December 11, 2007
Effects of the Interaction Between Ss-Carboline-3-Carboxylic Acid N- Methylamide and Polynucleotides on Singlet Oxygen Quantum Yield and DNA Oxidative Damage
By Garcia-Zubiri, Inigo X Burrows, Hugh D; de Melo, J Sergio Seixas; Pina, Joao; Monteserin, Maria; Tapia, Maria J
ABSTRACT The complexation of beta-carboline-3-carboxylic acid N- methylamide (betaCMAM) with the sodium salts of the nucleotides polyadenylic (Poly A), polycytidylic (Poly C), polyguanylic (Poly G), polythymidylic (Poly T) and polyuridylic (Poly U) acids, and with double stranded (dsDNA) and single stranded deoxyribonucleic acids (ssDNA) was studied at pH 4, 6 and 9. Predominant 1:1 complex formation is indicated from Job plots. Association constants were determined using the Benesi-Hildebrand equation. betaCMAM- sensitized singlet oxygen quantum yields were determined at pH 4, 6 and 9, and the effects on this of adding oligonucleotides, dsDNA and ssDNA were studied at the three pH values. With dsDNA, the effect on betaCMAM triplet state formation was also determined through triplet- triplet transient absorption spectra. To evaluate possible oxidative damage of DNA following singlet oxygen betaCMAM photosensitization, we used thiobarbituric acid-reactivity assays and electrophoretic separation of DNA assays. The results showed no oxidative damage at the level of DNA degradation or strand break.INTRODUCTION
beta-Carbolines are alkaloids present in many plants, and have interesting photophysical and pharmacological properties. Because of their luminescence characteristics, they have also been used as fluorescent standards (1,2). It is known that in the presence of oxygen, beta-carbolines can lead to phototoxicity, which is suggested to occur by photosensitized production of either superoxide radical anion (3,4) or singlet oxygen (3). The latter process, frequently termed Type II sensitization, is thought to involve energy transfer from long-lived triplet excited states of photosensitizers (5), and has been observed with porphyrins (6), fullerenes (C^sub 60^) (7), the neutral form of beta-carbolines (3,8) or other alkaloids (berberine and palmatine) (9), benzophenone (10), or other aromatic ketones (11), riboflavin (12), rose bengal (13), methylene blue (14), polyyines (acetylenes, thiophenes and related compounds), furanyl compounds and quinones (15).
It is believed that one of the major phototoxic processes in vivo involves reactions of singlet oxygen (^sup 1^O2), or other reactive oxygen species, with target molecules. When such oxidative damage happens with DNA, it may cause cell death, aging, cancer, mutagenesis or other human diseases (16,17). Damage to nucleotide bases commonly results in the formation of the 7,8-dihydro-8- oxoguanosine residue 8-0xoG from guanosine through a variety of mechanisms, suggested to involve attack by radicals or singlet oxygen, predominantly through a cycloaddition pathway (6,18-20). In addition to these deleterious aspects, singlet oxygen can be used in photodynamic therapy in a process that consists of combining a photosensitizing agent and focused irradiation to produce reactive species, such as ^sup 1^O2, which can destroy cancerous cells (21).
Due to the mutagenic and comutagenic properties of beta- carbolines (22-27), their potential in photodynamic therapy (28) and their antitumor properties (29-31), the interaction between beta- carbolines and DNA has been the focus of much attention (32-36) and has been studied with techniques, such as electronic spectroscopy (37), calorimetry and ^sup 1^H NMR spectroscopy (38).
In this work, we report the complexation of the beta-carboline-3- carboxylic acid N-methylamide (Fig. 1) with dsDNA and ssDNA at several pH values. This beta-carboline derivative, whose photophysical properties and acid-base behavior in the ground and first excited state in aqueous solution have previously been reported (39), is studied to determine the association behavior and analyze the role of intercalation on the betaCMAM-DNA interaction. The complexation with oligonucleotides (Poly A, Poly C, Poly G, Poly T and Poly U) is also considered to gain insight into which of the bases is most important for the interaction with DNA. The effect of DNA on the triplet state behavior of betaCMAM is studied with reference to the possible oxidative DNA damage induced by complexation. The betaCMAM-photosensitized singlet oxygen quantum yields are determined at pH 4, 6 and 9 alone and in the presence of polynucleotides. Finally, the possible effect of photosensitization by betaCMAM on the degradation of DNA is investigated by electrophoresis and thiobarbituric acid (TBA)-reactivity assay.
We carried out this singlet oxygen photosensitization study with betaCMAM because it is one of the most acidic beta-carboline derivatives (pK^sub a^ 3.8), so at pH values between 6 and 7 betaCMAM is in its neutral form (39), and singlet oxygen is mainly photosensitized by long-lived triplet state of the neutral form of beta-carbolines (8). On the contrary, most of the other beta- carboline derivatives have pK^sub a^ values around 7 (39), so at pH values between 6 and 7 (pH of highest biological interest), most of the beta-carbolines are in the cationic form, whose triplet state is apparently less efficient in singlet oxygen photosensitization.
Figure 1. Structure of betaCMAM.
The main novelty of this work is the fact that it combines the study of the singlet oxygen photosensitization by a beta-carboline derivative interacting with oligonucleotides and DNA (doubled and single stranded) with the study of the DNA-bound beta-carboline derivative triplet state and DNA degradation upon interaction with beta-carboline.
MATERIALS AND METHODS
Materials. beta-Carboline-3-carboxylic acid N-methylamide (betaCMAM), the sodium salts of the polyadenylic (Poly A), polycytidylic (Poly C), polyguanylic (Poly G), polythymidylic (Poly T) and polyuridylic (Poly U) acids and double strand deoxyribonucleic acid (dsDNA) from salmon testes were purchased from Sigma. A^sub 260^/A^sub 280^ ratio of dsDNA solutions (50 [mu]g mL^sup -1^) was between 1.8 and 2, indicating the absence of protein contamination. Solutions were prepared in methanol (Fluka, spectrophotometric grade), deuterium oxide (Merck) and Millipore MilliQ water. The pH of solutions was adjusted by the addition of appropriate volumes of aqueous solutions of sulfuric acid and sodium hydroxide.
