Photochernrriry and Phorohrology. Vol. 29. pp. 277 ZX I . 0 Pcrg.imon P r o * Lld.. 1979. Prinlcd in Grc.ii Britiiin
3-(a,a-DIMETHYLALLY L)-PSORALEN A LINEAR FUROCOUMARIN FORMING MAINLY 4’,5’-MONOFUNCTIONAL ADDUCTS WITH DNA D. VEDALDI,F. DALL’ACQUA*, S. CAFFIERI and G. RODIGHIERO Institute of Pharmaceutical Chemistry of the Padua University, Centro di Studio sulla Chimica del Farmaco e dei Prodotti biologicamente attivi del C.N.R.. 3510 Padova. Italy (Received 10 May 1978: accepted 6 July 1978)
Abstract-3-(a,~dimethylallyl)-p~oralen,a furocoumarin derivative recently isolated from Horriu arborea. has been studied with regard to its photoreacting (365nm) capacity with native DNA. This
compound photobinds to DNA to a much lower extent than psoralen, forming mainly 4.5’-monofunctional adducts, due to the presence in the 3-position of 3.2-dimethylallyl group. which exerts a considerable hindrance both for the formation of a molecular complex with native DNA and for the photoreaction of the 3,4-double bond. This site of the molecule photoreacts, and, only 2.5% of the bound substance forms cross-linkages.Therefore. 3-(z.~-dimethylallyl).psoralen behaves as a fairly pure monofunctional reagent.
INTRODUCTION
The molecular basis of the photobiological activity of furocoumarins is connected with their ability to photobind to DNA (Musajo and Rodighiero, 1972; Pathak et a/., 1974). As furocoumarins have two photoreactive sites (3.4- and 4‘3-double bonds, see Fig. l), both monofunctional and bifunctional adducts can be formed with the pyrimidine bases. Bifunctional adducts. forming interstrand cross-linkages, appeared to be more directly connected with some photosensitized effects [skin-erythema (Dall’Acqua e l al.. 1974); killing of bacteria (Cole, 1971)] than the monofunctional. However, the contribution of monofunctional adducts in producing biological effects has been well evidenced and studied (Bordin er al., 1975; Baccichetti rt a!.. 1976; Seki et a!.. 1978). Averbeck et a/. (1975) have recently demonstrated that the formation of bifunctional adducts with DNA induces mainly nuclear gene mutations in yeast cells, while the formation of monofunctional adducts is more effective in inducing cytoplasmic “petite” mutations. Therefore. it is of great interest to have both furocoumarin derivatives forming bifunctional adducts and derivatives forming only monofunctional adducts. While many linearly condensed furocoumarins are known to be capable of forming bifunctional adducts with DNA (cross-linkages; e.g. psoralen. 8-methoxypsoralen. 5-methoxypsoralen. 4,5’.8-trimethylpsoralen). only angelicin has been widely studied for its ability to form only monofunctional adducts with DNA (due to its angular structure) (Dall’Acqua et a/., 1971). Averbeck (1976) and Ou et al. (1978) also used 3-carbethoxy-psoralen and
* To whom correspondence should be sent.
a
CH3 ;,CH=CH,
0
5
Figure 1. Molecular structure of 3-(a,rr-dimethyIallyI)psoralen. dimethoxycoumarin, respectively, which are capable of forming only monofunctional adducts. In this connection, we have examined the naturally occurring 3-(a,adimethylaflyI)-psoralen (Fig. 1). In fact, the presence of the dimethylallyl group in the 3-position could decrease the accessibility of the 3,4-double bond which is the most reactive site for the photoreaction (Song et a/., 1971; Mantulin and Song, 1973). while the possibility of a cycloaddition at the level of 4‘3’-double bond should remain unchanged. In this paper we report on the results obtained concerning the possibility both of a complex formation with DNA in the dark and of a photoreaction.
MATERIALS AND METHODS
D N A . Calf-thymus DNA was purchased from Sigma Chemical Company; its hypochromicity. determined according to Marmur and Doty (1962). was higher than 40%. Furocoumarins. 3-(r,z-dimethylallyl)-psoralen. extracted from Hortiu arhorea (Delle Monache et a/.. 1976). was a kind gift of Prof. G . B. Marini Bettolo. University of Rome; psoralen was extracted from the leaves of Ficus carica (Dall’Acqua et al.. 1968); angelicin was a gift of the Franco-Indian Pharmaceutical Co.. Bombay. India. These were tritiated by the Radiochemical Centre of Amersham . (England) and purified according to a procedure described elsewhere (Marciani et al., 1969); they had the following
211 ’
D. VEDALIX F. DALL'ACQUA. S CAFFIERI and G . RODIGHIERO
278
