Type II photooxidation mechanism of biomolecules using chloro (5,10,15,20-Tetraphenylporphyrinato) indium (III) as a photosensitizer

Abstracts

The photooxidation mechanism of bovine serum albumin (BSA), L-tryptophan (Trp) and red blood cells (RBC) by chloro(5,10,15,20-tetraphenylporphyrinato)indium(III) (InTPP) was investigated. The photooxidation rate of Trp, BSA and RBC by InTPP was decreased in the presence of NaN3. The presence of D2O increases the photooxidation rate of Trp and BSA and decreases that of RBC. This decrease is probably related to a reduction of the binding constant between InTPP and RBC in the presence of D2O. No significant change in biomolecule fluorescence or in the percent of hemolysis was observed when radical quenchers (ferricyanide, mannitol and dismutase superoxide) were used. Experiments using electron paramagnetic resonance (EPR) show that only ¹O2 was generated by InTPP. A mechanistic model based on the preferential oxidation of Trp and BSA by singlet oxygen is proposed. The agreement between the experimental data and the kinetic model gives additional support to the predominance of a mechanism via ¹O2 in biomolecule photooxidation by InTPP.

chloro(5,10,15,20-tetraphenylporphyrinato)indium(III); photodynamic therapy; photooxidation mechanism; cancer; erythrocytes

Foi determinado o mecanismo de fotooxidação de albumina de soro bovino (BSA), L-triptofano (Trp) e células vermelhas do sangue (RBC) por cloro(5,10,15,20-tetrafenilporfirinato) de índio(III) (InTPP). A velocidade de fotooxidação de Trp, BSA e RBC por InTPP foi diminuída na presença de NaN3. A presença de D2O aumentou a velocidade de fotooxidação de Trp e BSA e diminuiu a de RBC. Esta diminuição provavelmente está correlacionada com a redução da constante de associação entre InTPP e RBC na presença de D2O. Não foi observada variação significativa na fluorescência das biomoléculas ou sobre a porcentagem de hemólise quando supressores de radicais (ferricianeto, manitol e superóxido dismutase) foram usados. Experimentos usando espectroscopia de ressonância paramagnética (EPR) mostraram que somente o ¹O2 foi gerado por InTPP. Foi proposto um modelo cinético para a fotooxidação de Trp e BSA. A concordância entre os resultados experimentais e este modelo corrobora a predominância do mecanismo via ¹O2 na fotooxidação das biomoléculas pelo InTPP.

ARTICLE

Type II photooxidation mechanism of biomolecules using chloro (5,10,15,20-Tetraphenylporphyrinato) indium (III) as a photosensitizer

André Romero da Silva^I; Joselito Nardy Ribeiro^II; Daniel Rettori^III; Renato Atilio Jorge^I,*

^IDepartment of Physical Chemistry, Institute of Chemistry, University of Campinas, 13084-862 Campinas-SP, Brazil

^IIDepartment of Physiological Sciences, Biomedical Center, Federal University of Espírito Santo, 29042-751 Vitória-ES, Brazil

^IIIFaculty of Pharmacy, Bandeirante University of São Paulo, 02071-013 São Paulo-SP, Brazil

ABSTRACT

The photooxidation mechanism of bovine serum albumin (BSA), L-tryptophan (Trp) and red blood cells (RBC) by chloro(5,10,15,20-tetraphenylporphyrinato)indium(III) (InTPP) was investigated. The photooxidation rate of Trp, BSA and RBC by InTPP was decreased in the presence of NaN₃. The presence of D₂O increases the photooxidation rate of Trp and BSA and decreases that of RBC. This decrease is probably related to a reduction of the binding constant between InTPP and RBC in the presence of D₂O. No significant change in biomolecule fluorescence or in the percent of hemolysis was observed when radical quenchers (ferricyanide, mannitol and dismutase superoxide) were used. Experiments using electron paramagnetic resonance (EPR) show that only ¹O₂ was generated by InTPP. A mechanistic model based on the preferential oxidation of Trp and BSA by singlet oxygen is proposed. The agreement between the experimental data and the kinetic model gives additional support to the predominance of a mechanism via ¹O₂ in biomolecule photooxidation by InTPP.

Keywords: chloro(5,10,15,20-tetraphenylporphyrinato)indium(III), photodynamic therapy, photooxidation mechanism, cancer, erythrocytes

RESUMO

Foi determinado o mecanismo de fotooxidação de albumina de soro bovino (BSA), L-triptofano (Trp) e células vermelhas do sangue (RBC) por cloro(5,10,15,20-tetrafenilporfirinato) de índio(III) (InTPP). A velocidade de fotooxidação de Trp, BSA e RBC por InTPP foi diminuída na presença de NaN₃. A presença de D₂O aumentou a velocidade de fotooxidação de Trp e BSA e diminuiu a de RBC. Esta diminuição provavelmente está correlacionada com a redução da constante de associação entre InTPP e RBC na presença de D₂O. Não foi observada variação significativa na fluorescência das biomoléculas ou sobre a porcentagem de hemólise quando supressores de radicais (ferricianeto, manitol e superóxido dismutase) foram usados. Experimentos usando espectroscopia de ressonância paramagnética (EPR) mostraram que somente o ¹O₂ foi gerado por InTPP. Foi proposto um modelo cinético para a fotooxidação de Trp e BSA. A concordância entre os resultados experimentais e este modelo corrobora a predominância do mecanismo via ¹O₂ na fotooxidação das biomoléculas pelo InTPP.

