Assay of Fractions of Bracken Fern (Pteris Aqullina) for Carcinogenic Activity'

[CANCER RESEARCH 30, 902-905, April 1970] Assay of Fractions of Bracken Fern (Pteris aqullina) for Carcinogenic Activity'

A. M. Pamukcu,J. M. Price,and GeorgeT. Bryan2 Department of Pathological Anatomy, College of Veterinary Medicine, University of Ankara, Ankara, Turkey fA. M. P.!; Scientific Divisions, Abbott Laboratories.North Chicago,Illinois 60064 [email protected]. P.1; and The Division of ClinicalOncology, University of WisconsinMedicalSchool, Madison, Wisconsin53706 1G. T. B.J

SUMMARY fractions of bracken fern-fed cattle and of cows bearing spontaneous urinary bladder tumors contained some Fresh bracken fern (Pteris aquilina) was extracted with cold carcinogenic substance(s) which induced tumors in the bladder and hot methanol followed by ethyl ether. Six fractions of the calf, dog, rat, or mouse and tumors in the skin of mice. designated as G, E, F, K, H, and M were obtained. The residue It is not known whether this carcinogenic activity is obtained after removal of the solvent was mixed with 4 times attributable to bracken fern itself or to its metabolites. It has its weight of cholesterol, and the mixture was converted to been shown (22) that the carcinogenicity of the crude cycad pellets with the aid of a pellet press. Pellets containing these material and of cycasin in rats depends on an intestinal residues were surgically implanted into the bladder lumens of bacterial flora capable of providing the enzyme @3-glucosidase, Swiss albino female mice, and a control group of mice received necessary for the liberation of the aglycone, only pure cholesterol pellets. A 53% and 56% incidence, methylazoxymethanol. In germ-free rats, cycasin per se, respectively, of bladder carcinomas was observed in the mice whether present in the intestinal tract or absorbed and of Groups G and H. This incidence was statistically greater excreted, is innocuous. It was desirable to determine whether a than that seen in the control group. A lower incidence of similar mechanism operates in the carcinogenicity of crude urinary bladder carcinomas was obtained in the mice of the bracken fern, or whether the fresh bracken fern itself contains other groups. These results suggest that one or more the carcinogenic substance before digestion in the animal carcinogenic substances occur in bracken fern itself. body. Therefore, it was decided to test fractions of bracken fern for carcinogenic activity on the bladder epithelium of the mouse. The results of these experiments are the subject of this INTRODUCTION report.

Studies have clearly demonstrated that bracken fern (Pteris MATERIALS AND METhODS aquilina) contains carcinogenic substance(s). A low level of bracken fern fed in the daily ration produced urinary bladder Preparation of Fractions. Fresh bracken fern obtained from cancer in cattle (19, 29, 32). Bracken fern also produced the vicinity of Bolu, Turkey, was extracted (Chart 1) with an tumors at different sites in the body and demonstrated a wide excess of cold methanol until the green color of the bracken range of action in different species. Recently, it was reported fern became light brown. The combined methanolic that bracken fern induced intestinal adenocarcinoma in rats extractions were evaporated under reduced pressure at low (13, 3 1, 32) and in Japanese quail (Coturnix coturnix japanica) temperature (36Â°), and the resulting dark green residue was (35), pulmonary adenomas in mice, and bladder tumors in designated as Fraction G. The remaining bracken fern was guinea pigs (1 2, 14, 35). Bracken fern can produce tumors in continuously extracted further with hot methanol. After this different sites in the body of the same species of animal. methanolic fraction was cooled overnight, a white solid Pamukcu and Price (3 1) and YalÃ§iner (36) were able to material was obtained. This material was filtered and produce a high percentage of simultaneously occurring fractionated with ether into ether-soluble (Fraction F) and intestinal and urinary bladder carcinomas in rats by feeding ether-insoluble (Fraction E) fractions. The ether-insoluble bracken fern. fraction could be crystallized with acetic acid, and showed a Biological tests (16â€”18, 30) also demonstrated that urine melting point of the resulting crystals at 75Â°. The soluble methanol fraction was evaporated to a reduced volume and cooled overnight. A solid material was obtained and filtered from the methanolic solution. This solid fraction was called

