95
J. Photochem. Photobiol. B: Bill., 14 (1992) 95-104
Synthesis and photobiological methylpsoralen derivatives 0. Gist E. Uriartett, Magno ’
G. Zagotto,
activity of new
F. Baccichetti,
C. Antonello
and S. Marciani-
Department of Pharmaceutical Sciences of Padua University, via Manolo 5, 35131 Padova (Italy)
(Received December 4, 1991; accepted January 31, 1992) Abstract
The synthesis and the photobiological activity of two new derivatives of psoralen (3,4’dimethylpsoralen and 3,4’,8-trimethylpsoralen) has been described. They are congeners of the monofunctional linear furocoumarin 3,4’-dimethyl-8-methomsoralen. Both compounds bind very efficiently to DNA, the extent of this process being modulated by the nature of substituents at position 8. The number of photolesions is linearly related to adenine-thymine content of the nucleic acid which indicates lack of specificity for particular sequences of the nucleic acid. The structural arrangement of DNA (single stranded, double stranded, nucleosomes and chromatin) plays an additional role in affecting the photobinding process. Unlike their 8-methoxy congener the new derivatives cross-link DNA to a substantial extent. Their photobiological properties, including erythema formation, reflect very closely those of 8-methoxypsoralen (8-MOP). The conclusion can be drawn that 3,4’-dimethyl-8-MOP represents a unique derivative in its family.
Keywords: Furocoumarins,
synthesis inhibition,
methylpsoralens, skin phototoxicity.
DNA photobinding,
cross-linking,
DNA
1. Introduction
Psoralens are bifnnctional photoreagents which form monoadducts and diadducts with pyrirnidine bases of DNA upon irradiation with long wavelength UV light [l-3]. The linear furocoumarins, such as S-methoxypsoralen (&MOP), 5methoqsoralen (SMOP) and 4,8,5’-trimethylpsoralen (TMP) are employed in the treatment of hyperproliferative skin diseases, especially psoriasis and mycosis fnngoides. Undesired side effects are observed during the therapeutic treatment, above all skin phototoxicity and risk of skin cancer [4]. Since the above effects appear to be mainly related to crosslink formation, the research in this field is aimed at modifying the psoralen molecular structure in order to reduce covalent diaddition to opposite strands of DNA. Recently we examined the dimethyl derivative of a furocoumarin, 3,4’-dimethyl-8-methoxypsoralen (DMe-8-MOP), which exhibits relevant photobinding ability to DNAand antiproliferative activity non-coupled to skin phototoxicity [5]. +Author to whom correspondence should be addressed. “Present address: Department of Organic Chemistry, University of Santiago de Compostela, Santiago de Compostela, Spain.
96
This behaviour conflicts with that of most known methylpsoralens, all of which have been shown to be capable of strongly inducing erythema 161. To gain insight into the molecular mechanism of action of DMe-g-MOP, we have investigated its noncovalent interaction and its covalent photoaddition to DNA. We suggested that a peculiar stereochemical arrangement could be present in the intercalated complex so as to prevent the compound from cross-linking to DNA [7, 81. To understand whether DMe-g-MOP is just an exception or its behaviour is characteristic for congener compounds too, we synthesized and investigated 3,4’-dimethylpsoralen and 3,4’,&trimethylpsoralen, which exhibit the 3,4’-dimethyl substitution pattern characteristic of DMe-g-MOP.
2. Materials and methods 2.1. Chemicals g-MOP was purchased from Chinoin SPA, Milan, Italy. 3,4’-Dimethylpsoralen (compound 7) and 3,4’,8-trimethylpsoralen (compound 8) were prepared and purified as described in Section 2.2. Radioactive samples of the furocoumarins were prepared by Amersham Radiochemical Centre (Amersham, UK) and purified on thin layer chromatography (TLC) plates as described in ref. 9. DNA from calf thymus was purchased from Sigma Chemical Company (St. Louis, MO). Its hypochromicity, determined according to Marmur and Doty [lo], was over 35%. RNA from bakers’ yeast was purchased from Sigma Chemical Company (St. Louis, MO). Single-stranded DNA from ox lung having a molecular weight of about 30000 D was a generous gift of Crinos SpA (Como, Italy); DNA from Micrococcus lysodeicticus and DNA from Clostridium perfrngens were purchased from Sigma Chemical Company (St. Louis, MO). Chromatin samples and 175bp nucleosomes were prepared from calf thymus glands according to the literature [ll-131. 2.2. Chemical synthesk Melting points (m.p.s), uncorrected, were determined using a Buchi 510 apparatus. IR spectra were recorded on a Perkin-Elmer 1640 FT spectrometer (KBr discs). ‘H nuclear magnetic resonance (NMR) spectra were recorded on a Varian CFl-20 (80 MHz) using tetramethylsilane as internal standard (chemical shifts in 6 values). W spectra were recorded with a Perkin-Elmer 554 instrument. Elemental analyses were obtained on a Perkin-Elmer 240 B instrument. Preparative column chromatography (CC) was performed with silica gel (Merck 60, 70-230 mesh) and analytical TLC was performed on pre-coated silica gel plates (Merck 60 F254, 0.25 mm). 2.2.1. 3-Methyl-7-hydroxycoumarin (3) A mixture of 2,4_dihydroxybenzaldehyde (1 g; 7.2 mmol), propionic anhydride (2.5 ml; 19.5 mmol), sodium propionate (1.5 g; 15.6 mmol) and piperidine (0.7 ml; 7.2 mmol) was heated for 6 h to reflux. The mixture was then poured into 3 N HCl (800 ml) and left overnight. A solid separated, which was recovered by filtration, washed with water and purified by CC using toluene-ethyl acetate (4~1) (yield, 4.9 g (38.5%) of 3; m.p., 221 “C). ‘H NMR (dimethyl sulphoxide (DMSO)) S: 2.03 (d, 3H, CHa, J=l.O Hz); 6.62-6.82 (m, 2H, H6+H8); 7.40 (d, lH, H5, 3=8.2 Hz); 7.71 (g, lH, H4, J= 1.0 Hz); 10.32 (s, lH, HO). IR: 3194(0H); 1684 (CO); 1610, 1569 and 1508 (C=C); 1259,1153; 1008 and 758 cm -r. CrJIsOa, calculated: C, 68.17; H, 4.57. CJIaO~, found: C, 68.33; H, 4.48.
