- Email: [email protected]
Characterization of the genotoxicity of anthraquinones in mammalian cells
Biochimica et Biophysica Acta 1428 (1999) 406^414 www.elsevier.com/locate/bba
Characterization of the genotoxicity of anthraquinones in mammalian cells Stefan O. Mueller
1;
*, Helga Stopper
Department of Toxicology, University of Wu«rzburg, 97078 Wu«rzburg, Germany Received 9 March 1999; accepted 21 April 1999
Abstract Naturally occurring 1,8-dihydroxyanthraquinones are under consideration as possible carcinogens. Here we wanted to elucidate a possible mechanism of their genotoxicity. All three tested anthraquinones, emodin, aloe-emodin, and danthron, showed capabilities to inhibit the non-covalent binding of bisbenzimide Hoechst 33342 to isolated DNA and in mouse lymphoma L5178Y cells comparable to the topoisomerase II inhibitor and intercalator m-amsacrine. In a cell-free decatenation assay, emodin exerted a stronger, danthron a similar and aloe-emodin a weaker inhibition of topoisomerase II activity than m-amsacrine. Analysis of the chromosomal extent of DNA damage induced by these anthraquinones was performed in mouse lymphoma L5178Y cells. Anthraquinone-induced mutant cell clones showed similar chromosomal lesions when compared to the topoisomerase II inhibitors etoposide and m-amsacrine, but were different from mutants induced by the DNA alkylator ethyl methanesulfonate. These data support the idea that inhibition of the catalytic activity of topoisomerase II contributes to anthraquinone-induced genotoxicity and mutagenicity. ß 1999 Elsevier Science B.V. All rights reserved. Keywords: Anthraquinone; L5178Y mouse lymphoma cell ; Topoisomerase II; Loss of heterozygosity
1. Introduction Naturally occurring anthracene derivatives, 1,8-dihydroxyanthraquinones (Fig. 1) are pharmacologically active laxatives. A carcinogenic potential was reported for 1,8-dihydroxyanthraquinone, danthron (Fig. 1) in rodents [1,2]. However, for other more abundantly occurring anthraquinone derivatives
* Corresponding author. Fax: +1-919-541-7935; E-mail: [email protected] 1 Present address: National Institute of Environmental Health Sciences, MD E4-01, PO Box 12233, Research Triangle Park, NC 27709, USA.
even antimutagenic properties were found [3]. Previously, we showed a clear-cut DNA damaging potency for this class of compound [4]. According to their supposed planar structure, Swanbeck [5] postulated that they intercalate into DNA. A more recent study described spectroscopic evidence for the intercalation of anthraquinone derivatives into calf thymus DNA [6]. The widely used anti-cancer drugs of the anthracycline type are derived from the anthraquinone structure. They interact via intercalation into DNA with topoisomerase (topo) II and inhibit its catalytic function [7^9]. In a previously published study, we could show that three 1,8-dihydroxyanthraquinones, emodin, aloe-emodin, and danthron (Fig. 1) inhibited topo II activity [4]. Here, we quantitatively determined
0304-4165 / 99 / $ ^ see front matter ß 1999 Elsevier Science B.V. All rights reserved. PII: S 0 3 0 4 - 4 1 6 5 ( 9 9 ) 0 0 0 6 4 - 1
BBAGEN 24822 27-7-99
S.O. Mueller, H. Stopper / Biochimica et Biophysica Acta 1428 (1999) 406^414
their potency to inhibit topo II function. This catalytic inhibition of the topo II could be mediated by a competition of DNA binding sites of the topo II, as it is seen for the anthracyclines. Therefore, we investigated the potential of the anthraquinones to compete with the DNA dye bisbenzimide Hoechst 33342 for binding sites of isolated DNA and in intact mouse lymphoma L5178Y cells. Inhibition of topo II can lead to chromosomal DNA breakage [10,11]. Therefore, we further investigated whether the topo II inhibiting qualities are in agreement with the type and chromosomal extent of the mutagenic DNA damage induced by these compounds. For this purpose we analyzed mutants of L5178Y mouse lymphoma cells using a recently developed LOH-PCR method [12,13].
407
2. Materials and methods
33342 and incubated until £uorescence equilibrium was reached (30 min for calf thymus DNA; 4 h in living cells). Experiments were initiated by addition of the respective compound and the decrease of £uorescence was monitored over time. The excitation wavelength was set at 347 nm and the emission wavelength at 454 nm. Total £uorescence of cells with Hoechst 33342 in equilibrium was about 100 £uorescence units (FU). Non-speci¢c £uorescence of the compounds was always less than 1 FU. Concentrations at which the £uorescence of DNA bound Hoechst 33342 was reduced to 50% (IC50 ) were determined in at least three independent experiments using concentrations of the test compounds including the respective range of IC50 values. The viability and membrane integrity of the cells was determined with the £uoresceindiacetat/ethidiumbromide method as described [16] and was between 70 and 85%.
2.1. Nucleic acids, chemicals and cell culture
2.3. Topoisomerase II assay
Calf thymus DNA was purchased from Serva (Heidelberg, Germany). Aloe-emodin, danthron, emodin, ethyl methanesulfonate (EMS), m-amsacrine (m-AMSA), etoposide and bisbenzimide Hoechst 33342 were purchased from Sigma (Deisenhofen, Germany). Mouse lymphoma L5178Y cells were cultured in suspension in RPMI-1640 supplemented with antibiotics, 0.25 mg/ml L-glutamine, 107 Wg/ml sodium pyruvate, and 10% heat inactivated horse serum (all from Sigma, Deisenhofen, Germany). Cell cultures were grown in a humidi¢ed atmosphere with 5% CO2 in air at 37³C.
Topoisomerase II assays were performed according to the manufacturers' protocol (TopoGen, Columbus, OH, USA). The reactions contained 150 ng kDNA, 4 U of topo II (170 kDa form) and the test chemicals (dissolved in DMSO, ¢nal concentration 10%) in 20 Wl reaction bu¡er (30 mM Tris-HCl pH 7.6, 3 mM ATP, 15 mM 2-mercaptoethanol, 8 mM MgCl2 , 60 mM NaCl). The reactions were incubated for 15 min at 37³C and terminated with 1% of SDS, followed by proteinase K (20 Wg/ml) treatment for 15 min at 37³C. After addition of 0.1 vol. gel loading bu¡er (0.25% bromphenol blue, 50% glycerol), the samples were extracted once with an equal volume of chloroform:isoamylalcohol (24:1). The blue upper layer was loaded onto a 1% agarose gel. The products were resolved by gel electrophoresis in 1UTAE bu¡er for 30 min at 100 V, which separated the catenated kDNA from the decatenated DNA monomers. Gels were stained with SYBR Green (Molecular Probes Europe, The Netherlands) and photographed. The inhibition of the catalytic activity of topo II was quanti¢ed by measuring the reduction of the amount of the monocircle DNA band (corresponding to the product of the topo II catalytic reaction) in relation to the network DNA band (corresponding to the substrate of the
2.2. DNA competition assay with isolated DNA and intracellular accumulation This method is based on the elevated £uorescence of Hoechst 33342 when it binds non-covalently to DNA [14,15]. The addition of other DNA binding compounds causes a reduction of this £uorescence, presumably because of competition of binding sites of the DNA. Fluorescence was monitored by a Kontron spectro£uorometer in 1U1 cm cuvettes under continuous stirring at 37³C. Five hundred ng calf thymus DNA or a suspension of 106 L5178Y cells/ ml in 2 ml PBS/G were mixed with 2 WM Hoechst
BBAGEN 24822 27-7-99
408
S.O. Mueller, H. Stopper / Biochimica et Biophysica Acta 1428 (1999) 406^414
topo II). The program NIH Image 1.54 (NIH, USA) was used to measure the intensities of the gel bands. 2.4. Generation of cell mutants in the mouse lymphoma mutation assay Mutation assays were performed with the in situ procedure [17]. Cultures of mouse L5178Y cells were treated with methotrexate before each experiment to kill pre-existing tri£uorthymidine (TFT)-resistant cells. To accomplish this, cells were incubated for 12 h in culture medium plus methotrexate (0.3 Wg/ ml), thymidine (9 Wg/ml), hypoxanthine (15 Wg/ml) and glycine (22.5 Wg/ml). The cells were then incubated for at least 48 h in the same medium without methotrexate. To measure chemically induced mutants, cultures containing 1U106 cells in 5 ml medium were treated with DMSO (¢nal concentration 1%; vehicle control) or with the respective test compounds with pretested non-cytotoxic concentrations (20^100 WM). EMS was used as a positive control. Incubation was performed for 4 h in media with reduced serum content (1% v/v), then the cells were washed twice with fresh medium. After that,
0.5U106 cells from each tested culture were added to 50 ml of semi-solid culture medium containing 0.25% granulated agar (Baltimore Biological Laboratories, USA) and plated into two plastic 100 mm culture dishes and allowed to solidify at room temperature. TFT-resistant cells were selected by adding an overlay of TFT (¢nal concentration 8 Wg/ml) in 10 ml semi-solid medium after an expression time of 42 h. Cloning e¤ciency was determined by adding 600 cells to 100 ml of semi-solid medium in three 100 mm culture dishes. All plates were incubated for a total of 9^12 days at 37³C in 5% CO2 for colony growth. An automatic colony counter was used to count the number of TFT-resistant colonies. For cultivation of mutant clones, replica plates were analyzed 7^8 days after TFT selection with an invertmicroscope under sterile conditions. Mutant cell clones were checked visibly for necrosis and intact clones were picked randomly of the plates with a sterile pasteur pipette and cultivated until the cell culture reached a density of 1^2U107 cells in 20 ml medium (5^10 days). TFT resistance was checked by incubating an aliquot of each cell clone culture with TFT. Only clones which showed TFT resistance were used for LOH analysis.