Single strand DNA (ssDNA) stock solutions of concentrations around 1.1 x 10^sup -2^ M (epsilon^sub 260^ = 6600 mol^sup -1^ L cm^sup -1^) (40) were prepared by heating dsDNA solutions at 85- 90[degrees]C for 15 min and dipping immediately into ice for fast cooling to prevent renaturation (41). To avoid re-absorption, solutions for spectroscopic measurements were prepared by diluting stock solutions of betaCMAM (10^sup -2^ M in methanol) 100-fold with Millipore water to give final concentrations around 10^sup -4^ M in 1% methanol-water (vol/vol). Appropriate amounts of polynucleotide solutions were added to this betaCMAM solution in a spectrometer cuvette to give polynucleotide concentrations: 6.1 x 10^sup -3^ M (Poly A), 4.2 x 10^sup -3^ M (Poly C), 8.6 x 10^sup -3^ M (Poly G), 1.3 x 10^sup -3^ M and 1.5 x 10^sup -3^ M (Poly U). Polynucleotide concentrations were calculated from their molar absorption coefficients: Poly A (epsilon^sub 256^ = 10400 mol^sup - 1^ L cm^sup -1^) (42), Poly C (epsilon^sub 269^ = 6700 mol^sup -1^L cm^sup -1^) (42), Poly G (epsilon^sub 254^ = 9900 mol^sup -1^L cm^sup -1^) (43), Poly T (epsilon^sub 264^ = 8520 mol^sup -1^L cm^sup -1^) (44) and Poly U (epsilon^sub 260^ = 8900 mol^sup -1^L cm^sup -1^) (42). The pH of these solutions were adjusted to 5.5-6 by addition of a few microliters of aqueous NaOH solution.
For photophysical studies, solutions were degassed by bubbling with argon for around 15 min. For singlet oxygen experiments, solutions were prepared using deuterium oxide because of the longer lifetime of ^sup 1^O2 in this solvent, and the final betaCMAM concentration was 10^sup -4^ M in ca 5% methanol/D^sub 2^O.
Methods: Electrophoretic separation of damaged DNA. Electrophoresis of two sets of dsDNA-betaCMAM samples (freshly prepared and incubated) was carried out using 1% (wt/vol) agarose gels. These were observed under ultraviolet light using a transilluminator and photographed. Solutions with DNA (200 [mu]g mL^sup -1^, 6.1 x 10^sup -4^ M) were exposed to the action of betaCMAM. The betaCMAM/DNA concentration ratios ranged from 1.6 x 10^sup -4^ to 1. Five hundred microliters of freshly prepared solutions were frozen and the same volume of solution was incubated for 12 h at 25[degrees]C in the presence of light. As control sample, DNA was exposed to the action of hydroxyl radicals generated by a mixture of ascorbic acid (10^sup -3^ M) and copper sulfate (II) (10^sup -4^ M) in aqueous solution. Samples (50 [mu]L) of these two sets of solutions were used for electrophoresis.
Thiobarbituric acid-reactivity assay of oxygen radical damage. The release of TBA-reactivity as an indication of DNA degradation was measured following the procedure described by Quinlan and Gutteridge (45). Five hundred microliters of trichloroacetic acid (28%, wt/ vol) and 500 [mu]L of TBA (1%, wt/vol) were added to 450 [mu]L of the betaCMAM/DNA solution and control samples prepared for the electrophoresis experiments. These were thoroughly stirred, heated at 100[degrees]C for 30 min and allowed to cool. The release of the TBA-reactive products was followed spectrophotometrically at 532 nm. Absorption spectra were recorded in 1 cm quartz cuvettes on a Shimadzu UV-2501 PC spectrophotometer. Steady-state fluorescence spectra were measured with a Shimazdu RF-5301 PC spectrofluorimeter with 3.0 nm excitation and 1.5 nm emission bandwidths with excitation at 335 nm. All spectra were registered at 298 K. The quantum yields ([straight phi]) of the neutral and cationic forms of betaCMAM were determined in aqueous solution using norharmane (Sigma) in benzene ([straight phi] = 0.30) (46) and norharmane in 0.05 m H^sub 2^SO^sub 4^ ([straight phi] = 0.60) as standards (1), respectively.
Fluorescence decays were measured using a home-built Time- Correlated Single Photon Counting apparatus consisting of an IBH NanoLED (lambda^sub exc^ = 339 nm) as excitation source, Jobin-Ivon monochromator, Philips XP2020Q photomultiplier, and Canberra instruments Time-to-amplitude converter and Multichannel Analyser. Alternate measurements (1000 counts per cycle), controlled by Decay(R) software (Biodinamica-Portugal), of the pulse profile at 339 nm and the sample emission were performed until 1-2 x 10^sup 4^ counts at the maximum were reached (47). The fluorescence decays were analyzed using the modulating functions method of Striker with automatic correction for the photomultiplier "wavelength shift" (48). All experiments were carried out at room temperature (293 K).
Room temperature singlet oxygen phosphorescence was detected at 1270 nm using a Hamamatsu R5509-42 photomultiplier, cooled to 193 K in a liquid nitrogen chamber (Products for Research model PC176TSCE- 005), following laser excitation of aerated solutions at 355 nm (OD at 355 nm = 0.20), with an adapted Applied Photophysics flash kinetic spectrometer. The modification of the spectrometer involved the interposition of a Scotch RG665 filter. A 600-line diffraction grating was used instead of the standard spectrometer to extend spectral response to the infrared. The filter employed is essential to eliminate from the infrared signal all the first harmonic contributions from the sensitizer emission in the 500-800 nm region. The singlet oxygen quantum yield, [straight phi]^sub Delta^, was determined using 1H-Phenalen-1-one (perinaphthenone) in acetonitrile ([straight phi]^sub Delta^ = 1) as standard (49).
