Ž .Mutation Research 425 1999 125–134

Mutagenicities of N-nitrosodimethylamine andN-nitrosodiethylamine in Drosophila and their relationship to

the levels of O-alkyl adducts in DNA

Yuki Goto a, Tomoko Matsuda a, Kazuo Ito a, Nam-ho Huh b, Jurgen Thomale c,¨Manfred F. Rajewsky c, Hikoya Hayatsu a, Tomoe Negishi a,)

a Faculty of Pharmaceutical Sciences, Okayama UniÕersity, Tsushima, Okayama 700-8530, Japanb Department of Biochemistry, Faculty of Medicine, Toyama Medical and Pharmaceutical UniÕersity, Sugitani, Toyama 930-0194, Japan

c Institute of Cell Biology, UniÕersity of Essen Medical School, D-45122 Essen, Germany

Received 29 May 1998; revised 28 August 1998; accepted 14 January 1999

Abstract

N-Nitrosodialkylamines are potent carcinogens in experimental animals. Previously, we reported that the mutagenicity ofŽ . Ž .N-nitrosodimethylamine NDMA was 10 times higher than that of N-nitrosodiethylamine NDEA in the Drosophila wing

spot test. To find out how to explain this difference, we have measured the levels of O-alkylated bases in the DNA ofexposed Drosophila larvae. Third instar larvae were fed for 3 or 6 h with NDMA or NDEA. Part of the treated larvae weregrown to adult flies to score their wings for the presence of mutant spots. From the remaining larvae, DNA was isolated and

Ž .digested to deoxyribonucleosides, and the digest fractionated by high-performance liquid chromatography HPLC . TheŽ .amounts of specific alkyldeoxyribonucleosides present in the fractions were quantified by a radioimmunoassay RIA using

monoclonal antibodies. Dose-dependent O6-methylguanine, O6-ethylguanine and O4-ethylthymine formations were found tobe correlated with the induction frequencies of mutant wing spots. At the same exposure dose, the values of O6-alkylde-oxyguanosiner106 deoxyguanosine were similar for NDMA and NDEA: on feeding 20 mmolr1.5 ml feeding solution, thevalues for NDMA were 4.0 with 3 h and 18.5 with 6 h of exposure; with 20 mmol NDEA, the corresponding values were5.4 with 3 h and 14.6 with 6 h of exposure. The wing spot frequencies were very different; however, with NDMA, the total

Ž . Ž . Ž . Ž .numbers of spotsrwing were 3.5 3 h and 15 6 h , and with NDEA 0.8 3 h and 0.9 6 h . Similar discrepancies exist aswell between the mutagenicities and the alkylation rates observed for O4-alkylthymidines. These results suggest that thedifference between the mutagenic potencies of NDMA and NDEA cannot be explained by the amounts of O-alkyl adductsformed. Different mechanisms are considered by which NDMA and NDEA may produce the genetic effects observed.q 1999 Elsevier Science B.V. All rights reserved.

Keywords: N-Nitrosodimethylamine; N-Nitrosodiethylamine; Drosophila somatic mutation and recombination; O-Alkyldeoxyguanosine;O-Alkylthymidine

) Corresponding author. Tel: q81-86-251-7946; Fax: q81-86-251-7927; E-mail: isaka@pheasant.pharm.okayama-u.ac.jp

0027-5107r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.Ž .PII: S0027-5107 99 00032-9

( )Y. Goto et al.rMutation Research 425 1999 125–134126

1. Introduction

N-Nitrosodialkylamines are alkylating agents andmany of them are proven carcinogens in experimen-

w xtal animals 1 . Some nitrosamines were detected inthe gastric juice of humans with oesophageal carci-

w xnoma 2 . It has been shown that nitrosation ofw xalkylamines can occur in the human body 3,4 . It is

commonly inferred that nitrosamines are relevant forcancer induction through the formation of mutagenicalkyl adducts in important genes involved in carcino-

Ž w x.genesis reviewed in Refs. 5–7 . There is evidencethat DNA alkylation, especially the formation of

6 Ž 6 . 4O -alkyldeoxyguanosine O -alkyl dGuo , O -al-Ž 4 . 2kylthymidine O -alkylThd and O -alkylthymidine

Ž 2 .O -alkylThd , is involved in mutagenesis and car-w x 6 Ž 6cinogenesis 8,9 . O -Methyldeoxyguanosine O -

.MedGuo , for instance, has been detected in theDNA of human esophageal tissues collected frompeople living in an area of China where the risk of

w x 4esophageal cancer is high 10 . Also, O -ethyl-Ž 4 .thymidine O -EtThd has been found in the DNA of

w xhuman liver tissues 11 .Cancer develops from a cascade of events includ-

ing somatic cell mutations. With respect to N-nitro-sodialkylamines, little is known about their capacityto cause mutations in somatic cells in vivo. New testsystems with transgenic rodents have allowed todemonstrate that N-nitrosodialkylamines are able to

w xinduce mutations in somatic tissues 12 ; however, ithas not been experimentally shown how efficient theDNA base alkylations are in the process of mutagen-esis in vivo. In the transgenic systems, only genemutations are detected. The Drosophila wing spottest, on the other hand, detects various genetic end-points including gene mutation, chromosomal recom-bination, segmental chromosomal deletions, and

w xnon-disjunction 13,14 . Analysis of DNA in thisorganism is expected to show quantitative relation-ships between the extent of specific base alkylationand the mutation-inducing effectiveness of the com-pounds. We have previously reported that N-nitro-

Ž .sodimethylamine NDMA is 10 times more muta-genic in the Drosophila wing spot test than N-

Ž . w xnitrosodiethylamine NDEA 15 . To find out whatthe reason for this difference may be, we determinedthe levels of O6-alkyl dGuo and O4-alkylThd inDrosophila DNA by a competitive radioimmunoas-

Ž .say RIA , using monoclonal antibodies against spe-cific adducts and compared the mutagenic effectswith the extents of O-alkylation in DNA.

