Ahaation Research. DNA Repait; 273 (1992) 243-251

243

~ 1992 Elsevier Scievce Publishers B.V. All rights reser~,c:!092!-8777/92/$~5.0(t

MUTDNA (16472

Sensitivity and single-strand D N A break repair in Chinese hamster m~atants exposed to the carcinogen aflatoxin B 1 epoxide and its dichloride model Elizabeth A. Martin and Raymond

Waters

Molecular Biology" Research Group, School of Biological Sciences, Unice~sity C~)llegeof Swansea, Sin~4e~on P,.~rk, Swansea, SA2 8PP, Wales (Grr'at Br#ain)

(Received 12 March 199t) (Revision received 15 July 1991) (Accepted 18 Ju~y 199t

Keywords: Aflatoxin BI epoxide; Aflatoxin Bi dichloride; Chir~esehamster mutants: DNA repair: DNA single-strand breaks

Summary Aflatoxin B1 (AFB~) is a potent carcinogen and mutagen, tt requires metabolic activation to be converted to the DNA-binding product aflatoxin B~ epoxide (AFt3~-epoxide). A model of this epoxide is af!atoxin B~ dichlorice (AF13.~CI ?). Both react a~ the N 7 position of guanine to form large adducts. The major adduct formed ca:~ either be rapidly removed to leave an apur'n:¢ site or can undergo ring opening of the imidazole ring to form a chemically stable adduct. A ;v.mber ef Chinese hamster D N A repair-deficient mutants have been screened for their sensitivity to AFB~-epoxide and AFB:C12. Some of the mutants screened belong to different CV comp!ementation groups. Human genes involved in nucleotide excision-repair correct deficiencies found in these complementadon groups. The mutants which were found to be most sensitive to AFB~ (V-C4 and V-H1) weie further ir,vestigated. Alkaline elution was used to measure AFB~-induced D N A single-strand break repair in the mutar.ts. V-H1 repaired completely in 24 h whereas V.C4 displayed only partial repair.

Aflatoxin B~ (AFB1), a potent hepatocarcinogen and mutagen is a secondary metabolite of the fungus Aspergillus flacus. Oxidative metabolism at the 8,9-double bond of the terminal furan ring is required for AFB~ to be converted to the ultimate carcinogen, 8,9-dihydro-8,9-epoxs'aflatoxin B~ (AFB1-epoxide) (Swenson et al., 1977).

Correspondence: Dr. E.A. Martin, Molecular Biology Research Group, School of Biological Sciences. University College of Swansea, Singleton Park, Swansea, SA2 8PP~ Wales (Great Britain).

The epoxide reacts exclusively al the N 7 position of guanine nucleotides in DNA, to form large adducts. The major adduct is 8,9-dihydro8(7-guanyl)-9-hydrox3'aflatoxin B~ (AFB~-N 7-G), both in vitro (Essigmann et al., 1977) and in vivo (Croy et al., 1978). This primary adduct enters one of two competitive pathways. The main pathway is the rapid removal of AFB~-NT-G both in vitro (Groopman et al., 1981) and in vivo (Croy and Wogan, 1981) to leave an apurinic site. The second pathway involves the opening of the imidazole ring of the guanine molecule to form chemically stable A F B t - f o r m a m i d o p y r i m i d i n e

244

adducts. A positive charge is induced in the imidazole ring leading to alkali-labile sites in DNA. The positive charge leaves the C ~ susceptible to mmleophilic attack b y hydroxide ions, this results in the imidazole ring opening, forming two ringopen products collectively refered to as the AFBi-fapy adducts (Hertzog et al., 1982). AFBi-8,9-dich!oride (AFBIC! 2) has been used as a model for the epoxide. This also reacts at the N v position of guanine to form bulky adducts which lie in the major groove of DNA. AFB~CI 2 has the same reactivity and mutagenic charcteristics as the epoxide (Wood et al, 19881. The same ring-closed and ring-open adducts are formed, except the AFB~CI 2 adducts have a chloride ion at position 9 of the aflatoxin B~ moiety. 2 h after dosing rats with AFB~ the major adduct is the AFB~-NT-G adduct (80%). The second major adduct is the ring-open (7%). During the next 24 h 7(1% of the ring-closed adducts are either removed to leave an apurinic site and 20% of ring-closed adducts are converted to the AFB~-fapy adducts. This means that the ringclosed adduct decreases in abundancy whereas