The experimental setup used to obtain triplet state absorption spectra consists of an Applied Photophysics laser flash photolysis apparatus pumped by an Nd:YAG laser (Spectra Physics). The detection system is at right angles to the excitation beam and uses a pulsed 150 W Xe lamp to analyze the transient absorption. The signal obtained is fed into an HP digital analyzer and transferred to an IBM RISC computer where the optical densities (OD) are collected at different wavelengths and different delays after flash using the appropriate software (Applied Photophysics). Transient triplet absorption spectra of betaCMAM were obtained by monitoring the optical density changes at intervals of 20 nm over the 300-800 nm range. First-order kinetics was observed for the decay of the lowest triplet state. Excitation was at 355 nm with an unfocused beam. Special care was taken to use diluted solutions (absorbance [asymptotically =] 0.2 in a 10 mm square cell) and low laser energy (=2 mJ) to avoid multiphoton events and T-T annihilation.
RESULTS AND DISCUSSION
We present initially a study of the interaction between betaCMAM and the polyelectrolytes (Poly A, Poly C, Poly G, Poly T, Poly U, and dsDNA and ssDNA) at three pH values. The effect of polynucleotide additions (oligonucleotides and DNA) on the betaCMAM singlet oxygen quantum yield is also studied, as well as the effect of adding DNA on the betaCMAM triplet absorption spectra and DNA oxidative damage upon complexation with betaCMAM.
Study of interactions
With oligonucleotides. The interaction of betaCMAM with the five polyelectrolytes at the probe natural pH (ca 6) was studied by both absorption and emission (steady-state and time-resolved) spectroscopies. Small amounts of NaOH were added to Poly G and Poly U solutions to keep the pH values between 5.5 and 6.
Job plots were obtained to probe the interaction between betaCMAM and oligonucleotides, and to estimate the binding ratio of the complex formed. This involved mixing various volumes of betaCMAM and oligonucleotide solutions (concentration 10^sup -4^ m) to give molar fractions of the two components ranging between 0 and 1. The absorbance at the maximum of betaCMAM (335 nm) or the changes in its emission intensity with oligonucleotide to fluorophore molar fractions ratio can be linearly fitted for low and high ratios. The crossing point of these two lines occurs for an oligonucleotide to betaCMAM concentration ratio of 1 indicating that a predominant 1:1 complex is formed for all the oligonucleotides studied in this work.
To study the interaction between betaCMAM and oligonucleotides, the absorption and emission spectra of samples were registered for various concentration of oligonucleotides (up to ca mM except for poly T, for which the maximum concentration possible was 4 x 10^sup -4^ M).
Figure 2. Absorption spectra betaCMAM with several oligonucleotide additions.
Absorption spectral studies: In the absorption spectra, the most important changes are observed upon addition of Poly G (Fig. 2) and PolyA (Fig. 2). In these two cases, the absorbance of the neutral form of betaCMAM with maxima at 335 and 347 nm diminishes and the absorbance for wavelengths longer than 355 nm increases. The presence of an isosbestic point at 355 nm indicates an equilibrium between two species. As the pH was kept approximately constant, these changes cannot be attributed to an acid-base equilibrium (39) but must correspond to equilibrium between complexed and free betaCMAM. With the other three oligonucleotides, Poly C, Poly U and Poly T, the changes in absorbance resulting from complexation upon addition of nucleotide are minor and can mainly be attributed to the increase in the spectrum background caused by light scattering.
We chose 347 nm as the titration wavelength because, in general, this is the wavelength for which a majority of significant spectroscopic changes are observed. For Poly G and Poly A the association constants obtained were 495 +- 51 and 283 +- 69 mol^sup - 1^ L, respectively.
Because of the effect of scattering following complexation and relatively minor spectroscopic changes observed it was not possible to determine reliable values for the association constants in the cases of Poly C. Poly T and Poly U.
Table 1. Benesi-Hildebrand association constants obtained from the emission data, K^sup BH^^sub E^; static Stern-Volmer constant, K^sup s^^sub SV^, and nonstatic Stern-Volmer constant, K^sup n- s^^sub SV^.
Good linear fittings were obtained plotting the ratio F^sub 0^/F versus [P] for all the oligonucleotides (from zero to millimolar concentrations) except for poly T. In the case of poly T, both the absorbance and the emission intensity slightly increases with the increase of the polynucleotide concentration (between zero and 4 x 10^sup -4^ m). Higher concentration values were not used due to the small amounts of material available. However, with the lifetime experiments, we observe that betaCMAM lifetime is approximately constant upon the addition of poly A and poly G, whereas with the other three oligonucleotides (Poly C, PolyU and PolyT), good Stern- Volmer plots are obtained plotting tau^sub 0^/tau versus [P]. This suggests that for the latter oligonucleotides, both static and dynamic quenching processes are present, although, in general (except for poly T), the slopes of the dynamic quenching plots are markedly smaller than those of the static ones, a situation that is compatible with the high density of the betaCMAM-polynucleotide solution.
The observation of the highest Stern-Volmer static quenching constants for Poly G and Poly A, and the lack of significant dynamic quenching are in agreement with strong ground state complexation in these systems, as seen by the large association constants obtained. The values for all the constants are summarized in Table 1.
With DNA. The interaction between betaCMAM with dsDNA and ssDNA at three pH (4, 6 and 9) values was also studied with absorption and emission spectroscopies. In the six DNA systems studied, Job plot indicates that, as with the oligonucleotides, a 1:1 complex is formed between betaCMAM and DNA in all cases.