2. Materials and methods

2.1. Materials

ŽN-Nitrosodimethylamine NDMA; CAS No. 62-. Ž75-9 and N-nitrosodiethylamine NDEA; CAS No.

. Ž .55-18-5 were purchased from Tokyo Kasei Tokyo .Ž .DNase 1 of bovine pancreas , snake venom phos-

Žphodiesterase and alkaline phosphatase of calf intes-.tine were obtained from Boehringer Mannheim

Ž . ŽMannheim, Germany , RNase T1 of Aspergillus. Žoryzae from Worthington Biochemical Freehold,

. Ž .NJ , RNase A of bovine pancreas , rabbit IgG,Ž . Žspermine CAS No. 71-44-3 , spermidine CAS No.. Ž .124-20-9 and bovin serum albumin fraction V

Ž .from Sigma St. Louis, MO , coformycin from Cal-Ž . Žbiochem La Jolla, CA , and proteinase K of Tri-

. Ž .tirachium album from Merck Darmstadt, Germany .( ) J4 qDrosophila melanogaster y; Dp 1;3 sc , y flr

rTM1, Me ri sbd2 males and y; mwh j Õ females´were used in the wing spot test. The Drosophila

Žstrains were gifts of Dr. K. Fujikawa Kinki Univer-. Žsity, Higashi-osaka and Dr. H. Ryo Osaka Univer-

.sity, Suita .w xFor the RIA 16 , we used four specific mono-

clonal antibodies which we have prepared by our-Ž . 6selves J.T. and M.F.R. ; ER-06 for O -EtdGuo

w x 6 w x 417 , EM-21 for O -MedGuo 18 , ER-01 for O -Et-w x 4 w xThd 17 , and EM-3-33 for O -MeThd 19 . Authen-

tic O-alkylated compounds were synthesized accord-w xing to Farmer et al. 20 . Radiolabeled alkyl adducts,

6 w X X 3 x Ž . 6 w X XO -Me 1 ,2 - H dGuo 53 Cirmmol , O -Et 1 ,2 -3 x Ž . 4 w 3 x ŽH dGuo 32 Cirmmol , O -Et 6- H Thd 17

. 4 w 3 x Ž .Cirmmol and O -Me 6- H Thd 120 Cirmmolwere also synthesized. Labeled deoxyribonucleosideswere purified by high-performance liquid chro-

Ž .matography HPLC immediately before their useŽ .see below for details .

2.2. Treatments of larÕae and the wing spot test

ŽThe heterozygous larvae mwhrflr and mwhr.TM1 were obtained by mating virgin mwh-females

( )Y. Goto et al.rMutation Research 425 1999 125–134 127

Ž . Ž .Fig. 1. Toxicities of NDMA v and NDEA ` in DrosophilaŽ . Ž .3 h treatments . Survival % s100=the number of survivingadult flies treated with N-nitrosodialkylaminerthe number ofsurviving adult flies treated comparably but without N-nitrosodialkylamine.

w xwith flrrTM1-males 13 . The 3rd instar larvae werecollected 72–96 h after oviposition and introduced atapproximately equal numbers into tight-lid petri

Ž .dishes Toyo Roshi Kaisha, Tokyo; F 5 cm , whichcontained 1.5 ml of 0.25 M sucrose with or withoutan N-nitrosodialkylamine. Each dish had a hole cov-

ered with a nylon mesh for the exchange of air. After3 h or 6 h, all the larvae were still alive; part of themwas taken and reared to adult flies on a standard

Žmedium Formula 4-24, Carolina Biological Supply,.Burlington, NC . The remaining larvae were frozen

in liquid nitrogen and stored at y808C until theirDNA was extracted to quantify the alkyl adducts.Living Drosophila larvae and flies were kept at258C during these procedures. The wings of the flieswere screened for mutant spots under a microscope.The mutant spots were classified into small-single,large-single and twin spots according to Graf et al.w x13 .

2.3. DNA extraction and hydrolysis

The larval DNA was extracted according to Jowettw x21 with a slight modification. About 1 g of frozenlarvae was crushed into a fine powder using an agatepestle–mortar held in liquid nitrogen. The powder

Žwas homogenized in 9 ml of a lysis buffer 100 mMTris–HCl pH 8.0, 50 mM NaCl, 50 mM EDTA, 2%

.SDS, 0.15 mM spermine and 0.5 mM spermidine .