the ring-open adduct increases (Croy and Wogan, 1981). The rate of AFB cN 7-G adduct removal in vivo far exceeds that observed in vitro (Groopman et al., 1981). This is because in vivo spontaneous hydrolyses is facilitated by enzymatic removal (Croy and Wogan, 1981). Bulky adducts are formed in DNA by chemicals such as AFB~, N-acetoxy-acetylaminofluorene and 4-nitroquinoline 1-oxide (4NQO). These cause helical distortion in DNA. Such damage can be removed in part by the major DNA-repair pathway - - nucleotide excision repair which also operates on UV-induced DNA damage (Friedberg, 19851. In Escherichia coli the uc~vl, uvrB and ucrC genes code for the nucleotide excision-repair enzymes (HowardFlanders et at., 19661. E, coli mutant strains deficient in one or more of these genes ap" unable to repair DNA damage caused by bulky adduces and UV photoproducts (Lindahi, 1982). The ring-open fapy lesions maybe .removed by a second excision-repair pathway - - base excision repair. In E. coti the formamidopyrimidine-DNA glycosylase ( f a p y - D N A glycosylase) excises

TABLt?_ 1 I N C R E A S E FN S E N S I T I V I T Y a O F C H I N E S E H A M S FFR M U T A N T S T O D N A D A M A G I N G A G E N T S U V Comp. Group

Celt line

D N ' o damaging agent AFBt c epoxide

AFB 1 ~ Ct,

UV ~

~NQO h

3Me4NQO ~

X-ray h

MMC "

I 2 2 2 3 5 -

UV20 UV5 C-A6 V-HI 2%1 UVI35 V-H4 V-C4 V-E5

ND 2,3 ND 2,2 ND 2.6 1.2 ND ND

1.t 1,5 1.5 1.8 1.6 ND I. 1 2.8 1.3

5.5 d 5.5 d 2.6 10.0 7.0 t 5.5 ~ 1.3 2.4 1.4

2.7 i 2.1 i 4.5 5.3 1.0 t t.6 ' 2,7 2.4 1.4

ND ND 4,9 95 2,6 ND 3.0 1.5 ! .33

1.3 d l.,q d ND 1.4 1.0 r ND 1.6 2.9 2,7

112.0 5.0 2.3 !.7 2.5 5,0 33.3 2.2 3."

ND, no data available. " Increase in sensitivity is expressed as: D m of parental l i n e / D m of mutant, b Zdzienicka and Simons (t9871. Personal communication with M.Z, Zdzienicka. a Thompson e t a ! , ( 19801. This paper. i Zdzienicka and Simons (19861. g Thompson et al. (19821. l~ Thompson and Carrano (1983). i Yang et al. (1991}.

a a

I h

245 A F B ~ - f a p y adducts from D N A both in vivo and in vitro by cleavage of the glycosyl bond (Chetsanga and Frenette, 19831. Fapy D N A glycosylase is coded for by the fpg gene which has been cloned, sequenced and m a p p e d (Boiteux et al,~ 1987; Boiteux and Huisman, 19891. In recent years a number of DNA-repair-deficient mutants of Chinese hamster cells have been isolated, on the basis of their sensitivity to D N A damaging agents such as UV, mitomycin C, X-rays and Neomycin. Many of the UV-sensitive mutants have been placed into complementation groups. Currently 10 groups have been identified. These rodent mutants are being used to identity specific human genes involved in nucleotide excision repair. 6 human genes have been reported which efficiently and specifically correct the defect in the first 6 rodent UV-complementation groups, l'hese are ERCC1 (excision-repair crosscomplementing) to ERCC6. Most of these genes have been cloned and some have been characterized. Data showing an increase in sensitivity of some of these hamster mutants to different D N A damaging agents are showe in Table I. In tiffs p a p e r a number of these m~:tants have been screened for their sensitivity to AFB,CI 2 and AFB~-epoxide. Those that were sensitive to AFB~C12 have been assayed for their D N A - r e pair capacity using the alkaline elution technique. This method was selected because the positive charge induced in the imidazole ring of the adducted guanine molecule renders the adduct alkaline-labile. In addition, if the adducts have been removed either by depurination of the A F B ~-N 7-G, or by base excision of the A FB ~-~apy adduct an alkaline-labile apurinic site would be left.

( N U N C ) containing Dulbecco's modified Eagles medium ( D M E M ) plus 10% donor calf serum (DCS, Gibco). AA8, UV5, UV20 and UVt35 also contained 1% non essential amino acids (MEM, Gibco). The cell-~ were kept at 37 ° C in a sealed flask with 10% C O 2 / 9 0 % air a d d e d until the medium was a light red colour.

