Journal o f Photochemistry and Photobiology, B: Biology, 5 (1990) 179 - 195



Institut Curie-Biologie, CNRS UA 1292, 75231 Paris Cddex 05 (France) E. CUNDARI

Istituto di Mutagenesi e Differenziamento, CNR, Pisa (Italy) M. DARDALHON

Institut Curie-Biologie, CNRS UA 1292, 75231 Paris Cddex 05 (France) F. DALL'ACQUA and D. VEDALDI

Department o f Pharmaceutical Sciences, University o f Padova, 5 Via Marzolo, 35131 Padova (Italy) (Received March 30, 1989; accepted July 25, 1989)

Keywords. 8-Methoxypsoralen, pyranocoumarin, homopsoraien, yeast DNA mono-adducts, DNA interstrand cross-links, mutation, alkaline step elution, mitotic recombination.

Summary The relationship between DNA mono- and di-adducts and genetic effects induced b y the pyranocoumarin 8,8-desmethylxanthyletine (homopsoralen) HP and 365 nm radiation (UVA) was investigated in the diploid yeast strain D7 (Saccharomyces cerevisiae) taking 8-methoxypsoralen (8MOP) as a reference c o m p o u n d . The number of DNA cross-links (CLs) induced was determined using alkaline step elution analysis. The induction and removal of total photo-adducts was followed using radioactively labelled compounds. HP showed the same photobinding capacity as 8-MOP. As a function of UVA dose, it was less effective than 8-MOP for the induction of CLs and genetic effects. However, as a function of CLs induced, HP was shown to be more effective for the induction of lethal effects and mitotic recombination than 8-MOP but equally effective for the induction of mutations. The results suggest that, although CLs are recognized as genetically effective lesions, at a given number of CLs, HP induced mono-adducts efficiently contribute to the induction of lethal effects and mitotic recombination but less to the induction of mutations. Using a re-irradiation protocol, HP was brought to yield the same relative amounts of CLs at the same number of total adducts as single UVA exposures with 8-MOP. In these conElsevier Sequoia/Printed in The Netherlands

180 ditions, mutation induction and the kinetics for the removal of photoadducts were the same for both agents indicating that n o t only the removal of adducts but also mutation induction are highly dependent on the relative level of CLs induced.

1. Introduction Bifunctional furocoumarins, such as 8-methoxypsoralen (8-MOP), 5methoxypsoralen (5-MOP) and 4,5',8-trimethylpsoralen (TMP), are widely used in the p h o t o c h e m o t h e r a p y of various human skin diseases. Consequently, much work has been devoted to the determination of their genotoxic activity (see refs. 1 - 4 for reviews). In many biological systems, the high mutagenic and recombinogenic activity of bifunctional furocoumatins has been found associated with their capacity to induce DNA interstrand cross-links (CLs). The high mutagenic and recombinogenic activity of bifunctional furocoumarins is thought to be related to their capacity to photoinduce interstrand cross-links in DNA. For instance, in yeast, bifunctional furocoumarins, such as 8-methoxypsoralen (8-MOP), were shown to be more efficient than monofunctional furocoumarins [5] and increased yields of cross-links obtained either by a re-irradiation protocol [6] or by exposure at selective wavelengths [7, 8] correlated with increased induction of genetic damage. As with monofunctional furocoumarins, clear differences in photomutagenicity were observed also amongst bifunctional furocoumarins. For example, in yeast [9] and Chinese hamster V-79 cells [10], at equal levels of photoadducts induced in DNA, 5-MOP induced lesions were less mutagenic than those induced by 8-MOP. From their capacity to photobind to DNA and to form DNA interstrand cross-links, it was inferred that the results were due to different ratios of mono- and di-adducts and/or to different isomeric types of photo-adducts induced [10, 11]. Recently, several pyranocoumarins, intracyclic homologues of psoralen, have been synthesized [ 1 2 - 16] amongst which 8,8-desmethylxanthyletine or homopsoralen (HP) attracted much interest because of its relatively high photobiological activity on haploid yeast with respect to the induction of lethal effects and cytoplasmic " p e t i t e " mutations [13]. From this and from its molecular structure, it was deduced that HP is likely to react as a bifunctional c o m p o u n d inducing mono- and di-adducts in DNA at a ratio largely in favour of mono-adducts [13]. Indeed, recent studies on the photophysical and photochemical properties of HP in vitro have shown that HP behaves similarly to bifunctional furocoumarins; it is able to form a complex with DNA in the dark, to bind covalently to DNA in the presence of UVA (365 nm) radiation and it induces mono-adducts and interstrand cross-links in DNA in vitro [17]. The present study was undertaken in order to determine in diploid yeast the relationship between DNA mono- and di-adducts and genetic


effects induced by HP using 8-MOP as a reference compound. The number of DNA interstrand CLs induced was assessed by the alkaline elution technique [7, 8], and the induction and removal of total photo-adducts were followed using, as previously, radioactively labelled c o m p o u n d s [9]. The results reveal that the relative level of HP or 8-MOP induced CLs is an important parameter with respect to the removal of photo-adducts and the induction of mutagenic effects.