97
2.2.2. 3.8-Dimethyl-7-hydroxycoumarin (4) It was prepared from 2,4-dihydroxy-3-methylbenzaldehyde in the same manner as described above (yield, 40%; m.p., 197-198 “C!. ‘H NMR (DMSO) 6: 2.03 (d, 3H, CHs-, J= 1.0 Hz); 2.14 (s, 3H, CI-I& 6.82 (d, lH, H6, 5=8.5 Hz); 7.24 (d, lH, H5, J=8.5 Hz); 7.70 (g, lH, H4, J- 1.0 Hz); 10.21 (s, lH, HO). IR: 3368; 2924; 1726, 1300, 1145 and 809 cm-‘. CllHx003, 1700 and 1688 (CO); 1624 and 1600 (C-C); calculated: C, 69.46; H, 5.30. CllHr003, found: C, 69.33; H, 5.17. 2.2.3. 7-Acetonyloxy-3-methylcoumarin (5) To a solution of 3 (176 mg; 1 mmol) in anhydrous acetone (80 ml), chloroacetone (0.2 ml; 2.5 mmol) and potassium carbonate (1 g) were added. The reaction mixture was heated for 12 h to reflux. The precipitate was filtered oE, the acetonic solution was evaporated under vacuum and the solid residue purified by CC using toluene-ethyl acetate (9:l). Subsequent recrystallization from ethyl acetate-ligroin (3:l) yielded 187 mg (81%) of 5 (m.p., 138 “C). ‘H NMR (CDCl,) 6: 2.16 (d, 3H, CHTC=C, J=l.l Hz); 2.29 (s, 3H, CH,-CO); 4.61 (s, 2H, -CHT); 6.74 (d, lH, H8, 5=2.5 Hz); 6.82 (dd, lH, H6, J=2.5 and 8.6 Hz); 7.32 (d, lH, I-E, J=8.6 Hz); 7.44 (g, lH, H4, J=l.l Hz). C!r3Hr204, calculated: C, 67.23; H, 5.21. C13H1204, found: C, 67.04; H, 5.30. (6) 2.2.4. 7-Acetonyloq-3,&dimethylcoumarin It was prepared from 4 in the same manner as described for the compound 5 (yield, 71%; m.p., 191-192 “C). ‘H NMR (CDCQ 6: 2.17 (d, 3H, CHs-, J= 1.3 Hz); 2.32 (s, 3H, CI&--); 2.38 (s, 3H, CH-); 4.61 (s, 2H, -CH,-); 6.63 (d, lH, H6, J=8.5 Hz); 7.18 (d, lH, H5, J-8.5 Hz); 7.42 (g, lH, H4, J- 1.3 Hz). Cr4Hr404, calculated: C, 68.27; H, 5.73. C14H1404, found: C, 68.31; H, 5.66. 2.2.5. 3,4’-Dimethylfiro[3,2-g]coumarin (7) To a solution of 5 (232 mg; 1 mmol) in ethyl alcohol (60 ml), 0.1 N NaOH (60 ml) was added. The mixture was heated for 3 h to reflux. It was then acidified with HCl, evaporated to half-volume and left overnight. A solid precipitated, which was recovered by filtration, washed with water and purified by CC using toluene. It was recrystallized from ethyl alcohol-ligroin. An analytical sample was obtained by sublimation (at 200 “C and 0.2 Torr) (yield, 177 mg (83%) of 7; m.p., 189 “C). ‘H NMR (CDCl,) 6: 2.20 (d, 3H, CH3-, J=l.O Hz); 2.25 (d, 3H, Cl&-, J- 1.3 Hz); 7.34 (s, lH, H8); 7.45 (m, 2H, H5 +H5’); 7.57 (g, lH, H4, J- 1.0 Hz). IR: 3106; 1707 (CO); 1641, 1622 and 1579 (C-C); 1140,1069 and 758 cm-‘. Cr3Hr003, calculated: C, 72.88; H, 4.71. Cr3H1,,03, found: C, 72.50; H, 4.62. 2.2.6. 3,4’-8-Trimethyllfuro[3,2-g]coumati (8) It was prepared from 6 in the same manner as described for compound 7. Compound 8 was recrystallized from toluene (yield, 88%; m.p., 147-147.5 “C). ‘H NMR (CDQ) 6: 2.19 (d, 3H, CI&, J= 1.1 Hz); 2.23 (d, 3H, CH3-, J= 1.3 Hz); 2.53 (s, 3H, Cl&-); 7.27 (s, lH, H5); 7.43 (d, lH, H5’, J= 1.3 Hz); 7.52 (g, lH, H4, J- 1.1 Hz). IR: 3099, 2965 and 2924; 1716 (CO), 1600 (C-C); 1364, 1122 and 759 cm-‘. G4Hr203, calculated: C, 73.66; H, 5.29. C1d&03, found: C, 73.68; H, 5.26. 2.3. Photochemical and photobiological methodr 2.3.1. UV light source Irradiation were performed by means of Philips HPW 125 lamps equipped with a filter emitting over 90% at 365 nm; the irradiation intensities determined by a
98
potassium ferrioxalate chemical actinometer on Ehrlich ascites and skin photosensitization
[14] were 5.5 J s-l mm2 for experiments and 9.34 J s-l rnb2 for DNA irradiations.
2.3.2. Radioactivity measurements In DNA synthesis studies, the radioactivity of the acid-insoluble fraction, collected on a filter, was determined by putting the dried filter into a toluene-based scintillator (2,5-diphenyloxazole, 5 g; 1,4-bis-2-(4-methyl-5-phenyloxazolyl)-benzene, 0.25 g; toluene up to 1 1 of solution; 5 ml per sample). Counting was performed using a Packard A 300 CD liquid scintillation spectrometer: the efficiency of the apparatus, checked using a sample of standard tritiated water (Packard Chemicals, Downers Grove, IL, USA), was in the range 20-25% for the acid-insoluble precipitates and 35-40% for the aqueous solutions. 2.4. Interaction with DNA in vitro 2.4. I. Photobinding experiments The experiments with DNA from calf thymus and with RNA were performed in the following way: small measured volumes of concentrated ethanolic solutions of the compounds were added to aqueous 0.05% solutions (containing 10 mM NaCl and 0.5 mM ethylenediaminetetraacetic acid (EDTA) (pH 7.0)) of DNA to achieve a DNAto-compounds ratio of about 75. Aliquots of these solutions were introduced into calibrated glass tubes, immersed in a thermostatically controlled bath and then irradiated for different times. After irradiation, DNA was precipitated with ethyl alcohol (EtOH) and NaCl and the pellets were dissolved in the same buffer. Radiochemical measurements were then performed on a Packard A 300 CD liquid scintillation spectrometer. Thevarious DNA preparations (chromatin, nucleosomes, DNA fromikf. Zysodeicticus, and DNA from C. perj&genr) were instead adjusted to yield concentrations of the order of 1.5 x 10e4 M in 10 mM NaCl and 0.5 mM EDTA (pH 7.0). The DNA-todrug ratio was about 5. After irradiation, each sample was extensively extracted with chloroform to remove unreacted material and low molecular weight byproducts. Appropriate control experiments showed complete removal of tritiated drug in the absence of irradiation without loss of macromolecular species. 2.4.2. Cross-links evaluation in vitro After irradiation in the presence of the tritiated furocoumarin, DNA solutions were denatured thermally (95 “C for 15 min) and rapidly cooled to room temperature. Chromatographic separation of the denatured and renatured polynucleotide was achieved by means of a Perkin-Elmer liquid chromatograph model 3B, equipped with a Biogel high performance hydroxylapatite column (Biorad). Briefly the column was eluted with a linear gradient of phospate buffer (50-400 mM, pH 6.8) at a flow rate of 0.8 ml min-’ at room temperature. 2.5. Photobiolqgical effects 2.5.1. DNA synthesis inhibition in Ehrlich ascites cells Ehrlich ascites tumour (Lcttre strain) was routinely transferred by injecting intraperitoneally 2X 106 cells per animal into NCL mice. For the experiments, the tumour cells, collected on the sixth or seventh day after the transplant, were diluted to 2 X 10’ cells ml-’ with Hank’s solution containing the compound to be tested (18 PM); they were kept in the dark at room temperature for 15 min; they were then irradiated in Petri dishes (diameter, 5 cm; samples, 5 ml). Just after irradiation, the cells were incubated for 30 min at 37 “C in Hank’s solution (0.5 ml samples) in the
99
presence of 1 &i of 3H thymidine. The reaction was stopped by chilling in ice and adding 1 ml of a 5 mM solution of cold thymidine in physiological saline. The cells were then collected by filtration on Wathman GFK glass filters (catalogue No. 1822025; diameter, 2.5 cm), washed three times with saline and treated with 10 ml of ice-cold 10% trichloroacetic acid. After 1 min the samples were filtered and washed six times with 10 ml of 1% trichloroacetic acid; the filters were then dried and counted. The DNA contents were determined by the diphenilamine reaction [U]. The results, calculated as the percentage of the radioactivity incorporated into DNA of untreated control cells (about 25000-30000 dpm pg-‘), were subjected to probit analysis and expressed as IDS,,, i.e. the WA radiation dose which is able to induce a 50% inhibition of DNA synthesis in the presence of an 18 PM concentration of the compound studied. Filtration was carried out using a sample manifold apparatus (Millipore Corporation, Bedford, USA). 2.5.2. Skin phototoxicity The ability to induce skin erythema was studied in albino guinea pigs (outbred Dunkin-Hartley strain); the compounds were applied topically on the depilated skin as a 0.1% solution in methanol. The animals were kept for 45 min in a dark room and then the treated skin was irradiated with a total dose of 9.9 x lo4 J m-‘; erythema was scored after 48 h. 3. Results 3.1. Chemical .synthesis The synthesis of the methylpsoralen derivatives 7 and 8 was performed to the pathways shown in Scheme 1 and described in Section 2.