Table 1Wing spot induction by NDMA and NDEA in mwhrflr heterozygotes

Ž . Ž .N-Nitrosodialkylamine Number Number of mutant spots spotsrwing Survival %Ž .mmolrdish of wings

Small singles Large singles Twins Total

3-h TreatmentŽ . Ž . Ž . Ž .None 104 50 0.48 3 0.03 1 0.01 54 0.52 100.0Ž . Ž . Ž . Ž .NDMA 10 64 114 1.78 )) 46 0.72 )) 6 0.09 ) 166 2.59 )) 111.4Ž . Ž . Ž . Ž .20 84 189 2.25 )) 96 1.14 )) 10 0.12 )) 295 3.51 )) 107.2Ž . Ž . Ž . Ž .30 77 529 6.87 )) 242 3.14 )) 26 0.34 )) 797 10.35 )) 101.6Ž . Ž . Ž . Ž .NDEA 20 62 38 0.61 13 0.21 )) 0 0.00 51 0.98 )) 107.2Ž . Ž . Ž . Ž .50 53 46 0.87 )) 4 0.08 0 0.00 50 0.94 )) 103.2Ž . Ž . Ž . Ž .80 62 63 1.02 )) 27 0.44 )) 1 0.02 91 1.47 )) 94.4

6-h TreatmentŽ . Ž . Ž . Ž .None 40 23 0.58 3 0.03 0 0.00 26 0.65 100.0Ž . Ž . Ž . Ž .NDMA 5 20 106 5.30 )) 41 2.05 )) 3 0.15 ) 150 7.50 )) 111.7Ž . Ž . Ž . Ž .10 20 236 11.80 )) 97 4.85 )) 3 0.15 ) 336 16.80 )) 93.5Ž . Ž . Ž . Ž .20 20 222 11.10 )) 73 3.65 )) 4 0.20 ) 299 14.95 )) 89.0Ž . Ž . Ž . Ž .30 20 226 11.30 )) 110 5.50 )) 7 0.45 )) 343 17.15 )) 108.1Ž . Ž . Ž . Ž .NDEA 10 20 15 0.75 2 0.10 0 0.00 17 0.85 111.2Ž . Ž . Ž . Ž .20 20 16 0.80 2 0.10 0 0.00 18 0.90 102.8Ž . Ž . Ž . Ž .50 20 32 1.60 )) 8 0.40 )) 2 0.10 42 2.10 )) 107.7Ž . Ž . Ž . Ž .80 20 43 2.15 )) 7 0.35 ) 2 0.10 52 2.60 )) 60.1

) P-0.05; )) P-0.01: significantly different from the corresponding negative controls.

( )Y. Goto et al.rMutation Research 425 1999 125–134128

Ž . Ž . Ž . Ž .Fig. 2. Mutagenicities of NDMA v and NDEA ` in Drosophila: treatments for 3 h A and 6 h B . )) P-0.01, significantlydifferent from the control where no nitrosamine was added.

The homogenate was treated with 0.2 mgrml pro-teinase K for 1 h at 508C. Nucleic acids were thenisolated by phenol–chloroform extraction and precip-itated with ethanol. RNA was removed by digestion

Ž . Žwith RNase T1 100 Urml and RNase A 500.mgrml at 378C overnight. DNA was precipitated

Žwith ethanol and dissolved in TE buffer 10 mM.Tris–HCl, 1 mM EDTA, pH 8.0 . The solution wasŽplaced on a nitrocellulose filter pore size, 0.025

. Ž .mm Nihon Millipore, Yonezawa, Japan and dia-lyzed against TE buffer overnight to remove anyremaining ribonucleotides. We obtained 1 to 1.4 mgDNA from 10 g samples of larvae. Purified DNAŽ . X4–8 A units was hydrolyzed to 2 -de-260

oxyribonucleosides by consecutive digestion withŽ .DNase 1 400 mgrml, 378C, 1 h and then with a

Ž .mixture of phosphodiesterase 200 mgrml and alka-Ž .line phosphatase 340 Urml, 378C, 1.5 h . The DNA

Ž .Fig. 3. Wing spot induction by N-nitrosodimethylamine and N-nitrosodiethylamine in Drosophila treated for 3 h: NDMA-in-Ž .duced spots on the wings of mwhrflr heterozygotes; NDMA-induced spots on the wings of mwhrTM1 heterozygotes;

Ž . Ž .NDEA-induced spots on the wings of mwhrflr heterozygotes; NDEA-induced spots on the wings of mwhrTM1Ž . Ž .heterozygotes. The insert shows the recombinogenicities of NDMA and NDEA .

( )Y. Goto et al.rMutation Research 425 1999 125–134 129

hydrolysis was carried out in the presence of 10y5

M coformycin to avoid degradation of O6-methyl-w xguanine residues 22 . After removal of enzyme pro-

teins by ethanol-precipitation, the deoxyribonucleo-side mixture was analyzed by HPLC separation.

2.4. Separation of modified deoxynucleosides and( )RIA HPLC–RIA

Four hundred fifty microliters of the hydrolysate,spiked with 1 mM caffeine as an internal standard,was injected into a 4=250 mm reverse phase col-

Žumn Lichrosorb RP18, 10 mm; Merck, Darmstadt,. ŽGermany in an HPLC system Waters M600 multi-

solvent delivery system linked to a 490E multi-wave-.length detector; Millipore, Milford, MA . The col-

umn was eluted at 408C with 0.1 M ammoniumŽ .acetate pH 5.0 which contained methanol, with a

flow rate of 1 mlrmin. Fractions were collectedevery 30 s. The methanol concentration was linearlyincreased from 10% to 20% during a period of 20min from the start, then from 20% to 50% during thenext 15 min. Finally, the column was eluted with100% methanol for 10 min. Aliquots of fractionscorresponding to the retention times of authenticnucleosides and also the neighboring fractions weresubject to the competitive immunoprecipitation as-say. Each fraction was immunoprecipitated by addi-tion of a specific monoclonal antibody in the pres-

w3 xence of a defined quantity of H -labeledŽ .alkyldeoxyribonucleoside tracer . The radioactivity

remaining in the supernatant was measured, and theinhibition of tracer–antibody binding by the alky-lated nucleosides derived from the larval DNA in thefraction was calculated. Fractions showing inhibi-