The absorption spectra of the six betaCMAM:DNA systems are shown in Fig. 3. For dsDNA at the three pH values and for ssDNA at pH 4 and 6, the absorbance between 326 and 355 nm decreases, while that above 355 nm increases. An isosbestic point is observed at 355 nm, indicative of an equilibrium between free and DNA-complexed betaCMAM. From the changes in the absorption spectra shown in Fig. 3, it seems clear that the interaction of betaCMAM is stronger with dsDNA than with ssDNA, which seems to corroborate the hypothesis that beta-carbolines bind duplex DNA by intercalation (37), as suggested by a 25 ~ 65 nm bathochromic shift in the absorption bands upon binding to the nucleic acids (53). The fact that the bathochromic shift is not effectively seen for duplex DNA at pH 9 is in agreement with intercalation, as DNA is partially denatured at this basic pH (54). However, as the bathocromic shift is also observed when betaCMAM interacts with ssDNA (Fig. 3) and Poly G (Fig. 2), it must be accepted that the bathochromic shift is not necessarily indicative of beta-carboline intercalation. Furthermore, intercalation does not seem to be the only mechanism of interaction between this dye and the polynucleotides studied in this work. As with the oligonucleotides, Benesi-Hildebrand method has been applied to obtain quantitative information on betaCMAM association with DNA for 1:1 complex (Eq. 2). The calculations have been done at 347 nm (titration wavelength) as at this wavelength we obtain the largest changes in absorbance for most of the DNA systems studied. The results are summed up in Table 2. Due to the small spectroscopic changes and problems with light scattering (Fig. 3), no reliable value of the association constant was obtained for ssDNA at pH 9. For the other systems the results obtained are in reasonable agreement with the observed spectroscopic changes.
Figure 3. Absorption spectra of betaCMAM (10^sup -4^ M) with various concentrations of dsDNA at: pH 4 (0,4.3 x 10^sup -4^, 10^sup -3^, 1.6 x 10^sup -3^, 2.5 x 10^sup -3^ M), pH 6 (0, 2.5 x 10^sup -4^, 5.6 x 10^sup -4^, 10^sup -3^, 2.3 x 10^sup -3^ M), pH 9 (0, 10^sup -4^, 8.5 x 10^sup -4^, 1.7 x 10^sup -3^, 2.7 x 10^sup -3^ M) and ssDNA at: pH 4 (0, 10^sup -4^, 9.1 x 10^sup - 4^, 1.7 x 10^sup -3^, 2.8 x 10^sup -3^, 3.4 x 10^sup -3^ M), pH 6 (0, 5.9 x 10^sup -4^, 1.8 x 10^sup -3^, 3.0 x 10^sup - 3^, 3.7 x 10^sup -3^ M) and pH 9 (0, 10^sup -4^, 9.4 x 10^sup - 4^, 2.1 x 10^sup -3^, 2.1 x 10^sup -3^, 3.6 x 10^sup -3^ M).
The interaction is stronger with dsDNA than with ssDNA at pH 6 and 9 and fairly similar at pH 4. According to the association constants obtained for dsDNA (pH 6 > pH 4 > pH 9), intercalation can be expected to play an important role in betaCMAM-dsDNA interaction. This could explain that, in this case, higher association constant is obtained at pH 6 than at pH 4, as at pH 4, beta-carboline is likely to be protonated (39), and it is known that in the protonated betaCMAM form, the norharmane skeleton plane is twisted by 19.58[degrees] with respect to the plane carbon 3 substituent (39) while for the neutral form they are coplanar (39). The partial denaturation of dsDNA at basic pH values explains the low association constant observed at pH 9, as intercalation is less favored under these conditions.
Table 2. Association constants by Benesi-Hildebrand method for a 1:1 complex, K^sub A^^sup BH^.
However, the association constants with ssDNA follows the order (pH 4 > pH 6 > pH 9) indicating that steric hindrance does not affect the complexation of betaCMAM and ssDNA, and that it is favored for the protonated form.
It is worth pointing out that as the spectroscopic changes observed upon the addition of DNA are very similar to those observed upon the addition of Poly G (Figs. 2 and 3), and that no other oligonucleotide shows a similar effect, the interaction of betaCMAM with DNA probably takes place mainly through the guanine base.
It has been previously reported that when beta-carboline derivatives bind to DNA, they show a high selectivity for the G-C base pair (55); moreover, molecular modeling and experimental results obtained with various techniques also indicate that the planar beta-carboline ring system makes a good match with the G-C base pairs and for some beta-carboline derivatives (structurally related to betaCMAM), hydrogen bonds are formed between the terminal -NH2 of carboxamide side chain substituents of carbon 3 of the beta- carboline ring (such as with betaCMAM) and N^sub 7^ of the guanine at the intercalation site (56). Although charge transfer interactions between the beta-carbolines and nucleotide bases are possible, this type of hydrogen-bonding interaction along the groove of the DNA helix can stabilize the beta-carboline-DNA complex with these derivatives (56). This can explain why the highest association constant for ssDNA is observed for pH 4, when betaCMAM is protonated because the hydrogen bond with guanine base is favored and the plane twist (substituent-beta-carboline ring) upon the dye protonation (39) does not provide any steric hindrance for single strand DNA.
Table 3. Benesi-Hildebrand association constants obtained from the emission data, K^sup BH^^sub E^; static Stern-Volmer constant, K^sup s^^sub SV^, and dynamic Stern-Volmer constant, K^sup n-s^^sub SV^.
Table 4. phi^sub Delta^ of singlet oxygen production sensitized by free betaCMAM at three different pH values.
As with the oligonucleotides, the interaction of betaCMAM with dsDNA and ssDNA was studied with steady-state and time-resolved fluorescence. Rayleigh scattering for all these systems increased upon adding DNA, and in many cases a change in the slope of the increase was observed for DNA concentrations close to that of betaCMAM, supporting the formation of 1:1 complex.
From the quenching of emission, the Benesi-Hildebrand association constants, K^sup BH^^sub E^ (Eq. 3), were obtained and are shown in Table 3. They follow the same general tendency for the association constants obtained from the absorption data by Eq. (2). The association constants from the emission data are higher with dsDNA than with ssDNA, and the effect of pH is different in both cases. For dsDNA the association constants follow the order (pH 6 > pH 4 > pH 9) while for ssDNA they follow the order (pH 4 [much greater than] pH 6
The static and dynamic quenching of betaCMAM emission was studied with Eqs. (4) and (5), respectively. As in the case of the oligonucleotides, the static Stern-Volmer quenching constants are much higher than the dynamic ones, indicating that complexation between betaCMAM and DNA is the main quenching process. It is also noteworthy to see the good agreement between the Benesi-Hildebrand association constants K^sup BH^^sub E^ and the Stern-Volmer constant K^sup s^^sub SV^.