Introduction

Photodynamic therapy (PDT) is a two-step therapeutic modality in which the topical or systemic delivery of photosensitizing drugs is followed by irradiation with visible light, at doses that are not in themselves harmful.¹ For treatment of cancer the photosensitizer is retained preferentially by the tumor and when excited by irradiation generates reactive oxygen species (ROS) that have a cytotoxic effect to the neoplasm.² These ROS can be generated by two mechanisms, known as type I and type II.^3,4

In type I the photosensitizer in the excited triplet state can interact directly with the substrate and/or solvent, through an electron transfer reaction or hydrogen transfer, generating radical ions or neutral radicals, which quickly react with oxygen molecules producing ROS such as superoxide radical (O₂ ^{• -} ), hydrogen peroxide (H₂O₂) and hydroxyl radical (^•OH), capable of oxidizing a variety of biomolecules. In the type II mechanism, the photosensitizer in the excited triplet state can interact with ground state oxygen molecules (³O₂) generating singlet oxygen (¹O₂) through an energy transfer process.⁵ Studies in the literature suggest that the photooxidative mechanisms via singlet oxygen are often more efficient than radical processes due to the higher diffusibility of ¹O₂ and the higher reaction rate constants with substrates.^6,7 Nonetheless, other studies show that photooxidations occur largely by a type I mechanism.^8,9 Some researchers have proposed that the type I mechanism is favored in polar solvents since the radicals generated through electron transfer would be stabilized in solvents with a high dielectric constant.^10,11 In contrast, the solubility and the lifetime of singlet oxygen are greater in lipophilic solvents and the type II mechanism is favored in hydrophobic systems. Recently, Vakrat-Haglili et al.¹² showed that the dye's environment determines the mechanism of photooxidation, yield, fate and efficacy of the species involved. Junqueira et al.¹³ showed that the negatively charged micelle interface of sodium dodecyl sulfate (SDS) induces dimer formation of methylene blue and shifts a type II to type I mechanism. Other studies also have shown that some radicals such as superoxide and hydroxyl can be generated simultaneously with singlet oxygen by PDT.^14-16

Tetraphenylporphines are compounds of interest in PDT due to their photophysical properties which favor singlet oxygen production.^17-19 Moreover the presence of indium(III) in the core of chloro[13²-(dimethoxycarbonyl)-pheophorbidato methyl ester]indium (III) enhances in vitro and in vivo photosensitizing efficacies due to the heavy atom effect.^20,21 Chloro[13²-(dimethoxycarbonyl)-pheophorbidato methyl ester]indium (III) encapsulated in liposomes of egg yolk phosphatidylcholine recently entered into phase II clinical trials for ocular photodynamic therapy in choroidal neovascular membranes.²² Recent studies carried out in our laboratory have shown that the singlet oxygen quantum yield of chloro(5,10,15,20-tetraphenylporphyrinato)indium(III) (InTPP) (Figure 1) in dimethylsulfoxide (DMSO) (Φ_Δ= 0.72) was higher than that of 5,10,15,20-tetraphenylporphyrin (TPP) (Φ_Δ = 0.52) and that InTPP was an excellent photosensitizer in the photooxidation of tryptophan, bovine albumin and erythrocytes.²³ Even with this high value of Φ_Δ for InTPP, type I photosensitization pathways cannot be ruled out.¹⁴ This work evaluates the main mechanism (type I or II) that acts in the photooxidation of tryptophan, bovine albumin and erythrocytes using InTPP or if both mechanisms proceed simultaneously in the photooxidation of the biomolecules and the cells. For this purpose, the influences of deuterium oxide and singlet oxygen quenchers were studied during the photooxidations. Electron paramagnetic resonance spectroscopy was also used to confirm if ¹O₂ is the main reactive intermediate generated by InTPP for Trp and BSA photooxidation. An explicit mechanistic model for the photooxidation of the biomolecules is proposed based on the elementary steps reported by Rosenkranz et al.,²⁴ considering the experimental results for the main mechanism of photooxidation.

Experimental

Chemicals

Chloro(5,10,15,20-tetraphenylporphyrinato)indium(III) was purchased from Frontier Scientific, Inc. (Logan, UT, USA), with 95% purity. Bovine serum albumin, polyoxyethylene-sorbitan monolaurate (Tween® 20), dimethylformamide (DMF), potassium monobasic and dibasic phosphates, sodium chloride, potassium chloride, deuterium oxide, sodium azide, mannitol, dismutase superoxide, potassium ferricyanide, ethylenediaminetetraacetic acid (EDTA) and 2,2,6,6-tetramethyl-4-piperidone (TEMP) were purchased from Sigma Chemical Company (St. Louis, MO, USA) while Na₂HPO₄•7H₂O and tryptophan was from Merck (Whitehouse Station, NJ, USA). The red blood cells used in the hemolysis assays were donated by the Clinical Hospital of the University of Campinas, according to the Ethical Protocols for Human Experimentation of the Faculty of Medical Sciences of the University of Campinas.

