CIENCIA 16(3), 370 - 378, 2008 Maracaibo, Venezuela Correlations between theoretical calculations and experimental photochemical reactivity of hydroxyanthraquinone drugs present in Aloe vera Fernando Ruette1*, Franklin Vargas2 y Lourdes Rivas1 1Laboratorio de Química Computacional, 2 Laboratorio de Fotoquímica, Centro de Química, Instituto Venezolano de Investigaciones Científicas (Ivic), Apartado 21827, Caracas 1020-A, Venezuela.

Recibido: 29-10-07 Aceptado: 13-09-08

Abstract The phototoxic drugs (S) (aloe-emodin (1), emodin (2), and rhein (3)) were studied with the parametric method CATIVIC in order to find correlations between calculated theoretical proper- ties and the experimental photodegradation rate. These correlations can be explained through the formation of free radical intermediates, singlet oxygen and stable photoproducts. A corres- pondence between calculated dipolar moment and heat of formation with photolysis rate and f quantum yield ( fl) of anthraquinones (1, 2 and 3) was found. In addition, HOMO energy correla- l - 1 tes with the wavelength of excitation ( exc). Formation of S , SH°, and O2 intermediates were eva- - luated and confirmed as feasible. The O2 anion is found to be the most probable species in the photooxidation of biological substrates. Key words: anthraquinones, Aloe vera, photolability, CATIVIC, theoretical correlations, photochemical reactivity. Correlaciones entre cálculos teóricos y la reactividad fotoquímica experimental de drogas hidroxiantraquinonas presentes en la Aloe vera

Resumen Las drogas fototóxicas (S) aloe-emodina (1), emodina (2) y rhein (3) fueron estudiadas con el método paramétrico CATIVIC con la finalidad de encontrar las correlaciones entre las propie- dades calculadas teóricamente y la velocidad de fotodegradación experimental. Estas coorrela- ciones pueden explicarse a través de la formación de intermediarios de radicales libres, oxígeno singlete y fotoproductos estables. Se encontró una correspondencia entre el momento bipolar y l el calor de formación, calculados con la velocidad de fotólisis y el rendimiento cuántico ( exc)de - 1 las atraquinonas (1, 2 y 3). La formación de los intermediarios S , SH°, y O2 fue evaluada y con- firmada como factible. Se encontró que el anión es la especie más probable en la fotoionización de los sustratos biológicos. Palabras claves: antraquinonas, Aloe vera, fotolabilidad, CATIVIC, correlaciones teóricas, reactividad fotoquímica.

* Autor.. para la correspondencia. E-mail: [email protected]

Scientific Journal of the Experimental Faculty of Sciences, at the Universidad del Zulia Volume 16 Nº 3, July-September 2008 F. Ruette et. al. / Ciencia Vol. 16, Nº 3 (2008) 370 - 378 371

