[CANCER RESEARCH 47, 6565-6571, DecemberlS, 1987]

Avarol-induced DNA Strand Breakage in Vitro and in Friend ErythroleukemiaCells1

Werner E. G. Müller,2Dusan Sladic, Rudolf K. Zahn, Karl-Heinz Bässler, Nikola Dogovic, Horst Gerner,

M in is lav J. Gasic, and Heinz C. SchröderInstitut fir Physiologische Chemie [W. E. G. M., R. K. Z., K-H. B., H. G., H. C. S.], Universität,Duesbergweg 6, 6500 Mainz, West Germany, and Faculty of Science[D. S., N. D., M. J. G.], Department of Chemistry, Belgrade, Yugoslavia

ABSTRACTThe hydroquinone-containingcytostatic compoundavarol inhibits pre

dominantlygrowth of those cell lines which have a low level of Superoxidedismutasi-.The substrate of this enzyme, the Superoxideanión,was foundto be formed during the in vitro oxidation reaction of avarol to itssemiquinone radical in the presence of oxygen. Under the same incubationconditions plasmid DNA (pBR322) was converted from the fully super-coiled circular form mainly to the nicked circular form, indicating thatthe compoundcauses primarily single-strand breaks. Using Frienderyth-roleukemia cells (FLC) it was found that avarol induces a dose-dependentDNA damage; the maximum numberof DNA strand breaks was observedat 5 h after addition of the compound to the cells. Removal of avarolresulted in a rapid DNA rejoiningwith biphasic repair kinetics [first half-time, 8 min (90% of the breaks) and a second half-time, 40 min (10% ofthe breaks)). When the degree of avarol-induced DNA damage in FLCwas compared with the drug-caused inhibition of cell growth a closecorrelation was established. Avarol displayed no effect on dimethylsulfoxide-induced erythrodifferentiation of FLC as determined by thebenzidine reaction and by dot blot hybridization experiments. Fromincubation studies of FLC with |3H|avarol no hint was obtained for the

formation of an adduct between DNA and the compound.The subcellulardistribution of |3H]avarol was studied in liver cells after i.v. application

of the compound. The predominantamount of the compoundwas presentin the cytosolic fraction; little avarol was associated with plasma membranes, nuclei, and mitochondria. Using (a) oxidative phosphorylationand (/>)oxygen uptake as parameters for mitochondria!function, no effectof the compound on the activity of this organelle was determined.

These results suggest that avarol forms Superoxide anions (and inconsequence possibly also hydroxyl radicals) especially in those cellswhich have low levels of Superoxide dismutase. Moreover, evidence isprovided that the active oxygen species cause DNA damage resulting inthe observed cytotoxic effect.

INTRODUCTION

Avarol (Fig. 1, /) is a sesquiterpenoid hydroquinone (1, 2)which has been isolated in large quantities from the spongeDysidea avara (2). Recently it was found that the compounddisplays a strong cytostatic activity against the lymphoma cellL5178y (3) and a lower antiproliferating activity towards human and murine lymphocytes (4). Moreover it was establishedthat avarol as well as its quinone derivative avarone shows aselective antiviral activity in the human immunodeficiency virus(human T-cell leukemia virus III,,) Hg cell system (5). On thesubcellular level, avarol/avarone acts as an antimitotic agent(6) and causes an inhibition of SOD3 in vitro as well as in an

intact cell system (7).

Received 4/13/87; revised 7/31/87, 9/10/87; accepted 9/18/87.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1Supported in part by a grant from the Deutsche Krebshilfe e.V. (W.E.G.M.)and by a grant from the Bundesministerium fürForschung und Technologie(German-Yugoslavic cooperation program) under the coordination of the Internationales BüroGesellschaft fürKernenergieverwertung in Schiffbau und Schiffahrt Geesthacht. This paper is dedicated to Prof. Dr. Hamao Umezawa.

*To whom requests for reprints should be addressed.1The abbreviations used are: SOD, Superoxide dismutase; FLC, Friend eryth-

roleukemia cells; MejSO, dimethyl sulfoxide; I I*<<.,concentration which causesa 50% inhibition of cell growth in vitro, when compared with the control; P/O(ratio), fiatoms of 1', esterified per Catoni of oxygen uptake.

It is the aim of the present investigation to elucidate thatmolecular mechanism which is responsible for the cytotoxicityof avarol. Together with earlier data (7), which showed thatavarol is converted into its corresponding quinone derivativeavarone (Fig. 1, //) via the semiquinone free radial, we provideexperimental evidence that Superoxide radical anions (Oï)areproduced by avarol in the presence of oxygen (02). This radicalspecies and/or hydroxyl radicals (OH'), which may be formed

from Oíand H2O2, induced DNA damage both in vitro and inFLC. The interaction of these oxygen radicals with DNA,resulting in the formation of DNA breaks, has been documentedearlier (8). Simultaneously, we studied if the drug-free super-oxide radical anions could be detoxified within the cells by theSODs. In view of published data (9), that free radicals areinvolved in the process of cellular differentiation and derepres-sion of genes, we studied the influence of avarol on the differentiation of FL cells. It is known that a series of metabolicinhibitors are potent inducers of erythroid differentiation in FLcells (10). Finally, we studied the effect of avarol on mitochondria! function. This series of experiments was necessary in viewof our previous findings (7) in which a half-wave potential [EVi]for avarol/avarone, ranging from —210mV (at pH 10) to +185

mV (at pH 3), was determined. Hence an interaction of avarol/avarone with enzyme complex(es) of the respiratory chain appeared to be possible.

MATERIALS AND METHODS

Materials. Hexokinase (from yeast; 140 units/ing) was obtained fromBoehringer, Mannheim (Germany); nitro blue tetrazolium was fromSigma, St. Louis, MO.

Superoxide dismutase (from beef erythrocytes; specific activity, 300units/mg) was a gift of Firma GrünenthalGmbH, Stolberg (Germany).