Fig. 1. Chemical structures of investigated compounds.
BBAGEN 24822 27-7-99
S.O. Mueller, H. Stopper / Biochimica et Biophysica Acta 1428 (1999) 406^414
2.5. Isolation of genomic DNA Genomic DNA of the mutant clones for the PCR analysis was prepared using the QIAamp BloodKit from Qiagen (Dassel, Germany). The cell cultures (1^ 2U107 cells/20 ml medium) were centrifuged, washed twice with PBS/CMF and the pellet resuspended in 200 Wl sterile water. DNA isolation was performed according to the Qiagen protocol with 200 Wl cell suspension. Isolated DNA was stored in TE 10/1 (10 mM Tris, pH 8.0, 1 mM EDTA) at 320³C and diluted for PCR analysis. 2.6. Molecular analysis of mutants by LOH-PCR Primers for the microsatellite sequences were obtained from Research Genetics (Huntsville, AL, USA) and used as a 6.6 WM solution. Thirty ng DNA from each mutant clone was ampli¢ed with four characteristic primer pairs as shown in Fig. 4. As controls, DNA from wild-type tk=3 L5178Y mouse lymphoma cells and from a mutant that is known to show LOH along the entire chromosome 11+ was used. Twenty Wl reactions for the PCR ampli¢cation were performed. Ten Wl 2UPCR Master (20 mM Tris-HCl, 100 mM KCl, 3 mM MgCl2 , 0.01% Brij 35, 400 WM dATP, dCTP, dGTP, dTTP each, 0.05 U/Wl Taq DNA polymerase; Boehringer Mannheim, Germany) was mixed with 3.3 WM of each primer and 30 ng template DNA. PCR was performed in a PTC100 (MJ Research, USA). To avoid unspeci¢c primer binding a touch-down PCR was carried out. For the primer pairs Mit2, Mit59 and Mit49 the following parameters were used: initial denaturation at 94³C for 2 min and 20 s followed by 20 cycles with 30 s, 94³C, 30 s, 61³C (touch down 30.5³C/cycle) and 20 s, 72³C. Then 10 additional cycles with the parameters 30 s, 94³C, 30 s, 50³C and 20 s, 72³C were performed followed by a ¢nal extension for 5 min at 72³C. For the primer pair Agl2 the following parameters were used: initial denaturation at 94³C for 2 min and 20 s followed by 14 cycles with 30 s, 94³C, 30 s, 72³C (touch down 30.5³C/cycle) and 25 s, 72³C. Then 15 additional cycles with the parameters 30 s, 94³C, 30 s, 65³C and 25 s, 72³C were performed followed by a ¢nal extension for 5 min at 72³C. PCR samples were stored at 4³C. Ten Wl of the reaction mixture were
409
loaded on 10% polyacrylamide gels (1UTBE) in 10% loading bu¡er (0.1% bromphenolblue, 50% glycerol). Electrophoresis was performed for 4^8 h at constant 200 V. A 50^2000 bp DNA ladder (Gibco, Germany) was used to assess the fragment lengths of the detected bands. Gels were stained with SYBR Green (Molecular Probes, The Netherlands) and photographed. 3. Results 3.1. Intracellular accumulation of anthraquinones We ¢rst analyzed whether the anthraquinones, emodin, aloe-emodin and danthron are able to compete with the non-covalent DNA binding compound bisbenzimide Hoechst 33342 (H33342) for DNA binding sites. We employed an indirect competition assay with isolated calf thymus DNA, measuring the quenching of the £uorescence of non-covalent DNA bound H33342 by the test compounds [18]. The wellknown DNA intercalator and topo II poison m-amsacrine (m-AMSA) (Fig. 1) [19] was used as a positive control. Etoposide (Fig. 1) as a topo II poison and a non-DNA binding compound was used as a negative control and showed no detectable e¡ects. The data were determined by using linear regression analysis. Concentration ranges of the respective compound including the dose resulting in a 50% reducTable 1 IC50 values of di¡erent compounds for the non-covalent DNA binding of Hoechst 33342 to DNA in solution (500 ng/ml calf thymus DNA) and in mouse lymphoma cells (1U106 cells/ml) Compound
IC50 (WM) with DNA in IC50 (WM) in intact solution cells
Emodin Aloe-emodin Danthron m-AMSA Etoposide
12 þ 4 11 þ 3 14 þ 4 10 þ 1 ra
23 þ 2 23 þ 1 28 þ 5 22 þ 3 ra
Data represent a compilation of the results shown in part in Fig. 2 obtained by linear regression analysis. IC50 values represent the concentrations at which the £uorescence of DNA bound Hoechst 33342 was reduced to 50%. Data are means þ S.D. (Nvthree independent experiments). a r means no inhibition measurable at any tested concentration (IC50 E500 WM (highest concentration tested)).
BBAGEN 24822 27-7-99
410
S.O. Mueller, H. Stopper / Biochimica et Biophysica Acta 1428 (1999) 406^414
Fig. 2. Inhibition of £uorescence of non-covalent DNA bound Hoechst 33342 by the investigated compounds in mouse lymphoma L5178Y. Tests were performed as described in Section 2. Arrow indicates the timepoint at which the respective compound (25 WM each) was added to the cell suspension (1U106 /ml PBS/G) after a steady state of £uorescence of DNA-bound Hoechst 33342 (2 WM; set to 100%) was reached.
tion of £uorescence (IC50 ) were used (data not shown). This measure re£ects the capability of the respective compound to inhibit the non-covalent binding of H33342 to DNA in solution (calf thymus DNA; ¢rst column in Table 1). The anthraquinones showed IC50 values similar to m-AMSA. To analyze whether the anthraquinones are also capable of reaching the DNA in intact cells we measured IC50 values with mouse lymphoma L5178Y cell suspensions, using H33342, readily capable to reach DNA in intact cells [15]. An intact cellular membrane is a prerequisite to assess a potential intracellular accumulation under nearby physiological conditions. Therefore, the £uoresceine/ethidiumbromide assay [16] was employed to control their viability. Only assays were analyzed in which the viability was more than 70%. In Fig. 2 we show a typical experiment with mouse lymphoma L5178Y cells. InC
Fig. 3. Decatenation assay for topoisomerase II activity and its inhibition. Assays were performed as described in Section 2. All lanes were loaded with 150 ng kDNA. Lanes 2^14: after incubation with 4 U topoisomerase II. Lanes 1 and 2: without inhibitor ; lanes 3^5: +100, 50, and 10 WM m-AMSA; lanes 6^8: +10, 1, 0.1 WM emodin; lanes 9^11: +1000, 750 and 500 WM aloe-emodin; lanes 12^14: +100, 50 and 5 WM danthron. NW: network DNA; MC: monocircle DNA.