The quantum yield of singlet oxygen production (phi^sub Delta^), sensitized by free betaCMAM, was determined at three pH values (6, 4 and 9), as well as in the presence of various concentrations of oligonucleotides and double and single DNA at the same pH values and are summarized in Table 4. The values obtained are of the same order of magnitude of those obtained for harmine in comparable solvents (DC1/D^sub 2^O in the presence of 2% methanol, phi^sub Delta^ = 0.028) (8), but are lower than the values observed with other beta- carboline derivatives (norharmane, harmane and harmine) in organic solvents, such as benzene and acetonitrile (0.35) (57). At pH 4, the phi^sub Delta^ value is lower than at pH 6 and 9 and (within experimental error) can be considered constant at the two latter pHs. These results are compatible with the decrease in singlet oxygen photosensitization observed on the protonation of beta- carbolines, as seen with other systems, such as norharmane, harmane and harmine (8,57).
It is believed that singlet oxygen is mainly photosensitized by long-lived triplet state of the neutral form of beta-carbolines (8). The triplet yield of the neutral forms of various beta-carbolines- norharmane, harmane and harmine-in methanol are 0.34, 0.55 and 0.55, respectively, but for the cationic forms in the same solvent with hydrochloric acid, the transient triplet-triplet absorption spectra is so weak that the triplet yields of the cationic forms were not determined (57). As S^sub 1^ to T^sub 1^ intersystem crossing is less efficient for beta-carboline cationic forms, the singlet oxygen photosensitization by the cationic beta-carbolines is also less efficient (8,57).
To study the effect of the complexation of betaCMAM with polynucleotides at pH [asymptotically =] 6 and dsDNA and ssDNA at the three pH values on the quantum yield of singlet oxygen production, phi^sub Delta^ of solutions with betaCMAM concentration 10^sup -4^ M and several concentrations of oligonucleotides were measured at pH 6, and dsDNA and ssDNA at three pH values (4, 6 and 9). For all the systems, the phi^sub Delta^ values decrease linearly with the oligonucleotide and DNA concentrations, as shown in Fig. 4 for the DNA systems. The slopes of the linear fittings are given in Table 5.
It seems clear that betaCMAM complexation with oligonucleotides and DNA reduces singlet oxygen efficiency. However, the decrease in singlet oxygen efficiency does not seem to be directly related to the strength of the complex formed.
The decrease in the singlet oxygen emission quantum yield upon the addition of DNA could be due to a decrease in the betaCMAM triplet state formation efficiency due to complexation with DNA and/ or to singlet oxygen reacting with the nucleotides or DNA, leading to oxidative damage. If the singlet oxygen is formed and then reacts with DNA, the initial phosphorescence signal would be the same but the decay would be faster upon the addition of DNA. However, under the experimental conditions studied, the signals were not good enough to check for these two possibilities. So, to evaluate these two pathways of decreasing singlet oxygen, the triplet state betaCMAM absorption spectra were registered under various conditions, and the extent of DNA oxidative damage was evaluated.
Figure 4. Singlet oxygen quantum yield of solutions with several additions of oligonucleotides at pH 6 and dsDNA and ssDNA at three pH values.
Table 5. Slopes of the linear fitting of singlet oxygen quantum yield decrease versus the polynucleotide molar concentration.
Figure 5. Transient triplet-triplet absorption spectra for betaCMAM (1.1 x 10^sup -4^M) at three different pH values (a) and with various concentrations of dsDNA (b).
The transient triplet-triplet absorption spectra of RCMAM at acid, natural and basic pH are shown in Fig. 5a. The cationic form of betaCMAM has an absorption maximum at around 465 nm and the neutral form at around 505 nm. Other beta-carboline derivatives, such as norharmane and harmine in methanol, also show triplet- triplet absorption maxima at around 500 nm (8,57). The absorbance of the band at 505 nm decreases upon the addition of DNA. This suggests that upon complexation with dsDNA at natural pH, the intersystem crossing process is less efficient in betaCMAM, which can also justify the observed decrease in the singlet oxygen quantum yield.
DNA oxidative damage. In addition to the observed decrease in intersystem crossing, a second pathway for the decrease in the observed singlet oxygen quantum yield would be oxidative damage to dsDNA. The strand breaks produced by the reaction of betaCMAM with dsDNA was determined by the TBARs assay, and DNA degradation was followed by electrophoresis.
To look for oxidative damage of dsDNA, electrophoresis of dsDNA samples at natural pH with several concentrations of betaCMAM (dye to DNA ratio up to 1:1) freshly prepared (Fig. 6) and incubated at 25[degrees]C for 12 h under light (Fig. 7) was carried out in 1% (wt/ vol) agarose gel. In both cases, dsDNA treated with Cu(II)-ascorbic acid, to provoke oxidative damage in DNA, was used as reference. As shown in Figs. 6 and 7, DNA treated with this oxidant mixture is the only species that diffuses appreciably. The samples with betaCMAM and DNA alone do not diffuse in the gel, indicating that no oxidative damage is being induced. To observe the diffusion, ethidium bromide was added to samples without betaCMAM.
The oxidative degradation of dsDNA in the presence of betaCMAM was studied by the release of TBA-reactive products, which was followed spectrophotometrically at 532 nm according to the method described in the Materials and Methods section. Only DNA treated with Cu(II)-ascorbic acid mixtures were pink, while the samples with betaCMAM were completely transparent, supporting the idea that no oxidative damage is provoked in DNA upon complexation with betaCMAM.
Figure 6. Agarose gel electrophoresis separation of oxidative damaged DNA and the effect of betaCMAM: DNA alone (1); DNA plus Cu (II)-ascorbic acid (10^sup -3^10^sup -4^ M, respectively) (2); DNA plus betaCMAM 10^sup -7^ M (3); 5 x 10^sup -6^ M (4); 10^sup -5^ M (5); 10^sup -4^ M (6); 3 x 10^sup -4^ M (7); and 6 x 10^sup -4^ M (8). Samples (1) and (2) contain ethidium bromide.