2. Experimental details

2.1. Photoreactive compounds We used chromatographically pure 8-MOP (Chinoin, Italy) and 8,8desmethylxanthyletine, a pyranocoumarin, (HP) (see Fig. 1 for chemical structures) synthetized by Faulques et al. [13]. The molar extinction coefficients e36snm of 8-MOP and HP in ethanol-water solution were 1367 and 3273 M -1 cm -1, respectively. All studies were performed at a final concentration of 50 pM by adding ethanolic stock solutions of the c o m p o u n d s to the cell suspension (final concentration of ethanol 1.3 %).

~~"o O C H 3 B-MOP

~~"o HP

Fig. 1. Chemical structures of 8-methoxypsoralen (8-MOP) and pyranocoumarin (homopsoralen) (HP).

The radioactively labelled c o m p o u n d s in methanolic solution had the following characteristics: (3H)--HP prepared and purified according to Vedaldi et al. [17] (spec. act. 23 Ci mmo1-1, concentration 3 × 10 -2 M), and (3H)-8-MOP (spec. act. 74 Ci mmo1-1, concentration 1.35 X 10 -s M) was a commercial preparation TRK-778 Batch 1 from Amersham (U.K.). The radioactive compounds were chromatographically pure.

2.2. Yeast strain The diploid strain D7 of Saccharomyces cerevisiae (a/s, ilvl-92/ilvl-92, trp5-12/trpS-27, ade2-40/ade2-119) (Zimmermann et al., 1975) was used to determine survival, the induction of reverse mutations (ILV+), mitotic gene conversion (TRP +) and genetically aberrant colonies including mitotic crossing over between the centromere and the ade2 locus, as previously described [9]. 2.3. Exposure to UVA As previously [9], for exposure to UVA, a HPW 125 Philips lamp was used {peak emission at 365 nm) at a dose rate of 20 J m -2 s-x, as determined


by a digital radiometer J-260 (Ultraviolet Products, Inc., San Gabriel, CA 91778, U.S.A.). Exposures were performed at room temperature. The reirradiation procedure was the same as that previously described [7], except for using exposures to UVA. The biological data are expressed as a function of incident dose as well as a function of photolesions induced (total adducts and CLs). The latter includes all possible corrections for differences in the physico-chemical properties of the c o m p o u n d s including differences in UV absorption.

2.4. Media and culture conditions Cells were grown to stationary phase in liquid complete medium (1% yeast extract (Difco), 2% bacto peptone (Difco), 2% glucose distilled water at 30 °C). At a concentration of 2 × l 0 s cells m1-1, t h e y were harvested, washed twice, resuspended at a concentration of 2 × 107 cells m1-1 in distilled water and incubated in the presence of the photoreactive c o m p o u n d at 50 ttM for 15 min at room temperature in the dark. Subsequently, cells were exposed to UVA (365 nm) radiation. The determinations of survival and genetic effects were done as previously [9]. Experiments were performed at least twice. Figures represent results from typical experiments. The experimental variation from one experiment to the other was less than 15%.

2.5. Determination o f DNA photobinding and removal of photoadducts Cells were treated with 3H-labelled 8-MOP and HP (see above), and the radioactivity b o u n d to the DNA was determined as previously described, immediately [9] and after different times of post-treatment incubation in liquid complete medium [18]. 2.6. Determination of DNA cross-links by alkaline step elution analysis We used a slightly modified method [7, 8] based on that described by Zuk et al. [19]. Cells were radioactively labelled by growth in minimal medium BYNB without amino acids (Difco) and 4% glucose for 16 h at 30 °C in the presence of 5 ttCi m1-1 o f (8-3H)-adenine (spec. act. 23 Ci mmol-1). After harvest, cells were resuspended in fresh non-radioactive minimal medium and incubated under shaking for 30 min at 30 °C. Spheroplasts were prepared according to Zuk et al. [19]. The alkaline step elution technique consisted in lysing the yeast spheroplasts on appropriate filters (Millipore, type BSWP, 25 mm, 2 tzm pore size) in the presence of 1 mg m1-1 proteinase K (BBH Chemicals Ltd., Poole, England) at room temperature followed by a selective stepwise separation of single stranded DNA fragments of different molecular weight eluted by alkaline solutions (tetrapropylammonium hydroxide, Fluka A.G. Buchs, Switzerland) of increasing pH values. After treatment with 8-MOP or HP and UVA, cells were exposed to 6°Co-7-irradiation (1000 Gy) in order to obtain DNA fragments containing no more than one single photo-adduct per DNA fragment. The presence of one CL in a DNA fragment being sufficient to


prevent its elution, the number of CLs was determined by calculating the number of DNA fragments retained on the filters [ 8].