InTPP solutions

Stock solutions of InTPP were prepared in DMF due to the low solubility of the photosensitizer in water. To obtain the desired concentration of InTPP in the biomolecules (Trp and BSA) photooxidation assays, aliquots of the stock solution were added to a photooxidation medium composed of phosphate buffer (pH 7.2, 0.07 mol L^-1 K₂HPO₄ and 0.03 mol L^-1 KH₂PO₄) and Tween® 20 (8.9 mmol L^-1). The final percentage of DMF in the photooxidation medium was maintained at 5%, v/v.

In the photohemolysis assays, aliquots of the InTPP stock solution were added to the photooxidation medium containing the erythrocyte solution (1.89 × 10¹⁰ cells L⁻¹), phosphate-buffered saline (PBS, pH 7.4, 171.1 mmol L^-1 NaCl, 3.4 mmol L^-1 KCl, 12.7 mmol L^-1 Na₂HPO₄, 2.2 mmol L^-1 KH₂PO₄), Tween® 20 (0.45 mmol L⁻¹) and DMF (1.6%, v/v).

Irradiation source

The irradiation system consisted of a mercury lamp (Phillips® HPLN 80W), a water jacket compartment (which absorbs infrared radiation emitted by the lamp), and a 400/600 nm bandpass filter (Oriel BG 38). The cuvette containing the solutions to be irradiated was placed 5 cm from the source lamp. The irradiance emitted by the mercury lamp, between 400 to 600 nm, was measured with a spectroradiometer (LI-1800, LI-Cor). The value measured was 62.7 W m^-2. All experiments were carried out in a dark room to prevent the influence of surrounding radiation.

The value of the irradiance absorbed by the photosensitizer solutions (I_photo,n), expressed as mol photons × m^-2 × s^-1 at each frequency, was calculated by equation 1:

where I_o,ν is the irradiance emitted by the mercury lamp at each frequency, A is the absorbance of InTPP, h is Planck's constant and N_A is Avogadro's constant. The total irradiance absorbed by the photosensitizer (I_abs) was calculated from: I_abs = Σ_i I_photo,ν. To convert I_abs to mol photons × m^-3 × s^-1 the value of I_abs was multiplied by width of the cuvette (10^-2 m).

Fluorescence measurements

Trp and BSA solutions were, respectively, excited at 281 and 279 nm, and fluorescence intensities were monitored at 357 and 326 nm with an ISS® PC1^TM-Photon Counting spectrofluorimeter from ISS (Champaign, IL, USA).

Evaluation of the mechanism involved in the photooxidation experiments

Trp and BSA photooxidation

Basically, the following photochemical reaction was utilized to determine photooxidation rate constants (k_p):

The overall reaction is pseudo first-order with respect to biomolecules because the flux of photons and the concentration of photosensitizer are constant. Considering the concentration of the biomolecules to be proportional to their fluorescence intensity, the following expression is obtained:

where F is the fluorescence intensity of the biomolecules at time t, and F_o is the initial fluorescence before irradiation. Graphs of ln(F/F_o) versus irradiation time allow the determination of k_p.

N₂- or O₂-saturated phosphate buffer solutions (pH 7.2) containing InTPP (20 µmol L^-1), Trp (175 µmol L^-1) or BSA (116 µmol L^-1), plus Tween® 20 (8.9 mmol L^-1) and DMF (5% v/v), were irradiated for periods from 30 to 60 min. The effects of D₂O (10-50% v/v) or NaN₃ (0.01-2.0 mmol L^-1) and the influence of radical quenchers (potassium ferricyanide (50-2800 µmol L^-1), mannitol (20-2800 µmol L^-1) or dismutase superoxide (12.5-100.0 µg mL^-1)) was investigated to analyze the involvement of type II and I mechanisms, respectively, in biomolecule photooxidations. Oxygen-saturated phosphate buffer solutions (pH 7.2) containing InTPP (1.1 or 20 µmol L^-1), Trp (150 m_symbomol L^-1) or BSA (23 or 116 µmol L^-1), plus Tween® 20 (8.9 mmol L^-1), DMF (5%, v/v), and D₂O, NaN₃ or a radical quencher were irradiated for 30-60 min. Fluorescence intensities were measured at different times.

