Photochemistry of 2-hydroxy-4-trifluoromethylbenzoic acid, major metabolite of the photosensitizing platelet antiaggregant drug triflusal

Article abstract of DOI:10.1562/0031-8655(2001)0730463POHTAM2.0.CO2

Triflusal is a platelet antiaggregant drug with photoallergic side effects. However, it is considered a prodrug since it is metabolized to 2-hydroxy-4-trifluoromethylbenzoic acid (HTB) - the pharmacologically active form. HTB was found to be photolabile under various conditions. Its major photodegradation pathway appears to be the nucleophilic attack at the trifluoromethyl moiety. The involvement of the triplet state in the photodegradation has been unequivocally proved by direct detection of this transient in laser flash photolysis and by quenching experiments with oxygen, cyclohexadiene and naphthalene. Finally, the photobinding of HTB to proteins such as bovine serum albumin has been demonstrated using ultraviolet-visible (UV-Vis) and fluorescence spectroscopy. Nucleophilic groups present in the protein appear to be responsible for the formation of covalent drug photoadducts, which is the first step involved in the photoallergy shown by triflusal.

Full text of DOI:10.1562/0031-8655(2001)0730463POHTAM2.0.CO2

Photochemistry and Photobiology, 2001, 73(5): 463–468
Photochemistry of 2-Hydroxy-4-trifluoromethylbenzoic Acid, Major
Metabolite of the Photosensitizing Platelet Antiaggregant Drug Triflusal
¶
F. Bosc a´ , M. C. Cuquerella, M. L. Mar ´ı n and M. A. Miranda*
Departamento de Qu ´ı mica/Instituto de Tecnolog ´ı a Qu ´ı mica UPV-CSIC, Universidad Polit e´ cnica de Valencia,
Valencia, Spain
Received 27 October 2000; accepted 14 February 2001
ABSTRACT
droxy-4-trifluoromethylbenzoic acid (HTB)†, which is in
fact the active form of the drug (Fig. 1). As HTB is also the
entity able to reach the skin and absorb sunlight, it was con-
sidered interesting to study this metabolite instead of triflusal
itself, in order to understand the side effects shown by the
drug.
In the present paper, mechanistic insight into the photo-
allergic disorders elicited by triflusal has been achieved by
studying the photochemical behavior and the photophysical
properties of HTB, together with its photobinding interaction
to proteins such as bovine serum albumin (BSA).
Triflusal is a platelet antiaggregant drug with photoal-
lergic side effects. However, it is considered a prodrug
since it is metabolized to 2-hydroxy-4-trifluoromethyl-
benzoic acid (HTB)—the pharmacologically active form.
HTB was found to be photolabile under various condi-
tions. Its major photodegradation pathway appears to be
the nucleophilic attack at the trifluoromethyl moiety. The
involvement of the triplet state in the photodegradation
has been unequivocally proved by direct detection of this
transient in laser flash photolysis and by quenching ex-
periments with oxygen, cyclohexadiene and naphthalene.
Finally, the photobinding of HTB to proteins such as bo-
vine serum albumin has been demonstrated using ultra-
violet–visible (UV–Vis) and fluorescence spectroscopy.
Nucleophilic groups present in the protein appear to be
responsible for the formation of covalent drug photoad-
ducts, which is the first step involved in the photoallergy
shown by triflusal.
EXPERIMENTAL PROCEDURES
Chemicals. Triflusal and HTB were from Laboratorios Uriach (Bar-
celona, Spain). Benzophenone, naphthalene (NPH), cyclohexadiene