Chemicals Aflatoxin B 1 ( A F B t) was purchased from Sigma. Aflatoxin B i dichloride (AFB1C12) was synthesized by bubbling chiorine gas through a solution of AFB~ in dichloromethane (CH2CI 2) for 4 - 5 min. The reaction was complete after about 5 mm. The dichloride is stable in 19:1 C H , C I 2 / a c e t o n e at - 20°C for several weeks. The CH2CI 2 and acetone were removed by rotary evaporation and the AFBICI 2 dissolved in D M S O prior to use. Aflatoxin B~ epoxide (AFBl-cpoxide) was synthesized following a method by Baertschi et al., 1988. In brief, the A F B t was dissolved in acetone and the oxidizing agent dimethyldioxirane (Murray and J e y a r a m ~ , !9851 was a d d e d for 15 min at room temperature. The solvent and excess dimethyldioxirane were removed by evaporation in a stream of nitrogen. The epoxide was dissoIved in acetone. The epoxide is stable at room t e m p e r a t u r e for 12 h or at - 1 0 ° C fcr long periods. For survival experiments the epoxide was used as a solution in acetone. For both the AFB~C12 and AFB~-epoxide, the concentrations used in the experiments are the co~centrations of the AFB~ originally dissolved in the solvent prior to chlorination or epoxidation. t~ is assumed that the conversions are approximately 1 : 1.

Surcicat experiments Materials and methods

Cell culture The parental cell lines V79 and CHO-9 and their respective mutants V°HI, V-C4, V-E5, V-H4 and 27-1 and C-A6 were provided by Dr. M.Z. Zdzienicka, Leiden. The parental cell line A A 8 and its mutants UV5, UV20 and UV135 were provided by Dro L.H. Thompson, Livermore. All cells were routinely cultured in 80-cm z flasks

Ceils growing exponentially were trypsinized and 200-300 cells (depending on the cell line) were plated into 90-ram dishes and allowed to attach for 4 h at 37°C in an atmosphere containing 5% C 0 2 / 9 5 % air. In the control samples this gave a frequency of about I00 colonies/dish. The medium was removed and the cells were washed with serum-free medium, this was removed and 10 ml of ficsh serum-free medium was added. The AFB1CI 2 was a d d e d as a solution in D M S O

246

so as no more than 0 5 % DMSO was present at the highest AFBICI z concentration. AFB~epoxide was added as a solution in acetone so as no more than 1.8% acetone was present at the highest AFB~-epoxide concentration. Each dose was performed iv. triplicate and the control plates were incubated with D M S O or acetone equivalent to the highest dose. The cultures were incubated as above for 1 h. The medium was removed and the dishes were washed twice with 10 ml of serum-free medium. 10 ml of growth medium were added and the cells were incubated for 7-10 days in the above atmosphere. The medium was removed and the dishes were rinsed twice with cold phosphate-buffered saline (PBS), then fixed in methanol for 20 rain. After air-drying the dishes were stained with 0.25% methylene blue for 15 rain and the colonies were counted. Colonies were identified as those containing 50 or more cells.

Surt,ival curw,s All survival experiments were carried out at teast 3 times. The curves presented are the mean of at least 3 experiments. The survival data is compared at D m (dose required to reduce the surdval to 10%). Alkaline ehaio, To assay the repair of single-strand breaks in D N A a modified method of alkaline elution developed by Kohn et at. (1981) was used. The parental ce!! . 2 2 "(79 and its mutants V-C4 and V-H1 were used. Cells growing exponentially were t~ypsinized and 2 × 105 cells were plated into 30-ram dishes Tile cells were allowed to attach for 4 h by incubating at 37 ° C in an atmosphere of 5% C O , / 9 5 % air. Following attachment tl4C]thymidine (54 m C i / m m o l e ) or [-~H]thymidine ( 5 C i / m m o l e ) was added to the growth medium at 0.02 and 0.1 , a C i / m l respectively. The cells were labelled for approximately 12 h. 3 h prior to treatment the labelled medium was removed and fresh growth medium was a d d e d to allow newly synthesized D N A to reach mature size (Kohn et al., 1974). The growth medium was removed and the [~4C]thymidine labelled cel!, were treated with 2.0 , a g / m l AFB~C12 in serumfree medium and incubated as above for 1 h.