specific activities: 3-(z.x-dimethylallyl)-psoralen 3.26 Ci/ mol: psoralen 2 Ci/mol and angelicin 1.5 Ci/mol. Irradiution procrdurr. A small measured volume of a concentrated alcoholic solution of 3H-3-(sc.sc-dimethylaIlyl)psoralen. or of 3H-psoralen or 'H-angelicin for comparison. was added to an aqueous solution (2.2 mM) of DNA containing 2 m M NaCI, up to a concentration of 53.8 pM. Samples (3 m / ) of the prepared solutions were introduced into calibrated glass tubes. immersed in a thermostatically controlled bath and irradiated for different times by means of two HPW 125 Philips lamps emitting almost exclusively at 365 nm. The total intensity incident on the 3 m/ solution. determined by using a chemical actinometer (Hatchard and Parker, 1956). corresponded to 1.75 x IO-'J s - ' . After irradiation. each sample was divided into two portions: 1.5 m/ were used for the crosslinkage determination (solution A : see later): from the remaining 1.5 m/ the macromolecule was precipitated by addition of 3m/ of absolute ethanol. washed with 80;b ethanol and redissolved in the initial volume of water (solution B). This solution has been used for the radioactivity and fluorescence measurements. Radioactioity rneasurenients. Small volumes (0.2 m/) of solution B. diluted with I m/ of water and added to 10 m/ ofdioxane base scintillator (P.P.O. 5 g, P.O.P.O.P. 0.075 g. naphthalene 120 g. dioxane up to 1000 m/ of solution) were counted for the evaluation of the radioactivity acquired by DNA during the irradiation. A Packard Model 3375 liquid scintillation spectrometer was used; the efficiency of the apparatus for counting tritium was within the range 26-34",. Fhrorirnrtric ~c~rc~rriiinatioris. I m/ of solution B (see above). acidified by addition of 0.1 m/ of Concentrated HCI. was introduced into vials. which were sealed and heated in a boiling water bath for 60 min. After neutralization with NaOH and addition of 0.8 m/ phosphate buffer. pH 6.98. to a volume of 2 m/. the fluorescence intensity was determined using a Perkin-Elmer model MPF-44 spectrophotofluorimeter: exciting radiation : 337 nm: fluorescent radiation: 395 nm. This procedure had previously been elaborated (Dall'Acqua et al.. 1978) for evaluating the extent of the fluorescent 4'.5'-cycloadducts formed in DNA as a consequence of irradiation in the presence of psoralen. avoiding the quenching due to the intercalation of the cycloadducts between two base pairs in double stranded DNA. Specrrophotometric measurements. For the determination of the UV absorption spectra of the compound in the presence and in the absence of DNA, a Perkin-Elmer Model 180 double beam spectrophotometer was used. Dtterrnination of rht cross-linkages. The determination of the cross-linkages formed in DNA was carried out on the basis of the evaluation of the renaturation capacity of cross-linked DNA after heat-denaturation (Dall'Acqua rt al.. 1974). 1 S m / of solution A (see above). heated for 10 min in a boiling water bath and then immersed in ice for 15 min. was chromatographed on a column (0.7 x 4 cm) of hydroxylapatite Bio-gel type (Bio-Rad Laboratories. CA. U.S.A.) and developed using a linear
gradient of 0.05-0.3 M phosphatc buffer pH 6.98. with a flow rate of 15 drops per min. Fractions of 3.5 m/ were collected and the absorbance at 260 nm was determined in each fraction. On the basis of the single stranded fraction (non-cross-linked) and of the double stranded fraction (renamed. cross-linked) of DNA separated and detcrmined in this way. the extent of cross-linking formation was evaluated according to Lawley and Brookes (1967). Equilibrium dialysis e.xprrinwnts. Cylindrical cells. having 1.5 m/ halfcell volume. were used. The two halfcells were divided by Visking dialysing membrane (Visking Corp.. U.S.A.). In one part of the cell the aqueous solution of the tritiated 3-(z.a-dimethylallyl)-psoralen. containing NaCl 2OmM and ETDA I mM was introduced. while in the other part of the cell was introduced the DNA solution (in the range 0.45-3.08mM) having the same ionic strength. The cells were mechanically shaken for 12 h in a thermostatically controlled bath at 25 & 0.05"C in the dark. After the 'shaking period. small measured volumes (0.2 m f ) of the two phases were utilized for radiochemical measurements; in such a way the concentrations of the psoralen derivative were determined. Skin-photosensiti_ariorr. Measured volumes of a 0. I"< methanol solution of the substance were placed by means of a micropipette on a 4cm' area of depilated albino guinea-pig skin. corresponding to an amount of 50pg/cm'. the solvent was rapidly evaporated by means of an air stream and the skin was irradiated for 1 h with a Philips HPW 125 lamp. emitting mainly at 365 nm (5.5 W/m'): the animals were observed 12, 24 and 48 h after irradiation according to a test (Musajo and Rodighiero. 1962) commonly used for evaluating the skin-photosensitizing activity of furocoumarins.
RESULTS AND DISCUSSION
Tkr molecular complex formation It is known that furocoumarins form molecular complexes with DNA through intercalation between two base pairs of the macromolecule (Dall'Acqua and Rodighiero, 1966; Dall'Acqua. 1977); the role of this complex formation is important for the successive photoaddition (Rodighiero er a/.. 1970: Dall'Acqua rt a/.. 1971). The possible formation of a molecular complex between 3-(!x.x-dimethylaIlyl)-psoralen and DNA has been studied by spectrophotometric and fluorimetric measurements and by equilibrium dialysis experiments. Very small modifications in the UV-absorption spectrum of the compound were observed in the presence of DNA. Fluorimetric measurements. too. showed small modifications of the flourescence intensity (about increase) when D N A was added t o a n aqueous solution of 3-(r.r-dimethylallyl)-psoralen.
