- Email: [email protected]
Lack of emodin genotoxicity in the mouse micronucleus assay
Mutation Research 393 Ž1997. 289–293
Lack of emodin genotoxicity in the mouse micronucleus assay U. Mengs a
a,)
, G. Krumbiegel a , W. Volkner ¨
b
Madaus AG, Ostmerheimer Strasse 198, D-51109 Cologne, Germany b CCR Cytotest Cell Research, D-64380 Rossdorf, Germany
Received 16 January 1997; revised 29 May 1997; accepted 29 May 1997
Abstract The study was performed to investigate the potential of emodin Ž1,3,8-trihydroxy-6-methylanthraquinone. to induce micronuclei in polychromatic erythrocytes ŽPCEs.. Mice of both genders received a single oral dose of 2000 mg emodinrkg and were killed 24 and 48 h later. Bone marrow cells were collected from 5 males and 5 females and 2000 PCEs per animal were scored for the presence of micronuclei. There was no enhancement in the frequency of micronuclei at both preparation intervals when compared to the negative controls. Blood level examinations confirmed the systemic availability of emodin. Plasma levels of up to 190 mg emodinrml represented concentrations being in the concentration range that induced positive responses in several genotoxicity cell culture assays. q 1997 Elsevier Science B.V. Keywords: Emodin; Anthraquinone; Genotoxicity; Micronucleus assay; In vivo
1. Introduction Emodin Ž1,3,8-trihydroxy-6-methylanthraquinone. is a naturally occurring anthraquinone found in fungi and higher plants belonging to the families of Cassia, Rhamnus, Rheum and Frangula. Preparations of these drugs are in worldwide use for the treatment of constipation. Since danthron, a synthetic laxative anthraquinone, has been shown to induce tumors in rats and mice w1,2x, the mutagenic potential of several anthraquinones has been investigated intensively w3– 5x. Among these, emodin has been shown to possess genotoxic activity in bacterial and mammalian cell culture systems w3,6–10x. In addition, a recent paper ) Corresponding author. Tel.: q49 Ž221. 8998251; fax: q49 Ž221. 8998740.
describes the clastogenicity of emodin in mouse lymphoma cells in vitro w11x. This is interesting with respect to the present work using the same end point in bone marrow cells in vivo. The positive results from different cell culture assays show emodin to be an in vitro mutagen capable of inducing point mutations and chromosomal aberrations. This is toxicologically important as emodin represents one of the major components in natural laxative drugs, e.g. rhubarb root and frangula bark. In senna pods and leaves, however, sennosides are the main active principle and emodin can be found only in traces w12x. The positive in vitro results with emodin and other anthraquinones, e.g. aloeemodin w3,5x, gave rise to discussions as to whether these drugs may pose a genotoxic or carcinogenic risk to humans taking anthraquinone laxatives. A reasonable basis for a discussion on risk assessment can be the inclusion of
1383-5718r97r$17.00 q 1997 Elsevier Science B.V. All rights reserved. PII S 1 3 8 3 - 5 7 1 8 Ž 9 7 . 0 0 1 1 3 - 7
290
U. Mengs et al.r Mutation Research 393 (1997) 289–293
data from in vivo experiments. However, in the case of emodin no such data have been published up to now. Therefore, we conducted a micronucleus assay, a validated and well-known animal model that is generally a part of the test strategy for the evaluation of the genotoxic potential of new drugs. In addition, we determined the blood plasma concentrations of emodin after oral administration to demonstrate systemic exposure. 2. Materials and methods 2.1. Chemicals The hydroxyanthraquinone emodin Ž1,3,8-trihydroxy-6-methylanthraquinone; CAS No. 518-82-1; batch 950101rBa. used in the present in vivo assay was of the highest purity Ž) 99%; HPLC, DSC. and was provided by Madaus AG ŽKoln, ¨ Germany.. As a reference mutagen, cyclophosphamide ŽCPA. was used ŽSigma–Aldrich, Deisenhofen, Germany.. For the kinetic investigations, acetonitrile, di-isopropyl ether, glacial acetic acid, concentrated hydrochloric acid Ž37%., methanol, sodium acetate Žanhydrous. and water in analytical grade quality were obtained from E. Merck ŽDarmstadt, Germany.. b-Glucuronidaserarylsulfatase Žfrom Helix pomatia. was supplied by Boehringer Mannheim GmbH ŽMannheim, Germany.. To prepare a sodium acetate buffer Ž1 molrl, pH 5.6., 41.0 g sodium acetate were dissolved in approx. 450 ml water, pH was adjusted with glacial acetic acid and the volume made up to 500 ml. For calibration, the same batch of emodin was used as in the animal study. Danthron Ž1,8-dihydroxyanthraquinone. was purchased from Aldrich Chemie GmbH ŽSteinheim, Germany.. 2.2. Micronucleus assay 2.2.1. Animals The experiments were carried out on male and female NMRI mice ŽBiological Research Laboratories, Fullinsdorf, Switzerland., weighing about 30 g. ¨ The animals were kept under normal standardised laboratory conditions, given food and tap water ad libitum. Five male and five female mice were used for each group.