(CHD), SA and BSA were provided by Aldrich (Steinheim, Ger-
many). Sephadex G-10 was from Pharmacia Biotech (Uppsala, Swe-
den). Methanol, acetonitrile (HPLC grade) and acetic acid were from
Scharlau (Barcelona, Spain). Other chemicals were of reagent grade
and used as received. Phosphate-buffered saline (PBS) consisted of
0.050 M phosphate buffer and 0.136 M NaCl solution.
Analytical instrumentation. UV spectra were recorded on a Shi-
madzu (Barcelona, Spain) ultraviolet–visible (UV–Vis) scanning
spectrophotometer (2101PC) with a slit width of 5 nm. The high-
performance liquid chromatography analyses were performed on an
HPLC Varian Systems (Middelburg, The Netherlands) equipped
with a 9012Q pump and a photodiode array (Varian 9065). Samples
were injected onto an analytical Kromasil (Eka Chemicals, Bohus,
INTRODUCTION
Several drugs are known to induce phototoxic or photoal-
lergic side effects after either systemic or topical application
(1–3). Phototoxicity is in general associated with cell degen-
Sweden) 100 C18 column (Tracer, 25
ϫ 0.4 cm, mean particle size
eration or cell death. It appears as a result of the photosen-
sitized formation of free radicals, singlet oxygen, photo-
products or other intermediates. Photoallergy, much less
common, is the result of covalent binding of a photosensi-
tizing drug, photochemical intermediate or photoproduct to
a protein resulting in the formation of a photoantigen (4–8).
In this context clinical evidence for photosensitizing dis-
orders such as phototoxicity and/or photoallergy (9–12) as-
sociated with triflusal (2-acetoxy-4-trifluoromethylbenzoic
acid) has been found. Triflusal is a salicylic acid (SA) de-
rivative used as a platelet antiaggregant (13–15). However,
it is considered a prodrug since it is metabolized to 2-hy-
5 ␮
m) using acetonitrile/water/acetic acid (30:68:2 vol/vol/vol) as
Ϫ
1
the mobile phase; flow rate 1 mL min . Gas chromatography (GC)
analyses were achieved with a Varian 3300 spectrometer equipped
with a HP-5 column (25 m
ϫ 0.32 mm). The proton nuclear mag-
1
netic resonance ( H NMR) spectra were measured by means of a
Varian Gemini 300 MHz instrument. Deuterated chloroform and
deuterated methanol were used as solvents and the signal corre-
sponding to the deuterated solvent in each case was taken as the
reference.
Absorption and fluorescence measurements. Concentration was
fixed by adjusting the absorbance of the solutions at the arbitrary
value of 0.4 at the excitation wavelength of 320 nm and emission
measurements were performed in the region 330–600 nm. Fluores-
†
Abbreviations: BSA, bovine serum albumin; CHD, cyclohexadi-
ene; HTA, hydroxyterephthalic acid; HTB, 2-hydroxy-4-trifluo-
romethylbenzoic acid; GC, gas chromatography; H NMR, proton
nuclear magnetic resonance; NPH, naphthalene; OD, optical den-
sity; PBS, phosphate-buffered saline; PMT, photomultiplier tube;
SA, salicylic acid; UV–Vis, ultraviolet–visible; ⑀, extinction co-
¶
*
Posted on the website on 28 February 2001.
1
To whom correspondence should be addressed at: Departamento
de Qu ´ı mica/Instituto de Tecnolog ´ı a Qu ´ı mica UPV-CSIC, Univ-
ersidad Polit e´ cnica de Valencia, Avenida de los Naranjos s/n,
46022 Valencia, Spain. E-mail: mmiranda@qim.upv.es
efficient; ␾ISC, intersystem crossing quantum yield; ␾flu, fluores-
᭧
2001 American Society for Photobiology 0031-8655/01
$5.00
ϩ0.00
cence quantum yield.
463