Avarol was isolated from the sponge Dysidea avara as described (2).For the /;/ vitro studies avarol was dissolved at a concentration of 10mg/ml in a solvent system consisting of 60 g glycerol (water free),100 g distilled water, and 969 g of polyethylene glycol 400. For the cellculture experiments, the compound was dissolved in Me..SO (10 mg/ml). The maximum concentration of Me2SO used was 0.1%; at thisconcentration the solvent was without influence on cell growth or thestate of differentiation. [3H]Avarol was obtained by catalytic exchangewith tritium-labeled water as described (11). The material was purifiedby high-pressure liquid chromatography (12); the specific radioactivitywas 24.2 Ci/mmol.

Analysis of DNA Strand Breakage in Vitro. The potential of avarolto cause DNA strand breakage was assayed by its ability to induce therelaxation of negatively supercoiled plasmid pBR322 (13). The reactions were terminated by adding 10 HIMEDTA and 0.1% sodiumdodecyl sulfate (final concentrations). The reaction products were analyzed by electrophoresis in 1% agarose gels followed by ethidiumbromide staining and visualization with UV (300 nm).

The standard assay mixture contained in a final volume of 20 /¿I,0.3ini of negatively supercoiled circular pBR322 DNA (containing lessthan 10% nicked circles) in 10 mM Tris-HCl (pH 8.0) and 100 HIMNaCl, IS pg/m\ of bovine serum albumin, and avarol at differentconcentrations. Reactions were performed at 30°Cfor 0-30 min, under

either nitrogen or oxygen while shaking.6565

Research. on November 27, 2020. © 1987 American Association for Cancercancerres.aacrjournals.org Downloaded from

AVAROL-INDUCED DNA DAMAGE

OH

I IIFig. 1. Structures of avarol (/) and avarone (//).

As a control, pBR322 DNA was incubated in the DNA topoisom-erase II assay (with 1 HIMATP), using 2 ¿ig/20//I of hen oviduct matrix-associated topoisomerase II as enzyme source, essentially as described(14). The incubation was performed for 30 min at 30*C.

Determination of Superoxide Anions. For the detection of Superoxideanions, the procedure described by Beauchamp and Fridovich ( 15) wasused. This method is based on the conversion of nitro blue tetrazoliumto blue formazan. The reduction was recorded at 560 nm. The reactionmixture (final volume, 0.5 ml) contained 0.14 M NaCl, 0.01 M sodiumphosphate buffer (of different pH), 0.15 mw nitro blue tetrazolium, anddifferent concentrations of avarol. Where indicated, Superoxide dismutase was also added to the assay. The reaction was performed at37*C for 0-10 min (linear part of the kinetics) while shaking under air

conditions. In one series of experiments, the reactions were performedunder nitrogen.

Culture of FLC. FLC derived from a clone of Friend virus-transformed 745A cells (16) were grown in Joklik minimal essential medium(without NaHCOj, but with 15 mM 4-(2-hydroxyethyl)-l-piperazine-ethanesulfonic acid) supplemented with 10% fetal calf serum. The cellswere routinely seeded at 104/ml and incubated at 37"C for 72 h; controls

showed a generation time of 11.4 h. Cell growth was determined bycell count with a Cytocomp counter (128-channel counter; Michaelissystem).

For the dose-response experiments to determine effects on cellgrowth, 5 ml cultures were initiated by inoculation of either IO4or 5 x10* cells/ml and were incubated at 37'C for 72 and 24 h, respectively.

The EDso values as well as the parameters describing the cell distribution curves (mean values) were estimated as described (4). The valuespresented are means of 10 parallel experiments each.

The effect of avarol on the MeaSO-induced differentation of FLCwas studied under the following routine conditions. The cells wereseeded at I(I4/ml and incubated for 120 h in the presence of 2.0%

Me2SO(10).Culture of Other Cells. The descriptions of the culture procedures

for the following cell lines were given earlier (3, 17): L5178y mouselymphoma cells; Ili-la cells (ATCC CCL 2.2); human malignant mel

anoma cells (ATCC CRL 1424); human fibroblasts (obtained fromforeskin), and macrophage-containing lymphocytes (from mouse

spleen).Assay of Hemoglobin-producing FLC. The number of differentiating

(hemoglobin-producing) cells was determined by scoring benzidine-positive cells (18). Five-10 min after addition of the benzidine reagent,an average of 300 cells was scored by 2 observers for benzidine positiv-ity.

Isolation of RNA and Dot Blot Hybridization. After the 120-h incubation period of FLC in the presence of different concentrations ofavarol, the cells were harvested and the RNA was extracted, essentiallyas described (19).

The dot blot hybridization assay was performed according to themethod of White and Bancroft (20) with the modifications describedin (21). Briefly, RNA was denatured by adding 7.5% formaldehyde and6x standard saline citrate (0.15 M sodium chloride; 0.015 M sodiumcitrate, pH 7.4) and heating to 56'C for 20 min. Aliquots obtained after

adjusting the sample to 13x SSC were applied to nitrocellulose sheets.The baked sheets were hybridized with the 32P-labeled /3-globin-specific

probe as described by Maniatis et al. (22) using a manifold suctiondevice (Schleicher & Solinoli, Dassel, Germany). For this study aplasmid containing the clone of the 390-base pair restriction fragmentD (which covered most of the 3'exon of the murine major /3-globin

gene) was used (23). Exposition of the dry nitrocellulose filters toKodak XAR-5 X-ray film (Eastman Kodak) backed by one intensifyingscreen was done at -70"C for 2-8 days. The concentration of RNA

was determined spectrophotometrically (1 A260= 40 pg/ml).Assay for DNA Strand Breakage in FLC. DNA damage analysis was

carried out using a fluorometric technique that measures the rate ofunwinding of cellular DNA on exposure to alkaline conditions (24). Inbrief, 35 ml of FLC suspension (initial concentration, 5x10* cells/ml)were treated with different concentrations of avarol for 0-24 h andthen chilled to O'C; the cells were collected by centrifugaron, washed,

and distributed equally to a set of 10 tubes. Cells were lysed with aurea/detergent solution. An alkaline solution was added, and DNAstrand unwinding was allowed to occur for a 60-min period (22V).Then, the samples were neutralized and the amount of residual double-stranded DNA was estimated by using ethidium bromide as the fluorescence dye. After calculating the rate of DNA unwinding, the valueswere converted to the number of strand breaks per cell by reference tothe effect produced by -y-rays. The results are given in units of Qj (25),

which is a measure of DNA damage; 1 Qd unit corresponds to about100 strand breaks/cell. Using this approach, single- and double-strandbreaks and alkali-labile lesions are detected but are not distinguishable(24).