2.2.2. Test procedures After a single administration of 2000 mg emodinrkg by gastric gavage Žvolume 20 mlrkg, using a 0.3% aqueous suspension of tragacanth., the mice were killed by cervical dislocation 24 and 48 h later. The femora were removed from each animal and bone marrow smears were made on slides. After staining with May GrunwaldrGiemsa, 2000 PCEs ¨ per animal were scored for micronuclei. To demonstrate a cytotoxic effect due to the treatment with the test compound, the ratio between PCEs and normochromatic erythrocytes ŽNCEs. was determined in the same sample and reported as the number of NCEs per 2000 PCEs. The negative control group received the vehicle and positive controls were treated once with 30 mg CPArkg orally. 2.2.3. Guidelines The study was performed in compliance with the OECD Principles of Good Laboratory Practice and by following the OECD Guidelines for Testing of Chemicals, Section 4, No. 474, adopted May 26, 1983, Micronucleus Test. A compound is considered mutagenic in this test system if at any of the preparation intervals and dose groups, a relevant andror statistically significant increase in the number of micronucleated PCEs is found in comparison to the negative controls. 2.3. Statistics The statistical significance at the five per cent level Ž p ) 0.05. is normally evaluated by means of the nonparametric Mann–Whitney test. However, the obtained results did not require a statistical evaluation as the mean micronucleus frequencies after treatment with the test compound did not exceed the mean negative control values. 2.4. Kinetic studies For the determination of emodin blood plasma concentrations as a proof of intestinal absorption, satellite animals Ž4 males, 4 females. were treated with the same dose of 2000 mg emodinrkg. The animals were killed 1 or 3 h later by cervical dislocation and bled. The blood was centrifuged to separate
U. Mengs et al.r Mutation Research 393 (1997) 289–293
cellular matter and the plasma supernatant was aspirated and stored at y808C. 2.4.1. Chromatography HPLC separations were carried out on Nucleosil w 120-3C18 from Macherey-Nagel ŽDuren, Germany. ¨ at 408C. The m obile phase was aceto nitrilermethanolracetic acid Ž1 molrl., mixed in a time-programmed ratio from an initial ratio of 37:7:56 Ž0 to 4 min. to a final ratio of 60:7:33 Ž6 to 13 min. with a linear increase of the acetonitrile proportion between 4 and 6 min after injection. The eluent was delivered at a flow rate of 0.4 mlrmin. Fluorescence spectrophotometric detection was made at lex s 430 nm and lem s 530 nm. Retention times were approx. 12.5 and 13.2 min for emodin and danthron, respectively. 2.4.2. Sample preparation Ten ml mouse plasma was made up with water to 1.0 ml. Then 1.0 ml sodium acetate buffer and 100 ng danthronr20 ml methanol was added and the sample hydrolyzed with 50 ml b-glucuronidaser arylsulfatase at 378C for 15 to 18 h. The hydrolysate was acidified with 0.5 ml hydrochloric acid Žapprox. 2 molrl, prepared from the 37% solution., extracted with 5.0 ml di-isopropyl ether for 20 min and centrifuged at 2000 = g. The upper layer was aspirated, evaporated to dryness with nitrogen and the solid residue dissolved in 40 ml methanol, and 10 ml were injected. 2.4.3. Calibration and calculation Human blank plasma aliquots of 1.0 ml were diluted to a final volume of 2.0 ml with sodium acetate buffer, spiked with 5 to 2500 ng emodinr50
291
ml methanol and 100 ng danthronr20 ml methanol, and hydrolysed with 50 ml b-glucuronidaser arylsulfatase at 378C overnight. After extraction, centrifugation and reconditioning to 40 ml, 10 ml were injected for separation. For calculation, peak–height ratios were used. There was a linear dependence of the relative peak–height and the concentration of emodin. The lower limit of quantitation was approx. 5 ng emodinrml plasma.
