Aristolochic acid induces 6-thioguanine-resistant mutants in an extrahepatic tissue in rats after oral application

Mutation Research, 143 (1985) 143-1 48

143

Elsevier

MRLelt. 0709

Aristolochic acid induces 6-thioguanine-resistant mutants in an extrahepatic tissue in rats after oral application P . Maier, H.P. Schawalder, B. Weibel and G. Zbinden Institute of Toxicology, S wiss Federal Institute of Technology and University of Zurich, Schorenstrasse 16, 8603 Sch werzenbach (Switzerland) (Accepted 11 March 1985)

Summary The mutagenic activity of the natural plant product aristolochic acid (AA) was tested in the Granuloma Pouch Assay, which detects gene mutations induced in a subcutaneous granuloma tissue of rat s. After direct expo sure of the target tissue, AA induced high frequencies of mutants at a relati vely low cyto static/ cytotoxic level. AA was more potent that N-methyl -N ' -nitro-N-nitrosoguanidine (MNNG) at equimolar do ses. After oral application of AA, a dose-dependent mutagenic activ ity was seen. In contrast a very weak and inconsistent mutagenic effect was seen after systemic application of MNNG. These observation s suggest that after oral appli cation AA is not detoxified efficiently and can exert its mutagenic activity in extrahepatic tissues whereas MNNG is detoxified to a large extent at the site of administration.

The extract isolated from the roots of the plant Aristolochia clematitis has been known as a medicine since antiquity. In modern times, it has been used mainly as an antiphlogistic agent in several pharmaceutical preparations (Mose, 1966). The active component is aristolochic acid (AA), a mixture of at least 6 components (pailer et al., 1955, 1966; Gracza, 1981). The ma in ingredients are two phenanthrene carboxylic acids which differ from each other by one methoxy group (Schmeiser et aI. , 1984). Recent toxicological evalu ations include studies on nephrotoxicity in humans (Thiele et al . 1967), on the cyto static activity in plant and mammalian cells (Moretti et al ., 1979) and on antifertility effects in rabbits (Pakrashi et al. , 1978). In a carcinogenicity study in rats, AA caused gastric carcinomas with a short latency period (Mengs et aI. , 1982). Since then, AA has been put through a number of in vitro mutagenicity tests. In

Salmonella (Robisch et aI. , 1982; Schmei ser et al., 1984), in human lymphocytes (Abel and Schim mer, 1983) and in Chinese hamster cells (Puri and Muller, 1984), AA induced gene mutations, sisterchromatid exchanges or chromosome aberrations. In Drosophila melanogaster, AA was found to be mutagenic and recombinogenic (Frei et al., 1984). However, the compound was ina ctive in the micronucleus test in mice (Biolog ische Heilmittel Heel, 1982). This led to the conclusion that for the most part the chemical acts directly , ha s a local activity and is detoxified efficiently in vivo . However , the micronucleus test onl y detects the loss of chromosomal material during mito sis of erythroblasts, usually as a result of a clastogenic or spindle disturbing activity of the test chemical (Schmid, 1976; Maier and Schmid, 1976). Therefore, it does not conclusively exclude a mutagenic risk in extrahepatic tissues after oral ap-

016 5·7992 / 85/ $ 03.30 © 1985 Elsevier Science Publishers B.V. (Biom ed ical Division)

144

plication. The negative outcome of the micronucleus test might be due to peculiarities of the genetic endpoint investigated. For this reason the present study was undertaken with the Granuloma Pouch Assay (GPA). This experimental model is a phase II mutagenicity test in which several genetic changes induced in vivo are detectable, including gene mutations (Maier, 1984). This test has the further advantages that the target, a subcutaneous granuloma tissue, can be exposed to the test chemical directly or via systemic routes. Therefore, AA tested in the GPA will elucidate the contradictionary results obtained in the mutagenicity tests and will demonstrate the usefulness of phase II mutagenicity tests. NMethyl-N' -nitro-N-nitrosoguanidine (MNNG) was chosen as a control mutagen with a high local and low systemic activity (Maier, 1984). Materials and methods