Žtions higher than 15% were pooled e.g., fractionsNo. 38, 39 and 40, around 19 min retention time in

.Fig. 4a . The quantitation of alkyldeoxyribonucleo-sides in the pooled fraction was carried out by the

w xcompetitive RIA as described previously 23 .In a separate experiment, the concentrations of the

corresponding unmodified deoxyribonucleosides inDNA were quantified. For this purpose, HPLC sepa-ration of 5 ml of the hydrolysate was performed on a

ŽTSK gel ODS-80TM column 4=150 mm; Tosoh,.Tokyo . Normal nucleosides were eluted with 0.1 M

Ž .ammonium acetate pH 5.0 which containedmethanol, and quantitated by the absorbance at 260

nm. The methanol content was increased from 5% toŽ .15% during the first 15 min linear increase , then to

40% in the following 5 min, and was 100% duringthe last 10 min. The adduct levels in DNA weredetermined as the ratios between the molar concen-trations of the alkyldeoxyribonucleoside and the re-spective unmodified counterpart.

3. Results

3.1. Toxicity

The survival curves for the 3-h treatment ofDrosophila with NDMA and NDEA are shown in

Fig. 4. Separation of O6-methyldeoxyguanosine and O6-ethylde-oxyguanosine by HPLC. Analysis of DNA hydrolysates from

Ž .Drosophila larvae exposed to 5 mmolrdish NDMA a and withŽ .30 mmolrdish NDEA b . The retention time of caffeine as

internal standard was 24 min. Deoxynucleosides and caffeine wereŽ .monitored by UV absorption at 260 nm , methyl

Ž . Ž .adducts – Ø – and ethyl adducts – ( – were detected by theradioimmunoassay, and their amounts are represented as percent-age of inhibition in tracer–antibody binding.

( )Y. Goto et al.rMutation Research 425 1999 125–134130

Fig. 1. The lethality is represented as the ratio be-tween the numbers of adult flies that survived fol-lowing the treatment with nitrosamine and the analogtreatment without it. The LD values for Drosophila50

larvae were found to be 70 mmol NDMArdish and150 mmol NDEArdish.

3.2. Somatic cell mutation and recombination

The results of the wing spot test at various dosesof N-nitrosodialkylamines are shown in Table 1; Fig.2. Each type of mutant spot increased in frequencywith the increasing doses of NDMA and NDEA,except the twin spots after NDEA treatment. The

Ž .linearity is clearly seen in Fig. 2A 3-h treatmentsŽ 2for both NDMA and NDEA r s0.85 for either

.reagent . In the linear range, the slopes of the dose–response curves for the totals of spots were 0.3spotsrmmol for NDMA and 0.01 for NDEA in the3-h treatments, and 1.6 for NDMA and 0.026 forNDEA in the 6-h treatments. These values indicatethat NDMA was about 30 times more effective ininducing wing spots than NDEA in the 3-h treat-ments and about 60 times in the 6-h treatments. Thedoses evaluated for wing spot induction by the N-

nitrosodialkylamines were not lethal to the larvaewith the exception of the highest dose of NDEA in

Ž .the 6 h treatment Table 1 . At doses higher than 50mmol, NDMA killed the larvae dose-dependentlyŽ .Fig. 1 . Mutant spots of mwhrflr individuals aredue to somatic events such as chromosomal recom-bination, gene mutation, segmental chromosomal

w xdeletions and nondisjunction 13 . On the other hand,mutant spots of mwhrTM1 individuals are due toevents other than chromosomal recombination, be-cause mitotic crossing-over between homologous 3rdchromosomes is suppressed in this genotype. Thesuppression results from heterozygosity for a com-plex of super-imposed large inversions on the TM1

w w xxbalancer chromosome see references in Ref. 36 .Recombinogenicity was estimated by subtracting thefrequency of mutant spots on the wings ofmwhrTM1-heterozygous flies from the number ofmutant spots on the wings of mwhrflr-heterozygous

Ž .flies Fig. 3 . The relative frequencies of spots in-duced by recombination were 67% with NDMA and53% with NDEA among the total spots. However,the difference between these two percentages maynot be significant.

Table 2O-Alkyldeoxynucleosides formed by NDMA and NDEA in DNA of Drosophila larvae

6 4N-Nitrosodialkylamine Larvae DNA O -AlkyldGuordGuo O -AlkylThdrThdy6 y6Ž . Ž . Ž . Ž . Ž .mmolrdish g wet weight mg =10 =10

3-h TreatmentNDMA 10 10.8 1.1 2.10 ND20 9.0 1.0 4.00 ND30 9.5 1.2 3.96 ND

NDEA 20 11.2 1.2 5.44 ND50 11.0 1.3 11.01 ND80 9.6 1.1 15.80 ND

6-h TreatmentNDMA 5 15.7 1.7 11.61 –10 18.1 1.9 17.43 –20 17.8 1.4 18.54 –30 15.8 1.5 15.69 1.68

NDEA 10 14.8 1.6 9.76 0.8420 15.5 1.7 14.57 0.6350 15.1 2.1 32.36 2.0280 14.6 1.8 65.84 3.88

ND: not done.–: Not detected

Ž . 6Detection limits for each alkyldeoxyribonucleoside in the radioimmunoassay concentration in HPLC-sample : O -MedGuo 0.38 pmolrml,O6-Et dGuo 0.28 pmolrml, O4-MeThd 9.03 pmolrml, O4-EtThd 0.95 pmolrml.

