172
Chem. Res. Toxicol. 1998, 11, 172-177
Drug-Photosensitized Protein Modification: Identification of the Reactive Sites and Elucidation of the Reaction Mechanisms with Tiaprofenic Acid/Albumin as Model System† Miguel A. Miranda,*,‡ Jose´ V. Castell,§ Daniel Herna´ndez,§ Marı´a J. Go´mez-Lecho´n,§ Francisco Bosca,‡ Isabel M. Morera,‡ and Zaideth Sarabia‡ Departamento de Quı´mica/Instituto de Tecnologı´a Quı´mica UPV-CSIC, Universidad Polite´ cnica de Valencia, Avda de los Naranjos s/n, E-46022 Valencia, Spain, and Centro de Investigacio´ n, Hospital Universitario La Fe (SVS), Avda de Campanar 21, E-46009 Valencia, Spain Received May 12, 1997
Certain drugs can photosensitive the formation of protein modifications, which are thought to be responsible for the occurrence of photoallergy. In the present work, the UV irradiation of serum albumin in the presence of tiaprofenic acid has been studied as a model system for drug-photosensitized protein modifications. The photolysates evidenced that His, Tyr, and Trp are the reactive sites of the protein. The experimental results strongly suggest that formal hydrogen abstraction from the OH or NH groups of Tyr or Trp by the excited drug is the key photochemical process. Competition between cage escape and in cage recombination of the resulting radical pairs governs the final outcome: protein photo-cross-linking versus drugprotein adduct formation. These findings are highly relevant to understand the process of photohapten formation, the first event in the onset of photoallergy.
Introduction A number of biomolecules (e.g., proteins, nucleic acids, unsaturated lipids, and other cell constituents) are known to undergo extensive changes upon irradiation in the presence of photosensitizers. In the case of proteins, their photosensitized modifications can lead to loss of biological function (inactivation of enzymes, hormones, etc.) (1, 2). Certain drugs can behave as photosensitizers producing protein modifications which are thought to be responsible for the occurrence of photoallergy and other undesired light-induced side effects (3-5). Obviously, a precise knowledge of the active sites and reaction mechanisms involved would definitely contribute to understanding the photosensitizing potential of new drug candidates. Unfortunately, the photosensitized reactions of proteins turn out to be very complex, which usually frustrates attempts to establish their course. In the present work, we have examined the chemical changes taking place upon irradiation of a model protein (purified bovine serum albumin) in the presence of a representative photosensitizing drug (tiaprofenic acid). Selection of this drug for the study was based on the fact that it appears to be associated with a high incidence of both photoallergic and phototoxic reactions (6).
Experimental Procedures Chemicals. Tiaprofenic acid (TPA)1 was extracted from TORPAS, produced by Rousell Laboratories (Caracas, Venezu† Dedicated to Professor Waldemar Adam on the occasion of his 60th birthday. ‡ Universidad Polite ´ cnica de Valencia. § Hospital Universitario La Fe.
ela). Decarboxytiaprofenic acid (DTPA) was prepared as previously described in detail (7, 8); tritium oxide (3H2O), 185 GBq/ mL (specific radioactivity 3.33 GBq/mmol, 90 mCi/mmol), was from ICN (Irvine, CA). The procedure for radiolabeling, based on alkaline exchange of the hydrogen atoms R to the carboxy group in 3H2O, was previously described (9). The radioactivity was adjusted to obtain final working stock solutions containing ca. 1 cpm/pmol. Butylated hydroxyanisole (BHA), p-cresol (PC), p-(trifluoromethyl)phenol, indole, histidine (His), tyrosine (Tyr), and the methyl esters of His, Tyr, and Trp were purchased from Aldrich (Steinheim, Germany). Bovine serum albumin (BSA) and tryptophan (Trp) were from Sigma Chemical Co. (St. Louis, MO). Sodium dodecyl sulfate (SDS) and dithiothreitol (DTT) were from Biochemical (Mannheim, Germany). Coomassie blue and 2-mercaptoethanol were from Merck (Darmstadt, Germany). Instrumentation. The 1H NMR spectra were taken with a Varian Gemini 300 instrument. Ultraviolet spectra were recorded by means of a Shimadzu UV/vis spectrophotometer (UV- 1 60A). GC/MS spectra were obtained on a HewlettPackard 5988A spectrometer. Fluorescence measurements were made on a Perkin-Elmer MPF-43A spectrofluorimeter. The laser flash photolysis system was as follows: A pulsed Nd:YAG SL404G-10 Spectrum Laser Systems instrument was used for excitation at 355 nm. The single pulses were ca. 10-ns duration, and the energy was ca. 10 mJ/pulse. A Lo255 Oriel Xenon lamp was employed as the detecting light source. The laser flash photolysis apparatus consisted of the pulsed laser, the Xe lamp, a 77200 Oriel monochromator, and an Oriel photomultiplier (PMT) system made up of a 77348 side-on PMT tube, 70680 PMT housing, and a 70705 PMT power supply. The oscilloscope was a TDS-640A Tektronix. The ouput signal from the oscilloscope was transferred to a personal computer for study. 1 Abbreviations: bovine serum albumin (BSA), tiaprofenic acid (TPA), phosphate-buffered saline (0.2 M, pH 7.2) (PBS), 2-benzoyl-5ethylthiophene (DTPA), butylated hydroxyanisole (BHA), p-cresol (PC), electron donors (ED).