Chemicals. AA was purchased from EGAChemie Steinheim, F.R.G. (Lot No. 061287 EH) and MNNG from Fluka AG Buchs, Switzerland. The relative amount of the different types of aristolochic acids was not defined. A mixture of dimethyl sulfoxide/fetal calf serum (1:9) was used as a solvent for AA and a mixture of acetone/water (1:4) for MNNG. Animals and treatment. A subcutaneous 25-ml air pouch was formed in randomly bred albino male Sprague-Dawley rats (SIV 50, Ivanovas, Kisslegg), weight 220-240 g (Maier, 1984). Two days later the chemicals were injected in volumes of 1 ml into the pouch, or intraperitoneally (MNNG) or applied by gavage (AA). The doses chosen were adapted from those used in carcinogenicity studies in rats. Gene mutation test. Gene mutations at the HGPRT - locus are assumed to be present in cells able to form colonies in a medium containing the purine base analogue 6-thioguanine (6-TG). The genetic events involved are undefined mutations (base-pair substitutions, small deletions, chromosomal rearrangements) which result in a lost, reduced or altered synthesis of the enzyme

hypoxanthine guanine phosphoribosyl transferase. The test was performed according to standardized procedures (Maier et al., 1983; Maier, 1984). Two days after treatment, the granulation tissue was removed by dissection and dissociated enzymically into single cells. These individual cells were cultured in vitro in Dulbecco's modified Eagle medium supplemented with 10% FCS and 0.50/0 gentamycin (expression period). After 3 days, cells were subcultured and after a second period of 3 days, cells were harvested and exposed to selective culture medium (15 ~M 6-TG) for 7 days. The total number of mutant colonies found in selective medium divided by the total number of cells plated and corrected for plating efficiency was expressed as mutant frequencies. Results

The spontaneous mutant frequency in this study was similar to the historical control value determined in the past two years (Maier, 1984); After single application into the air pouch, both chemicals induced a dose-dependent increase in

•,

, ,,

, AA , ,,, , ,,, I

, +, ,, I

~

,

>-

I

, ,,,

!

;: 100

I

:!

I

,

~

I I

E

~

OJ

, I

I I I

, I

234

"mo1/op Fig. I. Mutant frequencies induced after injection of MNNG and AA into the granuloma pouch.

145

mutants. A nearly linear exposure-effect relationship was obtained within the range tested (Fig. 1). AA showed a much steeper slope than MNNG. The cytostatic/cytotoxic activity of the two test chemicals is expressed in the primary cloning efficiency (I CE) of the freshly isolated two days after treatment (Table 1). At an equimolar expo sure or at a given mutant frequency, the cyto static/cyto-

toxic activities of AA were lower than those of MNNG (Table I, Fig. 1). After systemic application , both chemicals caused a dose-dependent increase in mutant frequencies. The exposure levels chosen were limited by the toxicity of the chemical to the whole animal. With MNNG (i.p.), the highest dose corresponded to half the LD 50, whereas at equimolar exposure

TABLE 1 MUTANT FREQUENCIES INDUCED AFTER INJECTION OF MNNG AND AA INTO THE GRANULOMA POUCH (GP) OR APPLIED SYSTEMICALLY Administration route and chemical

Exposure mg /kg (JIM/kg)

AA into the pouch

Jlg/GP (JiM/GP)

40 (0.12) 160 (0.47) 320 (0.94)

Number of Mutant frequency x 106 clone Ran ge forming cells Mean tested x 106

Number of animals tested

Cloning efficiency

2

30.2

83.4

4.0

10.7*

2.9- 19.0

3

10.2

64.7

3.6

172.5*

88.5-320.9

3

6.8

65.1

3.7

305.3*

156.9-510.50

2

20.2

87.9

4.1

18.1*

8.1- 27.8

2

21.3

72.2

2.5

54.5*

29.3- 60.1

5

9.5

46.9

2.8

68.1*

35.6-135 .0

4

8.1

75.8

4.7

112.9*

28.2-308.4

5

5.4

51.0

2.9

251.3*

70.5-557 .3

2

21.7

57.5

1.3

2.1

4

12.6

63.6

2.9

13.5*

0.9- 31.6

3

14.0

66.4

3.1

16.8*

0- 38.6

22

19.4

77.9

20.6

3.7

0- 14.6

I

CE

II

CE

Orall y

45 (131.86) 90

(263.71) MNNG into the pouch

Intraperitoneally

Control

100 (0.68) 200 (1.36) 600 (4.08)