Search for Possible DNA-Avarol Adducts. FLC were seeded at adensity of 10'/ml and 10-ml cultures were incubated in the presence of[3H]avarol (10 ¿iCi/ml;0.41 MM)for 0 and 30 min and 2 and 10 h under

standard conditions. Then, the cells were harvested and the DNA wasextracted by a hydroxyapatite procedure (26). Samples of the purifiedDNA were analyzed for DNA content using the fluorescent dye(H33258) adduci method (27). Herring sperm DNA served as a reference standard. Using this procedure, 6.39 ¿¿gof DNA were extractedfrom 10' cells [yield, 94%. That content of nonextracted DNA in FLC,

determined according to the method of Kissane and Robins (28) wasset to 100%]. The DNA was used either directly or after its enzymaticdigestion to deoxyribonucleosides (29) for determination of the possibleassociated radioactivity.

Subcellular Distribution of |3H|Avarol in Liver Cells of Rats. Rats(Sprague-Dawley; male; 200 g) were given injections of 10 mg/kg of[3H]avarol (94 /¿Ci/mg)into the tail vein i.v. (3). The animals were

killed 30 min or 12 h later and immediately used for subcellularfractionation. The procedure of Fleischner and Kervina (30) was usedfor the isolation of (a) plasma membranes, (¿>)mitochondria, (r) nuclei,(</) cell supernatant, and (<•)microsomes. The isolation of lysosomesand peroxisomes was performed as described ("Addendum" in Ref. 30).

The identity of the organelles and fractions was established by determination of the relative amounts of DNA or of the following markerenzymes, essentially as described (30, 31): 5'-nucleotidase (plasma

membrane); DNA (nuclei); succinate dehydrogenase (mitochondria);aryl sulfatase (lysosomes); catalase (peroxisomes); and glucose 6-phos-phatase (microsomes). Using this preparation procedure, we obtained90% of the total protein (present in the homogenate) in the differentsubfractions.

For determination of the radioactivity, the samples were treated in30% H2O2 with 5% NH4OH (37'C, 6 h) in combination with aquasol(NEN). Glacial acetic acid (0.01 vol of the scintillant's volume) was

added to suppress chemiluminescense of the alkaline solution.Mitochondria, used for the study of the function of this organelle,

were prepared from rat liver (32).The nuclear envelopes were obtained from nuclei by the procedure

of Kaufmann et al. (33).SOD Extraction and Assay. The determination of SOD activity was

performed basically as described (34) with some modifications givenearlier (2). The cells were homogenized and the cell-free extract wascollected; then the total SOD activity and the cyanide-insensitive man-ganese-SOD activity in the presence of 1.0 mM CN~ were determined

as described (7). The copper/zinc-SOD activity was estimated as thedifference between the total and the manganese-SOD activities. The

6566

Research. on November 27, 2020. © 1987 American Association for Cancercancerres.aacrjournals.org Downloaded from

AVAROL-INDUCED DNA DAMAGE

specific activity is given in units per mg of protein (34).Measurements of Mitochondria! Functions. Measurement of oxidative

phosphorylation was performed as described (35).The oxygen uptake was measured by the Warburg manometric

method (36). The reaction mixture in a Warburg flask (total volume, 3ml) had the following composition: 17 HIMsodium phosphate buffer(pH 7.4); 5 mM magnesium sulfate; 13 HIM DL-/3-hydroxybutyrate(substrate); IS mM glucose; 0.013 mg/ml hexokinase; 2 mM ADP; and0.5 ml mitochondria suspension (corresponding to 0.5 g of liver).Glucose, hexokinase, and ADP were pipetted into the side arm of theWarburg flask and added after the flask had been connected to themanometer and the reaction mixture had reached 30°C.A 1 "zoll"filu-r paper (soaked with 0.2 ml of 2 M NaOH) was placed into the

central well of the flask in order to absorb the liberated CO2. Incubationwas performed for 15 min at 30*C. Immediately after termination of

the incubation, 0.5 ml of the reaction mixture was withdrawn anddeproteinized. In the supernatant, glucose 6-phosphate was determinedenzymatically (36) and the formation of ATP from ADP was estimatedaccordingly. The oxygen uptake was calculated as described (36). Therelation of oxygen uptake (in moleatoms) to ester formation (in mole-atoms) was used to calculate the P/O ratio.

Protein was determined according to the method of Lowry et al.(37).

RESULTS

Correlation of Avarol Cytotoxicity with the SOD Content

A comparison of the levels of both copper/zinc- (cytosolicspecies) and manganese-containing SOD in a series of cell lineswith their corresponding sensitivity to avarol shows a strikingparallelism (Table 1). It becomes evident that the higher theSOD content the lower is the sensitivity of the respective cellline to the effect of avarol, as shown by the respective ED50value. From these data, we hypothesize that avarol inhibits cellgrowth by formation of Superoxide anions which are known tocause DNA damage both in vitro (22) and in intact cell systems(38, 39).

Effect of Avarol on DNA in Vitro

Degradation of pBR322 DNA in Vitro. Incubation of pBR322DNA in the presence of avarol caused a conversion of fullysupercoiled DNA (form I) to nicked circular molecules (formII) only in the presence of oxygen (Fig. 2, lanes a, c, and </);only little double-strand broken DNA was detected in the assayscontaining avarol and oxygen, after an incubation period of 30min (Fig. 2, lane d; form III). In the presence of nitrogen onlyminute amounts of nicked circular molecules were formed (Fig.2, lanes b, g, and h). As a control, pBR322 DNA was incubatedwith DNA topoisomerase II in order to identify the mode ofconversion of plasmid DNA (Fig. 2, lanes e and/).