1. Introduction reactive oxygen species or nucleic acids and amino acids oxidation by the excited triplet Anthraquinones (1,8-dihydroxyanthra state specie (12), as also it photoinduces quinones, (DHA)) are present in several cleavage on DNA by aloe-emodin (13). Photo- plants mainly as glycosides. These com- toxicity due to occasional or therapeutical pounds have both therapeutic and cytotoxic contact with chemicals is attracting clinical properties related with diverse types of hu- and research interest because of its increas- man affections. For example, aloe-emodin ing relevance. Progress in this field has (1,8-dihydroxy-3-hydroxy-methylanthra probably been hindered by its intrinsic com- quinone, (1)) is a compound present in Aloe plexity; therefore, the need for an interdisci- vera leaves, which exhibits antifungal (1), plinary approach and the lack of reliable in mutagenic (2), and tumorigenic properties vitro tests for assaying the photosensitizing (3). Although in a recent report aloe-emodin capability of xenobiotics in general. was postulated as a new lead antitumor drug (4); emodin (1,3,8-trihydroxy-6-methylan In connection with these facts, the cor- thraquinone, (2)) possesses anticancer, diu- relation between experimental data of UV-A retic, antibacterial, vasorelaxant (5) and and visible irradiations on aloe-emodin, anti-inflammatory effects (6); and rhein emodin, and rhein molecules and electronic (1,8-dihydroxyanthraquinone-3-carboxilic properties of these drugs and their free radi- acid, (3)) is the active metabolite of diace- cal intermediates formed during its photo- tylrhein that can be used as a drug for the degradation may be of great importance to treatment of patients with osteoarthritis. establish the molecular bases of their photo- stability and their phototoxicity. On the other hand, some synthetic and vegetable-origin anthraquinones have Theoretical calculations with the para- shown to be able to produce free radicals metric method CATIVIC were carried out to and singlet oxygen highly cytotoxic, when understand electronic and structural as- they are irradiated with visible light (7). Spe- pects of 1, 2 and 3 molecules and their re- cifically, phototoxicity indications have been spective radicals; see fig. 1. Correlation be- detected in vitro in Chinese hamster V79 tween electronic structure and reactivity has cells irradiated with light UVA and UVB in been used to understand some details of the presence of emodin (8). These findings these complex systems. A simple method suggest, in general terms, that the irradia- and new theoretical tools were employed to tion of DHAs can lead to the production of try to explain the behavior of the above men- reactive oxygen species able to produce toxic tioned molecules. effects on diverse cellular systems. Also, The present work is organized in the fol- some evidences in vivo confirm the impor- lowing manner. A brief description of the ba- tance of discoveries in vitro. For example, sis of the method employed is presented in two cases of photodermatitis have been re- section 2 with the corresponding tools to ported: the first one (9), caused by an- evaluate properties of interest. Section 3 thraquinone (9,10-anthracenedione) and deals with the discussion of experimental re- the other one by a preparation coming from sults and interpretation of theoretical calcu- Aloe vera (10). lations in order to propose an explanation of Recent photochemical works have experimental facts. Finally conclusions and shown a comparative study of the photosta- comments are presented in Section 4. bility and phototoxicity (photohemolysis on human erythrocytes) (11). Other researches 2. Theoretical method have also shown that emodic acid is poten- The theoretical calculations were car- tially phototoxic through photogeneration of ried out in the program CATIVIC (14) based

Scientific Journal of the Experimental Faculty of Sciences, at the Universidad del Zulia Volume 16 Nº 3, July-September 2008 372 Correlations between theoretical calculations and experimental photochemical reactivity

$= &= $ # &

"= ! =

" 

# % ' !  $ & " = #= '

'= '> $= % "=

Carbon

Oxygen

Hydrogen

Figure 1. Structure of anthraquinones aloe-emodin (1), emodin (2) and rhein (3) molecules. on simulation techniques of parametric traction, electron-electron repulsion, reso- methods (15-17). For example, the binding nance, and core-core repulsion, respectively energy functional for a molecular system of (15). It is easy to see that the total energy for N atoms can be defined as, a molecular system (E) is divided in diatomic e e ( XY ) and monoatomic ( X )energy terms 12/ æ 2 ö çå I - I ÷ min DE I Î ç EEexa pa ÷ [1] pa f è ø E =+å eeå [4] ISÎ XXY X XY>

I = 1({}{}{}{} Efjpa XXmmnn,,, k U X m nRes XY , The molecular and atomic parameters come from the semi-empirical method {}Rep,,,{}{} V g {}R ) [2] XY XYmmVV XY XY MINDO/SR (18) that is based on MINDO/3 (19). The bond strength was analyzed using where f is a family of functionals of bind- different tools: bond orders, diatomic ener- Y gies (DE) (20) and diatomic bonding energies ing energy. I is associated to the state I that belongs to an orthonormal set of eigenfunc- (DBE) (21). In the case of DBE, the X-Y bond {}Y strength is may be defined as, tions of Hexa that is complete in a sub- I space S of the Hilbert space (H). The E pa and I ()-=ee + () - +-()e DBE X YXY f X X Y X fXYY Y [5] E exa variables are parametric and exact en- ergy functionals for the state I, respectively. I and The term E pa depends on elementary of one-center {}{}{}jkUmmnn,, m n, two- ()-=eeå XX X fXXXYXY Y / [6] ¹ centers YX {}{}{}{}V ,,g Res , Rep para- XYmmnn XY XY m n XY m n 3. Investigations and results metric functionals, and the set of intera- There is sufficient evidence that etha- tomic distances {}R . These one-center XY nolic solutions of 1-3 compounds, shown in functionals correspond to electronic repul- fig. 1 are photolabile under aerobic condi- sion, exchange, and kinetic and attractive tions by irradiation with visible light (340- energy, respectively. On the other hand, 500 nm). The photolysis of the drugs 1-3 was two-center functionals are electron-core at- reported by Vargas et ál. (11) under these