BBAGEN 24822 27-7-99
S.O. Mueller, H. Stopper / Biochimica et Biophysica Acta 1428 (1999) 406^414
411
Fig. 4. Occurrence of LOH in anthraquinone-induced L5178Y mouse mutants. Cells were treated with emodin (40 WM), aloe-emodin (100 WM) and danthron (100 WM) as described in Section 2. On the left, localizations of investigated loci on chromosome 11 harboring the selectable intact tk allele (11 ) are shown. Distances of the loci are given distal to the centromere in centiMorgan (cM). Shown are the total numbers of mutants showing LOH at the indicated loci and in parentheses the respective percentage. Large LOH is de¢ned as LOH occurring at all loci including Mit2. Medium-sized LOH is de¢ned as LOH at loci Mit59 and/or Mit49 and/or Agl2. Mutants without any LOH at the investigated loci are given as intragenic mutants. Total numbers of analyzed mutant clones are given on the right.
cubation of the cells with Hoechst 33342 resulted in an increased, stable £uorescence (increased £uorescence was reached within 5^10 min, a complete occupation of binding sites resulting in a steady-state was reached within 3^4 h). The addition of the compounds was followed by a rapid decrease of H33342 DNA bound £uorescence (Fig. 2). As for DNA in solution, IC50 values for the anthraquinones were in the same range as for m-AMSA (Table 1 second column).
DNA in intact cells could also result in an inhibition of topo II activity. To test this hypothesis we assayed enzymatic inhibition by monitoring the topo II cataTable 2 IC50 values of di¡erent compounds for the activity of topoisomerase II Compound
IC50 (WM)
3.2. Inhibition of topoisomerase II
Emodin Aloe-emodin Danthron m-AMSA
7þ1 741 þ 272 61 þ 18 63 þ 7
Many known topo II inhibitors, including anthracycline derivatives bind non-covalently to DNA. This non-covalent DNA binding is not a prerequisite for e¤cient topo II poisoning or inhibition [10], but it probably leads to a competition with the topo II for DNA binding sites. Therefore, the observed capabilities of emodin, danthron, and aloe-emodin to reach
Data represent a compilation of the results shown in part in Fig. 3. IC50 values were analyzed by linear regression analysis. They represent the concentrations at which the catalytic activity of the topo II was reduced to 50%. IC50 values were calculated by measuring the integrated density of the topo II product in the gels (band MC in Fig. 3) relative to the integrated density of the substrate kDNA and product band (band MC+band NW in Fig. 3). Data are means þ S.D. (Nvthree independent experiments).
BBAGEN 24822 27-7-99
412
S.O. Mueller, H. Stopper / Biochimica et Biophysica Acta 1428 (1999) 406^414
lyzed decatenation of kinetoplast DNA (kDNA), a catenated network of mitochondrial DNA rings (NW, network DNA, Fig. 3) isolated from Crithidia fasciculata. The decatenation reaction was monitored by the appearance of 2.5 kilobase DNA monomers following gel electrophoresis (MC, monocircle DNA, Fig. 3). Again, m-AMSA was used as a positive control. As we have shown before [4], emodin, danthron, and aloe-emodin reduced the amount of monomer DNA generated by topo II, indicating that all three compounds were capable of inhibiting this enzyme. To quantify the inhibition of topo II, we measured dose-dependent reduction of the product band (MC) in relation to the substrate band (NW; Fig. 3) by measuring integrated densities using image analysis as described [4] and calculated the IC50 values (Table 2). Concentration ranges including the supposed IC50 values were determined before (data not shown). Emodin exhibited even greater topo II inhibiting qualities than m-AMSA. Danthron showed similar activity to m-AMSA and aloe-emodin was much less active.
(repetitive sequences) are often heteromorphic between homologous chromosomes. In this case, the homologous chromosomes can be discriminated after PCR ampli¢cation of these sequences following gel electrophoresis as two DNA fragments di¡erent in length. LOH can be detected as the loss of one of the two DNA fragments [12]. The used heteromorphic loci are located at 2 centiMorgan (cM) (Mit2), 61 cM (Mit59), 77 cM (Mit49) and 78 cM (Agl2; within the tk gene) distal to the centromere (Fig. 4). If a mutant clone does not show any LOH, the lesion is located within the tk gene, possibly including sequences distal to tk, for which heteromorphic microsatellite sequences are unknown (Fig. 4, `intragenic mutation'). In Fig. 5 the results for the primer pair Mit59 are given. A compilation of the results obtained at all investigated loci is given in Fig. 4. Emodin and aloe-emodin yielded no intragenic mutations. For a better understanding, we de¢ned LOH of 2.5^24% of the chromosomal length from the te-
3.3. Analysis of mouse lymphoma L5178Y mutants by LOH-PCR Next, we asked whether the above described modes of action are also re£ected by the extent of the induced chromosomal damage. Therefore, we applied a recently developed PCR based method to analyze mutant clones of mouse lymphoma L5178Y cells. L5178Y cells are heterozygous at the thymidinekinase gene (tk=3 ) on chromosome 11 [20]. Inactivation of the intact tk gene (tk ) leads to a loss of heterozygosity (LOH). This LOH results in resistance of the mutated cells against the selective agent tri£uorthymidine (TFT). We analyzed mutants induced by emodin, aloe-emodin, and danthron obtained at non-cytotoxic concentrations [4]. The relative mutation frequencies per 106 cells were 2.8 for 40 WM emodin (relative cloning e¤ciency was 0.5, emodin was cytotoxic at higher concentrations), 3.6 for 100 WM aloe-emodin (relative cloning e¤ciency was 1) and 6.2 for 100 WM danthron (relative cloning e¤ciency was 0.7). The qualitative investigation of the mutagenic lesions on mouse chromosome 11 of these mutants (tk3=3 or tk3 ) was performed by an analysis of the LOH using PCR [13]. DNA microsatellites
Fig. 5. Gel electrophoresis of a LOH-PCR analysis at Mit59. Analyses were performed as described in Section 2. All lanes were loaded with 20 Wl of each PCR sample. Lanes 1^4: mutants induced by emodin; lanes 5^8: mutants induced by aloeemodin; lane 9: control with heterozygous tk=3 (wild-type, wt; no LOH); lane 10: control with LOH (LOH at all loci); lanes 11^14: mutants induced by danthron. Given is the respective number of the analyzed cell mutant. Lane 15: 200^2000 base pair (bp) DNA ladder (marker, M). Mutants with LOH at Mit59 show only one band with an approximate length of 240 bp, the band with an approximate length of 254 bp is lost (indicated by arrows; lanes 4, 5, 8 and 10). NS, non-speci¢c ampli¢cation products.