Hemolysis of RBC

Photohemolysis was carried out as described by Silva et al.²³ Typically, human blood of a single donor was collected 48 hours before the hemolysis assays in a tube containing EDTA, which was used as an anticoagulant. The tube was kept in a freezer until the moment of use. The serum was separated from the erythrocytes by centrifugation at 780 × g for 10 min. The RBC were washed with NaCl solution (0.85%, m/v) using three times the blood volume and then the cells were centrifuged at 1760 × g for 10 min to reduce the anticoagulant and serum residues. This procedure was repeated three times. RBC solutions (4 mL, 1.89 × 10¹⁰ cells L⁻¹) containing PBS (pH 7.4), Tween® 20 (0.45 mmol L⁻¹) and DMF (1.6%, v/v) were incubated in the dark with InTPP (6.4 µmol L⁻¹) for 30 min and then irradiated for 60 min using the photooxidation system previously described. InTPP was initially solubilized in DMF and later added to the solution containing PBS, Tween® 20 and DMF. The involvement of type II and I mechanisms in photohemolysis was evaluated, respectively, by effects of D₂O (25-75%, v/v) or NaN₃ (50-150 µmol L^-1) and by the influence of radical quenchers (potassium ferricyanide (50-2800 µmol L^-1), mannitol (20-2800 µmol L^-1) or dismutase superoxide (12.5-100.0 µg mL^-1)). Samples of 550 µL were collected at intervals of 10 min and centrifuged at 500 × g for 5 min. The supernatant was analyzed using a Hewlett Packard 8453A Diode Array Spectrophotometer at a wavelength of 542 nm to measure the oxyhemoglobin chromophore released from the erythrocytes as a result of the destruction of the RBC by the photodynamic action of the photosensitizer. The percentage of hemolysis achieved with the photodynamic action was calculated by equation 4:

where A₁ is the absorbance of the supernatant from the solution that contains the photosensitizer, A₂ is the absorbance of the control (without photosensitizer), and A_T is the absorbance of total hemolysis, which is calculated from the lysis of the erythrocytes with an ultrasound cell disrupter (Ney ULTRAsonik^TM ultrasonic system).

Binding of sensitizers to erythrocytes

The binding constant of sensitizer to erythrocytes in the presence of D₂O was measured as described by Silva et al.²³ Typically, 1.0 mL of RBC at a total concentration ([P]_total) of 1.1 ' 10¹¹ cells L⁻¹ in a solution containing 75% D₂O, PBS (pH 7.4), Tween® 20 (0.45 mmol L⁻¹) and DMF (1.6%, v/v), were incubated in the dark in the presence of photosensitizer concentrations ranging from 1.0 to 8.0 mmol L^-1 for 30 min and then centrifuged at 500 × g for 5 min. The concentration of the free photosensitizer was determined spectrofluorimetrically (λ_excitation = 436 nm and λ_emission = 614 nm), based on a previously constructed calibration curve. Silva et al.²³ showed that the InTPP binding to erythrocytes is cooperative (there are four cooperative binding sites per cell). Then, we calculated the binding constant (K) of InTPP using the equation 5:²⁵

where n = [L]_bound/[P]_total = [sites occupied]/[sites total], n is the number of cooperative sites and [L] is the free concentration of the InTPP. [L]_bound in equation 5 was calculated by: [L]_bound = [L]_total - [L]. The parameters n and K of equation 5 were evaluated using the software Origin® 7.5 (OriginLab Inc, Northampton, MA, USA).

EPR measurements

According to Lion et al.,²⁶ the presence of singlet oxygen - in any system - can be confirmed by the detection of a triplet EPR signal assigned to the stable nitroxide radical 2,2,6,6-tetramethyl-4-piperidone-N-oxyl (TEMPO). TEMPO is generated by the reaction of TEMP with ¹O₂ (Scheme 1).²⁶ Oxygen-saturated solutions of TEMP (38 mmol L^-1) were irradiated for 30 min at room temperature in the presence or absence of InTPP (20 µmol L^-1), in phosphate buffer (pH 7.2) containing Tween® 20 (8.9 mmol L^-1) and DMF (5%, v/v). The irradiation system was the same described earlier. EPR spectra were recorded with an ELEXSYS-CW Bruker spectrometer using a rectangular cavity operating in TE₁₀₂ mode. Aliquots from irradiated solutions were promptly transferred to a flat quartz cell and measurements were carried out at room temperature with the following instrument settings: microwave power, 20 mW; microwave frequency, 9.685 GHz; field modulation frequency, 100 kHz; field modulation amplitude, 1 G; time constant, 21 ms; and scan time, 84 s.

Oxygen concentration measurements

The oxygen concentrations in the N₂- and O₂-saturated phosphate buffer solutions containing InTPP, Trp or BSA, plus Tween® 20 and DMF were measured before the photooxidation experiments using a Clark-type oxygen electrode connected to a computer-operated Oxygraph control unit (Hansatech Instruments, Norfolk, England) at 25 ºC.

Results and Discussion

Determination of the mechanism (type I or II) for biomolecule photooxidation

Photooxidations were carried out under both aerobic and anaerobic conditions to evaluate the involvement of molecular oxygen in the degradation of Trp and BSA. Figures 2a and 2b show a significant reduction of fluorescence intensities (> 70%) for irradiated solutions of Trp and BSA in the presence of O₂. In contrast, photodegradation was negligible under anaerobiosis (reduction < 34%). These results clearly show that oxygen is essential for the photooxidations. The small amount of degraded Trp and BSA, under anaerobic conditions, probably occurred due to the incomplete removal of oxygen from the solutions. Ericson et al.²⁷ reported that oxygen is not completely removed from solutions deoxygenated with N₂. Measurement of the oxygen concentration in the solutions used in these experiments shows that, in anaerobic conditions, the oxygen concentration (139.8 ± 0.8 µmol L^-1) was 5.2 times smaller than the oxygen concentration measured in aerobic conditions (723.7 ± 0.6 µmol L^-1).