After treatment the medium was removed and the dishes were .,,,shed with growth medium and then 2 ml of fresh growth medium was added. The cells wele incubated for 0, l, 3, 6 and 24 h. After tee repair incubation the growth medium was removed, the cells were washed with ice-cold PBS and then detached in 2 ml of ice-cold PBS with a rubber policeman. Immediately before the elution the [3H]thymidine labelled cells were washed with ice-cold PBS and irradiated with 6.9 Gy (GEC X-ray unit with a 50KVp tube Machlett, U.K.). This was operated at 15KVp 7.5 mA av`d the doses measured with a P T W - d i a m e t e r - D dose meter (Physikalisch-Technische Werk~tatten, Freiburg, F.R.G.). The irradiated cells were detached in 2 ml of ice-cold PBS with a rubber policeman. A n equal mixture of AFBIC12-treated cells and irradiated cells were layered onto a 25 mm diameter, 2 ,am pore size, polycarbonate filter (Nuclepore). (0.5-2.0 × 106 cells/filter). 5 mI of room t e m p e r a t u r e lysis solution (25 mg proteinase K (Boehringer), 4 ml 25% SDS (BDH) and 46 mt 0.02 M E D T A ) was a d d e d and allowed to drip through the filter. The filters were then washed with 3 ml of washing solution (0.02 M E D T A , p H 9.7-10.2). D N A was eluted in the dark from the filters in eluting solution (20 mM E D T A , 50 mM tetrapropylammonium hydroxide (Aldrich), p H 11.9) at a rate of 3 m l / h . 15 fractions were collected at 60-;7:in intervals. Each fraction was mixed with 10 mi of scintillation fluid (Packard pico-fluor 40) for scintillation counting. D N A remaining on the filters was hydrolysed by heating in 0.4 ml 1 N HCI at 60 ° C for 1 h then 2.5 ml of 0.4 N N a O H was added and the samples were left at room t e m p e r a t u r e for 1 h. 10 ml of scintillation fluid was added. All samples were counted in a Beckman LS3801 counter on a dual label program for 1 rain.

Results Sensitivity to aflatoxhz B~ dichloride Fig. 1 shows the survival of the wild-type V79 and the mutants V-H4, V-E5, V-H1 and V-C4 following exposure to AFBICI 2. V-H4 and V-E5 showed a slight sensitivity (about t.1 and 1.3 times respectively) (Table 1) compared to the parental cells in terms of Dl0 value. Both these

24? 10o-

g

~o

10

,,=, r~

o

o:~

o:2

o:a

&

gs

0

0625

0-675

0,'100

AFLATOXIN BI DICHLORIDE (~glmI)

A F L A T O X I N B 1 DICHLORIDE {ptglmt)

Fig. 1. Survivalcurves after aflatoxin BI dichloride treatment of V79 (~) and mutant cell lines; V-It4. B; V-E5, G; V-HI, ~ ; V-C4, ~.

0"()50

Fig. 2. Survivalcurves after a,qa~,{:xinB= dichloride treatment of AA8 (e) and mutant cel~~h~es;UV20, ~: UV5, A.

Sensitivigv to aflatoxin Bl-epoxide mutants are sensitive to the cross-linking agent mitomycin C. V-H1 (UV complemcntation group 2) which is hypersensitive to UV, 4NQO and its methyl derivative 3Me4NQO showe~ an increase in sensitivity, to AFB~C12 of about 1.8 times (Table 1). V-C4 showed an even larger increase in sensitivity to AFB~C12 of about 2.8 times (Table 1). V-C4 is sensitive to UV, X rays and 4NQO. Fig. 2 shows the smvival of the wild-type AA8 and the mutants UV20 and UV5 foli,,wing exposure to AFB~CI z. UV20 (UV comptementatio~ group t) which is hypersensitive to MMC and UV showed a very slight sensitivity to AFB~Ct 2 of about 1.1 times (Table 1). UV5 (UV camp. gp. 2) also sensitive to MMC and UV, was sensitive to AFBtC1 z. It showed an increase in sensitivity of about 1.6 times (Table 1). Fig. 3 shows the survival of the wild-type CHO-9 and the mutants C-A6 and 27-1 following AFBiCI z exposure. C-A6 (UV corer, gp. 2) is sensitive to 4NQO, 3Me4NQO, UV and MMC. 27-1 (UV comp. gp. 3) is sensitive to UV, 3Me4NQO and MMC. Both these mutants showed an increase in sensitivity to AFB :C! z similar to that of UV5 of about 1.6 times (Table 1).