2.5 (16.99) 12.5 (84.98) 25 (169.95)

I CE, primary cloning efficiency determined from fre shl y isolated cells. II CE, secondary cloning efficiency determined from cells after 6 day expression per iod (Maier, 1984). * Significantly different from control values at P < 0.05 (Kastenbaum and Bowman, 1970).

0-

4.1

146

... c

•or " ~ c ;

1

E

"'" ';"

.....

...... ..........

......... AA

__....,....::::s-_ MNNQ 100

200

p.o .

I.p.

300 !,moI/kg

Fig. 2. Mutant frequencies induced after systemic application of MNNG (i.p .) and AA (p .o .) .

no toxicity was observed with AA (p.o.). It was possible to induce higher mutant frequencies by increasing exposure levels (Fig. 2). Discussion After a single application into the air pouch, AA proved to be a more potent inducer of mutants than MNNG. One reason may be that AA, in contrast to MNNG, is a S-phase independent clastogen (Abel and Schimmel, 1983). Therefore, the mutagenic effect is not limited by the number of cells going through the S-phase. Most likely, the Sphase-independent activity is also responsible for the relatively low cytostatic or cytotoxic effect of AA. In the gene mutation test, this is of particular importance, since only mutant cells can be detected which divide and form clones in selective media. This process also has its significance in tumor formation in vivo, as cells with a high proliferative capacity have the highest chance of expressing their transformed phenotype. The second reason for high mutant frequencies might be the metabolic activation system present in

the target cells. AA is considered to be a directacting mutagen in Salmonella tests (Robisch, 1982), but nonetheless a metabolic conversion by microbial enzymes cannot be ruled out. As a nitroaromatic compound, nitroreductase are most likely involved in the reduction of the nitro group was shown to be a basic requirement for the mutagenic and carcinogenic activity of nitrofurans (McCalla, 1983). Indeed, in the nitro reductase deficient Salmonella strain TAlOONR, AA was non, or only weakly, mutagenic (Schmeiser, 1984). Because nitroreductases in bacteria are highly efficient, bacterial tests might give an exaggerated picture of the potency of nitro compounds . However, in mammalian tissue several factors, including efficient detoxification reactions and probably also the pOz in the tissue, are important parameters for determining the capacity of reductive enzyme systems (McCalla, 1978). The pOz is low in the subcutaneous tissue, (47.8 ± 9.4 mm Hg; Maier et aI., 1984). These conditions favor reductive chemical processes, most likely mediated by nitroreductase II (McCalla et aI., 1983). An alternative or additional activation of AA may occur by a cooxidation of AA by prostaglandin endoperoxide synthetase. This enzyme is present in the granulation tissue (Lang et aI., 1984). With nitrofurans moreover, it was shown that this cooxidative process leads to reactive metabolites which bind to DNA (Zenser et al., 1980, Kadlubar et al., 1982). Nitrofurans tested in the GP A were mutagenic (Fritschi et al., 1984). By comparing the mutant frequencies obtained in the GPA after local and systemic application, the pharmacokinetic behaviour of AA can be characterized . When AA was applied into the air pouch, about 2 g of the target tissue was immediately exposed to the chemicals. Assuming an uniform distribution after systemic application, 160 flg injected into the GP corresponds to a systemic exposure of 80 mg/kg body weight. After oral application of 90 mg/kg (Table I), the induced mutant frequency in the target tissue was only 3.2 times lower than after local application of 160 JLg. This indicates that the detoxification of AA after systemic application is inefficient.