Table 1 SOD level (both copper/zinc- and manganese-containing species) indifferent cell lines in comparison with their susceptibility towards an avarol effect

The degree of avarol-caused inhibition is shown by the respective ED$o value.All I-IK., determinations were evaluated from results of growth inhibition experiment; the drug exposure period was 72 h.

SOD activity (units/mg protein) Sensitivity to-wards avarol

Cell line Total Copper/zinc Manganese (1 I>•,,in MM)Friend leukemia cells 64.9 ±2.5° 43.5 ±1.3 21.4 ±1.8 1.61 ±0.25LSI78y lymphoma 58.3 ±2.3 39.1 ±1.2 19.2 ±1.6 0.93 ±0.13*

cellsMurine lymphocytesHeU cellsHuman fibroblastsMelanoma cells75.1

±2.7104.9 ±4.2105.2 ±4.1117.3 ±4.648.4

±1.675.6 ±2.270.6 ±2.278.7 ±2.426.7

±2.329.3 ±2.734.6 ±2.938.6 ±3.22.91

±0.23C12.4 ±2.2*16.9 ±2.9*37.8 ±5.9*°

Mean ±SD.* Taken from Ref. 3.' Taken from Ref. 4.

0' 0'

02 N2

Fig. 2. Agarose gel electrophoresis of pBR322 DNA following incubation withavarol at a drug to DNA base pair ratio of 1:220 (as described in "Materials andMethods"). Incubation was performed under oxygen (lanes a, c and d) or nitrogen

(lanes b, g, and A) for O min (lanes a and b) or 30 min (lanes c, d, g, and A). Inlanes e and/pBR322 DNA was incubated with DNA topoisomerase (Topo) II(see under "Materials and Methods"). FI, form I, or fully supercoiled circularDNA; / //. form II, or nicked circular molecules; /'///. form III, or linear molecule.

Table 2 Formation of Superoxide anions by avarolThe increase in absorbance reflects the reduction of nitro blue tetrazolium by

Superoxide anions. The standard incubation assay was supplemented with different concentrations of avarol and of SOD. The reactions were carried out atdifferent pH values and under shaking in air. Where indicated, the incubationwas carried out under nitrogen. The means ±SD of S independent experimentsare given.

Incubation conditions (standardassay)Control

incubationPlus

avarol, 10MMPlus

avarol1MM3

MM10

MMPlus

avarol, 10 MMand SOD (30units)Plus

avarol, 10 MM,under nitrogenpH8.26.07.48.29.58.28.28.28.28.2Change

in absorbanceat 560 nm/10min<

0.0020.01

2±0.0020.017±0.0020.065

±0.0090.051±0.0070.01

3±0.0020.038±0.0050.065±0.0090.021

±0.0030.009

±0.001

Formation of Superoxide Anions by Avarol. For the determination of Superoxide anions generated, a colorimetrie methodwas used (IS). As summarized in Table 2, avarol caused aconcentration- and pH-dependent increase in absorbance reflecting a reduction of the dye nitro blue tetrazolium by super-oxide anions; maximal formation of Superoxide was measuredat pH 8.2. To elucidate whether the products formed wereindeed Superoxide anions, the following 2 control assays wereperformed: (a) the addition of Superoxide dismutase to thereaction mixture inhibited the reduction of nitro blue tetrazolium; (b) the performance of the reaction under nitrogen insteadof air caused an almost complete inhibition of the reductionrate. From these experiments it must be concluded that avarolforms Superoxide aniónradicals in aqueous solution.

Effect of Avarol on Erythroleukemia Cells

Influence on Cell Growth and Me2SO-caused Differentation.When FLC were seeded at a density of 10" cells/ml and then

incubated for a period of 72 h, avarol inhibited cell growth to50% of control (=ED50 value) at a concentration of 1.61 ±0.25

6567

Research. on November 27, 2020. © 1987 American Association for Cancercancerres.aacrjournals.org Downloaded from

AVAROL-INDUCED DNA DAMAGE

(SD) íiM(=0.51 ±0.08 Mg/ml). Under the assay conditionsused for the determination of the influence of avarol on theintegrity of DNA (inoculum, 5 x 10s cells/ml; incubation

period, 24 h), the ED50 was 10.8 ±1.31 JIM.Avarol, in the absence of ML-..S(). did not cause an induction

of FLC erythroid differentiation. The cultures (inoculum, 10"

cells/ml; incubation for 120 h) were treated with 0.1, 0.3, 1, 3,and 10 n\i of avarol. At the end of the incubation the percentageof benzidine-positive cells was determined. In the avarol-treatedcultures less than 5% of the cells were benzidine positive. Inthe positive control assays, using 2% Me2SO, 88% of the FLCunderwent a differentiation process (Fig. 3). Next, it was studiedwhether avarol inhibited Me2SO-induced differentiation ofFLC. The results (Fig. 3) clearly show that the degree of theavarol-caused reduction of the percentage of benzidine-positivecells was directly correlated with the degree of the drug-inducedinhibition of cell growth.

The dose-dependent inhibition of FLC erythroid differentiation was further substantiated by analysis of the RNA extractedfrom FLC treated with avarol in the presence of 2% Me2SO(Fig. 4). Using the dot blot hybridization analysis with a ß-globin specific probe, a pronounced dose-dependent reductionof the j8-globin RNA content was observed. An approximate50% reduction was observed with RNA extracted from FLCtreated with 3 >IMof avarol (Fig. 4, lane c).

Production of DNA Strand Breaks in FLC. Avarol, if administered to FLC in culture, rapidly induces strand breaks inintracellular DNA in a concentration-dependent manner (Fig.5). The DNA damage was determined by a sensitive fluoro-metric technique (24). The number of strand breaks, expressedin g,, units as described (25), reached a maximum 5 h afteraddition of the compound. After this period of time the numberswere as follows: 5 pM, 33; 10 ¿IM,49; and 15 /IM, 75 Q¿units.