Table 1. Amounts of 3-(z.z-dimethylallyl)-psoralen bound with DNA. at various concentrations of the macromolecule. evaluated by equilibrium dialysis experiments. DNA concentration ( M of Pi x lo5) (1)
aqueous phase (2)
DNA-containing phase (3)
complexed with DNA (4)
ratio l/4
308 297 136 75 45
0.24 0.36 0.40 0.50 0.58
1.13 1.04 0.82 0.78 0.76
0.89 0.68 0.42 0.28 0.18
346 437 324 27 I 252
3-(z.z-dimethylallyl)-psoralenconcentration ( M x 10')
Monoadduct of furocoumarin to DNA
219
than those of psoralen and lower also than those of angelicin. Considering that under the experimental conditions used for the photoreaction, 3-(a,a-dimethylallyl)-psoralen is complexed within DNA only for 8.6%, we have calculated the rate constant of its photobinding (pseudo-first order reaction with respect to the furocoumarin, Dall’Acqua, 1977; Dall’Acqua et ul., unpublished results) on the basis of the initial concentration of the complexed fraction, instead of the total amount present in the solution. In this way the rate constant Photobinding to D N A calculated for 3-(a,a-dimethylallyI)-psoralen(9 x 10The amount of 3-(a,a-dimethylallyl)-psoralen which mol/l/min) was analogous to that of psoralen can photobind with DNA was studied by irradiating (8.5 x l o w 2mol///min) calculated also on the basis for increasing periods of time an aqueous solution of the complexed substance. Therefore, the low of 3H-compound (13.6pg/md) and of DNA (2.3mM amount of 3-(a,a-dimethylallyl)-psoralen photobindin respect to the fl content); after irradiation the ing to DNA is to be correlated to its low capacity macromolecule was precipitated by two volumes of to form the molecular complex with DNA rather than ethanol to remove the excess of unbound furocou- to its photoreacting capacity. marin. The precipitated DNA was washed with 80% ethanol and redissolved in water. This solution was 4’,5’-Fluorescent cycloudducts counted to evaluate the amount of the substance 4’,5’-Cycloadducts are formed by the cycloaddition covalently linked to DNA. of one 3-(a,a-dimethylallyl)-psoralen molecule, In Fig. 2 are reported the amounts (nmol) of 3-(a,a- through its 4,s-double bond, to one pyrimidine base dimethylally1)-psoralen bound to 1 mg of DNA as a of DNA. In the case of other furocoumarins studied, function of the time of irradiation. For comparison, this type of photoadduct, when excited with 365nm the amounts of psoralen and of angelicin bound to light, shows a strong violet fluorescence; by contrast, DNA, evaluated under the same experimental condi- 3,4-cycloadducts and bifunctional adducts, when irrations, are also reported. diated at 365 nm, are practically non-fluorescent The data obtained show that 3-(a,a-dimethylally1)- (Musajo et al., 1967; Dall’Acqua, 1977). Consequently, psoralen photobinds to DNA in amounts much lower the fluorescent properties acquired by DNA after irradiation at 365nm in the presence of a furocoumarin have been referred to the formation of 4‘,5‘-cycloadducts. In a recent note (Dall’Acqua et a!., 1978), we described a method for a quantitative evaluation of the 4‘,5’-cycloadducts formed by psoralen in DNA; the fluorescence intensity was determined after hydrolysis of the irradiated DNA to avoid the fluorescence quenching due to the intercalation of the 4,5’-cycloadducts in DNA. In the present study, the extent of 4,5’-monofunctional adducts formed in DNA by irradiation at 365 nm in the presence of 3-(a,a-dimethylallyI)-psoralen has been evaluated by measuring the fluorescence intensity acquired by DNA after irradiation. For the previously mentioned reason, the fluorescence intensity was measured after hydrolysis of the irradiated samples of DNA. The results obtained are reported in Fig. 3a in comparison with those obtained by irradiating DNA in the presence of equimolecular amounts of psoralen. In Fig. 3b are reported the values of the “specific fluorescence”, obtained by dividing the measured fluorescence intensities by the amounts of 3-(a,a-dimethylally1)-psoralen bound to DNA, determined by radioactivity measurements. As both fluorescent and 5 10 20 30 non-fluorescent adducts can be formed, this may be Time of irradiation (Minutes) a better indication of the extent of 4,5’-monofuncFigure 2. Photobinding of psoralen, angelicin and 3-(a,a- tional adducts than the simple fluorescence intensity. dimethylally1)-psoralen to calf thymus DNA. For a comparison, Fig. 3b also reports the “specific
Evidence for the complex formation was obtained by equilibrium dialysis experiments. The amounts of 3-(cc,cl-dimethyallyl)-psoralen bound to DNA at various concentrations of the macromolecule are reported in Table 1. Under the same experimental conditions, the amount of psoralen complexed with DNA was 5-8 times higher than that of 3-(a,a-dimethylally1)-psoralen. Therefore, these results indicate a slight ability of the substance t o form a molecular complex with DNA.
P&P
2912-1
280
D. VEDALDI,F. DALL’ACQUA. S. CAFFIERI and G. RODICHIERO
0)
U
-6
8. n
e
-4
: rn 0
ID
n 0
-2
5 10
20
30
Time of irradiation (Minutes)
Figure 3. Fluorescence (exciting wavelength 337 nm, fluorescent wavelength 395 nm) acquired by DNA under irradiation (365 nm) in the presence of psoralen and 3-(a,a-dimethylallyl)-psoralen after hydrolysis in acidic medium; (a): fluorescence intensity after irradiation in the presence of equimolecular amounts of the furocoumarins; (b): “specific fluorescence” obtained dividing the measured fluorescence by the amount of substance linked to DNA. fluorescence” acquired by DNA as a consequence of the irradiation in the presence of psoralen. Figure 3b shows that the “specific fluorescence” is much higher in the case of 3-(a,a-dimethylallyl)-psoralen than in the case of psoralen, indicating that 3-(a,a-dimethylaIlyl)-psoralen photobinds to DNA mainly through its 4‘,5’-double bond. An exact quantitative determination of the amounts of 4,5’-cycloadducts formed by 3-(a,a-dimethylallyl)psoralen in DNA would be possible only after the preparation in the pure state of the 4’,5’-cycloadducts between this furocoumarin and thymine or cytosine; in this case in fact, a calibration curve connecting the fluorescence intensities with the amounts of the cycloadducts could be prepared. We can consider, however, that the dirnethylallyl group does not significantly perturb the electronic structure of the furocoumarin molecule, as shown by the UV absorption spectra of psoralen and 3-(a,adimethylally1)-psoralen. Moreover, both excitation and fluorescence spectra observed in the present studies were superimposable with those of psoralen (Dall’Acqua et al., 1978). Therefore, assuming that the fluorescent properties of the 4,5’-cycloadduct between 3-(r,a-dimethylallyl)-psoralen and thymine has fluorescent properties analogous to those of the 4,S-cycloadduct between psoralen and thymine, we could make reference to a calibration curve prepared with the latter compound (Dall’Acqua et al., 1978) to evaluate the fluorescent cycloadducts formed in the photoreaction between 3-(a,a-dimethylallyl)-psoralen and DNA. If this assumption is considered possible, one may calculate that about 90% of the 3-(a,adimethylally1)-psoralen is bound to DNA forming 4,s’-cycloadducts.