Figure 7. Agarose gel electrophoresis separation of oxidative damaged DNA incubated at 25[degrees]C for 12 h and the effect of betaCMAM: DNA alone (1); DNA plus Cu (II)-ascorbic acid (10^sup -3^- 10^sup -4^ M, respectively) (2); DNA plus betaCMAM 10^sup -7^ M (3); 5 x 10^sup -6^ M (4); 10^sup -5^ M (5); 10^sup -4^ M (6); 3 x 10^sup -4^ M (7); and 6 x 10^sup -4^ M (8). Samples (1) and (2) contain ethidium bromide.
Although oxidation of guanine by singlet oxygen is possible, these latter results and the decrease in the intersystem crossing efficiency of betaCMAM upon complexation with polynucleotides seem to indicate that betaCMAM binding inhibits oxidation damage.
The betaCMAM complexes with oligonucleotides (Poly A, Poly C, Poly G, Poly T and Poly U) and ssDNA and dsDNA form predominantly a 1:1 species. The association constant can be determined by Benesi- Hildebrand method. The association constant with Poly G is similar to that obtained with DNA at acidic or natural pH, indicating that complexation with DNA takes place mainly through the guanine base. Hydrogen bonding between the amide group of betaCMAM and DNA is believed to facilitate association. Intercalation does not seem to be the only mechanism of interaction between this dye and DNA because relatively high association constants are also obtained with ssDNA.
The twist between the carbon 3 substituent plane and the beta- carboline plane upon betaCMAM protonation could justify the higher association constant obtained for dsDNA at pH 6. This steric hindrance is not important in the case of ssDNA, where protonation favors complexation, probably due to hydrogen bonding.
The emission of betaCMAM is quenched upon complexation with oligonucleotides and DNA. The quenching process is mainly static, although a small dynamic component is also detected in the time- resolved experiments. This is in good agreement with the complex formation.
Singlet oxygen quantum yields decrease linearly upon complexation. For dsDNA at natural pH this is shown to be due to the less efficient intersystem crossing process required to generate the betaCMAM triplet state, thus decreasing photosensitization of molecular oxygen.
No strand brake damage of DNA is observed in samples with betaCMAM incubated for 12 h at 25[degrees]C in the presence of light. This suggests that the photosensitized damage to cells caused by this beta-carboline derivative could involve other routes from the singlet oxygen attack on DNA which is relevant for phototherapy action.
Acknowledgements-Financial support from Spanish Junta de Castilla y Leon and European Union (Bu29/04) and Accion Integrada Hispano- Portuguesa (HP2003-0077, Acccao E/405) is gratefully acknowledged. We are indebted to Dr. Pilar Muniz and Dr. Vittorio Paiotta for their valuable help.
1. Pardo, A., D. Reyman, J. M. L. Poyato and F. Medina (1992) Some beta-carboline derivatives as fluorescence standards. J. Lumin. 51, 269-274.
2. Ghiggino, K. P., P. F. Skilton and P. J. Thistlewaite (1985) beta-Carboline as a fluorescence standard. J. Photochem. 31, 113- 121.
3. Chae, K. H. and H. S. Ham (1986) Production of singlet oxygen and superoxide anion radicals by beta-carbolines. Bull. Korean Chem. Soc. 7, 478-479.
4. Calle, P., A. Fernandez-Arizpe and C. Sieiro (1996) Photosensitization by harmine: An ESR spin trapping study on the generation of the superoxide anion radical. Appl. Spectrosc. 50, 1446-1451.
5. Schweitzer, C. and R. Schmidt (2003) Physical mechanisms of generation and deactivation of singlet oxygen. Chem. Rev. 103, 1685- 1757.
6. Clo, E., J. W. Snyder, N. V. Voigt, P. R. Ogilby and K. V. Gothelf (2006) DNA-programmed control of photosensitized singlet oxygen production. J. Am. Chem. Soc. 128, 4200-4201.
7. Yamakoshi, Y., N. Umezawa, A. Ryu, K. Arakane, N. Miyata, Y. Goda, T. Masumizu and T. Nagano (2003) Active oxygen species generated from photoexcited fullerene (C^sub 60^) as potential medicines: O^sup -.^^sub 2^ versus ^sup 1^O^sub 2^. J. Am. Chem. Soc. 125, 12803-12809.
8. Varela, A. P., H. D. Burrows, P. Douglas and M. da Graca Miguel (2001) Triplet state studies of beta-carbolines. J. Photochem. Photobiol. A, Chem. 146, 29-36.
9. Hirakawa, K., S. Kawanishi and T. Hirano (2005) The mechanism of guanine specific photooxidation in the presence of berberine and palmatine: Activation of photosensitized singlet oxygen generation through DNA-binding interaction. Chem. Res. Toxicol. 18, 1545-1552.
10. Morin, B. and J. Cadet (1994) Benzophenone photosensitization of 2'-deoxyguanosine: Characterization of the 2R and 2S diastereoisomers of 1-(2-Deoxy-Beta-D-Erythro-Pentofuranosyl)-2- methoxy-4,5-imidazolidinedione. A model system for the investigation of photosensitized formation of DNA-protein crosslinks. Photochem. Photobiol. 60, 102-109.
11. Marti, C., O. Jurgens, O. Cuenca, M. Casals and S. Nonell (1996) Aromatic ketones as standards for singlet molecular oxygen O2(^sup 1^Deltag) photosensitization. Time-resolved photoacoustic and near-IR emission studies. J. Photochem. Photobiol. A, Chem. 97, 11-18.
12. Kasai, H., Z. Yamaizumi, M. Berger and J. Cadet (1992) Photosensitized formation of 7,8-dihydro-8-oxo-2'-deoxyguanosine (8- hydroxy-2'-deoxyguanosine) in DNA by riboflavin: A nonsinglet oxygen- mediated reaction. J. Am. Chem. Soc. 114, 9692-9694.
13. Cadet, J., C. Decarroz, S. Y. Wang and W. R. Midden (1983) Mechanisms and production of photosensitized degradation of nucleic acids and related model compounds. Isr. J. Chem. 23, 420-429.