b
HO
4s :I
’
HO R
0
R
3. 4
1. 2
fi 5, 6
7, 6 R (1,3,5,7) = H
Scheme 1.
R (2,4,6,8) = CH,
CH3 0
according
100
The starting products were 2,4_dihydroxybenzaldehyde (1)and 2,4-dihydroxy-3methylbenzaldehyde (2) which led to the corresponding 7-hydroxy-3-methylcoumarins 3 and 4 by treatment with propionic anhydride and sodium propionate. From these 7-hydro+oumarins, the 7-acetonyl ethers 5 and 6 respectively have been prepared by treatment with chloroacetone in acetonic solution in the presence of potassium carbonate and then by cyclixation to the furan ring in alkaline solution to give the desired methylpsoralens 7 and 8, carrying methyl groups in the 3 and 4’ position. 3.2. Photobinding to nucleic acids 3.2.1. Photobinding to DNA from calf thymes The extent of photobinding of compounds 7 and 8 to DNA, as a function of irradiation time, is reported in Fig. 1. g-MOP is also included as a reference compound. The photoreactivity exhibited by both new compounds is substantially higher than the reference drug g-MOP, the trimethyl derivative being more efficient. 3.2.2. Photobinding to RNA from bakers’ yeast and to single-stranded DNA The results of the photoreaction with RNA are reported in Fig. 2. Clearly both derivatives are very efficient in photobinding to this arrangement of the nucleic acid. Similar data are obtained using single-stranded DNA instead of RNA (not shown). 3.2.3. Photobinding to DNA with different contents of adenine-thymine Thymine is by far the preferred pyrimidine base in the photoreaction between furocoumarins and DNA [16]. While the 5’-TpA sequence is favoured over the 5’ApT sequence, recent investigations have pointed out modulating effects of flanking sequences too. To evaluate possible specificity effects exhibited by compounds 7 and 8, we studied the covalent binding to bacterial DNAs (from M. Zysodeicticus and from C. [email protected]), with different base pair compositions. The amounts of examined compounds linked
ij
IO?
30----
8-
6-
40 IRRADIATION
TIME
lminl
0
40
20 IRRADIATION
TIME
60
lminl
Fig. 1. Photobinding of compounds 7, 8 and S-MOP to double-stranded DNA from calf thymus (nucleotide-to-drug ratio, 75) as a function of irradiation time. Fig. 2. Photobinding of compounds 7 and 8 to ribosomal RNA from bakers’ yeast (nucleotideto-drug ratio, 75) as a function of irradiation time.
101 to the various samples of DNA after irradiation are reported in Fig. 3. The data at 40 min irradiation are plotted as a function of adeninethymine (A-T) percentage content in Fig. 4. Increasing A-T content (28% in M. lysodeicticus, 60% in calf thymus, 72% in C. perjkingens) increases linearly the amount of drugs covalently bound to the polynucleotide. 3.2.4. Photobinding to chromatin and nucleosomes Since the native structure of eukaryotic DNA is organized with basic proteins to give the nucleosomal arrangement, which then folds into chromatin, useful information on the activity in viva can be obtained by investigating the extent of covalent photobinding to these DNA structures. Both new compounds show a remarkable decrease in photobinding ability to soluble chromatin and to nucleosomes when compared with “free” DNA. The results obtained for compound 8 are reported in Fig. 5; the amount 5-
A
4-
3-
2-
l-
65432l1 0
40 60 20 IRRADIATION TIME [min]
A-T
% CONTENT
Fig, 3. Photobinding pf compounds 7, 8 to DNAs having different base compositions (nucleotideto-drug ratio, 5) as a function of irradiation time for (A) compound 7 and (B) compound 8 (DNA concentration, 1.5 x lo-’ M): curves a, M. lysodeicticus; curves b, calf thymus; curves c, c. pe$ingens. Fig. 4. Photobmdmg of compounds 7 and 8 to DNA (nucleotide-to-drug of A-T percentage content.
ratio, 5) as a function
102 IRRADIATION
IRRADIATION
TIME
TIME
[min
)
[min)
Fig. 5. Photobinding of compound DNA (curve c) (nucleotide-to-drug 1.5 x 1O-4 M).
8 to chromatin (curve a), nucleosomes, (curve b) and “free” ratio, 5) as a function of irradiation time (DNA concentration,
Fig. 6. Cross-linking of compounds 7, 8 and 8-MOP to calf thymus DNA (nucleotide-to-drug ratio, 75) as a function of irradiation time.
TABLE
1
DNA synthesis inhibition by irradiation 8-methoxypsoralen as reference drug Compound
7 8 8-MOP
(365 nm) in the presence
of examined
compounds
and
;2m-‘) 3.80 2.87 3.85
UVA dose which induces 50% inhibition of DNA synthesis in the presence at 18 PM concentration.
of tested compounds
of drug covalently bound to nucleosomes is reduced to less than one third in comparison with “free” DNA. Even lower amounts of drug photobind to chromatin.