Cross-linkage formation The capacity of 3-(a,a-dimethylallyl)-psoralen to form cross-linkages in DNA was studied by evaluating the unique renaturation capacity of cross-linked DNA after heat denaturation. Quantitative evaluation was made according to Lawley and Brookes (1967). Figure 4 reports the amounts of cross-linkages formed in DNA when irradiated in the presence of 3-(a,adimethylally1)-psoralen for increasing periods of time. In the same figure the extents of cross-linkages
-
1.00
g
0.80
a z n
4
0.20
It II
OIMETHVLALLVLPSORALEN
~~
5
10
20
30
Time of irradiation (Minutes)
Figure 4. Amount of cross-linkages formed in DNA by psoralen and by 3-(a,a-dimethylallyl)-psoralen by irradiation at 365 nrn.
Monoadduct of furocoumarin to DNA formed in DNA by psoralen in the same experimental conditions are also reported. 3-(cc,r-dimethylallyl)-psoralen is capable of forming a very small number of cross-linkages. This ability is dramatically low compared with that of psoralen (see Fig. 4). This derives from the lesser ability of 3-(r,r-ditnethylallyl)-psoralen both to photobind t o D N A and t o form cross-linkages in respect to psor-
alen. Comparing these results with those reported in Fig. 2, it is possible t o calculate that only about 2.5% of
28 1
the 3-(a,a-dimethylallyl)-psoralen molecules linked t o D N A forms bifunctional adducts (cross-linkages).
Skin-photosensit ization
To assay the skin-photosensitizing properties of 3-(cc,a-dimethylallyl)-psoralen, the depilated skin of albino guinea pig was treated with the substance and irradiated at 365 n m (see Materials and Methods). No development of erythema or hyperpigmentation was observed.
REFERENCES
Averbeck, D. (1976) Abstracts, VII Int. Congress on Photobiology. Rome. p. 137. Averbeck. D., P. Chandra and R. K. Biswas (1975) Radiat. Enuiron. Biophys. 12, 241-252. Baccichetti. F., F. Bordin, S. Marciani, F. Dall’Acqua and G . Rodighiero (1976) Z . Naturforsch. 3 1 ~ 207-208. , Bordin, F., S. Marciani, F. Baccichetti, F. Dall’Acqua and G. Rodighiero (1975) I t a l . J. Biochem. 24, 258-267. Cole. R. S . (1971) J. Bacteriol. 107. 846852. Dall’Acqua, F. (1977) In Research in Photobiology (Edited by A. Castellani), pp. 245-255. Plenum Publishing, New York. Dall’Acqua, F., S. Caffieri and G. Rodighiero (1978) Photochern. Photobiol. 27, 77-79. Dall’Acqua, F., S. Marciani and G . Chiarelotto (1968) A f t i 1st. Ven. Sci. Lett. Arti. Cl. Sci. Mar. Natur. 126, 103-115. Dall’Acqua, F., S. Marciani, L. Ciavatta and G. Rodighiero (1971) Z . Naturjorsch. 26b. 561-569. Dall’Acqua. F., S. Marciani, D. Vedaldi and G. Rodighiero (1974) 2. Naturforsch. 29c, 635-636. Dall’Acqua, F. and G . Rodighiero (1966) Rend. Atti Accad. Naz. Lincei 40, 412-422. Delle Monache, F., F. Marletti, G . B. Marini-Bettolo, J. F. De Mello and 0.G . De Lima (1976) Guzz. Chirn. Ital. 106, 681-689. Hatchard, C. G. and C. A. Parker (1956) Proc. R . Soc. B 235, 518-536. Lawley, P. D. and P. Brookes (1967) J . Mol. B i d . 25, 143-160. Mantulin W. W. and P. S. Song (1973) J. Am. Chem. Soc. 95, 5122-5129. Marciani, S., F. Dall’Acqua and C. Colombini (1969) Ann. Chim. (Rome) 59, 1067-1074. Marmur. J. and P. Doty (1962) J. Mol. B i d . 5, 109-118. Musajo, L., F. Bordin and R. Bevilacqua (1967) Photochem. Photobiol. 6, 927-931. Musajo, L. and G . Rodighiero (1962) Experientia 18, 153-161. Musajo, L. and G . Rodighiero (1972) In Photophysiology, Vol. VII (Edited by A. C. Giese) pp. 11 5-147. Academic Press, New York. Ou, C. N., C. H. Tsai. J. K. Tapley and P. S. Song (1978) Biochemistry 17, 1047-1053. Pathak, M. A,, D. M. Kramer and T. B. Fitzpatrick (1974) In Sunlight und Man (Edited by M. A. Pathak, L. C. Harber, M. Seiji and A. Kukita) pp. 335-368. University of Tokyo Press, Tokyo. Rodighiero. G., L. Musajo. F. Dall’Acqua, S. Marciani, G. Caporale and L. Ciavatta (1970) Biochim. Biophys. Acta 217, 40-49. Seki T., K. Nozu and S. Kondo (1978) Photochem. Photobiol. 27, 19-24. Song P . 4 , M. H. Harter, T. A. Moore and W. C. Hendon (1971) Phorochem. Photobiol. 14, 521--530.