BBAGEN 24822 27-7-99
S.O. Mueller, H. Stopper / Biochimica et Biophysica Acta 1428 (1999) 406^414
lomere (LOH at Agl2, Mit49 and/or Mit59) as `medium-sized' LOH. Emodin and aloe-emodin yielded predominantly `medium-sized' LOH, 88% and 70%, respectively. LOH spanning at least 98% of the chromosome length (including Mit2) is de¢ned as `large' LOH. This was observed for 12% of emodin mutants and 29% of aloe-emodin mutants. Danthron yielded 40% intragenic mutations, 20% `medium-sized' and 40% `large' LOH. 4. Discussion All three tested anthraquinones, emodin, aloeemodin, and danthron inhibited non-covalent binding of H33342 to isolated DNA. Furthermore, these compounds showed strong capabilities to reach the DNA in intact cells. Bogush et al. [18] used the quenching of the £uorescence of anthracyclines when they intercalate into DNA as a direct measure for their DNA binding a¤nity. These data could also be con¢rmed by an indirect competition assay. We applied this indirect competition assay here, since the tested anthraquinones are not £uorescent. However, the reported IC50 values should not be used as a measure for the respective compounds' DNA binding a¤nity. The obtained data re£ect rather the intracellular accumulation and capability of emodin, aloe-emodin and danthron to inhibit the non-covalent DNA binding of H33342. Furthermore, we could show that the investigated anthraquinones exhibited strong topo II inhibiting qualities. It is of note, that the proposed enzyme inhibition is di¡erent from a topo II poisoning. Poisoning of topo II occurs via stabilization of the socalled ternary cleavable complex formed by DNA, topo II and drug (e.g. m-AMSA or etoposide) [7]. This interaction with the topo II results in enzymemediated DNA cleavage [7] and is thus assumed to confer to the compound's genotoxicity characterized by chromosomal deletions [10,11,21]. In contrast, an overall inhibition of the enzymes catalytic activity can be caused, for example, by prevention of the binding of topo II to DNA [7,10]. Therefore, we assume that the shown capability of the investigated anthraquinones to reach DNA in intact cells leads to the observed topo II inhibition. To analyze whether this proposed mechanism could also be relevant for
413
the anthraquinone-induced mutagenicity [4], we investigated their mutation pro¢les. Emodin and aloe-emodin yielded no intragenic mutations and predominantly (around 80%) `medium-sized' LOH. Danthron showed 40% intragenic mutations, 20% `medium-sized' and 40% `large' LOH. Spontaneously occurring mouse lymphoma L5178Y cell mutants (cells treated with the solvent DMSO only, 1% v/v) were analyzed by Liechty et al. (personal communication). These spontaneous mutants showed 30^40% LOH for each de¢ned category. In comparison, we found for etoposide (43 nM) and m-AMSA (5 nM) 0% (etoposide) and 6% (m-AMSA) intragenic mutations. m-AMSA showed predominantly `mediumsized' LOH (77%), whereas `large' LOH occurred in only 18% of the m-AMSA-induced mutants. In contrast, etoposide showed higher percentages of `large' (47%) and lower percentages for `mediumsized' LOH (53%) when compared to m-AMSA. The alkylator ethyl methanesulfonate (2 mM, EMS) showed 94% intragenic mutations, 6% `medium-sized' LOH and no `large' LOH. Thus, the investigated anthraquinones showed considerably more similarity to the topo II poisons m-AMSA and etoposide than to EMS as an intragenic mutagen. Furthermore, except for danthron there was a marked di¡erence to the background (solvent control) LOH, as DMSO yielded similar LOH at all investigated loci. One possible explanation for the di¡erent extent of LOH (`medium-sized' vs. `large' LOH) induced by various topo II inhibitors, including these anthraquinone, could be the distribution of topo II along the chromosome during mitosis. In the presence of topo II poisons like etoposide a maximum of the topo II content is found at the centromere [22]. Therefore, in this case chromosomal mutations (lesions) are expected close to the centromere, i.e. LOH at Mit2 (`large' LOH) as it is found for etoposide. The topo II concentration at the centromere is reduced in the presence of DNA intercalators. m-AMSA acts at high concentrations as an intercalator, rather than as a topo II poison [22]. This could possibly re£ect the found reduced LOH near the centromere. We could observe a similar LOH pattern for emodin and aloe-emodin, but not for danthron. Another hypothesis was postulated by Gaulden [23] who stated that an interference with topo II could lead to `chro-
BBAGEN 24822 27-7-99
414
S.O. Mueller, H. Stopper / Biochimica et Biophysica Acta 1428 (1999) 406^414
mosome stickiness' at mitosis, resulting in chromatin bridges and later in large chromosomal deletions. Such a `chromosome stickiness' was reported to occur at high concentrations of m-AMSA (about 10 nM) [23] and would be in accordance with the mutation pro¢le shown here. We also observed chromatin bridges after treatment with the investigated anthraquinones in mammalian cells (unpublished observation). Overall, emodin and aloe-emodin, but not danthron showed chromosomal deletion patterns more similar to the intercalator m-AMSA than to the non-DNA binding topo II poison etoposide. In conclusion, the presented data are in agreement with the idea that an indirect inhibition of the catalytic activity of topoisomerase II contributes to the genotoxic e¡ects induced by 1,8-dihydroxyanthraquinones. However, these data do not exclude other possible mechanisms, especially generation of reactive intermediates by redoxcycling [9,24], that may contribute to the genotoxicity induced by anthraquinones. Acknowledgements This work was supported by the Swiss Federal O¤ce of Public Health (BAG Grant number FE 316.95.0500). We thank Dr. W.K. Lutz, head of the Department of Toxicology, University of Wu«rzburg and Dr. J. Schlatter, monitoring scientist of the BAG, for valuable advice in all parts of this work and Ms. M. Gerhard for her expert technical assistance. References
[2] H. Mori, S. Sugie, K. Niwa, N. Yishimi, T. Tanaka, I. Hirono, Jpn. J. Cancer Res. 77 (1986) 871^876. [3] H.Y. Su, S.H. Cheng, C.C. Chen, H. Lee, Mutat. Res. 329 (1995) 205^212. [4] S.O. Mueller, I. Eckert, W.K. Lutz, H. Stopper, Mutat. Res. 371 (1996) 165^173. [5] G. Swanbeck, Biochim. Biophys. Acta 123 (1966) 630^ 633. [6] D.T. Breslin, G.B. Schuster, J. Am. Chem. Soc. 118 (1996) 2311^2319. [7] S.J. Froelich-Ammon, N. Oshero¡, J. Biol. Chem. 270 (1995) 21429^21432. [8] L.F. Liu, Annu. Rev. Biochem. 58 (1989) 351^375. [9] J.W. Lown, Pharmacol. Ther. 60 (1993) 185^214. [10] L.R. Ferguson, B.C. Baguley, Mutat. Res. 355 (1996) 91^ 101. [11] R.D. Anderson, N.A. Berger, Mutat. Res. 309 (1994) 109^ 142. [12] M.C. Liechty, Z. Hassanpour, J.C. Hozier, D. Clive, Mutagenesis 9 (1994) 423^427. [13] W. Caspary, H. Stopper, L. Davis, J. Hozier, M. Liechty, in: H.K. Mu«ller-Hermelink, H.-G. Neumann, W. Dekant, (Eds.), Risk and Progression Factors in Carcinogenesis, Vol. 143, Springer, Berlin, 1997, pp. 161^182. [14] T. Bogush, G. Smirnova, I. Shubina, A. Syrkin, J. Robert, Cancer Chemother. Pharmacol. 35 (1995) 501^505. [15] F. Belloc, F. Lacombe, P. Dumain, F. Lopez, P. Bernard, M.R. Boisseau, J. Reifers, Cytometry 13 (1992) 880^885. [16] G.H.S. Strauss, Mutat. Res. 252 (1991) 1^15. [17] D.L. Spencer, C.K. Hines, W.J. Caspary, Mutat. Res. 312 (1994) 85^97. [18] T. Bogush, J. Robert, Anticancer Res. 16 (1996) 365^368. [19] A.H. Corbett, N. Oshero¡, Chem. Res. Toxicol. 6 (1993) 585^597. [20] M.C. Liechty, H.S. Rauchfuss, M.H. Lugo, J.C. Hozier, Mutat. Res. 286 (1993) 299^307. [21] L.R. Ferguson, B.C. Baguley, Environ. Mol. Mutagen. 24 (1994) 245^261. [22] A.T. Sumner, Chromosome Res. 4 (1996) 5^14. [23] M.E. Gaulden, Mutagenesis 2 (1987) 357^365. [24] M. Kodama, Y. Kamioka, T. Nakayama, C. Nagata, N. Morooka, Y. Ueno, Toxicol. Lett. 37 (1987) 149^156.
[1] H. Mori, S. Sugie, K. Niwa, M. Takahashi, K. Kawai, Br. J. Cancer 52 (1985) 781^783.