Figures 3a and 3b show that increasing the concentration of D₂O to 50%, photooxidation rate constants for Trp and BSA increase 3.9 and 13.9 times, respectively. It is known that the lifetime of singlet oxygen is approximately one order of magnitude higher in D₂O than in H₂O. Thus, these results suggest that the biomolecules are photooxidized by a type II mechanism. Conversely, in the presence of 2.0 and 2.6 mmol L^-1 sodium azide, rate constants for Trp and BSA decrease, respectively, 4.2 and 3.3 times (Figures 4a and 4b). This decrease was higher at low concentrations of sodium azide, and lower at higher concentrations. It is known that azide quenches singlet oxygen by two processes (reversible and irreversible).²⁸ The latter is characterized by generation of superoxide and azide radicals which can oxidize biomolecules. Probably at high concentrations of azide (> 0.2 mmol L^-1 for Trp and > 1.0 mmol L^-1 for BSA photooxidations), the generation of radical species accounted for the slower decrease in the rate constants. Otherwise, when a low concentration of a quencher was used, a non-significant amount of radical species was generated and an abrupt decrease of rate constant was observed due to the quenching of singlet oxygen. These results are similar to data reported in other research and provide strong evidence to support the hypothesis that the photooxidation of Trp and BSA occurs by a type II mechanism.^10,29,30 However, these trials do not allow evaluation of the participation of other active oxygen intermediates during the photooxidation. Many studies have shown that radicals (e.g., superoxide and hydroxyl) can be generated by PDT simultaneously with the singlet oxygen.^14-16 Thus, we decided to evaluate the influence of known ROS scavengers on the InTPP-sensitized photooxidation of Trp and BSA. Photooxidations were carried out in the presence either of potassium ferricyanide (electron-scavenger), mannitol (^•OH quencher) or dismutase superoxide (O₂^{• -} suppressor). No alterations in the decay of the biomolecule fluorescence intensities were observed (results not shown). These results show that Trp and BSA are not oxidized by electron transfer, or by hydroxyl or superoxide radicals.

EPR experiments using TEMP as a singlet oxygen trapping agent were carried out to confirm if ¹O₂ is the main reactive intermediate generated by InTPP for biomolecules photooxidation.^26,31 The reaction of TEMP with singlet oxygen produces a stable nitroxide radical (TEMPO) (Scheme 1) readily detectable by EPR. Irradiation of an oxygen-saturated phosphate buffer (pH 7.2) solution containing InTPP (20 µmol L^-1) and TEMP (38 mmol L^-1) led to the generation of a triplet EPR signal (Figure 5a) characteristic of a nitroxide radical (TEMPO in this case).^16,26 The intensity of the signal increased rapidly during irradiation of the solution containing InTPP (Figure 5a). In the absence of the photosensitizer, no increase of the signal was observed (Figure 5b). A small triplet signal was present before the irradiation of TEMP due to impurities of TEMPO. In conclusion, these results show that the photooxidation of TEMP proceeds via a type II mechanism, confirming that Trp and BSA are also oxidized by a type II process.

Determination of the mechanism (type I or II) for the RBC photooxidation

Figure 6a shows the results of photohemolysis experiments carried out in phosphate-buffered saline with increasing D₂O concentrations. There was a marked decrease in the rate of photohemolysis when the concentration of D₂O is increased from 25 to 75%, v/v. For example, the irradiation time to inactivate 50% of the cell population increased from 29 to 34 min when the D₂O concentration increased from 0 to 25%, and to 44 and 51 min when 50 and 75% of D₂O were used. The same situation was observed by De Polis et al.³² If the photohemolysis proceeds by a singlet-oxygen mechanism, it would be expected to be faster in D₂O than in H₂O because the lifetime of singlet oxygen is greater in the former solvent.³³ However, the rate decrease observed should not be taken as an argument against a singlet-oxygen mechanism. As discussed by Valenzeno et al.³⁴ the D₂O effects on photohemolysis using other sensitizers are somewhat variable, ranging from no effect to an enhancement of over 200%. We have shown recently that the binding constant between RBC and InTPP is (2.40 ± 0.05) × 10⁷ L mol^-1, in aqueous solutions.²³ Therefore, InTPP has a high affinity for the erythrocyte membrane due to its lipophilicity.²³ It is thus expected that the InTPP molecules enter into the hydrophobic region of the lipid bilayer together with oxygen molecules where the singlet oxygen will be generated. However, the presence of D₂O in the photooxidation medium decreased the InTPP affinity for the erythrocytes membrane. The binding constant between RBC and InTPP was decreased to (0.92 ± 0.06) × 10⁷ L mol^-1 when 75% D₂O was used. Rosen and Klebanoff reported that D₂O may influence photochemical reactions by exchanging a deuterium for hydrogen in a C-H or an O-H bond that is involved in the reaction.³⁵ The D₂O may alter the observed reaction rate if a C-D or an O-D bond is involved in the rate-limiting reaction step. Besides, the exchange of deuterium for hydrogen at the binding sites can reduce the affinity of InTPP to bind to RBC. This smaller affinity probably contributed to decrease the photohemolysis in the D₂O solvent. In this situation, the increase of the D₂O concentration in the photohemolysis medium decreases the rate of the hemolysis and increases the time needed to inactive the RBC.