40-

20-

0 0.5 1 234

Avarol concentration (uM)

Fig. 3. Effect of avarol on Me2SO-treated FLC. The percentage of benzidine-positive (11+) cells and cell number were determined after 120 h of growth in thepresence of 2.0% Me2SO with the avarol concentration indicated.

a

Fig. 4. Dot blot hybridization analysis of RNA from FLC, treated with avarolfor 120 h in the presence of 2% Mi-.SO. A fixed amount of RNA, 1 «tgeach,from cultures treated with 0 (a), 2 (h), 3 (r) or 6 (d) I>Mof avarol was spotted.The RNA was blotted and hybridized with 32P-labeled plasmid containing a .;globin restriction fragment. The membranes were exposed to X-ray films.

6 12 18

Time of incubation (h)24

Fig. S. Effect of avarol on DNA damage in FLC. Cells were inoculated at adensity of 5 x 105/ml and incubated for 0-24 h in the presence of differentconcentrations of avarol, as described under "Materials and Methods." D, 0; x,

5; •,10; O, IS UMavarol. At the indicated times, samples for analysis of DNAstrand-break damage were taken and analyzed for DNA strand breaks (in unitsli(tjj) The means of 10 independent experiments are given; the SD was less than15%.

•= 10-

Time of incubation Imin)

Fig. 6. Rejoining of DNA strand breaks in FLC during posttreatment of FLCafter incubation for 5 h in the presence of 10 (O) or IS (•)pM of avarol. Furtherdetails are given in the legend to Fig. 5. The data of 10 parallel experiments aregiven. Bars, SD.

After a longer incubation period, the number of strand breaksdecreased and leveled off after 18-24 h to the following Q¿unitvalues (24 h): 5 /¿M,12; 10 /iM, 25; and 15 /¿M,28.

FLC have the ability to rejoin strand breaks rapidly afterexposure to avarol (Fig. 6); 50% of the breaks had been repairedalready during the first 5 min after removal of the compound,both in the assays with 10 and 15 pw, 90% of the breaks wererejoined in less than 30 min. The repair kinetics appears to bebiphasic, if the data are blotted semilogarithmically. In the firstcomponent (90%) the half-time of the repair kinetics is 8 min,and in the second component (10%) it is 40 min.

Determination of Possible DNA-Avarol Adducts. FLC wereincubated in the presence of 0.41 /IM [3H]avarol as describedunder "Materials and Methods." DNA was extracted and used

either directly or after enzymatic digestion to deoxyribonucleo-sides to determine the specific radioactivity. Fifty ng of a DNApreparation from cells, incubated for different lengths of time(30 min to 10 h) in the presence of [3H]avaroI, were used for

the determination. The results revealed a specific radioactivityof less than 5 dpm/50 ^g DNA. From this result we cancalculate that less than one molecule of avarol/1.56 x IO6

molecules of deoxyribonucleoside could have formed an adduct.Even after separation of the deoxyribonucleosides by high-pressure liquid chromatography as previously described (40),

6568

Research. on November 27, 2020. © 1987 American Association for Cancercancerres.aacrjournals.org Downloaded from

AVAROL-INDUCED DNA DAMAGE

we could not determine a significant amount of radioactivity tobe associated with the DNA fragments (details not given).

Subcellular Distribution of |3H)Avarol in Liver Cells

Applying the method of autoradiography, we established ina previous study that [3H]avarol accumulates in lymphocytesand in NIH-3T3 cells in the cytoplasmic compartment close tothe nucleus (41).

In the present study, we quantitated the subcellular distribution of avarol in liver cells after i.v. injection of the radiolabeledcompound into rats. The determinations were performed 30min and 12 h after injection (Table 3). After these 2 periods oftime avarol was found to be primarily localized in the cytosolicfraction (=cell supernatant), 64% with respect to the amountin the homogenate after 30 min and 46% after 12 h, respectively. A considerable amount was also associated with theplasma membranes (30 min, 16.3%; 12 h, 5.1%) and the nuclei(3.9 and 8.6%, respectively). Considering the previous observation that avarol is localized at the periphery of the nucleus(41), the nuclei were fractionated into the nuclear envelopesand the residual nuclear structures. Most of the compound wasfound in the nuclear envelope-containing fraction (30 min,nuclear envelopes, 83% and residual nuclear structures 13%;12 h, 79 and 14%, respectively) (100% refers to the total amountof [3H]avarol recovered in the nucleus).

Effect of Avarol on Mitochondria! Function

In order to investigate the effect, if any, of avarol on mitochondria! function, mitochondria! oxidative phosphorylationand oxygen uptake were measured in the presence of varyingconcentrations of avarol. Avarol, at concentrations below 50UM had no significant effect on oxidative phosphorylation. Inthe controls (without avarol), the P/O ratio was 2.96 ±0.62 (n= 5) and in the assays with 50 /¿Mof avarol it was 2.82 ±0.58.Only at the high concentration tested (250 /¿M)was the P/Oratio significantly reduced to 1.60 ±0.31.

The oxygen uptake rate of the mitochondria remained unchanged at avarol concentrations below 250 n\i: controls, 0.72±0.09 nmol O2 consumed/0.5 ml of mitochondrial suspension(containing 1 mg of protein)/15 min; and avarol-treated (250MM)mitochondria, 0.69 ±0.10 timo\ O2/0.5 ml/15 min. Theseconcentrations of avarol used are 30- to 160-fold higher thanthose causing 50% inhibition of growth of FLC (see Table 1).