BBAGEN 24822 27-7-99
Characterization of the genotoxicity of anthraquinones in mammalian cells Stefan O. Mueller
1;
*, Helga Stopper
Department of Toxicology, University of Wu«rzburg, 97078 Wu«rzburg, Germany Received 9 March 1999; accepted 21 April 1999
Abstract Naturally occurring 1,8-dihydroxyanthraquinones are under consideration as possible carcinogens. Here we wanted to elucidate a possible mechanism of their genotoxicity. All three tested anthraquinones, emodin, aloe-emodin, and danthron, showed capabilities to inhibit the non-covalent binding of bisbenzimide Hoechst 33342 to isolated DNA and in mouse lymphoma L5178Y cells comparable to the topoisomerase II inhibitor and intercalator m-amsacrine. In a cell-free decatenation assay, emodin exerted a stronger, danthron a similar and aloe-emodin a weaker inhibition of topoisomerase II activity than m-amsacrine. Analysis of the chromosomal extent of DNA damage induced by these anthraquinones was performed in mouse lymphoma L5178Y cells. Anthraquinone-induced mutant cell clones showed similar chromosomal lesions when compared to the topoisomerase II inhibitors etoposide and m-amsacrine, but were different from mutants induced by the DNA alkylator ethyl methanesulfonate. These data support the idea that inhibition of the catalytic activity of topoisomerase II contributes to anthraquinone-induced genotoxicity and mutagenicity. ß 1999 Elsevier Science B.V. All rights reserved. Keywords: Anthraquinone; L5178Y mouse lymphoma cell ; Topoisomerase II; Loss of heterozygosity
1. Introduction Naturally occurring anthracene derivatives, 1,8-dihydroxyanthraquinones (Fig. 1) are pharmacologically active laxatives. A carcinogenic potential was reported for 1,8-dihydroxyanthraquinone, danthron (Fig. 1) in rodents [1,2]. However, for other more abundantly occurring anthraquinone derivatives
* Corresponding author. Fax: +1-919-541-7935; E-mail: [email protected] 1 Present address: National Institute of Environmental Health Sciences, MD E4-01, PO Box 12233, Research Triangle Park, NC 27709, USA.
even antimutagenic properties were found [3]. Previously, we showed a clear-cut DNA damaging potency for this class of compound [4]. According to their supposed planar structure, Swanbeck [5] postulated that they intercalate into DNA. A more recent study described spectroscopic evidence for the intercalation of anthraquinone derivatives into calf thymus DNA [6]. The widely used anti-cancer drugs of the anthracycline type are derived from the anthraquinone structure. They interact via intercalation into DNA with topoisomerase (topo) II and inhibit its catalytic function [7^9]. In a previously published study, we could show that three 1,8-dihydroxyanthraquinones, emodin, aloe-emodin, and danthron (Fig. 1) inhibited topo II activity [4]. Here, we quantitatively determined
0304-4165 / 99 / $ ^ see front matter ß 1999 Elsevier Science B.V. All rights reserved. PII: S 0 3 0 4 - 4 1 6 5 ( 9 9 ) 0 0 0 6 4 - 1
BBAGEN 24822 27-7-99
S.O. Mueller, H. Stopper / Biochimica et Biophysica Acta 1428 (1999) 406^414
their potency to inhibit topo II function. This catalytic inhibition of the topo II could be mediated by a competition of DNA binding sites of the topo II, as it is seen for the anthracyclines. Therefore, we investigated the potential of the anthraquinones to compete with the DNA dye bisbenzimide Hoechst 33342 for binding sites of isolated DNA and in intact mouse lymphoma L5178Y cells. Inhibition of topo II can lead to chromosomal DNA breakage [10,11]. Therefore, we further investigated whether the topo II inhibiting qualities are in agreement with the type and chromosomal extent of the mutagenic DNA damage induced by these compounds. For this purpose we analyzed mutants of L5178Y mouse lymphoma cells using a recently developed LOH-PCR method [12,13].
407
2. Materials and methods
33342 and incubated until £uorescence equilibrium was reached (30 min for calf thymus DNA; 4 h in living cells). Experiments were initiated by addition of the respective compound and the decrease of £uorescence was monitored over time. The excitation wavelength was set at 347 nm and the emission wavelength at 454 nm. Total £uorescence of cells with Hoechst 33342 in equilibrium was about 100 £uorescence units (FU). Non-speci¢c £uorescence of the compounds was always less than 1 FU. Concentrations at which the £uorescence of DNA bound Hoechst 33342 was reduced to 50% (IC50 ) were determined in at least three independent experiments using concentrations of the test compounds including the respective range of IC50 values. The viability and membrane integrity of the cells was determined with the £uoresceindiacetat/ethidiumbromide method as described [16] and was between 70 and 85%.
2.1. Nucleic acids, chemicals and cell culture
2.3. Topoisomerase II assay
Calf thymus DNA was purchased from Serva (Heidelberg, Germany). Aloe-emodin, danthron, emodin, ethyl methanesulfonate (EMS), m-amsacrine (m-AMSA), etoposide and bisbenzimide Hoechst 33342 were purchased from Sigma (Deisenhofen, Germany). Mouse lymphoma L5178Y cells were cultured in suspension in RPMI-1640 supplemented with antibiotics, 0.25 mg/ml L-glutamine, 107 Wg/ml sodium pyruvate, and 10% heat inactivated horse serum (all from Sigma, Deisenhofen, Germany). Cell cultures were grown in a humidi¢ed atmosphere with 5% CO2 in air at 37³C.
Topoisomerase II assays were performed according to the manufacturers' protocol (TopoGen, Columbus, OH, USA). The reactions contained 150 ng kDNA, 4 U of topo II (170 kDa form) and the test chemicals (dissolved in DMSO, ¢nal concentration 10%) in 20 Wl reaction bu¡er (30 mM Tris-HCl pH 7.6, 3 mM ATP, 15 mM 2-mercaptoethanol, 8 mM MgCl2 , 60 mM NaCl). The reactions were incubated for 15 min at 37³C and terminated with 1% of SDS, followed by proteinase K (20 Wg/ml) treatment for 15 min at 37³C. After addition of 0.1 vol. gel loading bu¡er (0.25% bromphenol blue, 50% glycerol), the samples were extracted once with an equal volume of chloroform:isoamylalcohol (24:1). The blue upper layer was loaded onto a 1% agarose gel. The products were resolved by gel electrophoresis in 1UTAE bu¡er for 30 min at 100 V, which separated the catenated kDNA from the decatenated DNA monomers. Gels were stained with SYBR Green (Molecular Probes Europe, The Netherlands) and photographed. The inhibition of the catalytic activity of topo II was quanti¢ed by measuring the reduction of the amount of the monocircle DNA band (corresponding to the product of the topo II catalytic reaction) in relation to the network DNA band (corresponding to the substrate of the
2.2. DNA competition assay with isolated DNA and intracellular accumulation This method is based on the elevated £uorescence of Hoechst 33342 when it binds non-covalently to DNA [14,15]. The addition of other DNA binding compounds causes a reduction of this £uorescence, presumably because of competition of binding sites of the DNA. Fluorescence was monitored by a Kontron spectro£uorometer in 1U1 cm cuvettes under continuous stirring at 37³C. Five hundred ng calf thymus DNA or a suspension of 106 L5178Y cells/ ml in 2 ml PBS/G were mixed with 2 WM Hoechst
BBAGEN 24822 27-7-99
408
S.O. Mueller, H. Stopper / Biochimica et Biophysica Acta 1428 (1999) 406^414
topo II). The program NIH Image 1.54 (NIH, USA) was used to measure the intensities of the gel bands. 2.4. Generation of cell mutants in the mouse lymphoma mutation assay Mutation assays were performed with the in situ procedure [17]. Cultures of mouse L5178Y cells were treated with methotrexate before each experiment to kill pre-existing tri£uorthymidine (TFT)-resistant cells. To accomplish this, cells were incubated for 12 h in culture medium plus methotrexate (0.3 Wg/ ml), thymidine (9 Wg/ml), hypoxanthine (15 Wg/ml) and glycine (22.5 Wg/ml). The cells were then incubated for at least 48 h in the same medium without methotrexate. To measure chemically induced mutants, cultures containing 1U106 cells in 5 ml medium were treated with DMSO (¢nal concentration 1%; vehicle control) or with the respective test compounds with pretested non-cytotoxic concentrations (20^100 WM). EMS was used as a positive control. Incubation was performed for 4 h in media with reduced serum content (1% v/v), then the cells were washed twice with fresh medium. After that,
0.5U106 cells from each tested culture were added to 50 ml of semi-solid culture medium containing 0.25% granulated agar (Baltimore Biological Laboratories, USA) and plated into two plastic 100 mm culture dishes and allowed to solidify at room temperature. TFT-resistant cells were selected by adding an overlay of TFT (¢nal concentration 8 Wg/ml) in 10 ml semi-solid medium after an expression time of 42 h. Cloning e¤ciency was determined by adding 600 cells to 100 ml of semi-solid medium in three 100 mm culture dishes. All plates were incubated for a total of 9^12 days at 37³C in 5% CO2 for colony growth. An automatic colony counter was used to count the number of TFT-resistant colonies. For cultivation of mutant clones, replica plates were analyzed 7^8 days after TFT selection with an invertmicroscope under sterile conditions. Mutant cell clones were checked visibly for necrosis and intact clones were picked randomly of the plates with a sterile pasteur pipette and cultivated until the cell culture reached a density of 1^2U107 cells in 20 ml medium (5^10 days). TFT resistance was checked by incubating an aliquot of each cell clone culture with TFT. Only clones which showed TFT resistance were used for LOH analysis.