3. Results 3.1. Micronucleus test The results are given in Table 1. Following a single oral dose of 2000 mg emodinrkg, there was no statistically significant enhancement in the frequency of micronucleated PCEs in comparison to the negative controls at both preparation intervals. Further, the mean number of NCEs was not substantially increased as compared to the mean values of NCEs of the negative controls, indicating that emodin had no cytotoxic effects on bone marrow cells. CPA, used as a reference mutagen Žpositive control., significantly increased the micronucleus frequency at an oral dose of 30 mgrkg. 3.2. Kinetic studies Individual blood plasma concentrations of emodin are shown in Table 2. The aim of the kinetic investigations was to demonstrate that emodin could reach the target cells under the experimental conditions described. After a
Table 1 Summary of the results from the micronucleus assay Group
Preparation time Žh p.a..
Number of PCEs scored
Negative control Emodin Ž2000 mgrkg. Positive control
24 24 48 24
2000 2000 2000 2000
a
PCEs with micronuclei Ž%.
PCErNCE
0.08 0.07 0.04 1.35
1.15 1.11 1.09 1.14
PCEs: polychromatic erythrocytes; NCEs: normochromatic erythrocytes; negative control: tragacanth; positive control: cyclophosphamide ŽCPA.; p.a.s post applicationem. a 2000 PCEs were scored of each 5 males and 5 females.
U. Mengs et al.r Mutation Research 393 (1997) 289–293
292
Table 2 Individual concentration values of total emodin in mouse blood plasma Animal no.
Gender
1 2 3 4 5 6 7 8
m m f f m m f f
Total emodin 1 h p.a.
a
Žmgrml. 3 h p.a.
97 130 139 190 173 82 119 73
a
Total emodin means the sum of free Žunconjugated. and conjugated emodin. p.a.s post applicationem; m s male, f s female.
single dose of 2000 mg emodinrkg, the mice showed plasma concentrations ranging from 73 to 190 mgrml within 3 h.
4. Discussion Anthraquinones are widely distributed in medicinal plants, being mainly in use for the treatment of constipation. It is very important to clarify their genotoxic or carcinogenic potential, because several compounds of this chemical class have been found to be mutagenic in vitro, e.g. aloeemodin, emodin, chrysophanol, physcion and others w3,5x. An intensive discussion on the safety of anthraquinone containing laxatives began when the positive findings of two long-term studies with the laxative danthron on rodents had been published w1,2x. On the basis of these findings, it has generally been thought that anthraquinones are tumorigenic in humans w3,10x. However, with the exception of danthron, this conclusion has been drawn only from in vitro assays. This is also the situation with emodin, being the main active principle of frangula drugs. Emodin caused mutation in the Ames test and in the V79HPRT assay and induced UDS in primary rat hepatocytes. Furthermore, emodin showed transforming activity in C3HrM2 mouse fibroblasts in vitro and induced micronuclei in mouse lymphoma L51784Y cells in the absence of metabolic activation w3,11x. However, there were no data available from the