( )Y. Goto et al.rMutation Research 425 1999 125–134 131

Ž 6 Ž . 4 .Fig. 5. Amount of O-alkyl adducts in larval DNA formed by the treatment with NDMA closed symbols: v, ' O -MedG, B O -MeThdŽ 6 4 . Ž . Ž .and NDEA open symbols: `, ^ O -EtdGuo, I O -EtThd for 3 h A and 6 h B .

3.3. Measurement of DNA alkylation products

The DNA hydrolysates from NDMA- andNDEA-treated larvae were analyzed by HPLC–RIA.Both O6-MedGuo and O6-Et dGuo in the Drosophilasamples gave distinct peaks in the RIA screeningŽ . 6 4Fig. 4 . O -AlkyldGuo and O -alkylThd were

Ž .quantified by the competitive RIA Table 2 . Thelevels of DNA alkylation products increased with thedoses of N-nitrosodialkylamines and with the dura-

Ž . 4tion of treatment Fig. 5 . O -MeThd was detectableŽonly at the highest dose of NDMA 30 mmolrdish, 6

. 6 4h treatment . The O -alkyldGuo:O -alkylThd ratioin larval DNA was about 10:1 for NDMA and 15:1for NDEA exposure. The results show that the rates

Fig. 6. Correlation between the levels of O6-alkyldeoxyguano-sines in larval DNA and the number of wing spots induced by agiven dose of either NDMA or NDEA. Closed symbols for

6 Ž .O -MedGuo v 3 h treatment, ' 6 h treatment , open symbols6 Ž .for O -EtdGuo ^ 3 h treatment, ` 6 h treatment .

of alkylation at the O6-site of guanine and at theO4-site of thymine in DNA were rather similar for

Ž . Ž .NDMA ;9% and for NDEA ;6% .

3.4. Relationship between O-alkylation andmutagenicity

As shown in Fig. 6, both the O6-MedGuo and theO6-Et dGuo contents of DNA correlated well withwing spot induction by the respective alkylating

Ž 2 .agents r s0.75 for NDMA, and 0.86 for NDEA .In the same manner, the amounts of O4-EtThdformed by NDEA showed a good correlation with

Ž 2 . Žwing spot induction by NDEA r s0.88 data not.shown .

4. Discussion

Among the N-nitrosodialkylamines, NDMA andNDEA are the structurally simplest and the mostprevalent ones. These two nitrosamines have beenextensively investigated for their carcinogenic poten-

w xcies and their mechanisms of action 24–26 . BothNDMA and NDEA are strong carcinogens in ro-

w xdents, with similar potencies 1 . In one exceptionalcase, the observed TD of NDEA was lower than50

w xthat of NDMA 27 . To our knowledge, however,there are no reports on in vivo studies in which themutagenic potencies of these nitrosamines are com-pared to the levels of methyl and ethyl adductsformed in the DNA of exposed animals.

( )Y. Goto et al.rMutation Research 425 1999 125–134132

Our previous and present studies have shown thatNDMA is more mutagenic and toxic for Drosophilalarvae than NDEA, as determined by the wing spottest, the in vivo DNA repair assay, and the lethality

w x w xtest 15 . Vogel and Nivard 28 reported that NDMAwas more genotoxic than NDEA in the Drosophilaeye mosaic assay.

Ethyl adducts in DNA may cause greater geno-toxic effects than methyl adducts because of greater

w xsteric hindrance. However, Bignami et al. 29 re-ported that cultured CHO cells were more sensitive

Ž .to killing by N-methyl-N-nitrosourea MNU than byŽ .N-ethyl-N-nitrosourea ENU , while similar levels of

O6-alkylguanines were found in the DNA and simi-lar frequencies of mutation with respect to ouabain

w xresistance were observed. Beranek et al. 30 alsoreported that O6-MedGuo and O6-Et dGuo residuesformed in the DNA of CHO cells were equallyeffective in inducing 6-thioguanine-resistant muta-tions. In the present study, we used whole bodies ofDrosophila to examine the correlation between theO6-alkyl adduct formation in DNA and the muta-

Ž .genicity chromosomal recombinations included ofNDMA and NDEA. Equimolar exposure doses ofthese compounds produced similar amounts of O6-

Ž .alkylguanine in DNA Fig. 5 . However, the muta-genic potency of NDMA was 30–60 times greater

Ž .than that of NDEA Fig. 2 . We found that whentreated larvae were chased for 4 h in nitrosamine-freesucrose medium, there was no decrease in the levels

6 6 Žof either O -MedGuo or O -EtdGuo data not.shown . This suggests that the repair rate of EtdGuo

is not different from that of MedGuo.The difference in the mutagenic potency between

NDMA and NDEA may be not simply due to thedifference in the amount of O6-alkylguanine in DNA.There are several possible reasons for the difference

Ž .in the mutagenic potency. In Drosophila larvae, 1a methyl group at the O6 of guanine or O4 ofthymine in DNA may be more mutagenic than an