Fig. 1. Chemical structures of investigated compounds.
BBAGEN 24822 27-7-99
S.O. Mueller, H. Stopper / Biochimica et Biophysica Acta 1428 (1999) 406^414
2.5. Isolation of genomic DNA Genomic DNA of the mutant clones for the PCR analysis was prepared using the QIAamp BloodKit from Qiagen (Dassel, Germany). The cell cultures (1^ 2U107 cells/20 ml medium) were centrifuged, washed twice with PBS/CMF and the pellet resuspended in 200 Wl sterile water. DNA isolation was performed according to the Qiagen protocol with 200 Wl cell suspension. Isolated DNA was stored in TE 10/1 (10 mM Tris, pH 8.0, 1 mM EDTA) at 320³C and diluted for PCR analysis. 2.6. Molecular analysis of mutants by LOH-PCR Primers for the microsatellite sequences were obtained from Research Genetics (Huntsville, AL, USA) and used as a 6.6 WM solution. Thirty ng DNA from each mutant clone was ampli¢ed with four characteristic primer pairs as shown in Fig. 4. As controls, DNA from wild-type tk=3 L5178Y mouse lymphoma cells and from a mutant that is known to show LOH along the entire chromosome 11+ was used. Twenty Wl reactions for the PCR ampli¢cation were performed. Ten Wl 2UPCR Master (20 mM Tris-HCl, 100 mM KCl, 3 mM MgCl2 , 0.01% Brij 35, 400 WM dATP, dCTP, dGTP, dTTP each, 0.05 U/Wl Taq DNA polymerase; Boehringer Mannheim, Germany) was mixed with 3.3 WM of each primer and 30 ng template DNA. PCR was performed in a PTC100 (MJ Research, USA). To avoid unspeci¢c primer binding a touch-down PCR was carried out. For the primer pairs Mit2, Mit59 and Mit49 the following parameters were used: initial denaturation at 94³C for 2 min and 20 s followed by 20 cycles with 30 s, 94³C, 30 s, 61³C (touch down 30.5³C/cycle) and 20 s, 72³C. Then 10 additional cycles with the parameters 30 s, 94³C, 30 s, 50³C and 20 s, 72³C were performed followed by a ¢nal extension for 5 min at 72³C. For the primer pair Agl2 the following parameters were used: initial denaturation at 94³C for 2 min and 20 s followed by 14 cycles with 30 s, 94³C, 30 s, 72³C (touch down 30.5³C/cycle) and 25 s, 72³C. Then 15 additional cycles with the parameters 30 s, 94³C, 30 s, 65³C and 25 s, 72³C were performed followed by a ¢nal extension for 5 min at 72³C. PCR samples were stored at 4³C. Ten Wl of the reaction mixture were
409
loaded on 10% polyacrylamide gels (1UTBE) in 10% loading bu¡er (0.1% bromphenolblue, 50% glycerol). Electrophoresis was performed for 4^8 h at constant 200 V. A 50^2000 bp DNA ladder (Gibco, Germany) was used to assess the fragment lengths of the detected bands. Gels were stained with SYBR Green (Molecular Probes, The Netherlands) and photographed. 3. Results 3.1. Intracellular accumulation of anthraquinones We ¢rst analyzed whether the anthraquinones, emodin, aloe-emodin and danthron are able to compete with the non-covalent DNA binding compound bisbenzimide Hoechst 33342 (H33342) for DNA binding sites. We employed an indirect competition assay with isolated calf thymus DNA, measuring the quenching of the £uorescence of non-covalent DNA bound H33342 by the test compounds [18]. The wellknown DNA intercalator and topo II poison m-amsacrine (m-AMSA) (Fig. 1) [19] was used as a positive control. Etoposide (Fig. 1) as a topo II poison and a non-DNA binding compound was used as a negative control and showed no detectable e¡ects. The data were determined by using linear regression analysis. Concentration ranges of the respective compound including the dose resulting in a 50% reducTable 1 IC50 values of di¡erent compounds for the non-covalent DNA binding of Hoechst 33342 to DNA in solution (500 ng/ml calf thymus DNA) and in mouse lymphoma cells (1U106 cells/ml) Compound
IC50 (WM) with DNA in IC50 (WM) in intact solution cells
Emodin Aloe-emodin Danthron m-AMSA Etoposide
12 þ 4 11 þ 3 14 þ 4 10 þ 1 ra
23 þ 2 23 þ 1 28 þ 5 22 þ 3 ra
Data represent a compilation of the results shown in part in Fig. 2 obtained by linear regression analysis. IC50 values represent the concentrations at which the £uorescence of DNA bound Hoechst 33342 was reduced to 50%. Data are means þ S.D. (Nvthree independent experiments). a r means no inhibition measurable at any tested concentration (IC50 E500 WM (highest concentration tested)).
BBAGEN 24822 27-7-99
410
S.O. Mueller, H. Stopper / Biochimica et Biophysica Acta 1428 (1999) 406^414
Fig. 2. Inhibition of £uorescence of non-covalent DNA bound Hoechst 33342 by the investigated compounds in mouse lymphoma L5178Y. Tests were performed as described in Section 2. Arrow indicates the timepoint at which the respective compound (25 WM each) was added to the cell suspension (1U106 /ml PBS/G) after a steady state of £uorescence of DNA-bound Hoechst 33342 (2 WM; set to 100%) was reached.
tion of £uorescence (IC50 ) were used (data not shown). This measure re£ects the capability of the respective compound to inhibit the non-covalent binding of H33342 to DNA in solution (calf thymus DNA; ¢rst column in Table 1). The anthraquinones showed IC50 values similar to m-AMSA. To analyze whether the anthraquinones are also capable of reaching the DNA in intact cells we measured IC50 values with mouse lymphoma L5178Y cell suspensions, using H33342, readily capable to reach DNA in intact cells [15]. An intact cellular membrane is a prerequisite to assess a potential intracellular accumulation under nearby physiological conditions. Therefore, the £uoresceine/ethidiumbromide assay [16] was employed to control their viability. Only assays were analyzed in which the viability was more than 70%. In Fig. 2 we show a typical experiment with mouse lymphoma L5178Y cells. InC
Fig. 3. Decatenation assay for topoisomerase II activity and its inhibition. Assays were performed as described in Section 2. All lanes were loaded with 150 ng kDNA. Lanes 2^14: after incubation with 4 U topoisomerase II. Lanes 1 and 2: without inhibitor ; lanes 3^5: +100, 50, and 10 WM m-AMSA; lanes 6^8: +10, 1, 0.1 WM emodin; lanes 9^11: +1000, 750 and 500 WM aloe-emodin; lanes 12^14: +100, 50 and 5 WM danthron. NW: network DNA; MC: monocircle DNA.