literature showing that a genotoxic activity would also exist under in vivo conditions. We therefore conducted the micronucleus assay in mice to present the first in vivo data on emodin. The dose level used was 2000 mg emodinrkg, representing the maximum guideline-recommended dose and the maximum feasible dose for technical reasons. Emodin failed to induce micronuclei in PCEs of the bone marrow under the conditions described. The fact that plasma levels of emodin Žup to 190 mgrml. were found at two time points after administration Ž1 and 3 h p.a.., being multiples Ž) factor 10. of the minimum genotoxic concentration in vitro w3,11x, demonstrates that the substance must have been in contact with the bone marrow cells. However, the dose of 2000 mgrkg did not induce clear cytotoxic effects as determined by the PCErNCE ratio. The micronucleus data of this study did not indicate a clastogenic effect, either. This lends support to the assumption that data in a recent publication on the clastogenic potential of emodin in vitro w11x might be a special reaction of the mouse cell line L51784Y. Further, it should be kept in mind that our negative in vivo data were obtained under normal mammalian metabolic conditions while the in vitro study was performed under cell culture conditions and without metabolic additives. A similar situation is also present for aloeemodin which is mutagenic in vitro, too w3x, but failed to give positive responses in several in vivo assays including the micronucleus assay, chromosome aberration assay, mouse spot test, and UDS test ex vivo w5x. These results again demonstrate that in vitro data may not be a sufficient basis for the assessment of a possible genotoxic risk for humans. In conclusion, emodin did not induce micronuclei and is therefore considered to be non-genotoxic under the in vivo conditions described. The results further show that in the case of anthraquinones, risk assessments regarding genotoxicity should include valid in vivo results and relevant kinetic data such as blood plasma concentrations. References w1x H. Mori, S. Sugie, K. Niwa, T. Takahashi, K. Kawai, Induction of intestinal tumours in rats by chrysazin, Br. J. Cancer 52 Ž1985. 781–783.
U. Mengs et al.r Mutation Research 393 (1997) 289–293 w2x H. Mori, S. Sugie, K. Niwa, N. Yoshimi, T. Tanak, I. Hirono, Carcinogenicity of chrysazin in large intestine and liver of mice, Jpn. J. Cancer Res. 77 Ž1986. 871–876. w3x J. Westendorf, H. Marquardt, B. Poginsky, M. Dominiak, J. Schmidt, H. Marquardt, Genotoxicity of naturally occurring hydroxyanthraquinones, Mutat. Res. 240 Ž1990. 1–12. w4x A. Heidemann, H.G. Miltenburger, U. Mengs, The genotoxicity status of senna, Pharmacology 47 Žsuppl. 1. Ž1993. 178–186. w5x A. Heidemann, W. Volkner, U. Mengs, Genotoxicity of ¨ aloeemodin in vitro and in vivo, Mutat. Res. 367 Ž1996. 123–133. w6x F.C. Wehner, P.G. Thiel, M. du Rand, Mutagenicity of the mycotoxin emodin in the Salmonellarmicrosome system, Appl. Environ. Microbiol. 37 Ž1979. 658–660. w7x L. Tikkanen, T. Matsushima, S. Natori, Mutagenicity of anthraquinones in Salmonella preincubation test, Mutat. Res. 116 Ž1983. 197–304.
293
w8x H. Morita, M. Umeda, T. Masuda, Y. Ueno, Cytotoxic and mutagenic effects of emodin on cultured mouse carcinoma FM3A cells, Mutat. Res. 204 Ž1988. 329–332. w9x S. Krivobok, F. Seigle-Murandi, R. Steiman, D.R. Marzin, V. Betina, Mutagenicity of substituted anthraquinones in the AmesrSalmonella microsome system, Mutat. Res. 279 Ž1992. 1–8. w10x H. Helmholz, A. Ruge, A. Piasecki, S. Schroder, J. Westen¨ dorf, Genotoxizitat ¨ der Faulbaumrinde, Pharmaz. Z. 43 Ž1993. 3478–3480. w11x S.O. Muller, I. Eckert, W.K. Lutz, H. Stopper, Genotoxicity ¨ of the laxative drug components emodin, aloe-emodin and danthron in mammalian cells: Topoisomerase II mediated?, Mutat. Res. 371 Ž1996. 165–173. w12x W. Grimminger, K. Witthohn, Analytics of senna drugs with regard to the toxicological discussion of anthranoids, Pharmacology 47 Žsuppl. 1. Ž1993. 98–109.