Ž .ethyl group at these sites, 2 also alkyl adducts other6 4 Ž .than O -G or O -T e.g., N 7-alkylguanine may be

responsible in part for the mutagenicities of theseŽ .nitrosamines, and 3 damage other than alkyl-ad-

ducts may cause the wing spots.It is known that in chemical methylation of DNA,

N 7 of guanine is the major target site both in vitrow x 6and in vivo 7,31,32 . The ratio of N 7- to O -alkyla-

tions of guanine varies greatly among the alkylatingagents used. In vivo alkylation of rat liver DNA byNDMA led to a N 7rO6 alkylation ratio of 10

w xwhereas NDEA gave a ratio of 2–3 32,33 . Al-though N 7-alkylguanine itself is not expected to

w xalter the base-pairing mode 30,31 , Rodriguez-w xArnaiz et al. 34 suggested that the favored forma-

tion of N 7-alkylguanine may facilitate chromosomalrecombination in Drosophila. Apurinic sites andstrand breaks may arise in DNA following N-alkyla-tion. The role of N 7-dGuo alkylation in NDMA-and NDEA-induction of genetic damage inDrosophila remains to be studied.

w xEngelbergs et al. 35 have recently shown that inrat mammary cells, O6-Et dGuo but not O6-MedGuoresidues are preferentially repaired from active genesresulting in strongly reduced frequencies of muta-tions in the Ha-ras gene of mammary tumors in-duced by ethylnitrosourea. However, in Drosophila,cyclobutane dimers in transcribed and non-tran-

w xscribed genes are eliminated at the same rate 36 ,although in rodent cells, this damage is removed

w xselectively from transcriptionally active genes 37 .The mechanism of the gene- specific repair of alky-lation damage is not yet understood. There is thepossibility that the O6-Et dGuo in DNA detected inthe present experiments may have been preferentiallyrepaired before cell proliferation.

ŽAs shown in Fig. 3, recombinogenic activity re-.combinogenicity of NDMA was 20 times higher

than that of NDEA. In Drosophila, chromosomalrecombinations occur randomly, i.e., without gene-specificity. Therefore, the possibility that the higheractivity of NDMA is a result of a preferential repairof O6-Et dGuo at some specific genes seems un-likely. Since the chromosomal recombination-derivedtwin spots were below the detection limit in themwhrflr zygotes, it was not possible to determinewhether recombination to produce the twins were

Ž .indeed induced or not Table 1 . Twin spots arise ifrecombinations occur between the locus flr and thecentromere of the 3rd chromosome. However, theexpression of the flare phenotype is limited by thepartial lethality of flr-homozygous cells. As NDEAhas a low recombinogenicity, twin spots were belowthe detection limit. Part of the mwh single spots maybe caused by recombinations between the locus flrand the terminus of the 3rd-chromosome. The smaller

( )Y. Goto et al.rMutation Research 425 1999 125–134 133

occurrence of total spots in the mwhrTM1 zygotesthan in the mwhrflr may be due to the suppressionof recombination between the centromere and theleft arm terminus of chromosome 3 in the inversion-heterozygous genotype, mwhrTM1. As shown inFig. 3, the mutant spots induced by NDEA were lessin the mwhrTM1 zygotes than in the mwhrflr. Thisfact shows the recombination was induced by NDEAwith lower rate than by NDMA.

There is evidence suggesting that mismatch repairof O6-MedGuo plays a major role in methylation-in-

w x w xduced mutagenesis 38–43 . Zhang et al. 42 havereported that O6-MedGuo-induced intrachromoso-mal homologous recombination in human cells iscaused by single strand breaks occurring in the pro-cess of repairing O6-MedG:T mismatches. On the

w xother hand, Galloway et al. 40 reported that eth-ylated guanines in DNA are not targeted by themismatch repair system. In our present experiments,the recombinogenicity of NDMA and NDEA, sum-marized in Fig. 3, indicates that the activity ofNDEA is only 1r20 of that of NDMA. This suggeststhat methylated bases may more effectively inducechromosomal recombination than ethylated bases.

5. Conclusion

We speculate that NDMA had stronger mutageniceffects than NDEA in our experiments because thereis more mismatch repair with accompanying chro-mosomal recombination involved, if bases are meth-ylated instead of ethylated.

References

w x1 W. Lijinsky, Structure–activity relations in carcinogenesis byN-nitroso compounds, in: T.K. Rao, W. Lijinsky, J.L. EplerŽ .Eds. , Genotoxicology of N-Nitroso Compounds, Plenum,New York, 1984, pp. 189–231.

w x2 S.H. Lu, W.X. Yang, L.P. Guo, F.M. Li, G.J. Wang, J.S.Zhang, J.Z. Li, Determination of N-nitrosamines in gastricjuice and urine and a comparison of endogenous formation ofN-nitrosoproline and its inhibition in subjects from high- andlow-risk areas for oesophageal cancer, in: H. Bartsch, I.K.

Ž .O’Neill, R. Schulte-Hermann Eds. , The Relevance of N-Nitroso Compounds to Human Cancer: Exposures and Mech-anisms, IARC Scientific Publications No. 84, IARC, Lyon,1987, pp. 538–543.

w x3 H. Ohshima, H. Bartsch, Quantitative estimation of endoge-nous nitrosation in humans by monitoring N-nitrosoproline

Ž .excreted in the urine, Cancer Res. 41 1981 3658–3662.w x4 M. Tsuda, N. Frank, S. Sato, T. Sugimura, Marked increase

in the urinary level of N-nitrosothioproline after ingestion ofŽ .cod with vegetables, Cancer Res. 48 1988 4049–4052.

w x5 E. Vogel, A.T. Natarajan, The relation between reactionkinetics and mutagenic action of monofunctional alkylatingagents in higher eukaryotic systems: interspecies compar-

Ž .isons, in: F.J. de Serres, A. Hollaender Eds. , ChemicalMutagens, Vol. 7, Plenum, New York, 1982, pp. 295–336.

w x6 H. Bartsch, R. Montesano, Relevance of nitrosamines toŽ .human cancer, Carcinogenesis 5 1984 1381–1393.

w x7 D.E.G. Shuker, H. Bartsch, DNA adducts of nitrosamines, in:K. Hemminki, A. Dipple, D.E.G. Shuker, F.F. Kadlubar, D.