Fig. 4 shows the survival of the wild-type V79 and the mutants V-HI and V-H4 following epoxide exposure. The UV complementation group 2 mutant V-HI was about 2.2 times (Table t) more

10

0

LN

0'1

012

0'3

\

014

0"5

AFLATOX|N BI DICHLORIDE (pglm:~

Fig. 3. Surviva!curves aiter aflatoxin B~ dichloride treatment of CHO-9 (e) and mutant celt lines;C-A6: i : 27-1, ..~.

248

1 0 0 ~

% BREAK~

REMAINING

100

,,.a

80

60

V-C4

40 20 0

1

3

6

24

REPAIR TIME (HOURS) Fig. 6. P e r c e n t a g e of single-strand D N A breaks r e m a i n i n g after aflatoxin B Z dichloride t r e a t m e n t of V79 a n d the m u t a n t cell lines V-C4 a n d V - H I .

k 0

1'0

20

3'0

AFLATOXl81 NEPOXlD(jEag/ml)

46

Fig. 4. Survivalcurves after aflatoxin Bl epoxide treatment of V79 (e) and mutant cell lines; V-H4, m: V-HI. A.

sensitive to the epoxide than V79. The MMC mutant V-H4 was only very slightly sensitive to A F B r e p o x i d e (about 1.2 times) (Table 1).

Fig. 5 shows the survival of the wild-type A A 8 and the mutants UV5 and UV135 following exposure to A F B r e p o x i d e . Both the UV5 (comp. gp. 2) and UV135 (comp. gp. 5) are sensitive to A F B r e p o x i d e . The UV135 is slightly more sensitive than the UV5 (about 2.6 times compared with 2.3 times) (Table 1).

Repair of s&gie-strand DNA breaks

I00-

10.

x"~o~..,,,,~

~.=~ 0

16 AFLATOXIN

26 B.I E P O × I D E

30

40

QJglml)

Fig. 5. Survival curves after aflatoxin B r epoxide t r e a t m e n t of A A 8 (o) a n d n m t a n t celt lines; UV5, A; UV135, ©.

Fig. 6 shows a graph of perc~;nt single-strand breaks remaining against repair time of the parental cell line V79 and the mutants V-C4 and V-H1 following exposure to AFB~C12. In these three cell lines 2 . 0 / ~ g / m l of AFB1C12 induced a similar level of single-strand breakage, namely 22 b r e a k s / 1 0 u~ dalton. The graph shows that after 1 h repair time the normal cell line and the mutant V-C4 have about 70% breaks remaining, whereas the mutant V-H1 has 90% remaining. After 3 h repair time the V-H1 mutant has increased its repair rate and has 40% breaks remaining compared to the normal V79 which has about 50%. The mutant V-C4 has slowed down its repair rate and has about 60% remaining. After 6 h repair time the same trend is evident. V-C4 has about 50% breaks remaining, V79 has about 40% and the mutant V-H1 has about 20% remaining. A t 24 h after the treatment V-HI and the normal V79 have virtually no breaks remaining, whereas the mutant V-C4 has 40% of breaks still remaining.

249 Discussion In this study a number of DNA-repair-deficient rodent mutants were screened for their increased sensitivity to the carcinogens AFB~CI 2 and AFB~-epoxide. A F B v e p o x i d e is the ultimate carcinogen of A F B 6 AFB1CI 2 is a model of the epoxide. All the repairmeficient mutants used were isolated on the basis of their sensitivity to UV-irradiation except V-H4 which was isolated on the basis of its sensitivity to MMC. Two thirds of the mutants screened have been placed into UV complementation groups (see Table 1). Mutams from groups 1, 2, 3 and 5 were represented in this study. 3 mutants were negligibly sensitive to AFB~CI ~. These were UV20 from UV complementation group 1 and V-H4 and V-E5 which are not UVsensitive. V-H4 was not sensitive to A F B v e p o x i d e either. These results suggest that these mutants have no defect in excision repair of AFB~CI 2, and V-H4 has no defect in excision repair of either of the aflatoxin bulky addt~c~s. However, V-H4 is sensitive to the large adducts formed by 4 N Q O and 3 M e 4 N Q O (Tab:e 1). The mutant 27-1 (UV comp. gp. 3) was sensitive to the dichloride. 27-I had an increased sensitivity of abeut 1.8. The mutants in UV complementation group 2, UV5. "-H1 and C-A6 had abo,~t th,~ same sensitivity to the dichloride as each other, although V - H i showed the most sensitivity and C-A6 showed the least. The sensitivity of the group 2 and group 3 mutants to the dichloride was vepy similar. UV5 and V-H1 were more sensitive to the epoxide than they were to the dichloride; C-A6 was not screened for sensitivity with the epoxide. UV135 a UV complementation group 5 mutant was more sensitive to the epoxide than either of the two group 2 mutants, with an increased sensitivi~: of about 2.6. V-C4 which is UV-sensitive but has not been placed into a UV complementation group was the most sensitive mutant screened. It was only tested with the dichloride and showed an increase in sensitivity, of 2.8. In figs. ! and 3 the parental cell lines V79 and CHO-9 have approximately the same survival curve, whereas in Fig. 2 the parental ceti iine A A 8 appears to have a survival curve 5 times lower than that of V79 and CHO-9. This is due to