Lack of emodin genotoxicity in the mouse micronucleus assay U. Mengs a
a,)
, G. Krumbiegel a , W. Volkner ¨
b
Madaus AG, Ostmerheimer Strasse 198, D-51109 Cologne, Germany b CCR Cytotest Cell Research, D-64380 Rossdorf, Germany
Received 16 January 1997; revised 29 May 1997; accepted 29 May 1997
Abstract The study was performed to investigate the potential of emodin Ž1,3,8-trihydroxy-6-methylanthraquinone. to induce micronuclei in polychromatic erythrocytes ŽPCEs.. Mice of both genders received a single oral dose of 2000 mg emodinrkg and were killed 24 and 48 h later. Bone marrow cells were collected from 5 males and 5 females and 2000 PCEs per animal were scored for the presence of micronuclei. There was no enhancement in the frequency of micronuclei at both preparation intervals when compared to the negative controls. Blood level examinations confirmed the systemic availability of emodin. Plasma levels of up to 190 mg emodinrml represented concentrations being in the concentration range that induced positive responses in several genotoxicity cell culture assays. q 1997 Elsevier Science B.V. Keywords: Emodin; Anthraquinone; Genotoxicity; Micronucleus assay; In vivo
1. Introduction Emodin Ž1,3,8-trihydroxy-6-methylanthraquinone. is a naturally occurring anthraquinone found in fungi and higher plants belonging to the families of Cassia, Rhamnus, Rheum and Frangula. Preparations of these drugs are in worldwide use for the treatment of constipation. Since danthron, a synthetic laxative anthraquinone, has been shown to induce tumors in rats and mice w1,2x, the mutagenic potential of several anthraquinones has been investigated intensively w3– 5x. Among these, emodin has been shown to possess genotoxic activity in bacterial and mammalian cell culture systems w3,6–10x. In addition, a recent paper ) Corresponding author. Tel.: q49 Ž221. 8998251; fax: q49 Ž221. 8998740.
describes the clastogenicity of emodin in mouse lymphoma cells in vitro w11x. This is interesting with respect to the present work using the same end point in bone marrow cells in vivo. The positive results from different cell culture assays show emodin to be an in vitro mutagen capable of inducing point mutations and chromosomal aberrations. This is toxicologically important as emodin represents one of the major components in natural laxative drugs, e.g. rhubarb root and frangula bark. In senna pods and leaves, however, sennosides are the main active principle and emodin can be found only in traces w12x. The positive in vitro results with emodin and other anthraquinones, e.g. aloeemodin w3,5x, gave rise to discussions as to whether these drugs may pose a genotoxic or carcinogenic risk to humans taking anthraquinone laxatives. A reasonable basis for a discussion on risk assessment can be the inclusion of
1383-5718r97r$17.00 q 1997 Elsevier Science B.V. All rights reserved. PII S 1 3 8 3 - 5 7 1 8 Ž 9 7 . 0 0 1 1 3 - 7
290
U. Mengs et al.r Mutation Research 393 (1997) 289–293
data from in vivo experiments. However, in the case of emodin no such data have been published up to now. Therefore, we conducted a micronucleus assay, a validated and well-known animal model that is generally a part of the test strategy for the evaluation of the genotoxic potential of new drugs. In addition, we determined the blood plasma concentrations of emodin after oral administration to demonstrate systemic exposure. 2. Materials and methods 2.1. Chemicals The hydroxyanthraquinone emodin Ž1,3,8-trihydroxy-6-methylanthraquinone; CAS No. 518-82-1; batch 950101rBa. used in the present in vivo assay was of the highest purity Ž) 99%; HPLC, DSC. and was provided by Madaus AG ŽKoln, ¨ Germany.. As a reference mutagen, cyclophosphamide ŽCPA. was used ŽSigma–Aldrich, Deisenhofen, Germany.. For the kinetic investigations, acetonitrile, di-isopropyl ether, glacial acetic acid, concentrated hydrochloric acid Ž37%., methanol, sodium acetate Žanhydrous. and water in analytical grade quality were obtained from E. Merck ŽDarmstadt, Germany.. b-Glucuronidaserarylsulfatase Žfrom Helix pomatia. was supplied by Boehringer Mannheim GmbH ŽMannheim, Germany.. To prepare a sodium acetate buffer Ž1 molrl, pH 5.6., 41.0 g sodium acetate were dissolved in approx. 450 ml water, pH was adjusted with glacial acetic acid and the volume made up to 500 ml. For calibration, the same batch of emodin was used as in the animal study. Danthron Ž1,8-dihydroxyanthraquinone. was purchased from Aldrich Chemie GmbH ŽSteinheim, Germany.. 2.2. Micronucleus assay 2.2.1. Animals The experiments were carried out on male and female NMRI mice ŽBiological Research Laboratories, Fullinsdorf, Switzerland., weighing about 30 g. ¨ The animals were kept under normal standardised laboratory conditions, given food and tap water ad libitum. Five male and five female mice were used for each group.