147

In Salmonella mutagenicity tests, the addition of an S9 fraction resulted in onl y a slight decrease of the mutagenic activity of AA (Schleiser, 1984). This confirms that AA, in contrast to MNNG, cannot be detoxified directly by liver enzyme preparations. The more efficient local activity of AA and the more efficient con servation of the activity after systemic application might have two reasons. Either a sta ble intermediate is formed at the primary site of absorption, which is then subsequently distributed within the organism and transformed into the ultimate carcinogen intracellularly in the extrahepatic target tissues or, only part of the systemically applied chemical reacts at the site of application. The other fraction, because it is not detoxified in liver, is distributed unchanged in the body. This latter interpretation is supported by an excellent correspondence between the mutagenicity data obtained after systemic or local application in the GPA and the result s from carcinogenicity studies in rats: After feeding rats with AA , carcinomas were found in the stomach within 4 months (Mengs, 1983). Thi s shor t latency period was found to parallel GPA results , indicating an efficient induction of mutant in the target tissue after direct expo sure. In long-term studies with AA, carcinomas were also found in liver, kidney, bladder, lung and skin (Mengs et aI., 1982). This is in agreement with the mutagenic activity found after systemic application in the GPA. On the other hand, in carcinogenicity studies with MNNG, the chemical induced tumours in the glandular stomach and forestomach of rats after oral application (Sugimura and Fujimura, 1967). When injected subcutaneously (Schoenthal, 1966) or into the granuloma pouch (Fluckiger, 1983), fibrosarcomas developed only at the site of injection. These results correlate with the low mutagenic activity of MNNG found after systemic application and indicates that the chemical is detoxified for the mo st part even more than AA at the sites of application. In summary, the G PA proved to be an effective assa y for demonstrating the induction of gene mutations by AA in vitro. The mutagenic acti vity of AA in extrahepatic tissue after direct exposure

is mo re pronounced than that of MNNG at equimolar do ses. In contrast to MNNG, AA retains a substantial part of the mutagenic activity after systemic application. Therefore AA ha s to be cla ssified as a potent mutagenic and carcinogenic compound in extrahepatic tissue s in rat s. Acknowledgements Thi s research was supported by the Swiss National Science Foundation , contract 3.935-0.82.

References Abel , G ., and O . Schimmer (1983) Induction o f stru ctural chro mo some aberrations and sister chro ma tid exchanges in human lymph ocytes in vitro by ari stolochic acid , Hum . Genet ., 557, 1-3. Biologische Heilmittel Heel (1982) Bad en-Bad en , BRD, Contractor for a micronucleic study performed by the Lab or atory o f Mutagen icity Testin g, Darmstad t , BRO . Fluckiger , B. (1983) Wa chstum und chemisch indu zierte maligne Transfor mation von Zellen eines subkutanen Gran oloms unt er dem Einflu ss von Stimulatore n der Zellpro lifera tion, Dissertation No . 7292, Swiss Federal Institu te of Techn olo gy, Zurich, Switzerla nd . Frei , H ., F.E . Wiirgler , H. Juon, C. B. Hall and U. Graf (l 984) Ari stolochic acid is mutagenic a nd recombinogenic in Drosophil a genotoxicity tests, Arch . Toxicol. , in pre ss. Fr itschi , S., P . Maie r and G. Zbind en (1985) The mutagenic activity of nitrofurans in an extr ah epati c tissue in rats (Gr anuloma Pouch Assay), in preparat io n. Gracza, L., and P. Ruff (1981) Einfache Methode zur Bestimmung der Aristolochiasauren durch HPLC, Dtsch. Apoth. Ztg., 121,2817-2818. Kadlubar , F. F., C.B . Frederick, C. c. Weis and T.V. Zenser (1982) Pr ostaglandin endoperoxide syntheta se-mediated metabolism of carcinogenic aromat ic amines a nd their bind ing to DNA and protein, Biochem . Bioph ys. Res. Commun ., 108, 253-258. Kastenb aum , M.A ., a nd K.O . Bowman (1970) Tables for determining the statistical signifi can ce of mut ati on frequ encies , Mutat ion Res., 9, 527-549. Lang , B., P . Maier and G. Zbinden (1984) Metabolic activation o f aldrin in extrahepatic cells, Experientia, 40, 647. Maier , P . (1984) Th e Granuloma Pouch Assay, in: F.J . de Serres (Ed .), Ch emical Mutagens, Prin ciples a nd Methods for The ir Detection Vol. 9, Plenum, New York, pp, 233- 260. Maier , P ., and W. Schmid (1976) Te n model mutagens eva luated by the micronucleus test, Mut at ion Res., 40, 325-338 .