Scientific Journal of the Experimental Faculty of Sciences, at the Universidad del Zulia Volume 16 Nº 3, July-September 2008 F. Ruette et. al. / Ciencia Vol. 16, Nº 3 (2008) 370 - 378 373 conditions and was followed by monitoring ings can be explained in terms of the micro- the changes of their absorption and emis- environment effects on the spectroscopic sion bands. These authors concluded that properties of the molecules. Absorption the photolability of ethanolic solutions of spectra change dramatically with solvent aloe-emodin and emodin are higher than properties, such as polarity, polarizability, rhein under visible light and aerobic condi- hydrogen-bonding capacity, PH, and viscos- tions (i.e., 1 > 2 > 3). Fig. 2 shows an experi- ity. For example, molecule 3 has the lowest mental comparison of the photodegradation photoreactivity because of the high polarity velocity of the three anthraquinones studied of this anthraquinone produces a water sol- by measuring the absorbance at 430 nm as vation layer from the surrounding microen- a function of the time. vironment. This effect generates a sort of protection to absorbants, creating an obsta- In order to understand from the mo- cle in the formation of photoproducts and lecular point of view the order of photode- also enhances recombination of dissociated gradation velocity, we calculate energetic radical fragments. and electronic properties of selected mole- cules of 1,8-dihydroxyanthraquinones Radical formation may be related with aloe-emodin (1), emodin (2) and rhein (3) values of HOMO, LUMO, and energy differ- shown in fig. 1. Correlation between theo- ences (D(|H-L|)). For example, a high HOMO retical and experimental data were analyzed absolute value (a very negative and stable in terms of photodegradation velocity and HOMO) implies more difficulty to excite the calculated global molecular properties, molecule. Thus, the lowest HOMO value such as, total dipolar moment (DP), highest (-0.3279 au for 3) corresponds to the least occupied molecular orbital (HOMO), lowest active anthraquinone. However, 1 and 2 occupied molecular orbitals (LUMO), molecules show very close HOMO values, D(|HOMO-LUMO|) energies (D(|H-L|)), and but they present a converse behavior. It heat of formation (HF). Calculated values of means, |(HOMO(1)|>|HOMO(2)|) and the these properties are in Debye, au, and order should be |(HOMO(1)|<|HOMO(2)| < kcal/mol, as shown in table 1. The magni- |(HOMO(3)|. tude of molecular DM follows the order 1 < 2 With respect to LUMO values, negative < 3, in inverse order withexperimental pho- quantities indicate that these molecular sys- todegradation findings. A possible correla- tems have a non negligible electronic affin- tion of these results with experimental find- 0.8

0.7 0.6

0.5 Rhein 0.4 Emodin 0.3 Aloe emodin 0.2 Absorbance at 430 nm 0.1 0 0 100 200 300 400 Time (min)

Figure 2. Experimental photodegradation velof the anthraquinones (1), (2) and (3) (l= 430 nm).

Scientific Journal of the Experimental Faculty of Sciences, at the Universidad del Zulia Volume 16 Nº 3, July-September 2008 374 Correlations between theoretical calculations and experimental photochemical reactivity

Table 1 Global properties of dihydroxyanthraquinones (1, 2 and 3): dipolar moment (DM), highest occupied molecular orbitals (HOMO), lowest occupied molecular orbitals (LUMO), (|HOMO-LUMO|) energy difference (D| H-L|) and heat of formation (HF).