DISCUSSION

It is a characteristic feature of avarol to display a pronouncedcell-line dependent antiproliferative effect (3); e.g., growth ofL5178y mouse lymphoma cells is inhibited at 90-fold lowerconcentration than those required for inhibition of human

Table 3 Subcellutar distribution off'HJavarol in rat liver celts. Rats were giveninjections off'H/avarol and livers were taken as described ¡n"Materials and

Methods. "

Specificradio-FractionHomogenateMitochondriaPlasma

membranesNucleiCell

supernatantMicrosomesLysosomesPeroxisomesProtein

concentration(mg/mlfresh

livertissue)19.311.71.010.38.48.90.90.85Totalprotein(mg)22502107.954152518819.418.9activity

(dpm x10~3/mgprotein)30

min11.400.941.860.456.450.100.250.1012h8.140.700.420.703.220.181.750.11

gingival cells and a 40-fold lower concentration for the sameeffect on melanoma cells. Evidence has been presented that theantitumor activity of quinone/hydroquinone agents, such asAdriamycin, daunorubicin, trenimon, and aziridinylbenzoqui-none is at least partially caused by the production of DNAstrand breaks mediated by free radicals and active oxygenspecies (42, 43). Consequently, we determined the level of theenzyme SOD, which is known to detoxify the Superoxide radicalanión (Oí)to H2O2 (38), in several different cell lines. Theresults revealed a clearcut correlation between the level of SODand the sensitivity of different cell lines to avarol. The higherthe SOD concentration, the lower the inhibitory activity ofavarol on the respective cell line.

Next we studied whether avarol can form the free oxygenradical species, the Superoxide radical anión (O2), by a one-electron reduction of molecular oxygen (O2); Superoxide radicalanions are known to be the substrate of SOD (38). From aprevious study it was known that avarol is converted to avaronevia the semiquinone state (7). This finding already indicated aone-electron transfer mechanism. The data, presented here,unequivocally show that the electron which is liberated duringthe avarol to semiquinone oxidation process is transferred tooxygen; the resulting Superoxide radical anión serves as substrate for the SOD. In the absence of oxygen, no Superoxideanions are formed (data given) and also no semiquinone radicalsare formed.4 The finding that Superoxide anions are formed

from avarol and oxygen does not necessarily mean that this freeradical is the ultimate cytotoxic component. It is possible thatO2 and H2O2 interact with each other to generate highly reactivehydroxyl free radicals (OH'), as described (44).

Since other quinone/hydroquinone-containing cytostaticagents, e.g., Adriamycin and aziridinoquinone, appear to inducedamage in DNA both indirectly by free radical-mediated pathways (45) and directly by binding to DNA, either by intercalation [Adriamycin (46)] or by alkylation [aziridinoquinone (47)]it is important to determine which mechanism is prevalent inthe case of avarol. Based on the space-filling model of avarol itappears to be unlikely that this compound binds to DNA byintercalation. Experiments performed in our laboratory4 con

firm this suggestion; avarol (a) does not increase the meltingtemperature of DNA, (b) does not decrease the buoyant densityof DNA, (c) does not increase the intrinsic viscosity of DNA,and (¡I)does not change the sedimentation coefficient of DNA.The data given herein show that [3H]avarol does not interact

with DNA after incubation of FLC together with the compound.

In the central part of this study, it is demonstrated that avarolcauses DNA strand breaks both in vitro and in FLC. This effectis not caused by a direct influence but is mediated by Superoxideanions or hydroxyl radicals as can be deduced by the plasmidnicking assay. The DNA lesions formed are primarily single-strand breaks as can be deduced from the migration behaviorof the modified plasmids during agarose gel electrophoresis.Mainly single-strand broken (relaxed) DNA (form II) and verylittle double-strand broken (linear) DNA (form III) were detected after incubation of pBR322 DNA with avarol. The DNAstrand-break damage is also observed if FLC were incubatedwith avarol in vitro. Even after the short incubation period of 5h, a drastic increase of the number of intracellular DNA strandbreaks occurs. The breaks were rapidly rejoined after removalof the compound from the cultures. From the biphasic repairkinetics (first half-time, 8 min; second half-time, 40 min) it can

J Submitted for publication.

6569

Research. on November 27, 2020. © 1987 American Association for Cancercancerres.aacrjournals.org Downloaded from

AVAROL-INDUCED DNA DAMAGE

be deduced that 2 DNA repair mechanisms are involved in therejoining of breaks. It remains to be studied whether the double-strand breaks account for the approximately 10% of strandbreaks that are observed to be repaired slowly.

Two lines of evidence were presented suggesting that theDNA-damage induced by avarol is the cause of the drug-inducedcytotoxicity: (a) it was determined that the DNA damage wasdirectly correlated with the drug-caused inhibition of FLCgrowth; and (b) it was shown that avarol had no effect onMejSO-induced erythrodifferentiation of FLC but interferedwith cell growth only.

Previous autoradiographic studies in our laboratory revealedthat [3H]avarol is taken up by cells and accumulates in the

cytoplasm around the nuclear membrane (41). The data presented here, applying the technique of subcellular fractionation,confirmed this finding. The predominant amount of [3H]avarol,

taken up by liver cells in vivo, was recovered in the cytoplasmicfraction. The portion in the plasma membrane, mitochondria!or nuclear fractions was lower. Functional studies on the effectof avarol in mitochondria in vitro (determination of oxidativephosphorylation and oxygen uptake) gave no hint that thecompound alters the rate of ATP synthesis in isolated mitochondria. In view of the finding, avarol to be taken up bymitochondria, it is suggested that the Superoxide radicals whichmight be formed from avarol and oxygen are detoxified by themitochondria! SOD.

In summary, this study demonstrates that avarol forms oxygen radicals during its oxidation to its semiquinone form. TheSuperoxide free anions are very likely involved in the mechanismof avarol-caused induction of the cytotoxicity. This conclusionis based on the observation of a direct correlation between theinduction of DNA strand breaks and cytotoxicity. It remains tobe investigated, whether the observed anti-human immunodeficiency virus effect of avarol (5) is also due to the avarol-mediated production of Superoxide radical anions. This line ofstudy seems to be promising in view of the published datashowing that the antiviral state in a cell is correlated with thelevel of SOD (48).

ACKNOWLEDGMENTS

We are indebted to K. Wagner for valuable assistance.

REFERENCES

1. Minale, L., Riccio, R., and Sodano, G. Avarol, a novel sesquiterpenoidhydroquinone with a rearranged drimane skeleton from the sponge Disideaavara. Tetrahedron Lett., 3401-3404, 1974.