3.2.5. Cross-linking The DNA cross-linking ability of derivatives 7 and 8 is reported in Fig. 6. The trimethyl compound is more effective than 8-MOP, whereas the dimethyl congener forms bifunctional lesions to a remarkably lower extent. 3.3. Biological effects 3.3.1. DNA synthesis inhibition in Ehrlich cell The results of inhibition of DNA synthesis in Ehrlich ascites tumour cells, expressed in terms of IDSo, are reported in Table 1. Both compounds inhibit DNA synthesis comparably with the reference drug 8-MOP.
103
TABLE 2 Skin phototoxicity in guinea pigs Furocoumarins
Drug concentration (KS m-2)
WA dose (kJ m-‘)
Erythema formation’
7 8 S-MOP
5 5 5
20 20 20
++++++-
‘+ + -, moderate erythema.
3.3.2. Skin phototmicity The results related to erythema formation
in the skin of guinea pigs are reported in Table 2, the new compounds, unfortunately, are shown to be able to induce relevant erythema, as is known for linear furocoumarins [6].
4. Discussion The results obtained using the methyl derivatives of psoralen 7 and 8 show the classical trend of photobinding efficiency to DNA [6]. In fact, methyl substitution enhances the photoreaction, which is most prominent with compound 8, followed by compound 7. Enhanced reactivity is also shown toward RNA or single-stranded DNA. As for the preference for specific regions of the nucleic acid, the linear dependence of the photobinding upon A-T content seems to indicate a reduction in selectivity in comparison with other linear psoralens [16] thus far investigated and, in particular, DMe-&MOP [17]. Probably lack of specificity is coupled with prominent reactivity, which renders strong and weak sites barely distinguishable in a given sequence. The reduction in binding to nucleosomes and soluble chromatin is in line with our previous findings on a number of linear psoralens including 5-MOP and 8-MOP [18]. Indeed the tight structural organization of DNA provides a protective frame against photoreaction. This should be primarily due to a reduced formation of the reversible complex between drug and nucleic acid as is the case for 5-MOP and 8MOP and a number of other compounds reacting with DNA [18]. Unlike their 8-methoxy congener [7], both 7 and 8 are able to cross-link DNA to a remarkable extent. In turn their phototoxicity is substantially high and very comparable with that of 8-MOP. In conclusion the new derivatives behave as “classical” psoralens, while their 8-methoxy analogue exhibits an unusual behaviour, in terms both of cross-linking and of eIythema formation [7]. Thus it appears that a methoxy substitution at position 8 causes dramatic modifications in the photochemical and photobiological properties of 3,4’-dimethyl psoralen derivatives. As stated above, the new compounds are more reactive than DMe-8-MOP, and they show lower selectivity. Possibly a structural arrangement of the drug-DNA complex, which is preferred by the latter compound, is no longer predominant for its more efficient congeners, so that a levelling off of the peculiar properties occurs. Additionally or alternatively, the distribution of new drugs 7 and 8 along the genome could differ with reference to DMe-8-MOP as it might be expected from their modified (reduced) sequence preference. In any event, subtle stereochemical and electronic effects, modulating photoreactivity toward DNA, must be operating, which apparently make DMe-8-MOP a unique compound in its family.
104
Acknowledgment
funds
The authors gratefully acknowledge for cultural exchange.
the financial
support
from the Italian-Spanish
References 1 E. Ben-Hur and P. S. Song, The photochemistry and photobiology of furocoumarins (psoralens), Adv. Rudiut. BioL, II (1984) 131-171. 2 G. Rodighiero, F. Dall’Acqua and M. A. Pathak, Photobiological properties of monofunctional furocoumarin derivatives, in K. C. Smith (ed.), Topics in Photomedicine, Plenum, New York, 1984, pp. 319-398. 3 D. Karme, K. Straub, H. Rapoport and J. E. Hearst, Psoralen-deoxyribonucleic acid photoreaction. Characterization of the monoaddition products from 8-methoxypsoralen and 4,5’,8trimethylpsoralen, Biochemisby, 21 (1982) 861-871. 4 M. A. Pathak, J. A. Parrish and T. B. Fitzpatrick, Psoralens in photochemotherapy of skin disease, Famco, Ed SC& 36 (1981) 47-91. 5 F. Baccichetti, F. Carlassare, F. Bordin, F. Tamaro, 0. Gia, C. Antonello, S.Marciani Magno, Photobiological activity of 3,4’-dimethyl-S-MOP, Med. BioL Environ., 22 (1983) 361-366. 6 F. Dall’Acqua, S. Caffieri and G. Rodighiero, Photoreaction of furocoumarins (psoralens and angelicins), in R. V. Bensasson, G. Jori, E. J. Land and T. G. Truscott (eds.), Primary Photoprocesses in Biology and Medicine, Plenum, New York, 1985, pp. 259-272. 7 M. Palumbo, F. Baccichetti, C. Antonello, 0. Gia, A. Capozzi, S. Marciani Magno, Photobiological activity of 3,4’-dimethyl-8-methoxypsoralen, a linear furocoumarin with unusual DNA-binding properties, Photochem PhotobioL, 52 (1990) 533-540. 8 B. Norden and F. Tjemeld, Structure of methylene blue-DNA complexes studied by linear and circular dichroism spectroscopy, Biopolymem, 21 (1982) 1713-1734. 9 F. Carlassare, F. Baccichetti, A. Guiotto, P. Rodighiero, 0. Gia, A. Capozzi, G. Pastorini and F. Bordin, Synthesis and photobiological properties of acetylpsoralen derivatives, J. Photochem. PhotobioL B: Biol., 5 (1990) 25-39. 10 Maxmur and P. Doty, Determination of the base composition of deoxyribonucleic acid from its thermal denaturation temperature, J. Mol. BioL, 5 (1962) 109-118. 11 R. L Rill, B. R. Shaw and K. E. Van Holde, Isolation and characterization of chromatin subunits, Metho& Cell Biol. 18 (1978) 69-103. 12 D. M. Crothers, N. Dattagupta, M. Hogan, L. Klevan and K. S. Lee, Transient electric dichroism studies on nucleosomal particles, Biochemistry, 17 (1978) 4525-4533. 13 F. B. Schlessinger, N. Dattagupta and D. M. Crothers, Unfolding of 175 base-pair nucleosomes, Biochemktty, 21 (1982) 664-669. 14 C. G. Hatchard and C. A. Parker, A new sensitive chemical actinometer. II. Potassium ferrioxalate as a standard chemical actinometer, Proc. R Sot. London, Ser. B, 235 (1956) 518-536. 15 K Burton, A study of the conditions and mechanism of the diphenylamine reaction for the calorimetric estimation of deoxyribonucleic acid, Biochem J., 62 (1956) 315-323. 16 E. Sage and E. Moustacchi, Sequence context effects on 8-methoxypsoralen photobinding to defined DNA fragments, Biochemktry, 26 (1987) 3307-3314. 17 0. Gia, unpublished results, 1992. 18 0. Gia, G. Palii, M. Palumbo, C. Antonello and S. Marciani Magno, Photoreaction of psoralen derivatives with structurally organized DNA, Photochem PhotobioL, 45 (1987) 87-92.