S0893-228x(97)00082-9 CCC: $15.00 © 1998 American Chemical Society Published on Web 02/14/1998
Drug-Photosensitized Protein Modification Photosensitized Modification of BSA. Solutions of BSA [1 mg/mL in phosphate-buffered saline (PBS) 0.2 M, pH 7.2] and TPA or DTPA (0.3 mM) were irradiated for 3 h, using a medium pressure mercury lamp (OSRAM HQL, 125 W) in an N2 or O2 atmosphere. Samples received 1.7 × 10-4 J s-1 cm-2 during the 3-h exposure, as determined by ferrioxalate actinometry (10). Under these conditions, the absorbance of TPA at its longest wavelength band (314 nm) was very high (>2.5). The resulting photolysates were hydrolyzed with 6 N HCl for 18-20 h at 110 °C. After the exceeding acid was eliminated in vacuo, the residues were analyzed by ion-exchange chromatography and the eluates were allowed to react with ninhydrin, which forms deep-blue-colored adducts with absorption maxima at 570 nm (amino acid analyzer). The lability of Trp under acid conditions prevented its determination in protein hydrolysates. The resulting samples were directly analyzed by monitoring the characteristic Trp fluorescence properties (11). BSA irradiated in the absence of drugs was submitted to the same treatment and served as control. Photosensitized Modification of Isolated Amino Acids. For these experiments, the irradiation system was the same as that used for the photosensitized modification of BSA. (a) Histidine. Solutions of His (0.3 mM) and TPA or DTPA (0.3 mM) in PBS were irradiated for 1 h under O2 or N2 atmosphere. The resulting samples were treated as previously described (12) and then measured in a UV/vis spectrophotometer at 530 nm, after conversion of the unreacted amino acid to a colored azo derivative (modified Pauli reaction). (b) Tyrosine. Solutions of Tyr (1 mM) and TPA or DTPA (0.3 mM) in PBS were irradiated for 3 h under O2 or N2 atmosphere, in the presence or absence of BHA (1 µM). The resulting samples were treated as previously described (13) and then measured in a UV/vis spectrophotometer at 605 nm. In this case, the remaining amino acid was determined through quantification of the phenolic residues present in the photolysate. (c) Tryptophan. Solutions of Trp (1 mM) and TPA or DTPA (0.3 mM) in PBS were irradiated for 1 h under O2 or N2 atmosphere. The resulting samples were analyzed by following the decrease in Trp fluorescence, as stated above for the case of whole proteins. Photochemistry of DTPA in the Presence of p-Cresol or Indole. DTPA (250 µmol) and p-cresol or indole (500 µmol) were dissolved in methanol (25 mL) and placed in a Pyrex irradiation tube. The samples were irradiated at room temperature for 1 h with the light from a medium pressure mercury lamp as described above. After irradiation, the solvent was evaporated at reduced pressure, and the residues were analyzed by 1H NMR spectroscopy and GC/MS. Time-Resolved Quenching Studies. Triplet-state quenching by electron donors (ED) was performed with TPA and DTPA methanolic solutions (absorbance ) 0.2 at 355 nm) degassed by nitrogen purging. The expression Kq ) (K2 - K1)/([ED2] [ED1]) was used to obtain the rate constants for quenching of the triplet states, based on the facts that plots of the rate constants for decay of triplet excited state (Ki) vs the electron donor concentration [EDi] must be linear. The compounds used as electron donors were BHA, p-cresol, indole, and the methyl esters of His, Tyr, and Trp. Their concentration range was from 5 × 10-5 to 5 × 10-3 M. The Ki values were measured at 420 nm. Protein Photo-Cross-Linking and Drug-Protein Photobinding. Solutions of BSA (1 mg/mL) and TPA (eventually, [3H]TPA) or DTPA (0.3 mM) in PBS were irradiated for 3 h under O2 or N2 atmosphere as described above. Samples were heated for 5 min at 95 °C with 1% SDS and 1.9% DTT. Discontinuous polyacrylamide gel electrophoresis (PAGE) was performed as previously described (14), at pH 8.7, under denaturating conditions (1% SDS, 0.1% mercaptoethanol) on 5-20% gradient gel. The gels were stained with Coomassie blue. The protein bands were excised, and the radioactivity was measured in a liquid scintillation counter.