2. Müller,W. E. G., Zahn, R. K., Gasic, M. J., Dogovic, N., Maidhof, A.,Becker, <'.. Diehl-Seifen, B., and Eich, E. Avarol, a cytostatically active

compound from the marine sponge Dysidea arara. Comp. Biochem. Physiol.,«OC:47-52, 1985.

3. Müller,W. E. G., Maidhof, A., Zahn, R. K., Schröder,H. C, Gasic, M. J.,Heidemann, D., Bernd, A., Kurelec, B., Eich, E., and Seibert, G. Potentantileukemic activity of the novel cytostatic agent avarone and its analoguesin vitro and in vivo. Cancer Res., 45:4822-4826, 1985.

4. Müller,W. E. G., Geisert, M., Zahn, R. K., Maidhof, A., Bachmann, M.,and Umezawa, H. Potentiation of the cytostatic effect of bleomycin on1.517K\ mouse lymphoma cells by pepleomycin. Eur. J. Cancer Clin. Oncol.,19:665-670, 1983.

5. Sarin, P. S., Sun, D. Thornton, A., and Müller,W. E. G. Inhibition ofreplication of the etiological agent of AIDS (HTLV-III/LAV) by avarol andavarone. J. Nati. Cancer Inst., 78:663-666, 1987.

6. Müller,W. E. G., Dogovic, N., Zahn, R. K., Maidhof, A., Diehl-Seifert, B.,Becker, C., Sachsse, W., Gasic, M. J., and Schröder,H. C. Inhibition ofmitosis by avarol, a natural product isolated from the sponge Dysidea avara.Bas. Appi. Histochem., 29: 321-330, 1985.

7. Batke, E., Ogura, R., Vaupel, P., Hummel, K., Kallinowski, F., Gasic, M. J.,Schröder,H. C., and Müller,W. E. G. Action of the antileukemic and anti-IITI V III (anti-HIV) agent avarol on the level of Superoxide dismutase and

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

glutathione peroxidase activities in L5l78y mouse lymphoma cells. CellBiochem. Function, in press, 1987.Halliwell, B., and Gutteridge, M. C. Oxygen toxicity, oxygen radicals, transition metals and disease. Biochem. J., 219: 1-14, 1984.Zegarelli-Schmidt, E. C., and Goodman, R. The dipteran as a model systemin cell and molecular biology. Int. Rev. Cytol. 71: 245-314, 1981.Friend, C., and Freedman, II. A. Effects and possible mechanism of actionof dimethylsulfoxide on Friend cell differentiation. Biochem. Pharmacol.,27:1309-1313, 1978.Bloxsidge, J. P., Elvidge, J. A., Jones, J. R., Mane, R. B., Chambers, V. M.A., Evans, E. A., and Greenftade, D. [G-3H]Vinblastine and its analysis bytritium nuclear magnetic resonance and spectroscopy. J. Chem. Res., S: 42-43, 1977.Müller,W. E. G., Diehl-Seifert, B., Sobel, C., Bechtold, A., Kljajic, Z., andDorn, A. Sponge secondary metabolites: biochemical and structural localization of the antimitotic agent avarol in Dysidea avara. }. Histochem. Cyto-chem.,5* 1687-1690, 1986.Osheroff, N., Shelton, E. R., and Brutlag, D. L. DNA topoisomerase II fromDrosophila melanogaster (relaxation of supercoiled DNA). J. Biol. Chem.,258:9536-9543, 1983.Schröder,H. C., Trölltsch,D., Friese, U., Bachmann, M., and Müller,W.E. G. Mature mRNA is selectively released from the nuclear matrix by anA IT ilATI' dependent mechanism sensitive to topoisomerase inhibitors. J.Biol. Chem., 262: 8917-8925, 1987.Beauchamp, C., and Fridovich, I. Superoxide dismutase: improved assaysand an assay applicable to polyacrylamide gels. Anal. Biochem., 44: 276-287, 1971.Fourcade, A., Farki, J.-J., Bennoun, M., and Tapiero, H. Uptake, efflux, andhydrolysis of Aclacinomycin A in Friend leukemia cells. Cancer Res., 42:1950-1954, 1982.Müller,W. E. G., Sobel, C., Sachsse, W., Diehl-Seifert, B., Zahn, R. K.,Kljajic, Z., and Schröder,H. C. Biphasic and differential effects of thecytostatic agents avarone and avarol on DNA metabolism of human andmurine T and B lymphocytes. Eur. J. Cancer Clin. Oncol., 22: 473-476,1986.Singer, D., Cooper, M., Maniatis, G. M., Marks, P. A., and Rifkind, R. A.Erythropoietic differentiation in colonies of cells transformed by Friend virus.Proc. Nati. Acad. Sci. USA, 71: 2668-2670, 1974.Cathala, G., Savouret, J.-F., Méndez,B., West, B. L., Karin, M., Martial, J.A., and Baxter, J. D. Laboratory methods: a method for isolation of intact,translationally active ribonucleic acid. DNA, 2: 329-335, 1983.White, B. A., and Bancroft, F. C. Cytoplasmic dot hybridization. J. Biol.Chem., 257:8569-8572, 1982.Kindas-Mügge,I., and Sauermann, G. Transport of /3-globin mRNA fromnuclei of murine Friend erythroleukemia cells. Eur. J. Biochem., 148: 49-54, 1985.Maniatis, T., l ritsch. E. F., and Sambrook, J. Molecular Cloning. A Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory,1982.Hofer, E., and Darnell, J. E. The primary transcription unit of the mouse /(•major globulin gene. Cell, 23: 585-593, 1981.Birnboim, H. C., and Jevcak, J. J. Fluorimetrie method for rapid detectionof DNA strand breaks in human white blood cells produced by low doses ofradiation. Cancer Res., 41: 1889-1892, 1981.McWilliams, R. S., Cross, W. G., Kaplan, J. G., and Birnboim, H. C. Rapidrejoining and DNA strand breaks in resting human lymphocytes after irradiation by low doses of *°Coy rays or 14.6-MeV neutrons. Radiation Res.,94:499-507, 1983.Adriaenssens, P. I., Bixler, C. J., and Anderson, M. W. Isolation andquantitation of DNA-bound benzo(a)pyrene metabolites: comparison of hy-droxyapatite and precipitation procedures. Anal. Biochem., 123: 162-169,1982.Cesarone, C. G., Bolognesi, C., and Santi, L. Improved microfluorometricDNA determination in biological material using 33258 Hoechst. Anal.Biochem., 100: 188-197, 1979.Rissane, J. M., and Robins, E. The fluorimetrie measurement of deoxyribo-nucleic acid in animal tissue with special reference to the central nervoussystem. J. Biol. Chem., 233: 184-188, 1958.Baird, W. M., and Brookes, P. Isolation of the hydrocarbon-deoxyribonu-cleoside products from the DNA of mouse embryo cells treated in culturewith 7-methylbenz[a)anthracene-3H. Cancer Res., 33: 2378-2385, 1973.Fleischer, S., and Kervina, M. Subcellular fractionation of rat liver. MethodsEnzymol., 31:6-41, 1974.Graham, J. Isolation of subcellular organelles and membranes. In: D. Rick-wood (ed.) Centrifugaron, a Practical Approach, pp. 161-182. Oxford,England: IRL Press, 1984.Schröckert,W. M., and Bäffller,K. H., Messung der Gesamtaktivitätvon153Glutamat-Oxalacetat-Transaminase in Leberhomogenaten. Clin. Chim.Acta, 33: 55-60, 1971.Kaufmann, S. H., Gibson, W., and Shaper, J. H. Characterization of themajor polypeptides of the rat liver nuclear envelope. J. Biol. Chem., 258:2710-2719,1983.Kirby, T. W., and Fridovich, I. A picomolar spectrophotometric assay forSuperoxide dismutase. Anal. Biochem.. 727:435-440, 1982.Slater, E. C. Manometric methods and phosphate determination. MethodsEnzymol., 10: 19-29, 1967.Slater, E. C., and Holton, F. A. Oxidative phosphorylation coupled with the

