Product. Repeated recnystallizations from dosage, but the compound was not acutely toxic to rats weigh ing about 100 g, even when the administered dosage was as 5 T. Matsushima, H. Matsumoto, A. Shirai, M. Sawamura, and I. Sugimura. high as 1000 mg/kg body weight. The quantities injected in Mutagenicity of the naturally occurring carcinogen cycasin and synthetic meth ylazoxymethanol conjugates on Salmonella typhimurium, submitted for publica this series were 125, 250, 500, and 1000 mg/kg body weight, tion to Cancer Research. which would have been equivalent to 42, 84, 168, and 336 mg

AUGUST1979 3071

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1979 American Association for Cancer Research. H. Matsumoto et al. WAVELENGTH (MICRONS)

I.- z ‘U C-) ‘U A.

‘U C-) z

I—

C,, z 4 I-

WAVENUMBER (CM―) @ Chart 1. lR spectra of MAM-GIcUA and cycasin. MAM-GIcUA was synthesized by catalytic oxidation of cycasin. The appearance of the new 1715 cm band in the top spectrum confirms the oxidation of the primary alcohol of cycasin to a carboxylic acid group. The 1540 cm ‘bandin both spectra indicates the presence of the azoxy moiety in cycasin and MAM-GIcUA.

Table1 of free MAM if all the injected compound had been completely Proton magnetic resonance spectra of MAM-G!cUAinNo. ester hydrolyzed by tissue $-glucunonidase. The possibility that there CDCI3and in CDC!3plus 020methyl may be a difference in the fl-glucunonidase activity in rats of of different ages was tested by injecting 268 mg (1 mmol) MAM hy dro GIcUA pen kg body weight into rats weighing 200, 300, and gensAssignment1ppm Multiplicity 400 g. The compound was not acutely toxic either to these 00 MHz (CDCI3) larger rats onto 250-g female rats given injections of 300 mg/ 5.06 Broad singlet 2 kg body weight. MAM-GIcUA was nontoxic to Wistan rats of 4.9-4.4 Complex broad 5 H—C multiplet either sex and of varying ages. 4.06 Singlet 3 CH3—(O)N= Mutagenicity. MAM-GIcUA is mutagenicto S. typhimurium 3.75 Singlet 3 C(O)—O---CH3 when pneincubated with partially purified E. coli fl-glucunoni H—O—C13.95-3.3 Broad multiplet 3O—CH2—N= dase. There was no mutagenic effect when MAM-GIcUA was 00 MHz (CDCI3 plus 020) preincubated with bovine liver ,8-glucunonidase (Chart 2). 5.09 Broad singlet 2 4.2—3.2 Broad complex 5 H—C The specific activity [Fishman (4) units/mg of enzyme] of E. 4.06 Broad singlet 3 CH3—(O)N= coil /‘@-gIucunonidasewas5,500 while the bovine liven activity 3.76 Singlet 30—CH2—N=C(0)—0——CH3 was 11,100 when phenolphthalein glucunonide was used as

3072 CANCERRESEARCHVOL. 39

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1979 American Association for Cancer Research. Synthesis of MAM-GIcUA

have demonstrated that MAM-OAc inhibited DNA synthesis in the duodenum and colon of rats with cannulated bile ducts.

LU Matsubara et al. (12) have shown that rats subjected to single barreled colostomies and given injections of MAM-OAc devel _1 0@ oped tumors in the colon distal to the colostomy where the fecal stream did not contact the mucosa. Both groups have LU 0. concluded that MAM on MAM-OAc can reach the colon via the @ 1200 vascular system. The experiments, however, do not eliminate z the possibility that some MAM-GIcUA can be formed. The 0 synthesis of MAM-GIcUA now makes it possible to unequivo @ 1000 0 cally determine whether MAM reacts with glucuronic acid in I- the liver and whether the resulting MAM-GIcUA conjugate is z excreted with the bile or in the urine.