Chem. Res. Toxicol., Vol. 11, No. 3, 1998 173
Figure 1. Amino acid analysis of BSA after acid hydrolysis: (A) protein irradiated under aerobic conditions without drug, (B) mixtures of BSA/TPA irradiated under aerobic conditions, (C) mixtures of BSA/TPA irradiated under nitrogen.
Results and Discussion Modification of Amino Acids in BSA Photosensitized by TPA and DTPA. Solutions containing BSA and TPA in PBS were irradiated for 3 h with the Pyrexfiltered light of a medium pressure mercury lamp. Acid hydrolysis followed by automatized amino acid analysis evidenced that His and Tyr underwent a dramatic decrease (ca. 90%) with respect to a nonirradiated control (Figure 1). No effect was observed when BSA was irradiated under the same conditions. When N2 was bubbled through the solution prior to and during irradiation, such decrease (Tyr, His) was much less pronounced (ca. 10%). The lability of Trp under acidic conditions prevented its determination in protein hydrolysates. Hence, it was directly analyzed by monitoring the characteristic Trp fluorescence properties (11). As in the case of His and Tyr, a sharp decrease (ca. 90%) of Trp (measured by the emission at ca. 350 nm) was observed after irradiation under aerobic conditions. Again, pho-
174 Chem. Res. Toxicol., Vol. 11, No. 3, 1998
Figure 2. Photodegradation of His, Tyr, and Trp sensitized by TPA and its photoproduct DTPA under aerobic and anaerobic conditions.
todegradation of Trp was significantly reduced (ca. 50%) in the absence of oxygen. TPA is photolabile and undergoes extensive photodecarboxylation, to give DTPA as the major photoproduct (7). To ascertain whether the photoproduct could also participate in the observed photooxidation of amino acid residues, irradiation of BSA/DTPA mixtures was carried out under the above-mentioned conditions. Subsequent amino acid analysis again showed an almost complete disappearance of His, Tyr, and Trp in the presence of oxygen (data not shown). Drug-Photosensitized Photodegradation of His, Tyr, and Trp. Once established that these three amino acid residues are the key sites of drug-photosensitized protein damage, the next step was to perform the irradiation of TPA and DTPA in the presence of the pure amino acids. The results (Figure 2) were in excellent agreement with the observations on whole proteins. Under oxygen, His underwent efficient photodegradation (ca. 50%) with TPA and DTPA. This was much less so when the solution was purged with nitrogen prior to and during irradiation. When Tyr was used as a probe for the drug-photosensitized reaction, a similar effect was observed. In this case, the remaining amino acid was previously determined through quantification of the phenolic residues present in the photolysate (13). Under aerobic conditions, a marked Tyr consumption (ca. 60%) was detected after exposure to the combined action of TPA or DTPA and light. Likewise, aerobic irradiation of Trp led to almost complete consumption of the amino acid by both photosensitizers as determined by fluorescence at 350 nm. As expected, this occurred to a lesser extent under nitrogen. Photodegradation of His should be expected if TPA were acting as a singlet oxygen sensitizer (type II mechanism) (15, 16). We have been able to detect this active oxygen species by means of time-resolved nearinfrared emission and to determine its quantum yield in aqueous solutions (0.22). In the case of Tyr and Trp, besides the type II oxidation route, a further pathway involving free radicals (type I mechanism) could be operating (17, 18). Indeed, addition of the highly efficient radical scavenger BHA inhibited by 60% the TPA-
Miranda et al.