Molecule DM HOMO LUMO D| H-L| HF (D) (au) (au) (au) (kcal/mol) Aloe-emodin 3.6259 -0.3171 -0.0222 0.2949 -161.1 (1) Emodin 3.6508 -0.3150 -0.0204 0.2946 -174.1 (2) Rhein 7.9787 -0.3279 -0.0339 0.2940 -197.1 (3)

l D ity. This suggests that electronic transfer Table, energy HOMO values, exc, | H-L|, l from electron donator groups to these mole- emi, and DM are also displayed in order to cules is possible during the electronic exci- make easier comparisons with photoreactiv- l tation processes. However, there is not a ity. We found that the exc values (451, 465, correlation between LUMO energy and the and 437 nm) may be related with the corre- photodegradation velocity. Note that a sin- sponding HOMO energies (-0.3171, -0.3150, gle determinant function does not represent and -0.3279 au), respectively. In addition, a correctly unoccupied orbitals and the use of good correlation between excitation energies e l multiconfigurational approaches is re- ( exc =h/ ex) and HOMO energies is found, i.e. e e e quired (22) to calculate electron affinity. On exc (3)> exc (1)> exc (2) and HOMO(3)> D the contrary, the |H-L| energy difference HOMO(1) > HOMO(2). Note that high values presents very close values and follows the of excitation wavelengths are associated to order 1 > 2 > 3. The possibility to excite an low excitation energies and viceverse. electron from HOMO to LUMO shows an in- verse trends that the experimental photore- On the other hand, experimental activy rate; i.e., the less reactive compound maxima wavelengths of emission (11) are: l 3 has the lowest D|H-L| energy. emis= 537, 535, and 513 nm, correlate with the photoreactivity of 1, 2, and 3 com- Thermodynamic stability of these f pounds. The quantum yield ( fl= molecules is associated with the heat of for- l l l ( emis+ exc)/ emis) is calculated as: 0.16, 0.14, mation. A very stable molecule would be and 0.04 for the above compounds, respec- more difficult to transform in the corre- tively. It was found a direct correlation be- sponding radical. In this case, the order of tween the photolabilitiy and quantum yield photolability: HF(1) fl(2)> fl(3) in inverse concordance with the order of photoreactiv- order of dipolar moments, as shown in table ity (higher HF less photoreactive). Similar 2(DM(1)

Scientific Journal of the Experimental Faculty of Sciences, at the Universidad del Zulia Volume 16 Nº 3, July-September 2008 F. Ruette et. al. / Ciencia Vol. 16, Nº 3 (2008) 370 - 378 375

Table 2 Correlation between wavelength of excitation (lexc), wave length of emission (lemi), and quantum yield of fluorescence data (Ffl) and calculated properties of dihydroxyanthraquinones.

l D l F Molecule HOMO exc (|H-L|) emi DM fl (au) (nm) (au) (nm) (D) Aloe-emodin -0.3171 451 0.2949 537 3.6259 0.16 (1) Emodin -0.3150 465 0.2947 535 3.6508 0.14 (2) Rhein -0.3279 437 0.2939 513 7.9787 0.04 (3)

+ singlet excited states. These states can over- oxygen singlet excited state (1 å ) at about come an intersystem crossing process to g produce long-lived triplet states (24). Triplet 38 kcal/mol (25). This suggests that a faster energy transfer of the triplet sensitizer could sensitizer states can undergoes interactions + with molecules in their vecinities. An energy be first to the 1 å state and then it decay to transfer may occur to the oxygen molecule g the longest live state 1 D (45 min) through ground state (triplet state) to the excite state g (singlet state). Singlet state of molecular internal conversion process. oxygen is highly reactive and it is involved in Beside the mechanism of singlet oxy- photodynamic reactions that result in a gen photooxidation, S molecules may un- fully oxidized form of biological systems dergo reactions of electron transfer from the (substrate). Thus, we evaluate also total en- substrate (amino acids, proteins, nucleic ac- ergies of the triplet and singlet states for the ids, and other biological substances). The re- O molecule. The calculated excitation en- - 2 duced sensitizer (S ) can react with O2 to pro- ergy triplet ® singlet (1 D ) was about 27 - g duce superoxide radical (O2 ) that is highly kcal/mol that compares reasonable well active for oxidation of substrates (24). Thus, with experimental data of 22 kcal/mol (25). energy of negatively charged 1, 2 and 3 an- 2 - It would be expected that the energy re- thraquinones in the lowest doublet state ( S ) quired to form the triplet state would be re- were evaluated. Energy differences between lated with the formation of oxygen singlet. In charged anthraquinones and neutral ones D 2 - 1 order analyze these values excited triplet ( E( S - S)) show that anions are more stable states of 1, 2, and 3 sensitizers were calcu- than neutral systems, as displayed in the lated with respect to the correspondent third row of Table 3. This result may be ex- singlet states (ground states). Energy values plained by the electronic delocalization in of 40.2, 43.5, and 40.2 kcal/mol are dis- these aromatic compounds. Similar results played in the second row of table 3. The or- were reported in theoretical studies of the p der found (2 > 1 » 3) is contrary to that ex- anionic -radicals formation from polyaro- perimentally observed. Therefore, there is matic starting materials used for the car- not a correlation between energies of triplet bonization process (26). states formed from the photoactive molecule 2 - 3 Electron transfer from S to O2 is not and the photodegradation velocity. These feasible from the thermodynamic point of triplet states are close in energy to another 3 ® 2 - view because the energy gain of O2 +e O2 is only -10 kcal/mol (27) compared with -31