148 Maier, P., B. Weibel and G. Zbinden (1983) The mutagenic activity of 5-bromo-2' -deoxyuridine (Brdl.I) in vivo in rats, Environmental Mutagenesis 5, 695-703. Maier, P., B. Lang, H.P. Schawalder and G. Zbinden (1984) Influence of physiological p02 in vitro on growth and enzymatic activity in freshly isolated mesenchymal rat cells, in preparation. McCalla, D.R. (1983) Mutagenicity of nitrofuran derivatives, Environ. Mutagen., 5, 745-765. McCalla, D.R., C.R. Arlett and B. Broughton (1978) The action of AF-2 on cultured hamster and human cells under aerobic and hypoxic conditions, Chem.-Bio!. Interact. 21, 89-102. Mengs, U. (1983) On the histopathogenesis of rat forestomach carcinoma caused by aristolochic acid, Arch. Toxico!., 52, 209-220. Mengs, U., W. Lang and J .A. Poch (1982) The carcinogenic action of aristolochic acid in rats, Arch. Toxico!., 51, 107-119. Moretti, C., M. Rideau, J.c. Chenieux and C. Viel (1979) Isolement de l'acide aristolochique de deux aristoloches malgaches, Determination de sa cytotoxicite sur cellules vegetates, Comparaison avec les cellules animales, Planta Med., 35, 360-365. Mose, J.R. (1966) Weitere Untersuchungen uber die Wirkung der Aristolochiasaure, Arzneimittelforsch., 16, 118-122. Pailer, M., and P. Bergthaller (1966) Die Konstitution der Aristolochiasauren III and IlIa, Monatsh. Chern., 97, 484-493. Pailer, M., L. Belohlav and E. Simonitch (1955) Zur Konstitution der Aristolochiasauren, Monatsh. Chern., 86, 676-680. Pakrashi, A., and B. Chakrabarty (1978) Antifertility effect of

aristolic acid from Aristolochia indica (Linn.) in female albino rabbits, Experientia, 34, 1377. Puri, E., and D. Muller (1984) Zu den mutagenen Eigenschaften und zur Karzinogenitat der Aristolochiasaurc, Abstract GUM-Tagung 1984, Wuppertal, BRD. Robisch, G., O. Schimmer and W. Goggelmann (1982) Aristolochic acid is a direct mutagen in Salmonella typhimurium, Mutation Res., 105, 201-204. Schmeiser, H.H., B.L. Pool and M. Wiessler (1984) Mutagenicity of the two main components of commercially available carcinogenic aristolochic acid in Salmonella typhimurium, Cancer Lett., 23, 97-101. Schmid, W. (1976) The micronucleus test for cytogenetic analysis, in: A. Hollaender (Ed.), Chemical Mutagens, Principles and Methods for their Detection Vo!' 4, Plenum, New York, pp. 31-53. Schoenthal, R. (1966) Carcinogenic activity of N-methylN' -nitro-N-nitrosoguanidine, Nature (London), 209, 726-730. Sugimura, T., and S. Fujimura (1967) Tumor production in glandular stomach of rats by N-methyl-N' -nitro-Nnitrosoguanidine, Nature (London), 216, 943. Thiele, K.G., R.C. Muehrcke and H. Berning (1967) Nierenerkrankungen durch Medikamente, Dtsch. Med. Wochenschr., 92, 1632-1635. Zbinden, G., and P. Maier (1983) Single dose carcinogenicity of procarbazine in rats, Cancer Lett., 21, 155-161. Zenser, T.V., M.B. Mattammal and B.B. Davis (1980) Metabolism of N-(4-(5-nitro-2-furyl)-2-thiazolyl) formamide by prostaglandin endoperoxide synthetase, Cancer Res., 40, 114-118.