3-(a,a-DIMETHYLALLY L)-PSORALEN A LINEAR FUROCOUMARIN FORMING MAINLY 4’,5’-MONOFUNCTIONAL ADDUCTS WITH DNA D. VEDALDI,F. DALL’ACQUA*, S. CAFFIERI and G. RODIGHIERO Institute of Pharmaceutical Chemistry of the Padua University, Centro di Studio sulla Chimica del Farmaco e dei Prodotti biologicamente attivi del C.N.R.. 3510 Padova. Italy (Received 10 May 1978: accepted 6 July 1978)
Abstract-3-(a,~dimethylallyl)-p~oralen,a furocoumarin derivative recently isolated from Horriu arborea. has been studied with regard to its photoreacting (365nm) capacity with native DNA. This
compound photobinds to DNA to a much lower extent than psoralen, forming mainly 4.5’-monofunctional adducts, due to the presence in the 3-position of 3.2-dimethylallyl group. which exerts a considerable hindrance both for the formation of a molecular complex with native DNA and for the photoreaction of the 3,4-double bond. This site of the molecule photoreacts, and, only 2.5% of the bound substance forms cross-linkages.Therefore. 3-(z.~-dimethylallyl).psoralen behaves as a fairly pure monofunctional reagent.
INTRODUCTION
The molecular basis of the photobiological activity of furocoumarins is connected with their ability to photobind to DNA (Musajo and Rodighiero, 1972; Pathak et a/., 1974). As furocoumarins have two photoreactive sites (3.4- and 4‘3-double bonds, see Fig. l), both monofunctional and bifunctional adducts can be formed with the pyrimidine bases. Bifunctional adducts. forming interstrand cross-linkages, appeared to be more directly connected with some photosensitized effects [skin-erythema (Dall’Acqua e l al.. 1974); killing of bacteria (Cole, 1971)] than the monofunctional. However, the contribution of monofunctional adducts in producing biological effects has been well evidenced and studied (Bordin er al., 1975; Baccichetti rt a!.. 1976; Seki et a!.. 1978). Averbeck et a/. (1975) have recently demonstrated that the formation of bifunctional adducts with DNA induces mainly nuclear gene mutations in yeast cells, while the formation of monofunctional adducts is more effective in inducing cytoplasmic “petite” mutations. Therefore. it is of great interest to have both furocoumarin derivatives forming bifunctional adducts and derivatives forming only monofunctional adducts. While many linearly condensed furocoumarins are known to be capable of forming bifunctional adducts with DNA (cross-linkages; e.g. psoralen. 8-methoxypsoralen. 5-methoxypsoralen. 4,5’.8-trimethylpsoralen). only angelicin has been widely studied for its ability to form only monofunctional adducts with DNA (due to its angular structure) (Dall’Acqua et a/., 1971). Averbeck (1976) and Ou et al. (1978) also used 3-carbethoxy-psoralen and
* To whom correspondence should be sent.
a
CH3 ;,CH=CH,
0
5
Figure 1. Molecular structure of 3-(a,rr-dimethyIallyI)psoralen. dimethoxycoumarin, respectively, which are capable of forming only monofunctional adducts. In this connection, we have examined the naturally occurring 3-(a,adimethylaflyI)-psoralen (Fig. 1). In fact, the presence of the dimethylallyl group in the 3-position could decrease the accessibility of the 3,4-double bond which is the most reactive site for the photoreaction (Song et a/., 1971; Mantulin and Song, 1973). while the possibility of a cycloaddition at the level of 4‘3’-double bond should remain unchanged. In this paper we report on the results obtained concerning the possibility both of a complex formation with DNA in the dark and of a photoreaction.
MATERIALS AND METHODS
D N A . Calf-thymus DNA was purchased from Sigma Chemical Company; its hypochromicity. determined according to Marmur and Doty (1962). was higher than 40%. Furocoumarins. 3-(r,z-dimethylallyl)-psoralen. extracted from Hortiu arhorea (Delle Monache et a/.. 1976). was a kind gift of Prof. G . B. Marini Bettolo. University of Rome; psoralen was extracted from the leaves of Ficus carica (Dall’Acqua et al.. 1968); angelicin was a gift of the Franco-Indian Pharmaceutical Co.. Bombay. India. These were tritiated by the Radiochemical Centre of Amersham . (England) and purified according to a procedure described elsewhere (Marciani et al., 1969); they had the following
211 ’
D. VEDALIX F. DALL'ACQUA. S CAFFIERI and G . RODIGHIERO
278
specific activities: 3-(z.x-dimethylallyl)-psoralen 3.26 Ci/ mol: psoralen 2 Ci/mol and angelicin 1.5 Ci/mol. Irradiution procrdurr. A small measured volume of a concentrated alcoholic solution of 3H-3-(sc.sc-dimethylaIlyl)psoralen. or of 3H-psoralen or 'H-angelicin for comparison. was added to an aqueous solution (2.2 mM) of DNA containing 2 m M NaCI, up to a concentration of 53.8 pM. Samples (3 m / ) of the prepared solutions were introduced into calibrated glass tubes. immersed in a thermostatically controlled bath and irradiated for different times by means of two HPW 125 Philips lamps emitting almost exclusively at 365 nm. The total intensity incident on the 3 m/ solution. determined by using a chemical actinometer (Hatchard and Parker, 1956). corresponded to 1.75 x