Scientific Journal of the Experimental Faculty of Sciences, at the Universidad del Zulia Volume 16 Nº 3, July-September 2008 376 Correlations between theoretical calculations and experimental photochemical reactivity

Table 3 Energy differences (DE) of triplet excited state (3S), the negatively charged sensitizer (2S-), the hydrogenated radical in a doublet (2SH°), and the quadruplet (4S-) excited states with respect to the singlet (1S) ground state of sentitizers. In addition, excitation energies for 4S- and 4SH° radicals to their respective excited states 2S- and 2S-. Energy values are in kcal/mol.

DE Energy Difference (3S- 1S) (2S- - 1S) (4S- - 2S-)(4S- - 1S) (2SH° - 1S) (4SH° - 2SH°) Aloe-emodin 40.2 -31.1 54.1 23.0 -63.1 53.9 (1) Emodin 43.5 -31.4 46.1 14.7 -62.6 48.1 (2) Rhein 40.2 -41.7 50.0 9.3 -63.6 40.9 (3) kcal/mol for the 1S+e® 2S-. Because the 2S- Another process for the photodynamics systems are stable, excitations of the double of these studied compounds may involve an ground state to another doublet state of H transfer from the substrate to the sensi- higher energy may occur and then an inter- tizer to produce 2SH° radical. Then, it reacts system crossing mechanism leads to a with O2 to give hydrogen peroxide (H2O2). Cal- quadruplet excited state, see values in col- culations were carried out for 2SH° com- umn forth of table 3. On the other hand, re- pounds shown in fig. 1 adding H atoms to sults of total energy differences (D(4S--1S)) O(15), O(17) in 1 and 3 and to O(15), O(18) in between the excited quadruplet sensitizer 2, respectively. Results, displayed in column (4S-) and the ground state 1S may be the driv- six of table 3, indicate that formation of S-H ing force for the electron transfer reaction 4S- bond is very favorable (about -63 kcal/mol) 3 ® 1 2 - 4 - ® 1 1 + O2 S+ O2 . Results for the S Spro- with respect to the S sensitizer, because of cess are presented in column five of table 3. the formation of a new O-H bond. The order of these values correlates with the The formation of H O by H° transfer to experimental photoreactivity. It means that 2 2 4 - O from the hydrogenated sensitizer may be S , surrounded by O molecules, releases an 2 2 ® 1 produced in two steps: H° + O2 O2H° and electron as decays to S and as more energy ® O2H°+H° H2O2. The energy (E) gained in the is liberated more probability that an elec- 2 3 formation of O2H° (E( HO2)–E(O2) – E(H°)) is tron transfer to O2 occurs. The excited state 4 - only -51.1 kcal/mol. This is smaller than S may also transfer an electron by and in- that obtained from the formation of S-H termediate excited state as the following 2 4 - ® 3 bond. Therefore excitation of SH° to an- process ( S S + e), but it requires energy 3 1 4 - 3 1 other doublet state or O ® O must occur to because S is more stable than S. If O ex- 2 2 2 assure H transfer. Excited 2SH° can over- ists in some concentration, the formation of 4 2 - come intersystem crossing to a SH°. Re- O2 is also feasible because the energy differ- 4 2 1 2 - sults of ( SH° - SH°) energy differences are ence of the O ® O process would be about 2 2 presented in the last column of table 3. The 32 kcal/mol (25, 27). Thus the correlation order is concordant with experimental find- between experimental data and theoretical ings; i.e., higher release of energy implies findings results suggest that the formation 2 - higher photolability. Nevertheless, the en- of O maybe a controlling step in the photo- 2 ergy released in going from 4SH° to 2SH° is dynamics process of these S compounds.