two different batches of AFB1CI 2 being used. Survival experiments (data not shown) show that A A 8 and V79 have almost identical sensitiviG~ to killing by AFB~CI 2. Thus, in this p a p e r two thirds of the mutants (UV5, C-A6, V-H1, 27-t, UVt35 and V-C4) are sensitive to killing by AFB~. The same batches of AFB~C12 and AFB~-epoxide result in the increased lethality of the nucleotide excision-repair-deficient mutant, ut'#M, but not the fapy D N A glycosylase mutant of E. coil; a number of the tad3 group yeast mutants are also more sensitive than the wild-type, and xeroderma pigmentosum g~ 3up A and C cells exhibit increased sensitivities as well (Waters and Martin, unpublished observations). These results indicate that sensitivit5., of the cell lines to AFB~ is due to D N A damage and not some trivia! toxic effect. Furthermore. the Chinese hamster parental celt lines V79 and its mutants V-HI and V-C4, and the parental cell line A A 8 and its mutant UV5 exhibit the same initial level of D N A damage after treatment with a given dose of AFB~C12 or A F B v e p o x i d e (Martin and Waters. unpublished observations). This indicates that sensitMty of the mutants is not due to differential uptake o r D N A binding of the chemicals. The first five UV comp!ementation groups are known to be defective in the incision step of nucleotide excision repair of both UV lesions (Thompson et al.. t982) and bulky chemical adducts (Thompson et al., 1984). All the group 2, 3 and 5 mutants screened were sensitive to aflatoxin which forms bulk)' adducts in DNA. This implies that these mutants are se.nsitive because they are unable to perform the incision step to remove the AFB~ adducts. The mutant UV20 from group 1 was however unexpectedly not sensitive to AFB~Ct 2. This suggests that either the repair gene E R C C i which is required for the repair of UV and MMC is not required for the repair of aflatoxin bu!ky adducts, or that UV20 is an allele which produccs a mutant product able to repair AFB~ damage but not UV or M M C damage. Three mutants were treated with both the epoxide and the dichloride. In all cases the mutants were more sensiti~e to the epoxide. In the case of V-H4 it was negligible, but with the two

250 UV cornptementation group 2 mutants UV5 and V-H1 the difference was 0.7 and 0.4 increase respectively. The dichloride is an extremely good model of ihe epoxide. They produce the same' adducts which appear to behave comparably during incubation of modified D N A in mildly alkaline and mildly acidic environments (Wood et al., 1988). Therefore there is no immediate explanation for the difference in sensitivity to the two aflatoxin compounds. It could possibly be due to a difference in the solvent used; although the acetone used in the epoxide treatments for the mutant cell lines did not exceed 1.2% and for the parental cell line did not exceed 1,8% at the highest dose, and this was not toxic and did not effect colony formation, so this is unlikely. The alkaline elution results showed that the V-C4 mutant was the most inefficient at repairing the single-strand breaks induced by aflatoxin. After 24 h when the normal cell line had approximately 10% breaks remaining, V-C4 had about 40% remaining. These results show that this mutant is partially defective in the excision repair of aflatoxin induced damage. The results also suggest that the mutation in this cell line may be 'leaky' so that some aflatoxin adducts are detected and repaired by the cells, or that the repair genes mutated in this celt line are not fully required for repair. V-H1 showed complete repair of AFB~C1 z induced single-strand D N A breaks after 24 h. From approximately 2 h after the treatment onwards the V-H1 mutant repaired the breaks more efficiently than the the parental cell line. V-H1 is deficient in the production of the ERCC2 protein. It has been hypothesised that the ERCC2 protein has at least two functions (Thompson et al., 1989). V-HI and UV5 are defective in the removal of pyrimidine dimers but V-H1 unlike UV5 has intermediate repair of (6,4) photoproduct~ (Mitchell et a l , 19891. It has been shown that C H O celt survival correlates to repair of pyrimidine dimers in active genes (Bohr et al., 1987; 1988). Therefore, repair of nontranscribed regions of D N A may have little effect on the survival of the cell. One function of E R C C ? could therefore be the ability to interact with the repair complex, this is essential for repair and UV5 would lack this. The second function may be the