Ž .Segerback, H. Bartsch Eds. , DNA Adducts: Identification¨and Biological Significance, IARC Scientific PublicationsNo. 125, Lyon, 1994, pp. 73–89.

w x8 J. Doniger, R.S. Day, J.A. Dipaolo, Quantitative assessmentof the role of O6-methylguanine in the initiation of carcino-genesis by methylating agents, Proc. Natl. Acad. Sci. USA

Ž .82 1985 421–425.w x9 B. Singer, J.M. Essigmann, Site-specific mutagenesis: retro-

Ž .spective and prospective, Carcinogenesis 12 1991 949–955.w x10 D. Umbenhauer, C.P. Wild, R. Montesano, R. Saffhill, J.M.

Boyle, N. Huh, U. Kirstein, J. Thomale, M.F. Rajewsky,S.H. Lu, O6-Methyldeoxyguanosine in oesophageal DNAamong individuals at high risk of oesophageal cancer, Int. J.

Ž .Cancer 36 1985 661–665.w x11 N. Huh, M.S. Satoh, J. Shiga, M.F. Rajewsky, T. Kuroki,

Immunoanalytical detection of O4-ethylthymine in liver DNAof individuals with or without malignant tumours, Cancer

Ž .Res. 49 1989 93–97.w x12 T. Suzuki, T. Itoh, M. Hayashi, Y. Nishikawa, S. Ikezaki, F.

Furukawa, M. Takahashi, T. Sofuni, Organ variation in themutagenicity of dimethylnitrosamine in Big Blue mice, Envi-

Ž .ron. Mol. Mutagen. 28 1996 348–353.w x13 U. Graf, F.E. Wurgler, A.J. Katz, H. Frei, H. Juon, C.B.¨

Hall, P.G. Kale, Somatic mutation and recombination test inŽ .Drosophila melanogaster, Environ. Mutagen. 6 1984 153–

188.w x14 F.E. Wurgler, E.W. Vogel, In vivo mutagenicity testing using¨

somatic cells of Drosophila melanogaster, in: F.J. de SerresŽ .Ed. , Chemical Mutagens, Vol. 10, Plenum, New York,1986, pp. 1–72.

w x15 T. Negishi, T. Shiotani, K. Fujikawa, H. Hayatsu, Thegenotoxicities of N-nitrosoamines in Drosophilamelanogaster in vivo: the correlation of mutagenicity in thewing spot test with the DNA damages detected by the

Ž .DNA-repair test, Mutat. Res. 252 1991 119–128.w x16 R. Muller, M.F. Rajewsky, Immunological quantification by¨

high-affinity antibodies of O6-ethyldeoxyguanosine in DNAŽ .exposed to N-ethyl-N-nitrosourea, Cancer Res. 40 1980

887–896.w x17 J. Adamkiewicz, G. Eberle, N. Huh, P. Nehls, M.F. Rajew-

sky, Quantitation and visualization of alkyl-deoxynucleosidesin the DNA of mammalian cells by monoclonal antibodies,

Ž .Environ. Health Perspect. 62 1985 49–55.

( )Y. Goto et al.rMutation Research 425 1999 125–134134

w x18 J. Adamkiewicz, W. Drosdziok, W. Eberhardt, U. Langen-berg, M.F. Rajewsky, High-affinity monoclonal antibodiesspecific for DNA components structurally modified by alky-lating agents, in: B.A. Bridges, B.E. Butterworth, I.B. Wein-

Ž .stein Eds. , Indicators of Genotoxic Exposure, BanburyReport 13, Cold Spring Harbor Laboratory Press, New York,1982, pp. 265–276.

w x19 G. Ebele, Dissertation, Univ. Essen, 1989.w x20 P.B. Farmer, A.B. Foster, M. Jarman, M.J. Tisdale, Alkyla-

tion of 2X-deoxyguanosine and of thymidine with diazoalka-Ž .nes. O-Alkylation, Biochem. J. 135 1973 203–213.

w x21 T. Jowett, Preparation of nucleic acids, in: D.B. RobertsŽ .Ed. , Drosophila, A Practical Approach, IRL Press, Oxford,1986, pp. 275–286.

w x22 C.P. Wild, G. Smart, R. Saffhill, J.M. Boyle, Radioim-munoassay of O6-methyldeoxyguanosine in DNA of cells

Ž .alkylated in vitro and in vivo, Carcinogenesis 4 19831605–1609.

w x23 M.F. Rajewsky, R. Muller, J. Adamkiewicz, W. Drosdziok,¨Immunological detection and quantification of DNA compo-

Žnents structurally modified by alkylating carcinogens ethyl-. Ž .nitrosourea , in: B. Pullman, P.O.P. Ts’o, H. Gelboin Eds. ,

Carcinogenesis: Fundamental Mechanisms and Environmen-tal Effects, Reidel, Dordrecht, 1980, pp. 207–218.

w x24 J.A. Swenberg, D.G. Hoel, P.N. Magee, Mechanistic andstatistical insight into the large carcinogenesis bioassays onN-nitrosodiethylamine and N-nitrosodimethylamine, Cancer