1 Supported in part by Grants CA-08254 and CA-10017 from the Fraction K. The remaining methanol phase was further National Cancer Institute and by a grant from the Turkish Scientific evaporated under reduced pressure and at low temperature and Technical Council. (36 ), resulting in a residue designated as Fraction H. The 2Career Development Awardee of the National Cancer Institute remaining bracken fern residue obtained after both cold and (1-K4-CA-8245-O1A1). hot methanol extractions was dried and subjected to a Received June 11, 1969; accepted September 10, 1969. continuous extraction with hot water. The water-soluble

902 CANCER RESEARCH VOL.30

100 grams of chopped fresh bracken fern

Extracted with cold methanol with continuous stirring at room temperature until the color of bracken turned to light brown, then filtered

Combined methanolic extracts Bracken fern evaporated under reduced pressure at low temperature (36') Continuously extracted with hot methanol (Fraction G)

Methanolic fractions cooled Bracken f@rn overnight and filtered Continuously extracted with hot Solid Jxtracted Filtrate rLluced to water and filtered with ethyl ether 1/3 volume under reduced pressure and filtered Filtrate @vaporated Re@idue under reduced pressure discarded Ether-soIuble Ether-insoluble (Fraction M) portion (Fraction F) white material, Solid Filtrate evaporated crystallized from (FractionK) under reduced acetic acid, m.p. 75Â° pressure at 360 (Fraction E) (Fraction H)

Chart 1. Flow diagram illustrating preparation of fractions tested for carcinogenic activity. portion was dried under reduced pressure, and the residue was carcinogenicity. Probabilities of statistical significance were designated as Fraction M. evaluated by the exact method for 2 X 2 tables as used Preparation of Pellets. Each of Fractions G, E, F, K, H, and previously (6â€”11, 30). M was mixed with 4 times its weight of cholesterol (purified by recrystallization from ethyl alcohol just prior to use, m.p. RESULTS 15 10) by grinding thoroughly in a mortar. The mixture was compressed into spheroidal pellets, 0. 125 inch in diameter and The number of animals subjected to the surgical implantation weighing 22 to 26 mg, using a deep rounded-cup die in a of pellets, the total number of animals surviving 365 days or Eureka model pellet press (F. J. Stokes Corp., Philadelphia, more following surgery, and the incidence of microscopic Pa.). The dies were dusted frequently with fine magnesium changes observed in the mouse bladders are shown in Table 1. stearate powder to prevent fracturing of the pellet. Pellets of The number of mice that lived beyond 1 year varied from 28 comparable mass were also prepared from purified cholesterol. to 50%. About 30% of the animals died within 30 days after Animal Selection and Care. Swiss albino female mice surgery from impaction of the pellets into the urethra or from (obtained from the Institute of Bacteriology, Elazig, Turkey) the toxicity of the fractions, especially Fractions G and F. were housed in screen-bottomed metal cages with 5 Since the bladders of the mice were inspected after the mice animals/cage. The mice were 70 to 85 days old at the time of had survived more than I year after surgery, no data were surgery, and were fed a pellet diet (Yem Sanayi, Ankara, collected to determine the earliest time or the average time at Turkey) and water ad libitum. which carcinomas were noted. The incidence of carcinomas Carcinogenicity Studies. Pellets were surgically implanted observed ranged from 9% for Fraction F to 56% for Fraction into the urinary bladders of groups of 40 mice by the pellet H. implantation technique of Jull (20) as modified by Allen et aL The incidence of benign papillomas was essentially similar in (1) andasusedinpreviousstudiesfromtheselaboratoriesall groups of mice (Table 1). An examination of these data (6â€”11,30).Onegroupof40miceservedasanegativecontrolstrongly suggests that the carcinogenic substance(s) is soluble group and was implanted with pellets of pure cholesterol. in methanol, since a statistically significantly high incidence of After 52 to 55 weeks the surviving mice were sacrificed. The bladder carcinomas was observed in Groups G (53%) and H bladders were distended postmortem with Bouin's fixative (56%).Alowerincidenceofurinarybladdercarcinomaswas injected into the urethra. Gross and microscopic evaluation of obtained following other extraction procedures. This evidence the lesions was made according to the criteria of Bonser and indicates that bracken fern itself contains active carcinogenic Jull (5) and Roe (33). Carcinomas which were not observed to chemicals. invade the muscular layer of the bladder were classifIed as The histological changes in the mouse bladders showed great Grade I and those which invaded the muscle were classified as variation, including hyperplastic, metaplastic, and neoplastic Grade II. The incidence of carcinomas was used to assess changes. The histology of these lesions was similar to that