2.2.2. Test procedures After a single administration of 2000 mg emodinrkg by gastric gavage Žvolume 20 mlrkg, using a 0.3% aqueous suspension of tragacanth., the mice were killed by cervical dislocation 24 and 48 h later. The femora were removed from each animal and bone marrow smears were made on slides. After staining with May GrunwaldrGiemsa, 2000 PCEs ¨ per animal were scored for micronuclei. To demonstrate a cytotoxic effect due to the treatment with the test compound, the ratio between PCEs and normochromatic erythrocytes ŽNCEs. was determined in the same sample and reported as the number of NCEs per 2000 PCEs. The negative control group received the vehicle and positive controls were treated once with 30 mg CPArkg orally. 2.2.3. Guidelines The study was performed in compliance with the OECD Principles of Good Laboratory Practice and by following the OECD Guidelines for Testing of Chemicals, Section 4, No. 474, adopted May 26, 1983, Micronucleus Test. A compound is considered mutagenic in this test system if at any of the preparation intervals and dose groups, a relevant andror statistically significant increase in the number of micronucleated PCEs is found in comparison to the negative controls. 2.3. Statistics The statistical significance at the five per cent level Ž p ) 0.05. is normally evaluated by means of the nonparametric Mann–Whitney test. However, the obtained results did not require a statistical evaluation as the mean micronucleus frequencies after treatment with the test compound did not exceed the mean negative control values. 2.4. Kinetic studies For the determination of emodin blood plasma concentrations as a proof of intestinal absorption, satellite animals Ž4 males, 4 females. were treated with the same dose of 2000 mg emodinrkg. The animals were killed 1 or 3 h later by cervical dislocation and bled. The blood was centrifuged to separate
U. Mengs et al.r Mutation Research 393 (1997) 289–293
cellular matter and the plasma supernatant was aspirated and stored at y808C. 2.4.1. Chromatography HPLC separations were carried out on Nucleosil w 120-3C18 from Macherey-Nagel ŽDuren, Germany. ¨ at 408C. The m obile phase was aceto nitrilermethanolracetic acid Ž1 molrl., mixed in a time-programmed ratio from an initial ratio of 37:7:56 Ž0 to 4 min. to a final ratio of 60:7:33 Ž6 to 13 min. with a linear increase of the acetonitrile proportion between 4 and 6 min after injection. The eluent was delivered at a flow rate of 0.4 mlrmin. Fluorescence spectrophotometric detection was made at lex s 430 nm and lem s 530 nm. Retention times were approx. 12.5 and 13.2 min for emodin and danthron, respectively. 2.4.2. Sample preparation Ten ml mouse plasma was made up with water to 1.0 ml. Then 1.0 ml sodium acetate buffer and 100 ng danthronr20 ml methanol was added and the sample hydrolyzed with 50 ml b-glucuronidaser arylsulfatase at 378C for 15 to 18 h. The hydrolysate was acidified with 0.5 ml hydrochloric acid Žapprox. 2 molrl, prepared from the 37% solution., extracted with 5.0 ml di-isopropyl ether for 20 min and centrifuged at 2000 = g. The upper layer was aspirated, evaporated to dryness with nitrogen and the solid residue dissolved in 40 ml methanol, and 10 ml were injected. 2.4.3. Calibration and calculation Human blank plasma aliquots of 1.0 ml were diluted to a final volume of 2.0 ml with sodium acetate buffer, spiked with 5 to 2500 ng emodinr50
291
ml methanol and 100 ng danthronr20 ml methanol, and hydrolysed with 50 ml b-glucuronidaser arylsulfatase at 378C overnight. After extraction, centrifugation and reconditioning to 40 ml, 10 ml were injected for separation. For calculation, peak–height ratios were used. There was a linear dependence of the relative peak–height and the concentration of emodin. The lower limit of quantitation was approx. 5 ng emodinrml plasma.