J. Photochem. Photobiol. B: Bill., 14 (1992) 95-104
Synthesis and photobiological methylpsoralen derivatives 0. Gist E. Uriartett, Magno ’
G. Zagotto,
activity of new
F. Baccichetti,
C. Antonello
and S. Marciani-
Department of Pharmaceutical Sciences of Padua University, via Manolo 5, 35131 Padova (Italy)
(Received December 4, 1991; accepted January 31, 1992) Abstract
The synthesis and the photobiological activity of two new derivatives of psoralen (3,4’dimethylpsoralen and 3,4’,8-trimethylpsoralen) has been described. They are congeners of the monofunctional linear furocoumarin 3,4’-dimethyl-8-methomsoralen. Both compounds bind very efficiently to DNA, the extent of this process being modulated by the nature of substituents at position 8. The number of photolesions is linearly related to adenine-thymine content of the nucleic acid which indicates lack of specificity for particular sequences of the nucleic acid. The structural arrangement of DNA (single stranded, double stranded, nucleosomes and chromatin) plays an additional role in affecting the photobinding process. Unlike their 8-methoxy congener the new derivatives cross-link DNA to a substantial extent. Their photobiological properties, including erythema formation, reflect very closely those of 8-methoxypsoralen (8-MOP). The conclusion can be drawn that 3,4’-dimethyl-8-MOP represents a unique derivative in its family.
Keywords: Furocoumarins,
synthesis inhibition,
methylpsoralens, skin phototoxicity.
DNA photobinding,
cross-linking,
DNA
1. Introduction
Psoralens are bifnnctional photoreagents which form monoadducts and diadducts with pyrirnidine bases of DNA upon irradiation with long wavelength UV light [l-3]. The linear furocoumarins, such as S-methoxypsoralen (&MOP), 5methoqsoralen (SMOP) and 4,8,5’-trimethylpsoralen (TMP) are employed in the treatment of hyperproliferative skin diseases, especially psoriasis and mycosis fnngoides. Undesired side effects are observed during the therapeutic treatment, above all skin phototoxicity and risk of skin cancer [4]. Since the above effects appear to be mainly related to crosslink formation, the research in this field is aimed at modifying the psoralen molecular structure in order to reduce covalent diaddition to opposite strands of DNA. Recently we examined the dimethyl derivative of a furocoumarin, 3,4’-dimethyl-8-methoxypsoralen (DMe-8-MOP), which exhibits relevant photobinding ability to DNAand antiproliferative activity non-coupled to skin phototoxicity [5]. +Author to whom correspondence should be addressed. “Present address: Department of Organic Chemistry, University of Santiago de Compostela, Santiago de Compostela, Spain.
96
This behaviour conflicts with that of most known methylpsoralens, all of which have been shown to be capable of strongly inducing erythema 161. To gain insight into the molecular mechanism of action of DMe-g-MOP, we have investigated its noncovalent interaction and its covalent photoaddition to DNA. We suggested that a peculiar stereochemical arrangement could be present in the intercalated complex so as to prevent the compound from cross-linking to DNA [7, 81. To understand whether DMe-g-MOP is just an exception or its behaviour is characteristic for congener compounds too, we synthesized and investigated 3,4’-dimethylpsoralen and 3,4’,&trimethylpsoralen, which exhibit the 3,4’-dimethyl substitution pattern characteristic of DMe-g-MOP.
2. Materials and methods 2.1. Chemicals g-MOP was purchased from Chinoin SPA, Milan, Italy. 3,4’-Dimethylpsoralen (compound 7) and 3,4’,8-trimethylpsoralen (compound 8) were prepared and purified as described in Section 2.2. Radioactive samples of the furocoumarins were prepared by Amersham Radiochemical Centre (Amersham, UK) and purified on thin layer chromatography (TLC) plates as described in ref. 9. DNA from calf thymus was purchased from Sigma Chemical Company (St. Louis, MO). Its hypochromicity, determined according to Marmur and Doty [lo], was over 35%. RNA from bakers’ yeast was purchased from Sigma Chemical Company (St. Louis, MO). Single-stranded DNA from ox lung having a molecular weight of about 30000 D was a generous gift of Crinos SpA (Como, Italy); DNA from Micrococcus lysodeicticus and DNA from Clostridium perfrngens were purchased from Sigma Chemical Company (St. Louis, MO). Chromatin samples and 175bp nucleosomes were prepared from calf thymus glands according to the literature [ll-131. 2.2. Chemical synthesk Melting points (m.p.s), uncorrected, were determined using a Buchi 510 apparatus. IR spectra were recorded on a Perkin-Elmer 1640 FT spectrometer (KBr discs). ‘H nuclear magnetic resonance (NMR) spectra were recorded on a Varian CFl-20 (80 MHz) using tetramethylsilane as internal standard (chemical shifts in 6 values). W spectra were recorded with a Perkin-Elmer 554 instrument. Elemental analyses were obtained on a Perkin-Elmer 240 B instrument. Preparative column chromatography (CC) was performed with silica gel (Merck 60, 70-230 mesh) and analytical TLC was performed on pre-coated silica gel plates (Merck 60 F254, 0.25 mm). 2.2.1. 3-Methyl-7-hydroxycoumarin (3) A mixture of 2,4_dihydroxybenzaldehyde (1 g; 7.2 mmol), propionic anhydride (2.5 ml; 19.5 mmol), sodium propionate (1.5 g; 15.6 mmol) and piperidine (0.7 ml; 7.2 mmol) was heated for 6 h to reflux. The mixture was then poured into 3 N HCl (800 ml) and left overnight. A solid separated, which was recovered by filtration, washed with water and purified by CC using toluene-ethyl acetate (4~1) (yield, 4.9 g (38.5%) of 3; m.p., 221 “C). ‘H NMR (dimethyl sulphoxide (DMSO)) S: 2.03 (d, 3H, CHa, J=l.O Hz); 6.62-6.82 (m, 2H, H6+H8); 7.40 (d, lH, H5, 3=8.2 Hz); 7.71 (g, lH, H4, J= 1.0 Hz); 10.32 (s, lH, HO). IR: 3194(0H); 1684 (CO); 1610, 1569 and 1508 (C=C); 1259,1153; 1008 and 758 cm -r. CrJIsOa, calculated: C, 68.17; H, 4.57. CJIaO~, found: C, 68.33; H, 4.48.