APRIL 1970 903

Table 1

Incidence of changes in mouse bladder following implantation offracrions ofbracken fern

-GradeGradeFractions365 No.ofmice killed afterGlandularSquamousBenignCarcinomas daysmetaplasiametaplasiapapilomaIIITotal%pCholesterolalone2100503314G15115448530.02E20113325250.32F1101110190.82K18134347390.08H180353710560.008M19008235260.29

described by others (4, 5, 30, 33). Epithelial hyperplasia beyond reasonable doubt that bracken fern is carcinogenic for consisted of a loss of vacuolation of the cytoplasm of the basal the urinary bladder of cattle, rats, and guinea pigs; cells and an increase in the amount of eosinophilic cytoplasm carcinogenic for the intestinal mucosa of rats and Japanese in the superficial cells. The number of cell layers increased, quail; and carcinogenic for the lungs of mice. It was and exaggerated mucosal folding occurred. The normal demonstrated (30) that a metabolite(s) present in the urine epithelial lining of the mouse bladder consists of 2 to 4 cell fraction of bracken-fed cows is also carcinogenic for the layers, whereas in advanced epithelial hyperplasia the number urinary bladder epithelium of mice when implanted into the of cell layers increased to such a degree that the basal layers urinary bladder lumen. In the present experiments, it appears became folded into the subepithelial tissues forming small that the carcinogen(s) present in bracken fern need not be nests. These nests sometimes simulated adenoma. metabolized by a susceptible species to display its carcinogenic Two types of metaplasia were observed, squamous and activity. The results of these experiments (Table 1) strongly glandular. Squamous metaplasia with keratinization of the suggest that the carcinogenic substance is soluble in methanol, surface layers was seen in 9 cases. This change usually affected for a significantly high incidence of bladder carcinomas was a large part of the epithelial surface. Squamous metaplasia was observed in Groups G (53%) and H (56%). A lower incidence associated with subepithelial inflammation. Glandular of urinary bladder carcinomas was obtained following other metaplasia was observed only in 3 cases. extraction procedures. This evidence suggests that bracken Neoplastic changes consisted of papilloma and carcinoma of fern itself contains active carcinogenic chemicals. However, the bladder epithelium. Papillomas occurred in mice of all these results do not eliminate the possibility that the groups, with an incidence varying from 9 to 42%. Most of the metabolism by gastrointestinal flora or by the experimental solitary or multiple papillomas were sessile, but some were animal itself is necessary for expression of the carcinogenic pedunculated. Microscopically, they were of transitional cell activity by bracken fern. type. Malignant epithelial tumors of the bladder were either The nature of the carcinogenic substance(s) has not been papillary infiltrating or sessile infiltrating carcinomas. Papillary elucidated. A variety of chemicals such as astragalin, infiltrating carcinomas projected into the lumen, and the isoquercitrin , rutin (24), catecholtannins, pteraquilin , sugar, lamina propria of the papillary stalk was invaded with starch, aliphatic nondrying oil, and much pectose mucin (21) carcinoma cells. They were frequently solitary, but in a few have been identified in bracken fern. There is no indication cases the tumors were multiple and extensive. Sessile that these chemicals or their metabolites are bladder infiltrating carcinomas grew into underlying tissues from the carcinogens. start. The tumors were usually well differentiated and showed As bracken fern is used as a human food in greens or salads in a clear origin from transitional cells. Occasional carcinomas the United States, New Zealand, and especially Japan (2, 3, were of squamous type. Most of the carcinomas infiltrated 15, 23, 34), the high incidence of stomach cancer could be in into the muscle. A relatively high incidence of tumors and a part the result of eating bracken. Thus, it seems urgent that high degree of malignancy followed the implantation of the carcinogenic compound(s) present in bracken fern be cholesterol pellets containing bracken fern Fractions H and G. isolated and chemically characterized. Our investigations in No indication of metastasis was found during the 55 weeks of this direction are in progress. the experiment. ACKNOWLEDGMENTS DISCUSSION The assistance of Miss S. Wagner and Mrs. C. Schlotthauer with the preparation of the manuscript isgratefully acknowledged. lt appears that chronic ingestion of bracken fern is a significant factor in the etiology of urinary bladder cancer in REFERENCES the cattle of Turkey (19, 25â€”29, 32). The results of these studies and others (12â€”14, 31, 35, 36) have established 1. Allen, M. J., Boyland, E., Dukes, C. E., Horning, E. S., and Watson,