ability to determine active gene repair and V-H1 would lack this ability while retaining part of the overall repair function, by repairing unactive genes. This theory would mean that V-H1 is repairing AFB~CI 2 damage in nontranscribed regions of the DNA. However, repair in these regions does not aid cell survival and explains why the cell line is sensitive to aflatoxin in the survival experiments. Finally, one must bear in mind that the ringopen AFB~ adduct may be repairable via a f a p y D N A glycosylase. Hence ring-closed lesions not repaired by nucleotide excision may be later repairable once ring opening has occurred. This possible alternative would complicate the interpretation of data concerning the incidence of single-strand breaks. The probing of ring-closed and ring-open lesions in specific genes using the UvrABC excinuclease and th,v f a p y - D N A glycosylase respectively will help resolve this issue.

Acknowledgements This work was funded by a grant from the E E C Environmental Programme, The authors would like to thank Dr. L.H, Thompson and Dr. M.Z. Zdzienicka for supplying the cell lines used in this study.

References Baertschi. S.W., K.O. Raney, M.P. Stone and T.M. Harris (19881 Preparation of the 8.9-epoxide of the mycotoxin aflatoxin Bl: The ultin~zte carcinogenic species, J. Am. Chem. Soc., 110, 7929-7931. Bohr. V.A., D.H. Phillips and P.C. Hanawalt (19871 Heterogeneous DNA damage and repair in the mammalian genome, Cancer Res., 47. 6426-6436. Bohr, V.A., E.H.Y. Chu, M. Van Duin, P.C. Hanawalt and D.S. Okumoto (19881 Human repair, gene restores normal pattern of preferential DNA repair defective CHO cells, Nucleic Acids Res, 16, 7397-7403. Boiteux, S., and O. Huisman (1989) Isolation of a formamidopyrimidine-DNA glycosyt, ~ (Jpg) mutant of Escherichia coli KI2, Mol. Gen. Genet., 215, 300-305. Boiteux, S,, T.R. O'Connor and J. Laval (19871 Formamidopyrimidine-DNA glycosylase of Escherichia coli: Cloning and sequencing of the fpg structural gene and over production of the protein, EMBO J.. 6, 3177-3183. Chetsangar, C.J., and G.P. Frenette (19831 Excision of aflatoxin B~-imidazolering opened guanine adducts from DNA by formamidopyrimidine-DNA glycosylase,Carcinogenesis, 4, 997-1000.