97
2.2.2. 3.8-Dimethyl-7-hydroxycoumarin (4) It was prepared from 2,4-dihydroxy-3-methylbenzaldehyde in the same manner as described above (yield, 40%; m.p., 197-198 “C!. ‘H NMR (DMSO) 6: 2.03 (d, 3H, CHs-, J= 1.0 Hz); 2.14 (s, 3H, CI-I& 6.82 (d, lH, H6, 5=8.5 Hz); 7.24 (d, lH, H5, J=8.5 Hz); 7.70 (g, lH, H4, J- 1.0 Hz); 10.21 (s, lH, HO). IR: 3368; 2924; 1726, 1300, 1145 and 809 cm-‘. CllHx003, 1700 and 1688 (CO); 1624 and 1600 (C-C); calculated: C, 69.46; H, 5.30. CllHr003, found: C, 69.33; H, 5.17. 2.2.3. 7-Acetonyloxy-3-methylcoumarin (5) To a solution of 3 (176 mg; 1 mmol) in anhydrous acetone (80 ml), chloroacetone (0.2 ml; 2.5 mmol) and potassium carbonate (1 g) were added. The reaction mixture was heated for 12 h to reflux. The precipitate was filtered oE, the acetonic solution was evaporated under vacuum and the solid residue purified by CC using toluene-ethyl acetate (9:l). Subsequent recrystallization from ethyl acetate-ligroin (3:l) yielded 187 mg (81%) of 5 (m.p., 138 “C). ‘H NMR (CDCl,) 6: 2.16 (d, 3H, CHTC=C, J=l.l Hz); 2.29 (s, 3H, CH,-CO); 4.61 (s, 2H, -CHT); 6.74 (d, lH, H8, 5=2.5 Hz); 6.82 (dd, lH, H6, J=2.5 and 8.6 Hz); 7.32 (d, lH, I-E, J=8.6 Hz); 7.44 (g, lH, H4, J=l.l Hz). C!r3Hr204, calculated: C, 67.23; H, 5.21. C13H1204, found: C, 67.04; H, 5.30. (6) 2.2.4. 7-Acetonyloq-3,&dimethylcoumarin It was prepared from 4 in the same manner as described for the compound 5 (yield, 71%; m.p., 191-192 “C). ‘H NMR (CDCQ 6: 2.17 (d, 3H, CHs-, J= 1.3 Hz); 2.32 (s, 3H, CI&--); 2.38 (s, 3H, CH-); 4.61 (s, 2H, -CH,-); 6.63 (d, lH, H6, J=8.5 Hz); 7.18 (d, lH, H5, J-8.5 Hz); 7.42 (g, lH, H4, J- 1.3 Hz). Cr4Hr404, calculated: C, 68.27; H, 5.73. C14H1404, found: C, 68.31; H, 5.66. 2.2.5. 3,4’-Dimethylfiro[3,2-g]coumarin (7) To a solution of 5 (232 mg; 1 mmol) in ethyl alcohol (60 ml), 0.1 N NaOH (60 ml) was added. The mixture was heated for 3 h to reflux. It was then acidified with HCl, evaporated to half-volume and left overnight. A solid precipitated, which was recovered by filtration, washed with water and purified by CC using toluene. It was recrystallized from ethyl alcohol-ligroin. An analytical sample was obtained by sublimation (at 200 “C and 0.2 Torr) (yield, 177 mg (83%) of 7; m.p., 189 “C). ‘H NMR (CDCl,) 6: 2.20 (d, 3H, CH3-, J=l.O Hz); 2.25 (d, 3H, Cl&-, J- 1.3 Hz); 7.34 (s, lH, H8); 7.45 (m, 2H, H5 +H5’); 7.57 (g, lH, H4, J- 1.0 Hz). IR: 3106; 1707 (CO); 1641, 1622 and 1579 (C-C); 1140,1069 and 758 cm-‘. Cr3Hr003, calculated: C, 72.88; H, 4.71. Cr3H1,,03, found: C, 72.50; H, 4.62. 2.2.6. 3,4’-8-Trimethyllfuro[3,2-g]coumati (8) It was prepared from 6 in the same manner as described for compound 7. Compound 8 was recrystallized from toluene (yield, 88%; m.p., 147-147.5 “C). ‘H NMR (CDQ) 6: 2.19 (d, 3H, CI&, J= 1.1 Hz); 2.23 (d, 3H, CH3-, J= 1.3 Hz); 2.53 (s, 3H, Cl&-); 7.27 (s, lH, H5); 7.43 (d, lH, H5’, J= 1.3 Hz); 7.52 (g, lH, H4, J- 1.1 Hz). IR: 3099, 2965 and 2924; 1716 (CO), 1600 (C-C); 1364, 1122 and 759 cm-‘. G4Hr203, calculated: C, 73.66; H, 5.29. C1d&03, found: C, 73.68; H, 5.26. 2.3. Photochemical and photobiological methodr 2.3.1. UV light source Irradiation were performed by means of Philips HPW 125 lamps equipped with a filter emitting over 90% at 365 nm; the irradiation intensities determined by a
98
potassium ferrioxalate chemical actinometer on Ehrlich ascites and skin photosensitization
[14] were 5.5 J s-l mm2 for experiments and 9.34 J s-l rnb2 for DNA irradiations.
2.3.2. Radioactivity measurements In DNA synthesis studies, the radioactivity of the acid-insoluble fraction, collected on a filter, was determined by putting the dried filter into a toluene-based scintillator (2,5-diphenyloxazole, 5 g; 1,4-bis-2-(4-methyl-5-phenyloxazolyl)-benzene, 0.25 g; toluene up to 1 1 of solution; 5 ml per sample). Counting was performed using a Packard A 300 CD liquid scintillation spectrometer: the efficiency of the apparatus, checked using a sample of standard tritiated water (Packard Chemicals, Downers Grove, IL, USA), was in the range 20-25% for the acid-insoluble precipitates and 35-40% for the aqueous solutions. 2.4. Interaction with DNA in vitro 2.4. I. Photobinding experiments The experiments with DNA from calf thymus and with RNA were performed in the following way: small measured volumes of concentrated ethanolic solutions of the compounds were added to aqueous 0.05% solutions (containing 10 mM NaCl and 0.5 mM ethylenediaminetetraacetic acid (EDTA) (pH 7.0)) of DNA to achieve a DNAto-compounds ratio of about 75. Aliquots of these solutions were introduced into calibrated glass tubes, immersed in a thermostatically controlled bath and then irradiated for different times. After irradiation, DNA was precipitated with ethyl alcohol (EtOH) and NaCl and the pellets were dissolved in the same buffer. Radiochemical measurements were then performed on a Packard A 300 CD liquid scintillation spectrometer. Thevarious DNA preparations (chromatin, nucleosomes, DNA fromikf. Zysodeicticus, and DNA from C. perj&genr) were instead adjusted to yield concentrations of the order of 1.5 x 10e4 M in 10 mM NaCl and 0.5 mM EDTA (pH 7.0). The DNA-todrug ratio was about 5. After irradiation, each sample was extensively extracted with chloroform to remove unreacted material and low molecular weight byproducts. Appropriate control experiments showed complete removal of tritiated drug in the absence of irradiation without loss of macromolecular species. 2.4.2. Cross-links evaluation in vitro After irradiation in the presence of the tritiated furocoumarin, DNA solutions were denatured thermally (95 “C for 15 min) and rapidly cooled to room temperature. Chromatographic separation of the denatured and renatured polynucleotide was achieved by means of a Perkin-Elmer liquid chromatograph model 3B, equipped with a Biogel high performance hydroxylapatite column (Biorad). Briefly the column was eluted with a linear gradient of phospate buffer (50-400 mM, pH 6.8) at a flow rate of 0.8 ml min-’ at room temperature. 2.5. Photobiolqgical effects 2.5.1. DNA synthesis inhibition in Ehrlich ascites cells Ehrlich ascites tumour (Lcttre strain) was routinely transferred by injecting intraperitoneally 2X 106 cells per animal into NCL mice. For the experiments, the tumour cells, collected on the sixth or seventh day after the transplant, were diluted to 2 X 10’ cells ml-’ with Hank’s solution containing the compound to be tested (18 PM); they were kept in the dark at room temperature for 15 min; they were then irradiated in Petri dishes (diameter, 5 cm; samples, 5 ml). Just after irradiation, the cells were incubated for 30 min at 37 “C in Hank’s solution (0.5 ml samples) in the
99
presence of 1 &i of 3H thymidine. The reaction was stopped by chilling in ice and adding 1 ml of a 5 mM solution of cold thymidine in physiological saline. The cells were then collected by filtration on Wathman GFK glass filters (catalogue No. 1822025; diameter, 2.5 cm), washed three times with saline and treated with 10 ml of ice-cold 10% trichloroacetic acid. After 1 min the samples were filtered and washed six times with 10 ml of 1% trichloroacetic acid; the filters were then dried and counted. The DNA contents were determined by the diphenilamine reaction [U]. The results, calculated as the percentage of the radioactivity incorporated into DNA of untreated control cells (about 25000-30000 dpm pg-‘), were subjected to probit analysis and expressed as IDS,,, i.e. the WA radiation dose which is able to induce a 50% inhibition of DNA synthesis in the presence of an 18 PM concentration of the compound studied. Filtration was carried out using a sample manifold apparatus (Millipore Corporation, Bedford, USA). 2.5.2. Skin phototoxicity The ability to induce skin erythema was studied in albino guinea pigs (outbred Dunkin-Hartley strain); the compounds were applied topically on the depilated skin as a 0.1% solution in methanol. The animals were kept for 45 min in a dark room and then the treated skin was irradiated with a total dose of 9.9 x lo4 J m-‘; erythema was scored after 48 h. 3. Results 3.1. Chemical .synthesis The synthesis of the methylpsoralen derivatives 7 and 8 was performed to the pathways shown in Scheme 1 and described in Section 2.