904 CANCER RESEARCH VOL. 30

J. G. Cancer of the Urinary Bladder Induced in Mice with zur Feststellung der Anwesenheit Kanzerogener stoffe im Ham der Metabolites of Aromatic Amines and Tryptophan. Brit. J. Cancer, mit Heu aus Himaturie-gebieten gefÃ¼tterten KÃ¼he. Wien. ii: 212â€”228,1957. Tieraerztl. Monatsschr., 50: 589â€”595,1963. 2. Anonymous. Dietary Prejudices. J. Am. Dietet. Assoc., 19: 847, 19. G@ksoy, K. Si@irmesane tiim@6rlerinintesekkiiliinde E@reltiotunun 1948. (Pteris aquiina) rolii. Ph.D. Thesis, University of Ankara, Ankara, 3. Anonymous. Bracken, Cancer Link Seen. Sci. News Letter, 88: Turkey,1966. 402, December25, 1965. 20. Juil, J. W. The Induction of Tumours of the Bladder Epithelium in 4. Ball, J. K., Field, W. E. H., Roe, F. J. C., and Walters, M. The Miceby the Direct Application of a Carcinogen.Brit. J. Cancer,5: Carcinogenic and Co-Carcinogenic Effects of Paraffin Wax Pellets 328â€”330,1951. and Glass Beads in the Mouse Bladder. Brit. J. Urol., 36: 225â€”237, 21. Kwasniewski, V. Die lnhaltstoffe der Rhizome des Adlerfarns, 1964. Pteridium aquiinum (L.) Kuhn. Arch. Pharm., 288: 307â€”311, 5. Bonser, G. M., and Jull, J. W. The Histological Changes in the 1955. Mouse Bladder following Surgical Implantation of Paraffin Wax 22. Laqueur, G. L. The Induction of Intestinal Neoplasms in Rats with Pellets Containing Various Chemicals. J. Pathol. Bacteriol., 72: the Glycoside Cycasin and Its Aglycone. Arch. Pathol. Physiol., 489â€”495,1956. 340: 151â€”163,1965. 6. Bryan, G. T. Role of Tryptophan Metabolites in Urinary Bladder 23. Medsger, 0. P. Edible Wild Plants, p. 135. New York : The Cancer. Am. Ind. Hyg. Assoc. J., 30: 27â€”34,1969. MacMillan Co., 1939. 7. Bryan, G. T. Pellet Implantation Studies of Carcinogenic 24. Nakabayashi, T. Isolation of Astragalin and Isoquercitrin from Compounds.J.Natl.CancerInst.,43:255â€”261,1969. Bracken, Pteridium aquilinum. Bull. Agr. Chem. Soc. Japan, 19: 8. Bryan, G. T., Brown, R. R., and Price, J. M. Studies on the 104â€”109,1955. Etiology of Bovine Bladder Cancer. Ann. N. Y. Acad. Sci., 108: 25. Pamukcu, A. M. Investigations on the Pathology of Enzootic 924â€”937, 1963. Bovine Hematuria in Turkey. Zentr. Veterinaermed., 2: 409â€”429, 9. Bryan, G. T., Brown, R. R., and Price, J. M. Incidence of Mouse 1955. Bladder Tumors following Implantation Pellets Containing Certain 26. Pamukcu, A. M. Tumors of the Urinary Bladder in Cattle and Tryptophan Metabolites. Cancer Res., 24: 582â€”585,1964. Water Buffalo Affected with Enzootic Bovine Hematuria. Zentr. 10. Bryan, G. T., Brown, R. R., and Price, J. M. Mouse Bladder Veterinaermed., 4: 185â€”197,1957. Carcinogenicity of Certain Tryptophan Metabolites and Other 27. Pamukcu, A. M. Tumors of the Urinary Bladder in Cattle, with Aromatic Nitrogen Compounds Suspended in Cholesterol. Cancer Special Reference to Etiology and Histogenesis. Acta Unio Intern. Res., 24: 596â€”602,1964. Contra Cancrum, 18: 625â€”638,1962. 