BBAGEN 24822 27-7-99
S.O. Mueller, H. Stopper / Biochimica et Biophysica Acta 1428 (1999) 406^414
411
Fig. 4. Occurrence of LOH in anthraquinone-induced L5178Y mouse mutants. Cells were treated with emodin (40 WM), aloe-emodin (100 WM) and danthron (100 WM) as described in Section 2. On the left, localizations of investigated loci on chromosome 11 harboring the selectable intact tk allele (11 ) are shown. Distances of the loci are given distal to the centromere in centiMorgan (cM). Shown are the total numbers of mutants showing LOH at the indicated loci and in parentheses the respective percentage. Large LOH is de¢ned as LOH occurring at all loci including Mit2. Medium-sized LOH is de¢ned as LOH at loci Mit59 and/or Mit49 and/or Agl2. Mutants without any LOH at the investigated loci are given as intragenic mutants. Total numbers of analyzed mutant clones are given on the right.
cubation of the cells with Hoechst 33342 resulted in an increased, stable £uorescence (increased £uorescence was reached within 5^10 min, a complete occupation of binding sites resulting in a steady-state was reached within 3^4 h). The addition of the compounds was followed by a rapid decrease of H33342 DNA bound £uorescence (Fig. 2). As for DNA in solution, IC50 values for the anthraquinones were in the same range as for m-AMSA (Table 1 second column).
DNA in intact cells could also result in an inhibition of topo II activity. To test this hypothesis we assayed enzymatic inhibition by monitoring the topo II cataTable 2 IC50 values of di¡erent compounds for the activity of topoisomerase II Compound
IC50 (WM)
3.2. Inhibition of topoisomerase II
Emodin Aloe-emodin Danthron m-AMSA
7þ1 741 þ 272 61 þ 18 63 þ 7
Many known topo II inhibitors, including anthracycline derivatives bind non-covalently to DNA. This non-covalent DNA binding is not a prerequisite for e¤cient topo II poisoning or inhibition [10], but it probably leads to a competition with the topo II for DNA binding sites. Therefore, the observed capabilities of emodin, danthron, and aloe-emodin to reach
Data represent a compilation of the results shown in part in Fig. 3. IC50 values were analyzed by linear regression analysis. They represent the concentrations at which the catalytic activity of the topo II was reduced to 50%. IC50 values were calculated by measuring the integrated density of the topo II product in the gels (band MC in Fig. 3) relative to the integrated density of the substrate kDNA and product band (band MC+band NW in Fig. 3). Data are means þ S.D. (Nvthree independent experiments).
BBAGEN 24822 27-7-99
412
S.O. Mueller, H. Stopper / Biochimica et Biophysica Acta 1428 (1999) 406^414
lyzed decatenation of kinetoplast DNA (kDNA), a catenated network of mitochondrial DNA rings (NW, network DNA, Fig. 3) isolated from Crithidia fasciculata. The decatenation reaction was monitored by the appearance of 2.5 kilobase DNA monomers following gel electrophoresis (MC, monocircle DNA, Fig. 3). Again, m-AMSA was used as a positive control. As we have shown before [4], emodin, danthron, and aloe-emodin reduced the amount of monomer DNA generated by topo II, indicating that all three compounds were capable of inhibiting this enzyme. To quantify the inhibition of topo II, we measured dose-dependent reduction of the product band (MC) in relation to the substrate band (NW; Fig. 3) by measuring integrated densities using image analysis as described [4] and calculated the IC50 values (Table 2). Concentration ranges including the supposed IC50 values were determined before (data not shown). Emodin exhibited even greater topo II inhibiting qualities than m-AMSA. Danthron showed similar activity to m-AMSA and aloe-emodin was much less active.
(repetitive sequences) are often heteromorphic between homologous chromosomes. In this case, the homologous chromosomes can be discriminated after PCR ampli¢cation of these sequences following gel electrophoresis as two DNA fragments di¡erent in length. LOH can be detected as the loss of one of the two DNA fragments [12]. The used heteromorphic loci are located at 2 centiMorgan (cM) (Mit2), 61 cM (Mit59), 77 cM (Mit49) and 78 cM (Agl2; within the tk gene) distal to the centromere (Fig. 4). If a mutant clone does not show any LOH, the lesion is located within the tk gene, possibly including sequences distal to tk, for which heteromorphic microsatellite sequences are unknown (Fig. 4, `intragenic mutation'). In Fig. 5 the results for the primer pair Mit59 are given. A compilation of the results obtained at all investigated loci is given in Fig. 4. Emodin and aloe-emodin yielded no intragenic mutations. For a better understanding, we de¢ned LOH of 2.5^24% of the chromosomal length from the te-
3.3. Analysis of mouse lymphoma L5178Y mutants by LOH-PCR Next, we asked whether the above described modes of action are also re£ected by the extent of the induced chromosomal damage. Therefore, we applied a recently developed PCR based method to analyze mutant clones of mouse lymphoma L5178Y cells. L5178Y cells are heterozygous at the thymidinekinase gene (tk=3 ) on chromosome 11 [20]. Inactivation of the intact tk gene (tk ) leads to a loss of heterozygosity (LOH). This LOH results in resistance of the mutated cells against the selective agent tri£uorthymidine (TFT). We analyzed mutants induced by emodin, aloe-emodin, and danthron obtained at non-cytotoxic concentrations [4]. The relative mutation frequencies per 106 cells were 2.8 for 40 WM emodin (relative cloning e¤ciency was 0.5, emodin was cytotoxic at higher concentrations), 3.6 for 100 WM aloe-emodin (relative cloning e¤ciency was 1) and 6.2 for 100 WM danthron (relative cloning e¤ciency was 0.7). The qualitative investigation of the mutagenic lesions on mouse chromosome 11 of these mutants (tk3=3 or tk3 ) was performed by an analysis of the LOH using PCR [13]. DNA microsatellites
Fig. 5. Gel electrophoresis of a LOH-PCR analysis at Mit59. Analyses were performed as described in Section 2. All lanes were loaded with 20 Wl of each PCR sample. Lanes 1^4: mutants induced by emodin; lanes 5^8: mutants induced by aloeemodin; lane 9: control with heterozygous tk=3 (wild-type, wt; no LOH); lane 10: control with LOH (LOH at all loci); lanes 11^14: mutants induced by danthron. Given is the respective number of the analyzed cell mutant. Lane 15: 200^2000 base pair (bp) DNA ladder (marker, M). Mutants with LOH at Mit59 show only one band with an approximate length of 240 bp, the band with an approximate length of 254 bp is lost (indicated by arrows; lanes 4, 5, 8 and 10). NS, non-speci¢c ampli¢cation products.