The effect of sodium azide as a singlet oxygen quencher was also investigated in the photohemolysis. Figure 6b shows that the presence of NaN₃ in the photohemolysis medium affected the ability of InTPP to inactive the RBC. For example, after 60 min of irradiation the average hemolysis percentage of erythrocytes was reduced from 83 to 66% when the azide concentration was increased from 0 to 50 µmol L^-1, and to 55 and 33% for 100 and 150 µmol L^-1 of azide. These results suggest that the RBC are photooxidized by a type II mechanism.

Erythrocyte photohemolysis was also carried out in the presence of radical quenchers (potassium ferricyanide, mannitol or dismutase superoxide). No alterations in the hemolysis percentage were observed (results not shown). These results show that erythrocytes are not oxidized by electron transfer, or by hydroxyl or superoxide radicals.

Comparing the photooxidation of BSA and RBC by InTPP

The rate constant for the pseudo first-order photooxidation of the BSA using 6.4 mmol L^-1 of InTPP is k_p = (2.0 ± 0.15) x 10^-4 s^-1. This result was obtained from a graph of ln(F/F_o) versus irradiation time (graph not shown). The half-life of BSA calculated from k_p is (57.8 ± 3.5) min. The time to inactivate 50% of the cell population of RBC, obtained with 0% D₂O (Figure 6a), is (29.2 ± 1.9) min. The association constant for BSA is (1.15 ± 0.07) ' 10⁵ L mol^-1 and the binding sites are independent.²³ For erythrocytes, the association constant is (2.40 ± 0.05) ' 10⁷ L mol⁻¹ and there are four cooperative binding sites per cell.²³ These results suggest that the higher photodynamic activity of InTPP with RBC may be associated with a higher affinity of InTPP for RBC than for BSA.

Kinetic model

As the photooxidation of Trp and BSA occurred via a singlet oxygen-mediated mechanism, we propose a mechanistic model consisting of ten elementary reactions (Table 1) for biomolecule photooxidation mediated by InTPP. The kinetic steps of this mechanism are similar to those proposed by Rosenkranz et al.²⁴ for the photodynamic inactivation of lysozyme.

P_o, ¹P, and ³P, denote respectively the ground, excited singlet, and excited triplet states of the dye. O₂, and ¹O₂, are the ground (triplet) and the excited singlet states of molecular oxygen. S is the substrate (biomolecule) and S-O₂ the product of photooxidation. In their work, Rosenkranz et al.²⁴ showed that the enzymatic activity of solutions of lysozyme containing acridine orange decreased by an overall pseudo first-order process during irradiation with visible light. Applying quasi-stationary conditions for [¹O₂] and [³P], the rate constant (k_p) for the process is given by the following expression:²⁴

where k_p is a constant when the sum k₇ + k₈[S] + k₉[S] is approximately independent of time.²⁴ This is always the case when k₇ >> k₈[S] + k₉[S]. &oslash _T is the triplet quantum yield of the photosensitizer and it represents the relation between the constants k₃ and k₂,: &oslash _T = k₃/(k₃+k₂).¹⁹ Kinetic simulations of the InTPP-sensitized photooxidation of Trp were run with the software Gepasi 3.21.^38,39 Theoretical k_p values calculated by software Gepasi were obtained from plots of ln[Trp]/[Trp]_oversus irradiation time. [Trp] and [Trp]_o are, respectively, the concentrations of tryptophan during at different time points and that before irradiation. The Trp concentrations at different times were provided by the software. We also considered that quenching of the excited singlet state (k_s) of the photosensitizer occurred by radiative (fluorescence decay 'f') and by non-radiative processes (internal conversion 'ic' and intersystem-crossing 'isc'). The rate constant for step (2), k₂, was calculated using the experimental values of k_f, k_isc and &oslash _f (fluorescence quantum yield) for InTPP.³⁶ The following equations were used: k₂ = k_s – k_isc, k₃ = k_isc and k_s = k_f/&oslash _f.¹⁹ The constant k₄ was obtained from our own measurement of the triplet lifetime.²³