25t Croy, R.G., and G.N. Wogan 11981) Temporal pat*erns of covalent DNA add,,v~ in _~ctliver after single and multiple doses of aflatoxin B I, Cancer Res., 41, 197-203. Croy, R.G., J.M. Essigmann. V.N. Reinhold and G.N. Wogan (1978) Identification of the principal aflatoxin B~-DNA adduct formed in vitro in rat liver, Proc. Natl. Acad. Sci. (U.S.A.), 75, 1745-1749. Essigmann, J.M., R.G. Croy, A.M. Nadzan, W.F. Busby Jr., V.N. Reinhold, G. Buehi and G.N. Wogan (t977,~ Structural identification of the major DNA adduct formed by aflatoxin B I in vitro, Proc. Natl. Acad. Sci. (U.S.A.), 74. t 870-1874. Friedberg, E.C. (1985) DNA repair, Freeman, New York. Groopman, J.D., R.G. Croy and G.N. Wogan (1981) In vitro reactions of aflatoxin B l-adducted DNA, Proc. Natl. Acad. Sci. (U.S.A.), 78, 5445-5449. Hertzog, P.J., J.R. Lindsay-Smith and R.C. Garner (1982) Characterization of the imidazole ring-opened forms of trans-8.9-dihydro-8-(7-guanyl)9-hydro~cj aflatoxin Bp Carcinogenesis. 3. 723-725. Howard-Flanders, P., R.P. Boyce and L Theriot (1966) Three loci in Escherichia coli K-12 that control the excision of pyrimidine dimers and certain other mutage~ products from DNA. Genetics, 53, tlt9-1166. Kohn, K.W., C.A. Friedman, R.A.G. Ewig and Z.M. tqbal (19741 DNA chain growth during replication of asynchronous LI210 cells Alkaline elution of larger DNA segments from cells lysed on filters. Biochemistry, 13, 4134-4!39. Kohn. K.W., R.A.G. Ewig, L.C. Erickson and L.A. Zelting (198t) Measurement of strand breaks and cross-links by alkaline elation, in: E.C. Yriedberg and P.C. Hanawalt (Eds.), DNA Repair: A Labarato~' Manual of Research Procedures, Vol. 1, Part B. Mar el Dekker. New York, pp. 379-401. Lindahl, T. (1982) DNA repair enzy'mes, Anna. Rev, Biochem.. 52, 61-87. ~iitchell. D,L., M.Z. Zdzienicka. A.A. van Zeeland and R.S. : Nairn (1989) Intermediate 16-4~ photoproduct repair in Chinese hamster V79 mutant V-HI correlates with intermediate levels of DNA incit:.ion and repair replication. Mutation Res,, 266, 43-47. Murray, R.W.. and R. Jeyaraman (!985} Dioxiranes: Synthesis and reactions of methyldioxiranes. J. Or g. Chem., 50. 2847-2853. Swenson, D.H.. J-K. Lin, E.C. Miiter and J.A. Mi!k r ,q9771

Aftatoxin B~-2.3-oxide as a probable intermediate in the covalent binding nf aflatoxins B 3 and B 2 to rat liver DNA and ribosomal RNA in vivo, Cancer Res.. 37, 172-18I. Thompson, L.H,, and A.V. Carrano (1983) Analysis of mammMian cell mutagenesis and DNA repair using in vitro selected CHO cel: mutants. In: E.C. Friedbe,:g and B,A, Bridges (Eds,). Cellular Responses to DNA Damage. Liss, New York. pp. 125-143. Thompson, LH., J.S. Rubin. J.E. Cleaver. G.F. Whitmore and K. Brookman (1980) A screening method for isolating DNA repair-deficient mutants of CHO cells, Somat. Cell Genet,. 6. 391-405, Thompson, L.H,. K.W. Brookman. L.E. D~tleha,v. C.L. Mooney and A.V. Carrano f 1982) H vpersensitivi~' to mutation and sister-chromatid-exehange induction in CHO cell mutants defective in incising DNA containing UV lesions, Somat, Cell Genet., 8, 759-773. Thompson. L,H,. K.W. Brookma~ and C.L. Mooney (19841 Repair of DNA adducts in asynchronous CHO cells and the role of repair in cell killing and mutation induction in synchronous calls treated with 7-bromomethytbenz[a]anthracene. Somat. Celt Mok Genet., t0. 183-194. Thompsom L.H., C,A. Weber and N.J. Jones (19891 lluman DNA repair and recombination genes, In: M. kamben and J. Laval (EdsJ. DNA Repair Mechanisms and Their Biological Implications i_~ Mammalian Cells, Plenum, New York. pp. 537-561. Wood. M.L., J.R. Lindsay-Smith and R.C. Garner (I988) Structural characterization of the major adducts obtained after reaction of an u!~_imate carcinogen aflatoxin B~-dichloride with calf thymus DNA in ,Atro. Cancer Res., 48, 5391-5396. Yang, A-L.. M.Z. Zdzienicka. J.W.LM. 3imons and R. Waters (1991~ The repair of ~nitroquinoline-t-exide induced DNA adducts in hypersensRive Chinese hamster mutants: tack of repair of UV induced (6.4) photoproduct correlates with red~ced repair of adducts at the N-" of guano~ine. Mutage~esis. in press. Zdzienid~a. M.Z.. and J.W,LM. Simons (t986) Anakvsis of repai~ process by the determination of the induction of cetl killing and mutations in p,vo repair-deficient Chinese hamster ovary cell lines. Mutation Res.. 166. 59-69. Zdzienicka, M.Z., and J.W.t.M. Simons (t9871 Mutagen-sensitire cell lines are obtained with a high frequency in V79 Chinese hamster ceils. Mutation Res.. 178. 235-244.