6570

Research. on November 27, 2020. © 1987 American Association for Cancercancerres.aacrjournals.org Downloaded from

AVAROL-INDUCED DNA DAMAGE

oxidation of a-ketoglutarate by heart-muscle sarcosomes. Biochem. J., 55: 43. Begleiter, A., and Blair, G. W. Quinone-induced DNA damage and its530-544, 1953. relationship to antitumor activity in L5178y lymphoblasts. Cancer Res., 44:

37. Lowry, O. H., Rosebrough, N. J., Fair, A. R., and Randall, R. J. Protein 78-82, 1984.measurements with the Polin phenol reagent. J. Biol. Chem., 193:265-275, 44- Halliwell, B. The effects of the Superoxide radical and its products. Bull. Eur.1951 Physiopathol. Resp., / 7 suppl.: 21-28, 1981.

38. Fridovich, I. The biology of oxygen radicals. Science (Wash. DC), 201: 875- 4S- L0»"1-J-w- Sim- S- K- Majumdar, K. C., and Chang, R. Y. Strand scission880 1978 of DNA by bound Adnamycin and daunorubicin in the presence of reducing39. Birn'boim.H.C., and Kanabus-Kaminska, M. The production of DNA strand , agents. Biochem. Biophys Res. Commun. 76: 705-710, 1977

breaks in human leukocytes by Superoxide anión may involve a metabolic 46' D'^arco, A Arcamene F and Zunino, F. Daunorubicin and Adnamycmnrrx-i-M Prrw- Nati Ac-art Si-i I ISA Ä2-fi82O-A824 I«M and structural analogues: biological activity and mechanism of action. In: ).

An v^ •M J A ? ;; \v c, -^ til ; . , w- Cockran and F. H. Hahn (eds.), Antibiotics III, Mechanism of Action of40. Kulkirm. M and Anderson, M W. Persistence of benzo(fl)-pyrene metab- Antimicrobial and Antitumor Agents, pp. 101-108. Berlin: Springer-Verlag,olite: DNA adducts in lung and liver of mice. Cancer Res., 44:97-101, 1984. 1975

41. Müller,W. E. G., Sobel, C., Diehl-Seifert, B., Maidhof, A., and Schroder, 47 Akmar M H 8«.^^ A.f Johnson, D., LOW,,, j. W., McLaughlin, L.,H. C. Influence of the antileukemic and antihuman immunodeficiency virus and sjm s K Sludies re|ated ,„antitumor antibiotics. Part VI. Correlationagent avarol on selected immune responses in vitro and in vivo. Biochem. of covalent cross-linking to DNA by bifunctional aziridinoquinones withPharmacol., 36: 1489-1494, 1987. their antineoplastic activity. Can. J. Chem., 55: 2891-2905, 1976.

42. Bachur, N. R., Gordon, S. L., and Gee, M. V. A general mechanism for 48. Pottathil, R., Chandrabose, K. A., Cuatrecasas, P., and Land, D. J. Estab-microsomal activation of quinone anticancer agents to free radicals. Cancer lishment of the interferon-mediated antiviral state: possible role of SuperoxideRes., 3«:1745-1750, 1978. dismutase. Proc. Nati. Acad. Sci. USA, 78:3343-3347, 1981.

6571

Research. on November 27, 2020. © 1987 American Association for Cancercancerres.aacrjournals.org Downloaded from

1987;47:6565-6571. Cancer Res   Werner E. G. Müller, Dusan Sladic, Rudolf K. Zahn, et al.   Erythroleukemia Cells

and in Friendin VitroAvarol-induced DNA Strand Breakage

  Updated version

  http://cancerres.aacrjournals.org/content/47/24_Part_1/6565

Access the most recent version of this article at:

   

   

   

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

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

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

.http://cancerres.aacrjournals.org/content/47/24_Part_1/6565To request permission to re-use all or part of this article, use this link

Research. on November 27, 2020. © 1987 American Association for Cancercancerres.aacrjournals.org Downloaded from