Scientific Journal of the Experimental Faculty of Sciences, at the Universidad del Zulia Volume 16 Nº 3, July-September 2008 F. Ruette et. al. / Ciencia Vol. 16, Nº 3 (2008) 370 - 378 377 not enough to overcome the stability of 2S-H mechanism of intersystem crossing between + bond. The whole process, however, is ther- 3S and 3O to give 1Ointhe1 å state and it modynamically feasible because the forma- g ® decays to the longest life state, 1 D . tion of O-H bonds in H° + O2 O2H° process g contribute with about -51 kcal/mol. None- 5) Formation of a 2S- intermediate is fea- theless, the formation of H2O2 is less prob- 2 sible because the electronic charge can be able than the formation of SH°, because it delocalized in the aromatic systems. Excita- requires the transfer of two H° atoms. tion of this species to doublet and through Finally energy formation of radicals an intersystem crossing process to generate 4 - and bond strengths (diatomic binding ener- S is required to energetically allow the elec- 3 2 - gies from Eq. (5)) for all O-H bonds in the an- tron transfer to the O2 to form O2 . There is a thraquinones were evaluated. For example, correlation between excitation energies and the strength of O(16)-H, O(18)-H, O(20)-H experimental findings and therefore, maybe 2 - bonds in compound 1 (see labels shown in the formation of O2 is an intermediate in the 1 fig. 1) were calculated. No correlations be- photodynamic of these sensitizers. If O2 ex- - tween the photodegradation rate and the ists in some concentration around S radi- radical stability and O-H bond strengths cals, it also would contribute to the forma- 2 - were found. tion of O2 . 6) Hydrogen transfer mechanism to 4. Conclusions and comments form 2SH° radicals is highly probable for substrates that have weak H atoms because 1) Qualitative parametric methods can the stability of these radicals. Excitation of be used, for establishing correlation be- 2SH° to doublets and the decay to quadru- tween experimental data of phototoxicity of plets may occur in order to make possible aloe-emodin, emodin, and rhein and some the transference of H to 3O to form O H°. theoretical calculated properties. 2 2 There is concordance with experimental 2) Calculated total dipolar moment and findings; i.e., a high excitation energy entails heat of formation of the compounds 1, 2, high photolability. However, this mechanism and 3, used as sensitizers, correlate with ex- involves the transfer of two H atoms to form perimental photodegradation velocity. It H2O2 and for that reason it is less probable. was found that a higher stability and polar- 7) Calculations of relative stability of ity of the sensitizer implies less photoactiv- radicals from 1, 2 and 3 and O-H bond ity. High stability of sensitizer means more strength do not correlate with the photolysis difficulty to generate the correspondent rate of these anthraquinones. radical and high dipolar moment signifies a less effective radical formation because of 8) It is good to note that one single de- solvation effects. terminant and parametric methods are quite approximated approaches. Therefore, re- 3) Correlation of calculated HOMO en- sults shown are qualitative and further ergy of compounds 1, 2 and 3 and experi- studies of excited states are required in or- mental wavelength of excitation (l )isob- exc der to gain a deeper understanding of the served. photodynamic processes of these anthraqui- 4) Different mechanisms of photoxida- nones. tion were explored: formation of singlet oxy- gen, hydrogen transfer, and creation of Acknowledgments negatively charged sensitizer. Singlet oxy- gen formation energy does not correlate with This research was supported by a grant the photolability experimental findings. A from FONACIT G-9700667 project.