b
HO
4s :I
’
HO R
0
R
3. 4
1. 2
fi 5, 6
7, 6 R (1,3,5,7) = H
Scheme 1.
R (2,4,6,8) = CH,
CH3 0
according
100
The starting products were 2,4_dihydroxybenzaldehyde (1)and 2,4-dihydroxy-3methylbenzaldehyde (2) which led to the corresponding 7-hydroxy-3-methylcoumarins 3 and 4 by treatment with propionic anhydride and sodium propionate. From these 7-hydro+oumarins, the 7-acetonyl ethers 5 and 6 respectively have been prepared by treatment with chloroacetone in acetonic solution in the presence of potassium carbonate and then by cyclixation to the furan ring in alkaline solution to give the desired methylpsoralens 7 and 8, carrying methyl groups in the 3 and 4’ position. 3.2. Photobinding to nucleic acids 3.2.1. Photobinding to DNA from calf thymes The extent of photobinding of compounds 7 and 8 to DNA, as a function of irradiation time, is reported in Fig. 1. g-MOP is also included as a reference compound. The photoreactivity exhibited by both new compounds is substantially higher than the reference drug g-MOP, the trimethyl derivative being more efficient. 3.2.2. Photobinding to RNA from bakers’ yeast and to single-stranded DNA The results of the photoreaction with RNA are reported in Fig. 2. Clearly both derivatives are very efficient in photobinding to this arrangement of the nucleic acid. Similar data are obtained using single-stranded DNA instead of RNA (not shown). 3.2.3. Photobinding to DNA with different contents of adenine-thymine Thymine is by far the preferred pyrimidine base in the photoreaction between furocoumarins and DNA [16]. While the 5’-TpA sequence is favoured over the 5’ApT sequence, recent investigations have pointed out modulating effects of flanking sequences too. To evaluate possible specificity effects exhibited by compounds 7 and 8, we studied the covalent binding to bacterial DNAs (from M. Zysodeicticus and from C. [email protected]), with different base pair compositions. The amounts of examined compounds linked
ij
IO?
30----
8-
6-
40 IRRADIATION
TIME
lminl
0
40
20 IRRADIATION
TIME
60
lminl
Fig. 1. Photobinding of compounds 7, 8 and S-MOP to double-stranded DNA from calf thymus (nucleotide-to-drug ratio, 75) as a function of irradiation time. Fig. 2. Photobinding of compounds 7 and 8 to ribosomal RNA from bakers’ yeast (nucleotideto-drug ratio, 75) as a function of irradiation time.
101 to the various samples of DNA after irradiation are reported in Fig. 3. The data at 40 min irradiation are plotted as a function of adeninethymine (A-T) percentage content in Fig. 4. Increasing A-T content (28% in M. lysodeicticus, 60% in calf thymus, 72% in C. perjkingens) increases linearly the amount of drugs covalently bound to the polynucleotide. 3.2.4. Photobinding to chromatin and nucleosomes Since the native structure of eukaryotic DNA is organized with basic proteins to give the nucleosomal arrangement, which then folds into chromatin, useful information on the activity in viva can be obtained by investigating the extent of covalent photobinding to these DNA structures. Both new compounds show a remarkable decrease in photobinding ability to soluble chromatin and to nucleosomes when compared with “free” DNA. The results obtained for compound 8 are reported in Fig. 5; the amount 5-
A
4-
3-
2-
l-
65432l1 0
40 60 20 IRRADIATION TIME [min]
A-T
% CONTENT
Fig, 3. Photobinding pf compounds 7, 8 to DNAs having different base compositions (nucleotideto-drug ratio, 5) as a function of irradiation time for (A) compound 7 and (B) compound 8 (DNA concentration, 1.5 x lo-’ M): curves a, M. lysodeicticus; curves b, calf thymus; curves c, c. pe$ingens. Fig. 4. Photobmdmg of compounds 7 and 8 to DNA (nucleotide-to-drug of A-T percentage content.
ratio, 5) as a function
102 IRRADIATION
IRRADIATION
TIME
TIME
[min
)
[min)
Fig. 5. Photobinding of compound DNA (curve c) (nucleotide-to-drug 1.5 x 1O-4 M).
8 to chromatin (curve a), nucleosomes, (curve b) and “free” ratio, 5) as a function of irradiation time (DNA concentration,
Fig. 6. Cross-linking of compounds 7, 8 and 8-MOP to calf thymus DNA (nucleotide-to-drug ratio, 75) as a function of irradiation time.
TABLE
1
DNA synthesis inhibition by irradiation 8-methoxypsoralen as reference drug Compound
7 8 8-MOP
(365 nm) in the presence
of examined
compounds
and
;2m-‘) 3.80 2.87 3.85
UVA dose which induces 50% inhibition of DNA synthesis in the presence at 18 PM concentration.
of tested compounds
of drug covalently bound to nucleosomes is reduced to less than one third in comparison with “free” DNA. Even lower amounts of drug photobind to chromatin.