3. Results 3.1. Micronucleus test The results are given in Table 1. Following a single oral dose of 2000 mg emodinrkg, there was no statistically significant enhancement in the frequency of micronucleated PCEs in comparison to the negative controls at both preparation intervals. Further, the mean number of NCEs was not substantially increased as compared to the mean values of NCEs of the negative controls, indicating that emodin had no cytotoxic effects on bone marrow cells. CPA, used as a reference mutagen Žpositive control., significantly increased the micronucleus frequency at an oral dose of 30 mgrkg. 3.2. Kinetic studies Individual blood plasma concentrations of emodin are shown in Table 2. The aim of the kinetic investigations was to demonstrate that emodin could reach the target cells under the experimental conditions described. After a
Table 1 Summary of the results from the micronucleus assay Group
Preparation time Žh p.a..
Number of PCEs scored
Negative control Emodin Ž2000 mgrkg. Positive control
24 24 48 24
2000 2000 2000 2000
a
PCEs with micronuclei Ž%.
PCErNCE
0.08 0.07 0.04 1.35
1.15 1.11 1.09 1.14
PCEs: polychromatic erythrocytes; NCEs: normochromatic erythrocytes; negative control: tragacanth; positive control: cyclophosphamide ŽCPA.; p.a.s post applicationem. a 2000 PCEs were scored of each 5 males and 5 females.
U. Mengs et al.r Mutation Research 393 (1997) 289–293
292
Table 2 Individual concentration values of total emodin in mouse blood plasma Animal no.
Gender
1 2 3 4 5 6 7 8
m m f f m m f f
Total emodin 1 h p.a.
a
Žmgrml. 3 h p.a.
97 130 139 190 173 82 119 73
a
Total emodin means the sum of free Žunconjugated. and conjugated emodin. p.a.s post applicationem; m s male, f s female.
single dose of 2000 mg emodinrkg, the mice showed plasma concentrations ranging from 73 to 190 mgrml within 3 h.
4. Discussion Anthraquinones are widely distributed in medicinal plants, being mainly in use for the treatment of constipation. It is very important to clarify their genotoxic or carcinogenic potential, because several compounds of this chemical class have been found to be mutagenic in vitro, e.g. aloeemodin, emodin, chrysophanol, physcion and others w3,5x. An intensive discussion on the safety of anthraquinone containing laxatives began when the positive findings of two long-term studies with the laxative danthron on rodents had been published w1,2x. On the basis of these findings, it has generally been thought that anthraquinones are tumorigenic in humans w3,10x. However, with the exception of danthron, this conclusion has been drawn only from in vitro assays. This is also the situation with emodin, being the main active principle of frangula drugs. Emodin caused mutation in the Ames test and in the V79HPRT assay and induced UDS in primary rat hepatocytes. Furthermore, emodin showed transforming activity in C3HrM2 mouse fibroblasts in vitro and induced micronuclei in mouse lymphoma L51784Y cells in the absence of metabolic activation w3,11x. However, there were no data available from the