Scientific Journal of the Experimental Faculty of Sciences, at the Universidad del Zulia Volume 16 Nº 3, July-September 2008 378 Correlations between theoretical calculations and experimental photochemical reactivity

References 13. VATH P., WAMER W.G., FALVEY D.E. Photo- chem. Photobiol. 75: 346-352, 2001. 1. AGARWAL S.K., SINGH S.S., VERMA S., KU- 14. RUETTE F., SÁNCHEZ M., MARTORELL G., MAR S. J. Ethnopharmacol. 72: 43-46, GONZÁLEZ C., AÑEZ R., SIERRALTA A., 2000. RINCÓN L., MENDOZA C. Int. J. Quantum 2. WESTENDORF J., MARQUARDT H., POGIN- Chem. 96: 321-332, 2004. SKY B., DOMINIAK M., SCHMIDT J., MAR- 15. ROMERO M., SÁNCHEZ M., SIERRAALTA A., QUARDT H. Mutat. Res. 240: 1-12, 1990. RINCÓN L., RUETTE F. J. Chem. Infor. Comp. 3. WOLFLE D., SCHMUTTE C., WESTENDORF Sci. 39: 543-549, 1999. J., MARQUARDT H. Cancer Res. 50: 6540- 16. PRIMERA J.R., ROMERO M., SÁNCHEZ M., SI- 6544, 1990. ERRALTA A., RUETTE F. J. Mol. Struct. Theo- 4. PECERE T., GAZZOLA M.V., MUCIGNAT C., chem. 9: 177-190, 1999. PAROLIN C., DALLA VECCHIA F., CAVAG- 17. RUETTE F., GONZÁLEZ C., OCTAVIO A. J. GIONI A., BASSO G., DIASPRO A., SALVATO Mol. Struct. (Theochem.) 537: 17-25, 2001. B., CARLI M., PALÙ G. Cancer Res. 60: 2800-2804, 2000. 18. BLYHOLDER G., HEAD J., RUETTE F. Theor. 5. CHANG L.C., SHEU H.M., HUANG Y.S., TSAI Chim. Acta 60: 429-444, 1982. T.R., KUO K.W. Biochem. Pharmacol. 58: 19. BINGHAM R.C., DEWAR M.J.S., LO D.H. J. 49-57, 1999. Am. Chem. Soc. 97: 1285-1293, 1975. 6. KUO Y.C., MENG H.C., TSAI W.J. Inflamm. 20. SANCHEZ M., RUETTE F.: J. Mol. Struct. Res. 50: 73-82, 2001. (Theochem.) 254: 335-342, 1992, and refer- 7. INBARAJ J.J., GANDHIDASAN R., MURUGE- ences therein. SAN R. Free Rad. Biol. Med. 26: 1072-1078, 21. RUETTE F., POVEDA F. M., SIERRAALTA A., 1999. RIVERO J. Surf. Sci. 349: 241-247, 1996. 8. KERSTEN B., ZHANG J., BRENDIER- 22. BETOWSKI L.D., ENLOW M., RIDDICK L. SCHWAAB S.Y., KASPER P., MULLER L. Mu- Comput. Chem. 26: 371-377, 2002. tat. Res. 445: 55-71,1999. 23. AGUIAR C., VARGAS F., CANUDAS N., RU- 9. BRANDAO F.M., VALENTE A. Contact Der- ETTE F. Mol. Eng. 4: 451-463, 1995. matitis 18: 171, 1988. 24. GILBERT A., BAGGOTT J. Essentials of Mo- 10. DOMÍNGUEZ-SOTO L. Int. J. Dermatol. 31: lecular Photochemistry, Blackwell Scientific 372-373, 1992. Publications, Oxford (Great Britain) Chap. 1, 11. VARGAS F., FRAILE G., VELÁSQUEZ M., 2, 4, 1991.

CORREIA H., FONSECA G., MARÍN M., MAR- 25. FRIMER A.A. (Ed.) Singlet O2. CRC Press, CANO E., SÁNCHEZ Y. Die Pharmazie 57: Boca Ratón, Florida (USA), Vols. 1-4, 1985. 399-404, 2002. 26. RUETTE F., SIERRAALTA A., CASTELL V., 12. RAHIMIPOUR S., BILKIS I., PÉRON V., GES- LAYA M. Carbon 31: 645-677, 1993. CHEIDT G., BARBOSA F., MAZUR Y., KOCH 27. CRC. Handbook of Chemistry and Physics. Y., WEINER L., FRIDKIN M. Photochem. Pho- 79th ed. CRC Press, New York (USA), 1998. tobiol. 74: 226-236, 2001.

Scientific Journal of the Experimental Faculty of Sciences, at the Universidad del Zulia Volume 16 Nº 3, July-September 2008