IO-'J s - ' . After irradiation. each sample was divided into two portions: 1.5 m/ were used for the crosslinkage determination (solution A : see later): from the remaining 1.5 m/ the macromolecule was precipitated by addition of 3m/ of absolute ethanol. washed with 80;b ethanol and redissolved in the initial volume of water (solution B). This solution has been used for the radioactivity and fluorescence measurements. Radioactioity rneasurenients. Small volumes (0.2 m/) of solution B. diluted with I m/ of water and added to 10 m/ ofdioxane base scintillator (P.P.O. 5 g, P.O.P.O.P. 0.075 g. naphthalene 120 g. dioxane up to 1000 m/ of solution) were counted for the evaluation of the radioactivity acquired by DNA during the irradiation. A Packard Model 3375 liquid scintillation spectrometer was used; the efficiency of the apparatus for counting tritium was within the range 26-34",. Fhrorirnrtric ~c~rc~rriiinatioris. I m/ of solution B (see above). acidified by addition of 0.1 m/ of Concentrated HCI. was introduced into vials. which were sealed and heated in a boiling water bath for 60 min. After neutralization with NaOH and addition of 0.8 m/ phosphate buffer. pH 6.98. to a volume of 2 m/. the fluorescence intensity was determined using a Perkin-Elmer model MPF-44 spectrophotofluorimeter: exciting radiation : 337 nm: fluorescent radiation: 395 nm. This procedure had previously been elaborated (Dall'Acqua et al.. 1978) for evaluating the extent of the fluorescent 4'.5'-cycloadducts formed in DNA as a consequence of irradiation in the presence of psoralen. avoiding the quenching due to the intercalation of the cycloadducts between two base pairs in double stranded DNA. Specrrophotometric measurements. For the determination of the UV absorption spectra of the compound in the presence and in the absence of DNA, a Perkin-Elmer Model 180 double beam spectrophotometer was used. Dtterrnination of rht cross-linkages. The determination of the cross-linkages formed in DNA was carried out on the basis of the evaluation of the renaturation capacity of cross-linked DNA after heat-denaturation (Dall'Acqua rt al.. 1974). 1 S m / of solution A (see above). heated for 10 min in a boiling water bath and then immersed in ice for 15 min. was chromatographed on a column (0.7 x 4 cm) of hydroxylapatite Bio-gel type (Bio-Rad Laboratories. CA. U.S.A.) and developed using a linear
gradient of 0.05-0.3 M phosphatc buffer pH 6.98. with a flow rate of 15 drops per min. Fractions of 3.5 m/ were collected and the absorbance at 260 nm was determined in each fraction. On the basis of the single stranded fraction (non-cross-linked) and of the double stranded fraction (renamed. cross-linked) of DNA separated and detcrmined in this way. the extent of cross-linking formation was evaluated according to Lawley and Brookes (1967). Equilibrium dialysis e.xprrinwnts. Cylindrical cells. having 1.5 m/ halfcell volume. were used. The two halfcells were divided by Visking dialysing membrane (Visking Corp.. U.S.A.). In one part of the cell the aqueous solution of the tritiated 3-(z.a-dimethylallyl)-psoralen. containing NaCl 2OmM and ETDA I mM was introduced. while in the other part of the cell was introduced the DNA solution (in the range 0.45-3.08mM) having the same ionic strength. The cells were mechanically shaken for 12 h in a thermostatically controlled bath at 25 & 0.05"C in the dark. After the 'shaking period. small measured volumes (0.2 m f ) of the two phases were utilized for radiochemical measurements; in such a way the concentrations of the psoralen derivative were determined. Skin-photosensiti_ariorr. Measured volumes of a 0. I"< methanol solution of the substance were placed by means of a micropipette on a 4cm' area of depilated albino guinea-pig skin. corresponding to an amount of 50pg/cm'. the solvent was rapidly evaporated by means of an air stream and the skin was irradiated for 1 h with a Philips HPW 125 lamp. emitting mainly at 365 nm (5.5 W/m'): the animals were observed 12, 24 and 48 h after irradiation according to a test (Musajo and Rodighiero. 1962) commonly used for evaluating the skin-photosensitizing activity of furocoumarins.
RESULTS AND DISCUSSION
Tkr molecular complex formation It is known that furocoumarins form molecular complexes with DNA through intercalation between two base pairs of the macromolecule (Dall'Acqua and Rodighiero, 1966; Dall'Acqua. 1977); the role of this complex formation is important for the successive photoaddition (Rodighiero er a/.. 1970: Dall'Acqua rt a/.. 1971). The possible formation of a molecular complex between 3-(!x.x-dimethylaIlyl)-psoralen and DNA has been studied by spectrophotometric and fluorimetric measurements and by equilibrium dialysis experiments. Very small modifications in the UV-absorption spectrum of the compound were observed in the presence of DNA. Fluorimetric measurements. too. showed small modifications of the flourescence intensity (about increase) when D N A was added t o a n aqueous solution of 3-(r.r-dimethylallyl)-psoralen.