photosensitized oxidation of Tyr. The same was essentially true for Trp, where addition of BHA produced a significant inhibition of amino acid oxidation. Use of p-Cresol and Indole as Models for Tyr and Trp. In order to elucidate the mechanism of drugphotosensitized degradation of amino acids, model studies were undertaken. For this purpose, p-cresol (PC) and indole were chosen as simple analogues of Tyr and Trp (19, 20), while DTPA was used instead of TPA for practical reasons (higher solubility in organic solvents and easier purification of its possible photoproducts by chromatographic methods). Upon irradiation of DTPA/ PC under oxygen, both compounds underwent photodegradation. A complex mixture resulted, from which both diastereomers of the two possible DTPA hydrodimers, as well as C/C and C/O PC dimers arising from oxidative coupling, were identified (Scheme 1). DTPA hydrodimers were also obtained upon irradiation in the presence of indole, but in this case indole dimers could not be detected. Time-Resolved Studies. From the above series of experiments, only preliminary conclusions concerning the reaction mechanism may be inferred. For instance, analysis of the product distribution cannot allow a clear distinction between type I and type II reaction pathways (21). In principle, it should be possible to obtain more reliable direct information on the mechanistic aspects from kinetic data. For these reasons, flash photolysis studies were carried out on TPA and DTPA. In methanol, under anaerobic conditions, transient species with λmax ca. 600 nm were formed. Their spectra (not shown) were similar to those reported for TPA and DTPA in PBS (22). First-order kinetics was observed for the decays of 3TPA and 3DTPA in methanol, under anaerobic conditions, with lifetimes of ca. 3 µs in both cases. The rate constants of TPA and DTPA triplet quenching by oxygen, PC, indole, BHA, and the methyl esters of His, Tyr, and Trp were determined in the same solvent. The results are given in Table 1. In the case of TPA, the highest values (ca. 2 × 109 M-1 -1 s ) were obtained for the methyl ester of Trp and its simple analogue indole. The methyl ester of Tyr and its model compound p-cresol were also efficient quenchers of the TPA triplet (quenching rate constants between 1 × 109 and 0.3 × 109 M-1 s-1). A similar value was obtained for BHA. By contrast, quenching of the excited drug by the methyl ester of His was much slower. A similar trend was observed in the case of DTPA. Figure 3 shows typical decay traces of triplet TPA in a methanol solvent in the presence and absence of the three amino acid methyl esters. Again, it becomes evident that at the employed amino acid concentrations (1.5 × 10-3 M) there is an efficient quenching of 3TPA by Trp and Tyr, while addition of His does not produce significant changes. Reaction Mechanism. The above results indicate that His is not able to quench the triplet state of TPA or DTPA, while direct quenching by Tyr and Trp is a feasible process, as suggested by the time-resolved studies with their methyl esters. These data, combined with the rate constants reported for quenching of singlet oxygen by the three amino acids (His, 4.6 × 107; Tyr, 0.5 × 107; Trp, 3.2 × 107 M-1 s-1) (23), strongly suggest that His undergoes TPA-photosensitized oxidation via a type II mechanism, while in the case of Tyr a type I mecha-
Drug-Photosensitized Protein Modification
Chem. Res. Toxicol., Vol. 11, No. 3, 1998 175
Scheme 1. Photochemical Reactions between DTPA and p-Cresol
Table 1. Rate Constants for Quenching of 3TPA, 3DTPA, and 1O2 by Electron Donor Kq (M-1 s-1) quencher
3TPA
O2 BHA p-cresol indole histidinea tyrosinea tryptophana
2.1 × 0.4 × 109 0.3 × 109 2.0 × 109