We assumed that singlet oxygen was formed exclusively by energy transfer from the triplet state of the photosensitizer to ground state molecular oxygen. The proportion of ³P quenched by O₂ is approximately equal to one because k₅ + k₆ >> k₁₀. Quenching of ¹P by O₂ was not considered because of the very short lifetime of the first excited singlet state of InTPP.³⁶ Then, k₅ was calculated using the following equations:⁷

where &oslash _Δ is the quantum yield of singlet oxygen. According to the literature, values for k₆ are found in the range 1-3 × 10⁹ L mol^-1 s^-1, therefore, we assumed k₆ = 1 × 10⁹ L mol^-1 s^-1.⁵ Singlet oxygen quantum yields (&oslash _Δ= 0.72) were obtained by laser flash photolysis experiments using pheophorbide-a as a reference.²³ An oxygraph apparatus was used to determine the concentration of oxygen ([O₂] = 723.7 ± 0.6 µmol L^-1) as described earlier. The remaining constants in Table 1 were obtained from the literature.^24,37

Comparing results with the kinetic model

Figure 7a shows a linear relation between ln(F/F_o) and the irradiation time for Trp in the presence of different concentrations of InTPP, which indicates that the sum k₇ + k₈[S] + k₉[S] is approximately independent of time, probably due to the fact that k₇ >> k₈[S] + k₉[S]. The substitution of I_abs for k₁[P_o] in equation 1 provides a modified expression for k_p:

Therefore, a linear relation between k_p and [P_o] must be maintained if photooxidations occur by the mechanism suggested in Table 1.

For the photooxidation of Trp, experimental k_p values increase linearly with [InTPP] (Figure 7b). The same results were observed for the photooxidation of BSA (data not shown). Theoretical values of k_p (for Trp) were determined by kinetic simulations (using the software Gepasi 3.21) and plotted for comparison with the experimental data (Figure 7b). The theoretical values are in good agreement with experimental data, validating the photooxidative mechanistic model consisting of ten elementary reactions (Table 1, where ¹O₂, in step (8), is the only species responsible for the consumption of Trp). Therefore, kinetic modeling also confirms that InTPP photosensitized oxidation of Trp occurs in accordance with a type II process and Trp or BSA photooxidation occurs by the same elementary steps proposed by Rosenkranz et al.²⁴ for the photodynamic inactivation of lysozyme. To our knowledge, this is the first work showing that the photooxidation of biomolecules by photosensitizer may occur by the same elementary steps as for the inactivation of lysozyme.

Conclusions

In this work, we showed that a type II photodynamic mechanism explains the InTPP-mediated photooxidation of Trp and BSA, because the photooxidation rate constants (k_p) increased in the presence of deuterium oxide and decreased in the presence of sodium azide. The same mechanism also explains red blood cells photooxidation since the hemolysis percentage was reduced in the presence of sodium azide. Evidence in favor of a type II mechanism was also found when potassium ferricyanide, mannitol or dismutase superoxide were used as radical suppressors, because no variations on k_p values and percent hemolysis were observed. EPR studies corroborated the proposal that singlet oxygen was the principal oxidant generated by InTPP during the photochemical reactions, due to the detection of a triplet EPR signal assigned to TEMPO. Finally, kinetic simulations for the photooxidation of Trp were performed taking into account a series of 10 proposed elementary reactions (Table 1). Theoretical k_p values obtained from the simulations were in good agreement with experimental data, confirming, once more, that photooxidations mediated by InTPP occur by a type II mechanism.

Acknowledgments

We thank FAPESP and CNPq for financial support, and Professors P. J. G. Pagliuso and C. Rettori from the Institute of Physics Gleb Wataghin, University of Campinas for the EPR measurements.

References

1. Calzavara-Pinton, P. G.; Venturini, M.; Sala, R.; J. Eur. Acad. Dermatol. 2007, 21, 293.

2. Juzeniene, A.; Nielsen, K. P.; Moan, J.; J. Environ. Pathol. Toxicol. Oncol. 2006, 25, 7.

3. Gudgin Dickson, E. F.; Goyan, R. L.; Pottier, R. H.; Cell. Mol. Biol. 2003, 48, 939.

4. Dolmans, D. E. J. G. J.; Fukumura, D.; Jain, R. K.; Nat. Rev. Cancer 2003, 3, 380.

5. Foote, C. S. In Free Radicals in Biology; Pryor, W. A., ed.; Academic Press: New York, 1976, vol. 2, ch. 3.

6. Jori, G.; In Lasers in Photomedicine and Photobiology; Pratesi, R.; Sacchi, C. A., eds.; Springer-Verlag: Berlin, 1980, vol. 22.

7. Wilkinson, F.; Brummer, J. G.; J. Phys. Chem. Ref. Data 1981, 10, 809.

8. Rossi, E.; van de Vorst, A.; Jori, G.; Photochem. Photobiol. 1981, 34, 447.

9. Skalkos, D.; Hampton, J. A.; Keck, R. W.; Wagoner, M.; Selman, S. H.; Photochem. Photobiol. 1994, 59, 175.

10. MacDonald, I. J.; Dougherty, T. J.; J. Porphyrins Phthalocyanines 2001, 5, 105.

11. Ochsner, M.; J. Photochem. Photobiol.,B 1997, 39, 1.

12. Vakrat-Haglili, Y.; Weiner, L.; Brumfeld, V.; Brandis, A.; Salomon, Y.; Mcllroy, B.; Wilson, B. C.; Pawlak, A.; Rozanowska, M.; Sarna, T.; Scherz, A.; J. Am. Chem. Soc. 2005, 127, 6487.