11. Bryan, G. T., and Springberg, P. D. Role of the Vehicle in the 28. Pamukcu, A. M. Epidemiologic Studies on Urinary Bladder Tumors Genesis of Bladder Carcinomas in Mice by the Pellet Implantation in Turkish Cattle. Ann. N. Y. Acad. Sci., 108: 938â€”947, 1963. Technic. Cancer Res., 26: 105â€”109,1966. 29. Pamukeu, A. M., G@ksoy, S. K., and Price, J. M. Urinary Bladder 12. Evans, I. A. The Radiomimetic Nature of Bracken Toxin. Cancer Neoplasms Induced by Feeding Bracken Fern (Pteris aquiline) to Res., 28: 2252â€”2261, 1968. Cows.Cancer Res.,27: 917â€”924,1967. 13. Evans, I. A., and Mason, J. Carcinogenic Activity of Bracken. 30. Pamukcu, A. M., Olson, C., and Price, J. M. Assay of Fractions of Nature,208: 913â€”914,1965. Bovine Urine for Carcinogenic Activity After Feeding Bracken 14. Evans, I. A., and Widdop, B. Carcinogenic Activity of Bracken. In: Fern (Pteris aquilina). Cancer Res., 26: 1745â€”1753, 1966. British Empire Cancer Campaign for Research. Annual Report, 31. Pamukcu, A. M., and Price, J. M. Induction of Intestinal and Part II, p. 377. Eastboume, U. K.: Sumfield & Day, Ltd., Urinary Bladder Cancer in Rats by Feeding Bracken Fern (Pteris 1966. aquiina). J. Nail. Cancer Inst., 43: 275â€”279,1969. 15. Femald, M. L., and Kinsey, A. C. Edible Wild Plants of Eastern 32. Price, J. M., and Pamukcu, A. M. The Induction of Neoplasms of North America, p. 71. Cornwall-on-Hudson, N. Y.: Idlewild Press, the Urinary Bladder of the Cow and the Small Intestine of the Rat 1943. by Feeding Bracken Fern (Pteris aquiina). Cancer Res., 28: 16. Georgiev, R:, and Antonov, S. Uber die Aetiologie der chronischen 2247â€”2251, 1968. vesikalen Hamaturie der Rinder. II. Mitteilung. Versuche zur 33. Roe, F. J. C. An Illustrated Classification of the Proliferative and Feststellung der Anwesenheit Kanzerogener Metaboliten im Ham Neoplastic Changes in Mouse Bladder Epithelium in Response to gesunder Kiihe aus einem H@maturiefreien Gebiet bei Fiitterung Prolonged Irritation. Brit. J. Urol., 34: 238â€”253,1964. mit Heu aus [email protected]. Monatsschr.,52: 34. Stratton R. Edible Wild Greens and Salads of Oklahoma. Bull. 90â€”94,1965. Oklahoma Agr. Mech. Coil., 40: 25, 1943. 17. Georgiev, R., Antonov, S., Vrigasov, A., Dimitrov, A., and 35. Widdop, B. Separation and Characterisation of Various Goranov, C. Uber die Aetiologie der chronischen vesikalen Toxicological Components of Bracken. Ph.D. Thesis, University of S@imaturiederRinder. I. Mitteilung.Zur Frage des Karzinomat@sen Wales,Bangor,Wales,1967. Charakters der chronischen vesikalen H@maturieder Rinder. Wien. 36. YalÃ§iner,S.E@reltiotunun (Pteris aquiina) kanserogeniketkisinin Tieraerzfl. Monatsschr., 51: 641â€”657,1964. sicanlar iizerinde arastirilmasi. Ph.D. Thesis, Ankara University, 18. Georgiev, R., Vrigasov, A., Antonov, S., and Dimitrov, A. Versuche Ankara, Turkey, 1969.

APRIL 1970 905

A. M. Pamukcu, J. M. Price and George T. Bryan

Cancer Res 1970;30:902-905.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/30/4/902

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/30/4/902. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.