BBAGEN 24822 27-7-99
S.O. Mueller, H. Stopper / Biochimica et Biophysica Acta 1428 (1999) 406^414
lomere (LOH at Agl2, Mit49 and/or Mit59) as `medium-sized' LOH. Emodin and aloe-emodin yielded predominantly `medium-sized' LOH, 88% and 70%, respectively. LOH spanning at least 98% of the chromosome length (including Mit2) is de¢ned as `large' LOH. This was observed for 12% of emodin mutants and 29% of aloe-emodin mutants. Danthron yielded 40% intragenic mutations, 20% `medium-sized' and 40% `large' LOH. 4. Discussion All three tested anthraquinones, emodin, aloeemodin, and danthron inhibited non-covalent binding of H33342 to isolated DNA. Furthermore, these compounds showed strong capabilities to reach the DNA in intact cells. Bogush et al. [18] used the quenching of the £uorescence of anthracyclines when they intercalate into DNA as a direct measure for their DNA binding a¤nity. These data could also be con¢rmed by an indirect competition assay. We applied this indirect competition assay here, since the tested anthraquinones are not £uorescent. However, the reported IC50 values should not be used as a measure for the respective compounds' DNA binding a¤nity. The obtained data re£ect rather the intracellular accumulation and capability of emodin, aloe-emodin and danthron to inhibit the non-covalent DNA binding of H33342. Furthermore, we could show that the investigated anthraquinones exhibited strong topo II inhibiting qualities. It is of note, that the proposed enzyme inhibition is di¡erent from a topo II poisoning. Poisoning of topo II occurs via stabilization of the socalled ternary cleavable complex formed by DNA, topo II and drug (e.g. m-AMSA or etoposide) [7]. This interaction with the topo II results in enzymemediated DNA cleavage [7] and is thus assumed to confer to the compound's genotoxicity characterized by chromosomal deletions [10,11,21]. In contrast, an overall inhibition of the enzymes catalytic activity can be caused, for example, by prevention of the binding of topo II to DNA [7,10]. Therefore, we assume that the shown capability of the investigated anthraquinones to reach DNA in intact cells leads to the observed topo II inhibition. To analyze whether this proposed mechanism could also be relevant for
413
the anthraquinone-induced mutagenicity [4], we investigated their mutation pro¢les. Emodin and aloe-emodin yielded no intragenic mutations and predominantly (around 80%) `medium-sized' LOH. Danthron showed 40% intragenic mutations, 20% `medium-sized' and 40% `large' LOH. Spontaneously occurring mouse lymphoma L5178Y cell mutants (cells treated with the solvent DMSO only, 1% v/v) were analyzed by Liechty et al. (personal communication). These spontaneous mutants showed 30^40% LOH for each de¢ned category. In comparison, we found for etoposide (43 nM) and m-AMSA (5 nM) 0% (etoposide) and 6% (m-AMSA) intragenic mutations. m-AMSA showed predominantly `mediumsized' LOH (77%), whereas `large' LOH occurred in only 18% of the m-AMSA-induced mutants. In contrast, etoposide showed higher percentages of `large' (47%) and lower percentages for `mediumsized' LOH (53%) when compared to m-AMSA. The alkylator ethyl methanesulfonate (2 mM, EMS) showed 94% intragenic mutations, 6% `medium-sized' LOH and no `large' LOH. Thus, the investigated anthraquinones showed considerably more similarity to the topo II poisons m-AMSA and etoposide than to EMS as an intragenic mutagen. Furthermore, except for danthron there was a marked di¡erence to the background (solvent control) LOH, as DMSO yielded similar LOH at all investigated loci. One possible explanation for the di¡erent extent of LOH (`medium-sized' vs. `large' LOH) induced by various topo II inhibitors, including these anthraquinone, could be the distribution of topo II along the chromosome during mitosis. In the presence of topo II poisons like etoposide a maximum of the topo II content is found at the centromere [22]. Therefore, in this case chromosomal mutations (lesions) are expected close to the centromere, i.e. LOH at Mit2 (`large' LOH) as it is found for etoposide. The topo II concentration at the centromere is reduced in the presence of DNA intercalators. m-AMSA acts at high concentrations as an intercalator, rather than as a topo II poison [22]. This could possibly re£ect the found reduced LOH near the centromere. We could observe a similar LOH pattern for emodin and aloe-emodin, but not for danthron. Another hypothesis was postulated by Gaulden [23] who stated that an interference with topo II could lead to `chro-
BBAGEN 24822 27-7-99
414
S.O. Mueller, H. Stopper / Biochimica et Biophysica Acta 1428 (1999) 406^414
mosome stickiness' at mitosis, resulting in chromatin bridges and later in large chromosomal deletions. Such a `chromosome stickiness' was reported to occur at high concentrations of m-AMSA (about 10 nM) [23] and would be in accordance with the mutation pro¢le shown here. We also observed chromatin bridges after treatment with the investigated anthraquinones in mammalian cells (unpublished observation). Overall, emodin and aloe-emodin, but not danthron showed chromosomal deletion patterns more similar to the intercalator m-AMSA than to the non-DNA binding topo II poison etoposide. In conclusion, the presented data are in agreement with the idea that an indirect inhibition of the catalytic activity of topoisomerase II contributes to the genotoxic e¡ects induced by 1,8-dihydroxyanthraquinones. However, these data do not exclude other possible mechanisms, especially generation of reactive intermediates by redoxcycling [9,24], that may contribute to the genotoxicity induced by anthraquinones. Acknowledgements This work was supported by the Swiss Federal O¤ce of Public Health (BAG Grant number FE 316.95.0500). We thank Dr. W.K. Lutz, head of the Department of Toxicology, University of Wu«rzburg and Dr. J. Schlatter, monitoring scientist of the BAG, for valuable advice in all parts of this work and Ms. M. Gerhard for her expert technical assistance. References
[2] H. Mori, S. Sugie, K. Niwa, N. Yishimi, T. Tanaka, I. Hirono, Jpn. J. Cancer Res. 77 (1986) 871^876. [3] H.Y. Su, S.H. Cheng, C.C. Chen, H. Lee, Mutat. Res. 329 (1995) 205^212. [4] S.O. Mueller, I. Eckert, W.K. Lutz, H. Stopper, Mutat. Res. 371 (1996) 165^173. [5] G. Swanbeck, Biochim. Biophys. Acta 123 (1966) 630^ 633. [6] D.T. Breslin, G.B. Schuster, J. Am. Chem. Soc. 118 (1996) 2311^2319. [7] S.J. Froelich-Ammon, N. Oshero¡, J. Biol. Chem. 270 (1995) 21429^21432. [8] L.F. Liu, Annu. Rev. Biochem. 58 (1989) 351^375. [9] J.W. Lown, Pharmacol. Ther. 60 (1993) 185^214. [10] L.R. Ferguson, B.C. Baguley, Mutat. Res. 355 (1996) 91^ 101. [11] R.D. Anderson, N.A. Berger, Mutat. Res. 309 (1994) 109^ 142. [12] M.C. Liechty, Z. Hassanpour, J.C. Hozier, D. Clive, Mutagenesis 9 (1994) 423^427. [13] W. Caspary, H. Stopper, L. Davis, J. Hozier, M. Liechty, in: H.K. Mu«ller-Hermelink, H.-G. Neumann, W. Dekant, (Eds.), Risk and Progression Factors in Carcinogenesis, Vol. 143, Springer, Berlin, 1997, pp. 161^182. [14] T. Bogush, G. Smirnova, I. Shubina, A. Syrkin, J. Robert, Cancer Chemother. Pharmacol. 35 (1995) 501^505. [15] F. Belloc, F. Lacombe, P. Dumain, F. Lopez, P. Bernard, M.R. Boisseau, J. Reifers, Cytometry 13 (1992) 880^885. [16] G.H.S. Strauss, Mutat. Res. 252 (1991) 1^15. [17] D.L. Spencer, C.K. Hines, W.J. Caspary, Mutat. Res. 312 (1994) 85^97. [18] T. Bogush, J. Robert, Anticancer Res. 16 (1996) 365^368. [19] A.H. Corbett, N. Oshero¡, Chem. Res. Toxicol. 6 (1993) 585^597. [20] M.C. Liechty, H.S. Rauchfuss, M.H. Lugo, J.C. Hozier, Mutat. Res. 286 (1993) 299^307. [21] L.R. Ferguson, B.C. Baguley, Environ. Mol. Mutagen. 24 (1994) 245^261. [22] A.T. Sumner, Chromosome Res. 4 (1996) 5^14. [23] M.E. Gaulden, Mutagenesis 2 (1987) 357^365. [24] M. Kodama, Y. Kamioka, T. Nakayama, C. Nagata, N. Morooka, Y. Ueno, Toxicol. Lett. 37 (1987) 149^156.
[1] H. Mori, S. Sugie, K. Niwa, M. Takahashi, K. Kawai, Br. J. Cancer 52 (1985) 781^783.
BBAGEN 24822 27-7-99