13. Junqueira, H. C.; Severino, D.; Dias L. G.; Gugliotti, M. S.; Baptista, M. S.; Phys .Chem. Chem. Phys. 2002, 4, 2320.

14. Martins, J.; Almeida, L.; Laranjinha, J.; Photochem. Photobiol. 2004, 80, 267.

15. Xu, S.; Zhang, X.; Chen, S.; Zhang, M.; Shen, T.; Photochem. Photobiol. Sci. 2003, 2, 871.

16. Xu, S.; Chen, S.; Zhang, M.; Shen, T.; J. Photochem. Photobiol., B 2003, 72, 61.

17. Nyman, E. S.; Hynninen, P. H.; J. Photochem. Photobiol., B 2004, 73, 1.

18. De Rosa, M. C.; Crutchley, R. J.; Chem. Rev. 2002, 233-234, 351.

19. Wilkinson, F.; Helman, W. P.; Ross, A. B.; J. Phys. Chem. Ref. Data 1993, 22, 113.

20. Chen, Y.; Zheng, X.; Dobhal, M. P.; Gryshuk, A.; Morgan, J.; Dougherty, T. J.; Oseroff, A.; Pandey, R. K.; J. Med. Chem. 2005, 48, 3692.

21. Rosenfeld, A.; Morgan, J.; Goswami, L. N.; Ohulchanskyy, T.; Zheng, X.; Prasad, P. N.; Oseroff, A.; Pandey, R. K.; Photochem. Photobiol. 2006, 82, 626.

22. Ciulla, T. A.; Criswell, M. H.; Danis, R. P.; Snyder, W. J.; Small, W.; Retin.-J. Retin. Vitr. Dis. 2004, 24, 521.

23. Silva, A. R.; Pelegrino, A. C.; Tedesco, A. C.; Jorge, R. A.; J. Braz. Chem. Soc. 2008, 19, 491.

24. Rosenkranz, P.; Al-Ibrahim, A.; Schmidt, H.; In Singlet Oxygen Reactions with Organic Compounds and Polymers; Ranby, B.; Rabek, J. F., eds.; New York, 1978, ch. 19.

25. Holde, V.; Kensal, D.; Physical Biochemistry; Prentice-Hall, Inc, Englewood Cliffs: New Jersey, 1971, ch. 3.

26. Lion, Y.; Delmelle, M.; van de Vorst, A.; Nature 1976, 263, 442.

27. Ericson, M. B.; Grapengiesser, S.; Gudmundson, F.; Wennberg, A. M.; Larko, O.; Moan, J.; Rosen, A.; Lasers Med. Sci. 2003, 18, 56.

28. Foote, C. S.; In Singlet Oxygen in Organic Chemistry; Wasserman, H. H.; Murray, R. W., eds.; Academic Sciences: New York, 1979, vol. 40, ch. 5; Gollnick, K.; Kuhn, H. J.; In Singlet Oxygen in Organic Chemistry; Wasserman, H. H.; Murray, R. W., eds.; Academic Sciences: New York, 1979, vol. 40, ch. 8.

29. Inoue, K.; Matsuura, T.; Saito, T.; Bull. Chem. Soc. Jpn. 1982, 55, 2959.

30. Spikes, J. D.; van Lier, J. E.; Bommer, J. C.; J. Photochem. Photobiol., A 1995, 91, 193.

31. Konovalova, T. A.; Lawrence, J.; Kispert, L. D.; J. Photochem. Photobiol., A 2004, 162, 1.

32. De Polis, A.; Chandra, S.; Charalambides, A. A.; Bonnett, R.; Magnus, I. A.; Biochem. J. 1985, 226, 757.

33. Merkel, P. B.; Kearns, D. R.; J. Am. Chem. Soc. 1972, 94, 7244.

34. Valenzeno, D. P.; Trudgen, J.; Hutzenbuhler, A.; Milne, M.; Photochem. Photobiol. 1987, 46, 985.

35. Rosen, H; Klebanoff, S. J.; J. Biol. Chem. 1977, 252, 4803.

36. Ohno, O.; Kaizu, Y.; Kobayashi, H.; J. Chem. Phys. 1985, 82, 1779.

37. Bellus, D.; In Singlet Oxygen Reactions with Organic Compounds and Polymers; Ranby, B.; Rabek, J. F., eds.; New York, 1978, ch. 9.

38. Mendes, P.; Comput. Appl. Biosci. 1993, 9, 563.

39. Mendes, P.; Trends Biochem. Sci. 1997, 22, 361.

Received: February 8, 2008

Web Release Date: August 11, 2008

FAPESP helped in meeting the publication costs of this article.

Publication Dates

• Publication in this collection
  03 Oct 2008
• Date of issue
  2008

History

• Accepted
  11 Aug 2008
• Received
  08 Feb 2008