3.2.5. Cross-linking The DNA cross-linking ability of derivatives 7 and 8 is reported in Fig. 6. The trimethyl compound is more effective than 8-MOP, whereas the dimethyl congener forms bifunctional lesions to a remarkably lower extent. 3.3. Biological effects 3.3.1. DNA synthesis inhibition in Ehrlich cell The results of inhibition of DNA synthesis in Ehrlich ascites tumour cells, expressed in terms of IDSo, are reported in Table 1. Both compounds inhibit DNA synthesis comparably with the reference drug 8-MOP.
103
TABLE 2 Skin phototoxicity in guinea pigs Furocoumarins
Drug concentration (KS m-2)
WA dose (kJ m-‘)
Erythema formation’
7 8 S-MOP
5 5 5
20 20 20
++++++-
‘+ + -, moderate erythema.
3.3.2. Skin phototmicity The results related to erythema formation
in the skin of guinea pigs are reported in Table 2, the new compounds, unfortunately, are shown to be able to induce relevant erythema, as is known for linear furocoumarins [6].
4. Discussion The results obtained using the methyl derivatives of psoralen 7 and 8 show the classical trend of photobinding efficiency to DNA [6]. In fact, methyl substitution enhances the photoreaction, which is most prominent with compound 8, followed by compound 7. Enhanced reactivity is also shown toward RNA or single-stranded DNA. As for the preference for specific regions of the nucleic acid, the linear dependence of the photobinding upon A-T content seems to indicate a reduction in selectivity in comparison with other linear psoralens [16] thus far investigated and, in particular, DMe-&MOP [17]. Probably lack of specificity is coupled with prominent reactivity, which renders strong and weak sites barely distinguishable in a given sequence. The reduction in binding to nucleosomes and soluble chromatin is in line with our previous findings on a number of linear psoralens including 5-MOP and 8-MOP [18]. Indeed the tight structural organization of DNA provides a protective frame against photoreaction. This should be primarily due to a reduced formation of the reversible complex between drug and nucleic acid as is the case for 5-MOP and 8MOP and a number of other compounds reacting with DNA [18]. Unlike their 8-methoxy congener [7], both 7 and 8 are able to cross-link DNA to a remarkable extent. In turn their phototoxicity is substantially high and very comparable with that of 8-MOP. In conclusion the new derivatives behave as “classical” psoralens, while their 8-methoxy analogue exhibits an unusual behaviour, in terms both of cross-linking and of eIythema formation [7]. Thus it appears that a methoxy substitution at position 8 causes dramatic modifications in the photochemical and photobiological properties of 3,4’-dimethyl psoralen derivatives. As stated above, the new compounds are more reactive than DMe-8-MOP, and they show lower selectivity. Possibly a structural arrangement of the drug-DNA complex, which is preferred by the latter compound, is no longer predominant for its more efficient congeners, so that a levelling off of the peculiar properties occurs. Additionally or alternatively, the distribution of new drugs 7 and 8 along the genome could differ with reference to DMe-8-MOP as it might be expected from their modified (reduced) sequence preference. In any event, subtle stereochemical and electronic effects, modulating photoreactivity toward DNA, must be operating, which apparently make DMe-8-MOP a unique compound in its family.
104
Acknowledgment
funds
The authors gratefully acknowledge for cultural exchange.
the financial
support
from the Italian-Spanish
References 1 E. Ben-Hur and P. S. Song, The photochemistry and photobiology of furocoumarins (psoralens), Adv. Rudiut. BioL, II (1984) 131-171. 2 G. Rodighiero, F. Dall’Acqua and M. A. Pathak, Photobiological properties of monofunctional furocoumarin derivatives, in K. C. Smith (ed.), Topics in Photomedicine, Plenum, New York, 1984, pp. 319-398. 3 D. Karme, K. Straub, H. Rapoport and J. E. Hearst, Psoralen-deoxyribonucleic acid photoreaction. Characterization of the monoaddition products from 8-methoxypsoralen and 4,5’,8trimethylpsoralen, Biochemisby, 21 (1982) 861-871. 4 M. A. Pathak, J. A. Parrish and T. B. Fitzpatrick, Psoralens in photochemotherapy of skin disease, Famco, Ed SC& 36 (1981) 47-91. 5 F. Baccichetti, F. Carlassare, F. Bordin, F. Tamaro, 0. Gia, C. Antonello, S.Marciani Magno, Photobiological activity of 3,4’-dimethyl-S-MOP, Med. BioL Environ., 22 (1983) 361-366. 6 F. Dall’Acqua, S. Caffieri and G. Rodighiero, Photoreaction of furocoumarins (psoralens and angelicins), in R. V. Bensasson, G. Jori, E. J. Land and T. G. Truscott (eds.), Primary Photoprocesses in Biology and Medicine, Plenum, New York, 1985, pp. 259-272. 7 M. Palumbo, F. Baccichetti, C. Antonello, 0. Gia, A. Capozzi, S. Marciani Magno, Photobiological activity of 3,4’-dimethyl-8-methoxypsoralen, a linear furocoumarin with unusual DNA-binding properties, Photochem PhotobioL, 52 (1990) 533-540. 8 B. Norden and F. Tjemeld, Structure of methylene blue-DNA complexes studied by linear and circular dichroism spectroscopy, Biopolymem, 21 (1982) 1713-1734. 9 F. Carlassare, F. Baccichetti, A. Guiotto, P. Rodighiero, 0. Gia, A. Capozzi, G. Pastorini and F. Bordin, Synthesis and photobiological properties of acetylpsoralen derivatives, J. Photochem. PhotobioL B: Biol., 5 (1990) 25-39. 10 Maxmur and P. Doty, Determination of the base composition of deoxyribonucleic acid from its thermal denaturation temperature, J. Mol. BioL, 5 (1962) 109-118. 11 R. L Rill, B. R. Shaw and K. E. Van Holde, Isolation and characterization of chromatin subunits, Metho& Cell Biol. 18 (1978) 69-103. 12 D. M. Crothers, N. Dattagupta, M. Hogan, L. Klevan and K. S. Lee, Transient electric dichroism studies on nucleosomal particles, Biochemistry, 17 (1978) 4525-4533. 13 F. B. Schlessinger, N. Dattagupta and D. M. Crothers, Unfolding of 175 base-pair nucleosomes, Biochemktty, 21 (1982) 664-669. 14 C. G. Hatchard and C. A. Parker, A new sensitive chemical actinometer. II. Potassium ferrioxalate as a standard chemical actinometer, Proc. R Sot. London, Ser. B, 235 (1956) 518-536. 15 K Burton, A study of the conditions and mechanism of the diphenylamine reaction for the calorimetric estimation of deoxyribonucleic acid, Biochem J., 62 (1956) 315-323. 16 E. Sage and E. Moustacchi, Sequence context effects on 8-methoxypsoralen photobinding to defined DNA fragments, Biochemktry, 26 (1987) 3307-3314. 17 0. Gia, unpublished results, 1992. 18 0. Gia, G. Palii, M. Palumbo, C. Antonello and S. Marciani Magno, Photoreaction of psoralen derivatives with structurally organized DNA, Photochem PhotobioL, 45 (1987) 87-92.