14. Tuite, E. M. and J. M. Kelly (1993) Photochemical interactions of methylene blue and analogues with DNA and other biological substrates. J. Photochem. Photobiol. B 21, 103-124.
15. Hudson, J. B. and G. H. N. Towers (1991) Therapeutic potential of plant photosensitizers. Pharmacol. Ther. 49, 181-222.
16. Hickerson, R. P., F. Prat, J. G. Muller, C. S. Foote and C. J. Burrows (1999) Sequence and stacking dependence of 8-oxoguanine oxidation: Comparison of one-electron vs singlet oxygen mechanisms. J. Am. Chem. Soc. 121, 9423-9428.
17. Bohne, C, K. Faulhaber, B. Giese, A. Hafner, A. Hofmann, H. Ihmels, A.-K. Kohler, S. Pera, F. Schneider and M. A. L. Sheepwash (2005) Studies on the mechanism of the photo-induced DNA damage in the presence of acridizinium salts-Involvement of singlet oxygen and an unusual source for hydroxyl radicals. J. Am. Chem. Soc. 127, 76- 85.
18. Duarte, V., D. Gasparutto, L. F. Yamaguchi, J.-L. Ravanat, G. R. Martinez, M. H. G. Medeiros, P. D. Mascio and J. Cadet (2000) Oxaluric acid as the major product of singlet oxygen-mediated oxidation of 8-oxo-7,8-dihydroguanine in DNA. J. Am. Chem. Soc. 122, 12622-12628.
19. Schneider, J. E., S. Price, L. Maidt, J. M. C. Gutteridge and R. A. Floyd (1990) Methylene blue plus light mediates 8-hydroxy 2'- deoxyguanosine formation in DNA preferentially over strand breakage. Nucleic Acids Res. 18, 631-635.
20. Devasagayam, T. P. A., S. Steenken, M. S. W. Obendorf, W. A. Schulz and H. Sies (1991) Formation of 8-hydroxy(deoxy)guanosine and generation of strand breaks at guanine residues in DNA by singlet oxygen. Biochemistry 30, 6283-6289.
21. Dougherty, T. J., C. J. Gomer, B. W. Henderson, D. Kessel, M. Korbelik, J. Moan and Q. Peng (1998) Photodynamic therapy. J. Natl. Cancer Inst. 90, 889-905.
22. Meester, C. (1995) Genotoxic potential of beta-carbolines: A review. Mutat. Res. 339, 139-153.
23. Boeira, J. M., A. F. Viana, J. N. Picada and J. A. P. Henriques (2002) Genotoxic and recombinogenic activities of the two beta-carboline alkaloids harman and harmine in Saccharomyces cerevisiae. Mutat. Res. 500, 39-48. 24. Majer, B. J., F. Kassie, Y. Sasaki, W. Pfau, H. Glatt, W. Meinl, F. Darroudi and S. Knasmuller (2004) Investigation of the genotoxic effects of 2-amino-9H- pyrido[2,3-b]indole in different organs of rodents and in human derived cells. J. Chromatogr. B 802, 167-173.
25. Picada, J. N., K. V. C. L. da Silva, B. Erdtmann, A. T. Henriques and J. A. P. Henriques (1997) Genotoxic effects of structurally related beta-carboline alkaloids. Mutat. Res. 379, 135- 149.
26. Wakabayashi, K., Y. Totsuka, K. Fukutome, A. Oguri, H. Ushiyama and T. Sugimura (1997) Human exposure to mutagenic/ carcinogenic heterocyclic amines and comutagenic beta-carbolines. Mutat. Res. 376, 253-259.
27. Sasaki, Y. F., H. Yamada, K. Shimoi, N. Kinae, I. Tomita, H. Matsumura, T. Ohta and Y. Shirasu (1992) Enhancing effects of heterocyclic amines and beta-carbolines on the induction of chromosome aberrations in cultured mammalian cells. Mutat. Res. 269, 79-95.
28. Guan, H., X. Liu, W. Peng, R. Cao, Y. Ma, H. Chen and A. Xu (2006) beta-Carboline derivatives: Novel photosensitizers that intercalate into DNA to cause direct DNA damage in photodynamic therapy. Biochem. Biophys. Res. Commun. 342, 894-901.
29. Laronze, M., M. Boisbrun, S. Leonce, B. Pfeiffer, O. Lozach, L. Meijer, A. Lansiaux, C. Bailly, J. Sapi and J.-Y. Laronze (2005) Synthesis and anticancer activity of new pyrrolocarbazoles and pyrrolo-beta-carbolines. Bioorg. Med. Chem. 13, 2263-2283.
30. Boursereau, Y. and I. Coldham (2004) Synthesis and biological studies of 1-amino beta-carbolines. Bioorg. Med. Chem. Lett. 14, 5841-5844.
31. Nguyen, C. H., E. Bisagni, F. Lavelle, M.-C. Bissery and C. Huel (1992) Synthesis and antitumor properties of new 4-methyl- substituted-pyrido [4,3-b] indoles (gamma-carbolines). And Cancer Drug Design 7, 219-233.
32. Toshima, K., Y. Okuno, Y. Nakajima and S. Matsumura (2002) beta-Carboline-carbohydrate hybrids: Molecular design, chemical synthesis and evaluation of novel DNA photocleavers. Bioorg. Med. Chem. Lett. 12, 671-673.
33. Nii, H. (2003) Possibility of the involvement of 9H- pyrido[3,4-b]indole (norharman) in carcinogenesis via inhibition of cytochrome P450-related activities and intercalation to DNA. Mutat. Res. 541, 123-136.
34. Cao, R., W. Peng, H. Chen, Y. Ma, X. Liu, X. Hou, H. Guan and A. Xu (2005) DNA binding properties of 9-substituted harmine derivatives. Biochem. Biophys. Res. Commun. 338, 1557-1563.
35. Todorovic, N., T. B. Phuong, P. Langer and K. Weisz (2006) DNA triplex stabilization by a delta-carboline derivative tethered to third strand oligonucleotides. Bioorg. Med. Chem. Lett. 16, 1647- 1650.