3. Results

3.1. D N A photobinding o f 8-MOP and HP Figure 2 illustrates the photobinding of 8-MOP and HP to yeast cell DNA after treatment of diploid yeast cells with radioactive compounds (at 50 pM) and UVA. Per unit dose of UVA, 8-MOP and HP show an about equal photobinding to DNA. For example, at 5 KJ m -2 of UVA, about five adducts per 104 base pairs (7 × 103 adducts per genome) are induced by both compounds.



~5, (,J

o 0


UVA DOSE (kJm 2)


Fig. 2. DNA photobinding of 8-methoxypsoralen (8-MOP) and pyranocoumarin (homopsoralen) (HP) at 50 pM in diploid yeast (D7) as a function of UVA radiation dose.

3.2. Induction o f D N A interstrand CLs Figure 3 shows the results obtained after treatment of diploid yeast cells with 8-MOP and HP (50 #M) and UVA. Per unit dose of UVA, HP photo-induces about 2.6 times fewer CLs per genome than 8-MOP; for instance, at 5 KJ m -2 UVA, HP induces about 32 CLs per genome whereas 8MOP induces about 82 CLs per genome. 3.3. Survival data When diploid yeast cells were treated with either 8-MOP or HP and UVA in the same conditions as above, shouldered survival curves were obtained (Fig. 4). As a function of dose, i.e. at equal photobinding to DNA, cell killing was clearly more effective with 8-MOP than with HP and UVA. At 10 % survival, the dose modification factor was about 2.

184 200-



O z O w 100v,


D ,-I ,

0 u


0 Y





UVA DOSE (kJm 2)

Fig. 3. Induction of DNA interstrand CLs by 50 pM 8-methoxypsoralen (8-MOP) or 50 pM pyranocoumarin (homopsoralen) (HP) and UVA radiation in diploid yeast (D7) as determined by alkaline elution (see Experimental details).

3.4. Induction o f mutation and mitotic recombination As shown in Fig. 5, per unit dose of UVA (i.e. equal photobinding), HP was less effective than 8-MOP for the induction of reverse mutation ( I L V + revertants) (Fig. 5a), mitotic gene conversion (TRP+convertants) {Fig. 5b) and other genetic alterations including mitotic crossing over (Fig. 5c). When plotting the results as a function of survival, HP was less efficient than 8-MOP for the induction of mutations (Fig. 6a) as well as genetic alterations including crossing over (Fig. 6c). However, it was equally as efficient as 8-MOP for the induction of mitotic gene conversion (Fig. 6b) indicating that gene conversion is closely related to cell survival. 3.5. Induction o f genetic effects as a function o f D N A CLs In order to illustrate the effectiveness of photo-induced CLs, we plotted in Fig. 7 the results obtained for the induction of genetic effects as a function of the number of CLs determined by alkaline elution analysis. At equal

185 UVA DOSE (kJrn -2)


1.0 I~----.,,..~ =





L9 Z



8-MOP 0.0001

Fig. 4. Survival in diploid yeast (D7) cells obtained following treatment with 50 btM 8methoxypsoralen (8-MOP) or 50 pM pyranocoumarin (homopsoralen) (liP) as a function of UVA radiation dose.

numbers of CLs induced, HP induced lesions are slightly more effective on cell killing (Fig. 7a) and on the induction of mitotic gene conversion (Fig. 7c) than 8-MOP induced lesions, whereas the two are equally efficient for the induction of I L V + reverse mutations (Fig. 7b). In order to better define the differences obtained between HP and 8-MOP photo-induced effects on the induction of point mutation and gene conversion as a function of CLs, a linear regression analysis was performed. The analysis of the theoretical regression lines confirm the observation stated above. For the induction of I L V + revertants by HP and 8-MOP, the regression constants were 1.48 + 0.11 and 1.23 + 0.33, and the coefficients 0.055 + 0.004 and 0.059 + 0.0006, respectively. For the induction of T R P ÷ convertants, the regression constants were 1.9 + 0.07 and 0.7 + 0.23, and the coefficients 0.06 + 0.0023 and







,.~ ~




~o~ •- ~








~ "~


Genetic effects and repair of DNA photo-adducts induced by 8-methoxypsoralen and homopsoralen (pyranocoumarin) in diploid yeast.

The relationship between DNA mono- and di-adducts and genetic effects induced by the pyranocoumarin 8,8-desmethylxanthyletine (homopsoralen) HP and 36...
647KB Sizes 0 Downloads 0 Views