Table 1. Amounts of 3-(z.z-dimethylallyl)-psoralen bound with DNA. at various concentrations of the macromolecule. evaluated by equilibrium dialysis experiments. DNA concentration ( M of Pi x lo5) (1)
aqueous phase (2)
DNA-containing phase (3)
complexed with DNA (4)
ratio l/4
308 297 136 75 45
0.24 0.36 0.40 0.50 0.58
1.13 1.04 0.82 0.78 0.76
0.89 0.68 0.42 0.28 0.18
346 437 324 27 I 252
3-(z.z-dimethylallyl)-psoralenconcentration ( M x 10')
Monoadduct of furocoumarin to DNA
219
than those of psoralen and lower also than those of angelicin. Considering that under the experimental conditions used for the photoreaction, 3-(a,a-dimethylallyl)-psoralen is complexed within DNA only for 8.6%, we have calculated the rate constant of its photobinding (pseudo-first order reaction with respect to the furocoumarin, Dall’Acqua, 1977; Dall’Acqua et ul., unpublished results) on the basis of the initial concentration of the complexed fraction, instead of the total amount present in the solution. In this way the rate constant Photobinding to D N A calculated for 3-(a,a-dimethylallyI)-psoralen(9 x 10The amount of 3-(a,a-dimethylallyl)-psoralen which mol/l/min) was analogous to that of psoralen can photobind with DNA was studied by irradiating (8.5 x l o w 2mol///min) calculated also on the basis for increasing periods of time an aqueous solution of the complexed substance. Therefore, the low of 3H-compound (13.6pg/md) and of DNA (2.3mM amount of 3-(a,a-dimethylallyl)-psoralen photobindin respect to the fl content); after irradiation the ing to DNA is to be correlated to its low capacity macromolecule was precipitated by two volumes of to form the molecular complex with DNA rather than ethanol to remove the excess of unbound furocou- to its photoreacting capacity. marin. The precipitated DNA was washed with 80% ethanol and redissolved in water. This solution was 4’,5’-Fluorescent cycloudducts counted to evaluate the amount of the substance 4’,5’-Cycloadducts are formed by the cycloaddition covalently linked to DNA. of one 3-(a,a-dimethylallyl)-psoralen molecule, In Fig. 2 are reported the amounts (nmol) of 3-(a,a- through its 4,s-double bond, to one pyrimidine base dimethylally1)-psoralen bound to 1 mg of DNA as a of DNA. In the case of other furocoumarins studied, function of the time of irradiation. For comparison, this type of photoadduct, when excited with 365nm the amounts of psoralen and of angelicin bound to light, shows a strong violet fluorescence; by contrast, DNA, evaluated under the same experimental condi- 3,4-cycloadducts and bifunctional adducts, when irrations, are also reported. diated at 365 nm, are practically non-fluorescent The data obtained show that 3-(a,a-dimethylally1)- (Musajo et al., 1967; Dall’Acqua, 1977). Consequently, psoralen photobinds to DNA in amounts much lower the fluorescent properties acquired by DNA after irradiation at 365nm in the presence of a furocoumarin have been referred to the formation of 4‘,5‘-cycloadducts. In a recent note (Dall’Acqua et a!., 1978), we described a method for a quantitative evaluation of the 4‘,5’-cycloadducts formed by psoralen in DNA; the fluorescence intensity was determined after hydrolysis of the irradiated DNA to avoid the fluorescence quenching due to the intercalation of the 4,5’-cycloadducts in DNA. In the present study, the extent of 4,5’-monofunctional adducts formed in DNA by irradiation at 365 nm in the presence of 3-(a,a-dimethylallyI)-psoralen has been evaluated by measuring the fluorescence intensity acquired by DNA after irradiation. For the previously mentioned reason, the fluorescence intensity was measured after hydrolysis of the irradiated samples of DNA. The results obtained are reported in Fig. 3a in comparison with those obtained by irradiating DNA in the presence of equimolecular amounts of psoralen. In Fig. 3b are reported the values of the “specific fluorescence”, obtained by dividing the measured fluorescence intensities by the amounts of 3-(a,a-dimethylally1)-psoralen bound to DNA, determined by radioactivity measurements. As both fluorescent and 5 10 20 30 non-fluorescent adducts can be formed, this may be Time of irradiation (Minutes) a better indication of the extent of 4,5’-monofuncFigure 2. Photobinding of psoralen, angelicin and 3-(a,a- tional adducts than the simple fluorescence intensity. dimethylally1)-psoralen to calf thymus DNA. For a comparison, Fig. 3b also reports the “specific
Evidence for the complex formation was obtained by equilibrium dialysis experiments. The amounts of 3-(cc,cl-dimethyallyl)-psoralen bound to DNA at various concentrations of the macromolecule are reported in Table 1. Under the same experimental conditions, the amount of psoralen complexed with DNA was 5-8 times higher than that of 3-(a,a-dimethylally1)-psoralen. Therefore, these results indicate a slight ability of the substance t o form a molecular complex with DNA.
P&P
2912-1
280
D. VEDALDI,F. DALL’ACQUA. S. CAFFIERI and G. RODICHIERO
0)
U
-6
8. n
e
-4
: rn 0
ID
n 0
-2
5 10
20
30
Time of irradiation (Minutes)
Figure 3. Fluorescence (exciting wavelength 337 nm, fluorescent wavelength 395 nm) acquired by DNA under irradiation (365 nm) in the presence of psoralen and 3-(a,a-dimethylallyl)-psoralen after hydrolysis in acidic medium; (a): fluorescence intensity after irradiation in the presence of equimolecular amounts of the furocoumarins; (b): “specific fluorescence” obtained dividing the measured fluorescence by the amount of substance linked to DNA. fluorescence” acquired by DNA as a consequence of the irradiation in the presence of psoralen. Figure 3b shows that the “specific fluorescence” is much higher in the case of 3-(a,a-dimethylallyl)-psoralen than in the case of psoralen, indicating that 3-(a,a-dimethylaIlyl)-psoralen photobinds to DNA mainly through its 4‘,5’-double bond. An exact quantitative determination of the amounts of 4,5’-cycloadducts formed by 3-(a,a-dimethylallyl)psoralen in DNA would be possible only after the preparation in the pure state of the 4’,5’-cycloadducts between this furocoumarin and thymine or cytosine; in this case in fact, a calibration curve connecting the fluorescence intensities with the amounts of the cycloadducts could be prepared. We can consider, however, that the dirnethylallyl group does not significantly perturb the electronic structure of the furocoumarin molecule, as shown by the UV absorption spectra of psoralen and 3-(a,adimethylally1)-psoralen. Moreover, both excitation and fluorescence spectra observed in the present studies were superimposable with those of psoralen (Dall’Acqua et al., 1978). Therefore, assuming that the fluorescent properties of the 4,5’-cycloadduct between 3-(r,a-dimethylallyl)-psoralen and thymine has fluorescent properties analogous to those of the 4,S-cycloadduct between psoralen and thymine, we could make reference to a calibration curve prepared with the latter compound (Dall’Acqua et al., 1978) to evaluate the fluorescent cycloadducts formed in the photoreaction between 3-(a,a-dimethylallyl)-psoralen and DNA. If this assumption is considered possible, one may calculate that about 90% of the 3-(a,adimethylally1)-psoralen is bound to DNA forming 4,s’-cycloadducts.