36. Duportail, G. (1981) Linear and circular dichroism of harmine and harmaline interacting with DNA. Int. J. Biol. Macromol. 3, 188- 192.
37. Tamura, S., T. Konakahara, H. Komatsu, T. Ozaki, Y. Ohta and H. Takeuchi (1998) Synthesis of beta-carboline derivatives and their interaction with duplex-DNA. Heterocycles 12, 2477-2480.
38. Yang, M., K. Wang, Ch-B. Zang, B.-H. Wang and Y.-M. Zhang (1994) Binding of carboline derivatives to calf thymus DNA- Determination of binding mode and binding strength. J. Chin. Pharm. Sci. 3, 51-58.
39. Tapia, M. J., D. Reyman, M. H. Vinas, A. Arroyo and J. M. L. Poyato (2003) An experimental and theoretical approach to the acid- base and photophysical properties of 3-substituted beta-carbolines in aqueous solutions. J. Photochem. Photobiol. 156, 1-7.
40. Costa, D., H. D. Burrows and M. da Graca Miguel (2005) Changes in hydration of lanthanide ions on binding to DNA in aqueous solution. Langmuir 21, 10492-10496.
41. Rosa, M., R. Dias, M. da Graca Miguel and B. Lindman (2005) DNA-cationic surfactant interactions are different for double- and single-stranded DNA. Biomacromolecules 6, 2164-2171.
42. King, A. M. Q. and B. H. Nicholson (1969) The interaction of aflatoxin B1 with polynucleotides and its effect on ribonucleic acid polymerase. Biochem. J. 114, 679-687.
43. Kononov, A. I. and V. M. Bakulev (1996) Red-shifted fluorescence from polyguanylic acid in aqueous solution at room temperature. J. Photochem. Photobiol. B, Biol. 34, 211-216.
44. Polak, M. and N. V. Hud (2002) Complete disproportionation of duplex poly(dT)-poly(dA) into triplexpoly(dT)-poly(dA)-poly(dT) and poly(dA) by coralyne. Nucleic Acids Res. 30, 983-992.
45. Quinlan, G. J. and J. M. C. Gutteridge (1987) Oxygen radical damage to DNA by rifamycin SV and copper ions. Biochem. Pharmacol. 36, 3629-3633.
46. Dias, A., A. P. Varela, M. da G. Miguel, A. L. Mancanita and R. S. Becker (1992) beta-Carboline photosensitizers. 1. Photophysics, kinetics and excited-state equilibria in organic solvents, and theoretical calculations. J. Phys. Chem. 96, 10290- 10296.
47. Seixas de Melo, J. and P. F. Fernandes (2001) Spectroscopy and photophysics of 4- and 7-hydroxycoumarins and their thione analogs. J. Mol. Struct. 565-566, 69-78.
48. Striker, G., V. Subramaniam, C. A. M. Seidel and A. Volkmer (1999) Photochromicity and fluorescence lifetimes of green fluorescent protein. J. Phys. Chem. B 103, 8612-8617.
49. Marti, C., O. Jurgens, O. Cuenca, M. Casals and S. Novell (1996) Aromatic ketones as standards for singlet molecular oxygen O2(1Delta^sub g^) photosensitization. Time-resolved photoacoustic and near-IR emission studies. J. Photochem. Photobiol. 97, 11-18.
50. Benesi, H. A. and J. H. Hildebrand (1949) A spectrophotometric investigation of the interaction of iodine with aromatic hydrocarbons. J. Am. Chem. Soc. 71, 2703-2707.
51. Balon, M., M. A. Munoz, C. Carmona, P. Guradado and M. Galan (1999) A fluorescence study of the molecular interactions of harmane with the nucleobases, their nucleosides and mononucleotides. Biophys. Chem. 80, 41-52.
52. Abdel-Shafi, A. A. (2001) Effect of beta-cyclodextrin on the excited state proton transfer in 1-naphthol-2-sulfonate. Spectrochim. Acta [A] 57, 1819-1828.
53. Behravan, G., M. Leijon, U. Sechlstedt, B. Norden, H. Vallberg, J. Bergman and A. Graslund (1994) The interaction of ellipticine derivatives with nucleic acids studied by optical and H-NMR spectroscopy: effect of size of the heterocyclic ring system. Biopolymers 34, 599-609.
54. Rock, C., P. A. Shamlou and M. S. Levy (2003) An automated microplate-based method for monitoring DNA strand breaks in plasmids and bacterial artificial chromosomes. Nucleic Acids Res. 31, e65.
55. Feigon, J., A. D. William, L. Werner and D. R. Kearns (1984) Interactions of antitumor drugs with natural DNA: Proton NMR study of binding mode and kinetics. J. Med. Chem. 11, 450-465.
56. Xiao, S., W. Lin, C. Wang and M. Yang (2001) Synthesis and biological evaluation of DNA targeting flexible side-chain substituted beta-carboline derivatives. Bioorg. Med. Chem. Lett. 11, 437-441.
57. Becker, R. S., L. F. V. Ferreira, F. Elisei, I. Machado and L. Latterini (2005) Comprehensive photochemistry and photophysics of land- and marine-based beta-carbolines employing time-resolved emission and flash transient spectroscopy. Photochem. Photobiol. 81, 1195-1204.
Inigo X. Garcia-Zubiri1, Hugh D. Burrows*2, J. Sergio Seixas de Melo2, Joao Pina2, Maria Monteserin1 and Maria J. Tapia*1
1 Departamento de Quimica, Universidad de Burgos, Burgos, Spain
2 Departamento de Quimica, Universidade de Coimbra, Coimbra, Portugal
Received 28 March 2007; accepted 22 May 2007; DOI: 10.1111/ j.1751-1097.2007.00187.x
(c) 2007 The Authors. Journal Compilation. The American Society of Photobiology 0031-8655/07
Copyright American Society for Photobiology Nov/Dec 2007
(c) 2007 Photochemistry and Photobiology. Provided by ProQuest Information and Learning. All rights Reserved.