LU > LU REFERENCES

0 1. Ames, B. N., McCann, J., and Yamasaki, E. Methods for detecting carcine cr gens and mutagens with the Salmonella/mammalian-microseme mutagen LU icity test. Mutat. Res., 31: 347—367,1975. 2. Druckrey, H. Production of colonic carcinomas by 1,2-dialykylhydrazines :D and azexyalkanes. In: W. J. Burdette (ed), Carcinoma of the Colon and z Antecedent Epithelium, pp. 267—279.Springfield, III.: Charles C Thomas, Publisher, 1970. 3. Fiala, E. S. Investigations into the and mode of action of the colon carcinogen 1,2-dimethylhydrazlne and azoxymethane. Cancer (Phila.), MAM-GLUCURONIDE (p MOL) 40: 2436—2445,1977. Chart 2. A comparison of mutagenicity of MAM-GIcUA preincubated with E. 4. Fishman, W. H., and Bernfeld, P. Glucuronidases. Methods Enzymoi., 1: coli (C) and bovine liver (0) $..glucuronidase. S. typhimurium strain G46 was 262—269,1955. preincubated with MAM-GIcUA and each of the 2 enzymes for 90 mm before the 5. Heyns, K., and Paulsen, H. Selective catalytic oxidation of carbohydrates mixture was poured on a minimal-glucose agar plate. The number of revertant employing platinum catalysts. Adv. Carbehydr. Chem., 17: 169—221,1962. colonies was counted after 2 days of incubation at 37°. 6. Kebayashi, A., and Matsumete, H. Studies en methylazexymethanol, the aglycene of cycasin: isolation, biological, and chemical properties. Arch. Biochem. Biophys., 110: 373—380,1965. substrate. However, when MAM-GIcUA was used as substrate, 7. Korsch, B. H., and Riggs, N. V. Proton magnetic resonance spectra of allphatlc azexy compounds and the structure of cycasin. Tetrahedron Left., the specific activity of E. coil f.1-glucunonidaseactivity was 330, 523-525, 1964. but there was no activity with bovine liven glucunonidase. The 8. Laqueur, G., and Matsumete, H. Neoplasms in female Fisher rats following assays were carried out with 2000 units and at the optimum intraperitoneal injection of methylazoxymethanol. J. Nat. Cancer Inst., 37: 217—232,1966. pH for each of the enzymes. 9. Laqueur, G. L., and Spatz, M. Oncogenicity of cycasin and methylazexy methanol. Gann Menogr., 17: 189—204,1975. 10. Levy, G. A., and Cenchie, J. $-Glucuronidase and the hydrolysis of glucu DISCUSSION renides. In: G. J. Dutton (ed.), Glucuronic Acid, Free and Combined, pp. 301 -364. New York: Academic Press, Inc., 1966. The oxidation of cycasin with gaseous oxygen in the pres 11. Marsh, C. A. Chemistry of D-glucurenic acid and its . In: G. J. Dutton (ed), Glucuronic Acid, Free and Combined, pp. 3-1 36. New York: ence of a platinum-on-carbon catalyst is suitable for this highly Academic Press, Inc., 1966. labile glucoside because the reaction is specific. The primary 12. Matsubara, N., Men, H., and Hirone, I. Effect of colostomy en Intestinal hydroxyl group of glucose is attacked preferentially to the carcinogenesis by methylazoxymethanol acetate in rats. J. Nat. Cancer Inst.,61:1161—1164,1978. secondaryhydroxylgroups(5,11). 13. Matsumote, H.. Nagahama, T., and Larson, H. 0. StudIes en methylazexy The mutagenicity test indicates that MAM-GIcUA is active methanol, the aglycene of cycasin: A synthesis of methylazexymethyl ace when preincubated with E. coil /@-glucuronidasebut not when tate. Biochem. J., 95: 13C-1 4C, 1965. 14. Matsumete, H., and Strong, F. M. The occurrence of methylazoxymethanel preincubated with bovine liven glucuronidase. The enzyme in Cycas circinalis L. Arch. Biochem. Biephys., 101: 299—310,1963. assays also indicate that MAM-GIcUA can be hydrolyzed by E. 15. Mehltretter, C. L., Alexander, B. H., Mellies, A. L., and Rist, C. E. A practical synthesis of D-glucuronic acid through the catalytic oxidation of 1,2-lsopre coil /3-glucuronidase but not by bovine liver glucuronidase, pylidlne-o-glucose. J. Am. Chem. Soc., 73: 2424—2427,1951. although both enzymes can cleave phenolphthalein glucuro 16. Nishlda, K., Kobayashi, A., and Nagahama,T. Cycasin a newtexic nide. Mammalian liver glucuronidase apparently has a low of Cycas revoluta Thunb., I., isolation and structure of cycasin. Bull. Agric. Chem. Soc. Jpn. 19:77—84,1955. affinity for MAM-GIcUA. The Km'5 of a given mammalian fi 17. Weisburger, J. H. Chemical carcinogens and their mode of action In celonic glucuronidase for various glucunonide substrates, as well as neoplasia. Dis. Colon Rectum, 16: 431 -437, 1973. the /3-glucuronidase activity of difterent organs, vary consid 18. Yeshlmura, J., Sate, T., and Ando, H. Aminosugars. xvll. synthesIs of benzyl 2-acetamlne-2-deexy-a-o-glucepyranesldurenlc acidamides and N- enably (10). The lack of acute toxicity of MAM-GIcUA injected (2-acetamine-2.deoxy-o-glucurenyl)-amlneaclds. Bull. Chem. Soc. Jpn., 42: i.p. into rats probably indicates that the compound is not acted 2352—2356,1969. 19. Zedeck, M. S., Grab, D. J., and Stemberg, S. 5. Differences in the acute upon by tissue glucuronidase. response of the various segments of rat intestine to treatment with the Indirect tests have shown that MAM can be transported to intestinal carcinogen, methylazoxymethanel acetate. Cancer Res., 37: 32— the colon by means other than with the bile. Zedeck et al. (19) 36, 1977.

AUGUST1979 3073

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1979 American Association for Cancer Research. Synthesis of the Glucuronic Acid Conjugate of Methylazoxymethanol

Hiromu Matsumoto, Ronald H. Takata and David Y. Komeiji

Cancer Res 1979;39:3070-3073.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/39/8/3070

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/39/8/3070. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1979 American Association for Cancer Research.