Cross-linkage formation The capacity of 3-(a,a-dimethylallyl)-psoralen to form cross-linkages in DNA was studied by evaluating the unique renaturation capacity of cross-linked DNA after heat denaturation. Quantitative evaluation was made according to Lawley and Brookes (1967). Figure 4 reports the amounts of cross-linkages formed in DNA when irradiated in the presence of 3-(a,adimethylally1)-psoralen for increasing periods of time. In the same figure the extents of cross-linkages
-
1.00
g
0.80
a z n
4
0.20
It II
OIMETHVLALLVLPSORALEN
~~
5
10
20
30
Time of irradiation (Minutes)
Figure 4. Amount of cross-linkages formed in DNA by psoralen and by 3-(a,a-dimethylallyl)-psoralen by irradiation at 365 nrn.
Monoadduct of furocoumarin to DNA formed in DNA by psoralen in the same experimental conditions are also reported. 3-(cc,r-dimethylallyl)-psoralen is capable of forming a very small number of cross-linkages. This ability is dramatically low compared with that of psoralen (see Fig. 4). This derives from the lesser ability of 3-(r,r-ditnethylallyl)-psoralen both to photobind t o D N A and t o form cross-linkages in respect to psor-
alen. Comparing these results with those reported in Fig. 2, it is possible t o calculate that only about 2.5% of
28 1
the 3-(a,a-dimethylallyl)-psoralen molecules linked t o D N A forms bifunctional adducts (cross-linkages).
Skin-photosensit ization
To assay the skin-photosensitizing properties of 3-(cc,a-dimethylallyl)-psoralen, the depilated skin of albino guinea pig was treated with the substance and irradiated at 365 n m (see Materials and Methods). No development of erythema or hyperpigmentation was observed.
REFERENCES
Averbeck, D. (1976) Abstracts, VII Int. Congress on Photobiology. Rome. p. 137. Averbeck. D., P. Chandra and R. K. Biswas (1975) Radiat. Enuiron. Biophys. 12, 241-252. Baccichetti. F., F. Bordin, S. Marciani, F. Dall’Acqua and G . Rodighiero (1976) Z . Naturforsch. 3 1 ~ 207-208. , Bordin, F., S. Marciani, F. Baccichetti, F. Dall’Acqua and G. Rodighiero (1975) I t a l . J. Biochem. 24, 258-267. Cole. R. S . (1971) J. Bacteriol. 107. 846852. Dall’Acqua, F. (1977) In Research in Photobiology (Edited by A. Castellani), pp. 245-255. Plenum Publishing, New York. Dall’Acqua, F., S. Caffieri and G. Rodighiero (1978) Photochern. Photobiol. 27, 77-79. Dall’Acqua, F., S. Marciani and G . Chiarelotto (1968) A f t i 1st. Ven. Sci. Lett. Arti. Cl. Sci. Mar. Natur. 126, 103-115. Dall’Acqua, F., S. Marciani, L. Ciavatta and G. Rodighiero (1971) Z . Naturjorsch. 26b. 561-569. Dall’Acqua. F., S. Marciani, D. Vedaldi and G. Rodighiero (1974) 2. Naturforsch. 29c, 635-636. Dall’Acqua, F. and G . Rodighiero (1966) Rend. Atti Accad. Naz. Lincei 40, 412-422. Delle Monache, F., F. Marletti, G . B. Marini-Bettolo, J. F. De Mello and 0.G . De Lima (1976) Guzz. Chirn. Ital. 106, 681-689. Hatchard, C. G. and C. A. Parker (1956) Proc. R . Soc. B 235, 518-536. Lawley, P. D. and P. Brookes (1967) J . Mol. B i d . 25, 143-160. Mantulin W. W. and P. S. Song (1973) J. Am. Chem. Soc. 95, 5122-5129. Marciani, S., F. Dall’Acqua and C. Colombini (1969) Ann. Chim. (Rome) 59, 1067-1074. Marmur. J. and P. Doty (1962) J. Mol. B i d . 5, 109-118. Musajo, L., F. Bordin and R. Bevilacqua (1967) Photochem. Photobiol. 6, 927-931. Musajo, L. and G . Rodighiero (1962) Experientia 18, 153-161. Musajo, L. and G . Rodighiero (1972) In Photophysiology, Vol. VII (Edited by A. C. Giese) pp. 11 5-147. Academic Press, New York. Ou, C. N., C. H. Tsai. J. K. Tapley and P. S. Song (1978) Biochemistry 17, 1047-1053. Pathak, M. A,, D. M. Kramer and T. B. Fitzpatrick (1974) In Sunlight und Man (Edited by M. A. Pathak, L. C. Harber, M. Seiji and A. Kukita) pp. 335-368. University of Tokyo Press, Tokyo. Rodighiero. G., L. Musajo. F. Dall’Acqua, S. Marciani, G. Caporale and L. Ciavatta (1970) Biochim. Biophys. Acta 217, 40-49. Seki T., K. Nozu and S. Kondo (1978) Photochem. Photobiol. 27, 19-24. Song P . 4 , M. H. Harter, T. A. Moore and W. C. Hendon (1971) Phorochem. Photobiol. 14, 521--530.