Ž .Res. 51 1991 6409–6414.w x25 R. Peto, R. Gray, P. Brantom, P. Grasso, Dose and time

relationships for tumor induction in the liver and esophagusof 4080 inbred rats by chronic ingestion of N-nitroso-diethylamine or N-nitrosodimethylamine, Cancer Res. 51Ž .1991 6452–6469.

w x26 A.M. Camus, O. Geneste, P. Honkakoski, J.C. Bereziat, C.J.Henderson, C.R. Wolf, H. Bartsch, High variability of ni-trosamine metabolism among individuals: role of cy-tochromes P450 2A6 and 2E1 in the dealkylation of N-nitrosodimethylamine and N-nitrosodiethylamine in mice and

Ž .humans, Mol. Carcinog. 7 1993 268–275.w x27 E.W. Vogel, A. Barbin, M.J.M. Nivard, H. Bartsch, Nucle-

ophilic selectivity of alkylating agents and their hyper-mutability in Drosophila as predictors of carcinogenic po-

Ž .tency in rodents, Carcinogenesis 11 1990 2211–2217.w x28 E.W. Vogel, M.J.M. Nivard, Performance of 181 chemicals

in a Drosophila assay predominantly monitoring interchro-Ž .mosomal mitotic recombination, Mutagenesis 8 1993 57–

81.w x29 M. Bignami, E. Dogliotti, G. Aquilina, A. Zijno, C.P. Wild,

R. Montesano, O6-Methyltransferase-deficient and -profi-cient CHO cells differ in their responses to ethyl- andmethyl-nitrosourea-induced DNA alkylation, Carcinogenesis

Ž .10 1989 1329–1332.w x30 D.T. Beranek, R.H. Heflich, R.L. Kodell, S.M. Morris, D.A.

Casciano, Correlation between specific DNA-methylationproducts and mutation induction at the HGPRT locus in

Ž .Chinese hamster ovary cells, Mutat. Res. 110 1983 171–180.

w x31 B. Singer, DNA damage: chemistry, repair, and mutagenicŽ .potential, Regul. Toxicol. Pharmacol. 23 1996 2–13.

w x32 D.T. Beranek, Distribution of methyl and ethyl adductsfollowing alkylation with monofunctional alkylating agents,

Ž .Mutat. Res. 231 1990 11–30.w x33 E.W. Vogel, Nucleophilic selectivity of carcinogens as a

determinant of enhanced mutational response in excisionrepair-defective strains in Drosophila: effects of 30 carcino-

Ž .gens, Carcinogenesis 10 1989 2093–2106.w x34 R. Rodriguez-Arnaiz, P.O. Soto, J.C.G. Oyarzun, U. Graf,

Analysis of mitotic recombination induced by several mono-and bifunctional alkylating agents in the Drosophila wing-

Ž .spot test, Mutat. Res. 351 1996 133–145.w x35 J. Engelbergs, J. Thomale, A. Galhoff, M.F. Rajewsky, Fast

repair of O6-ethyl, but not O6-methylguanine, in transcribedgenes prevents mutation of Ha-ras in rat mammary tumori-genesis induced by ethylnitrosourea in place of methylni-

Ž .trosourea, Proc. Natl. Acad. Sci. USA 95 1998 1635–1640.w x36 P.J.L. van der Helm, E.C. Keink, P.H.M. Lohman, J.C.J.

Eeken, The repair of UV-induced cyclobutane pyrimidinedimers in the individual genes Gart, Notch and white fromisolated brain tissue of Drosophila melanogaster, Mutat.

Ž .Res. 383 1997 113–124.w x37 I. Mellon, G. Spivak, P.C. Hanawalt, Selective removal of

transcription-blocking DNA damage from the transcribedŽ .strand of the mammalian DHFR gene, Cell 51 1987 241–

249.w x38 A. Kat, W.G. Thilly, W.-H. Fang, M.J. Longley, G.-M. Li, P.

Modrich, An alkylation-tolerant, mutator human cell line isdeficient in strand-specific mismatch repair, Proc. Natl. Acad.

Ž .Sci. USA 90 1993 6424–6428.w x39 G.T. Pauly, S.H. Hughes, R.C. Moschel, Response of repair-

competent and repair-deficient Escherichia coli to threeO6-substituted guanines and involvement of methyl-directedmismatch repair in the processing of O6-methylguanine

Ž .residues, Biochemistry 33 1994 9169–9177.w x40 S.M. Galloway, S.K. Greenwood, R.B. Hill, C.I. Bradt, C.L.

Bean, A role for mismatch repair in production of chromo-some aberration by methylating agents in human cells, Mu-

Ž .tat. Res. 346 1995 231–245.w x41 L.J. Rasmussen, L. Samson, The Escherichia coli MutS

DNA mismatch binding protein specifically binds O6-meth-Ž .ylguanine DNA lesions, Carcinogenesis 17 1996 2085–

2088.w x42 H. Zhang, T. Tsujimura, N.P. Bhattacharyya, V.M. Maher,

J.J. McCormick, O6-Methylguanine induces intrachromoso-mal homologous recombination in human cells, Carcinogene-

Ž .sis 17 1996 2229–2235.w x43 D.R. Duckett, J.T. Drummond, A.I.H. Murchie, J.T. Rear-

don, A. Sancar, D.M.J. Lilley, P. Modrich, Human MutSarecognizes damaged DNA base pairs containing O6-methyl-

4 Ž .guanine, O -methylthymine, or the cisplatin-d GpG adduct,Ž .Proc. Natl. Acad. Sci. USA 93 1996 6443–6447.