literature showing that a genotoxic activity would also exist under in vivo conditions. We therefore conducted the micronucleus assay in mice to present the first in vivo data on emodin. The dose level used was 2000 mg emodinrkg, representing the maximum guideline-recommended dose and the maximum feasible dose for technical reasons. Emodin failed to induce micronuclei in PCEs of the bone marrow under the conditions described. The fact that plasma levels of emodin Žup to 190 mgrml. were found at two time points after administration Ž1 and 3 h p.a.., being multiples Ž) factor 10. of the minimum genotoxic concentration in vitro w3,11x, demonstrates that the substance must have been in contact with the bone marrow cells. However, the dose of 2000 mgrkg did not induce clear cytotoxic effects as determined by the PCErNCE ratio. The micronucleus data of this study did not indicate a clastogenic effect, either. This lends support to the assumption that data in a recent publication on the clastogenic potential of emodin in vitro w11x might be a special reaction of the mouse cell line L51784Y. Further, it should be kept in mind that our negative in vivo data were obtained under normal mammalian metabolic conditions while the in vitro study was performed under cell culture conditions and without metabolic additives. A similar situation is also present for aloeemodin which is mutagenic in vitro, too w3x, but failed to give positive responses in several in vivo assays including the micronucleus assay, chromosome aberration assay, mouse spot test, and UDS test ex vivo w5x. These results again demonstrate that in vitro data may not be a sufficient basis for the assessment of a possible genotoxic risk for humans. In conclusion, emodin did not induce micronuclei and is therefore considered to be non-genotoxic under the in vivo conditions described. The results further show that in the case of anthraquinones, risk assessments regarding genotoxicity should include valid in vivo results and relevant kinetic data such as blood plasma concentrations. References w1x H. Mori, S. Sugie, K. Niwa, T. Takahashi, K. Kawai, Induction of intestinal tumours in rats by chrysazin, Br. J. Cancer 52 Ž1985. 781–783.
U. Mengs et al.r Mutation Research 393 (1997) 289–293 w2x H. Mori, S. Sugie, K. Niwa, N. Yoshimi, T. Tanak, I. Hirono, Carcinogenicity of chrysazin in large intestine and liver of mice, Jpn. J. Cancer Res. 77 Ž1986. 871–876. w3x J. Westendorf, H. Marquardt, B. Poginsky, M. Dominiak, J. Schmidt, H. Marquardt, Genotoxicity of naturally occurring hydroxyanthraquinones, Mutat. Res. 240 Ž1990. 1–12. w4x A. Heidemann, H.G. Miltenburger, U. Mengs, The genotoxicity status of senna, Pharmacology 47 Žsuppl. 1. Ž1993. 178–186. w5x A. Heidemann, W. Volkner, U. Mengs, Genotoxicity of ¨ aloeemodin in vitro and in vivo, Mutat. Res. 367 Ž1996. 123–133. w6x F.C. Wehner, P.G. Thiel, M. du Rand, Mutagenicity of the mycotoxin emodin in the Salmonellarmicrosome system, Appl. Environ. Microbiol. 37 Ž1979. 658–660. w7x L. Tikkanen, T. Matsushima, S. Natori, Mutagenicity of anthraquinones in Salmonella preincubation test, Mutat. Res. 116 Ž1983. 197–304.
293
w8x H. Morita, M. Umeda, T. Masuda, Y. Ueno, Cytotoxic and mutagenic effects of emodin on cultured mouse carcinoma FM3A cells, Mutat. Res. 204 Ž1988. 329–332. w9x S. Krivobok, F. Seigle-Murandi, R. Steiman, D.R. Marzin, V. Betina, Mutagenicity of substituted anthraquinones in the AmesrSalmonella microsome system, Mutat. Res. 279 Ž1992. 1–8. w10x H. Helmholz, A. Ruge, A. Piasecki, S. Schroder, J. Westen¨ dorf, Genotoxizitat ¨ der Faulbaumrinde, Pharmaz. Z. 43 Ž1993. 3478–3480. w11x S.O. Muller, I. Eckert, W.K. Lutz, H. Stopper, Genotoxicity ¨ of the laxative drug components emodin, aloe-emodin and danthron in mammalian cells: Topoisomerase II mediated?, Mutat. Res. 371 Ž1996. 165–173. w12x W. Grimminger, K. Witthohn, Analytics of senna drugs with regard to the toxicological discussion of anthranoids, Pharmacology 47 Žsuppl. 1. Ž1993. 98–109.