464 F. Bosc a´ et al.
Figure 1. Chemical structures of triflusal and its major metabolite
HTB.
Scheme 1. Photodegradation products of HTB under different con-
ditions.
cence quantum yields (⌽flu) were determined using integrated peak
areas by comparison with naproxen as standard (⌽flu 0.47 in aceto-
nitrile) (16,17) in absence of oxygen. The steady-state fluorescence
was obtained with an FS900 Edinburgh Analytical Instruments
RESULTS AND DISCUSSION
Photochemistry
(Abingdon, MD) apparatus, equipped with a 450 W xenon lamp.
The time resolved fluorescence determinations were performed with
an FL900 Edinburgh Analytical Instruments apparatus using a hy-
drogen lamp (0.7 ns pulse width) as the excitation source. The sam-
ples were placed into quartz cells of 1 cm pathlength and deoxy-
genation was made by bubbling nitrogen through them.
HTB was found to be photolabile. When irradiated at a 1
mM concentration in N -bubbled PBS solutions, HTA was
2
obtained as the major photoproduct (Scheme 1). When an
alcohol, such as methanol, was used as the solvent the cor-
responding ester HTME was obtained. The irradiation was
even performed in KOH (1 mM)/MeOH, whereby the mon-
omethyl ester HTME was again obtained as the photoprod-
uct. As SA does not show significant photosensitivity prop-
erties it seems that the photobiological activity of triflusal
could be associated with a nucleophilic photosubstitution at
the trifluoromethyl moiety, as postulated in Scheme 1 for the
hydroxylic solvents. To determine the excited species in-
volved in the photodegradation mechanism, it appeared in-
teresting to study the effect of quenchers such as oxygen,
CHD and NPH on the extent of the photoreactions as well
as on the nature of the photoproducts obtained.
Thus, the same photoproducts were obtained when the ir-
radiation was performed under aerobic conditions, although
the extent of the reaction was lower (see Fig. 2). For in-
stance, after 15 min of light exposure in PBS the conversion
was 30% under nitrogen whereas it dropped to 17% under
aerobic conditions.
Laser flash photolysis measurements. A pulsed Nd:YAG
SL404G-10 (Spectron Laser Systems, Warwick, RI) was used for
the excitation at 266 or 355 nm. The single pulses were
ϳ
10 ns
10 mJ pulse . A pulsed Lo255 Oriel
Stratford, CT) xenon lamp was employed as the detecting light
Ϫ
1
duration and the energy was
ϳ
(
source. The laser flash photolysis apparatus consisted of the pulsed
laser, the Xe lamp, a 77200 Oriel monochromator, an Oriel photo-
multiplier tube (PMT) system made up of a 77348 side-on PMT
tube, 70680 PMT housing and a 70705 PMT power supply. The
output signal from the oscilloscope (TDS-640A Tektronix, K o¨ ln,
Germany) was transferred to a personal computer. All the HTB so-
lutions studied had an absorbance of 0.5 at the wavelength used
(unless other conditions specified) and were deaerated (when spec-
ified) by bubbling nitrogen.
For determination of the triplet extinction coefficients a 355 nm
laser source was employed to excite KOH/MeOH solutions of HTB
with NPH. At this wavelength more than 99% of the absorbed light
is absorbed by HTB. Concentration of NPH was calculated to give
Ն99% quenching of the HTB triplet. On the other hand, for calcu-
lating the intersystem crossing quantum yield methanolic solutions
of HTB and benzophenone with the same absorbance at 355 nm
were excited at this wavelength and the relative intensities of the
resulting transients were compared.
Steady-state photolysis of HTB. Solutions of HTB (1 mM) in dif-
ferent solvents were irradiated with Pyrex-filtered light from an OS-
RAM-HLQ 125 W medium pressure Hg lamp located inside an im-
mersion well photoreactor. The course of the reaction was monitored
by reverse phase high-performance liquid chromatography. Using
the conditions given under ‘‘Analytical Instrumentation,‘‘ the reten-
tion times of HTB and its photoproducts hydroxyterephthalic acid
Also, when a deaerated solution of HTB in MeOH (1 mM)
saturated with CHD was irradiated under the usual condi-
tions the conversion was much lower. Monitoring the pho-
todegradation by GC showed the formation of the CHD di-
mers, with a dimer distribution typical of an energy transfer
(HTA) or methyl hydroxyterephthalate (HTME) were 31, 3, and 11
min, respectively.
Identification of photoproducts. HTA and HTME were synthe-
sized from triflusal or HTB, respectively, following the procedures
1
described in the literature (18). H NMR of the photoproducts ob-
tained from HTB were identical to the synthetic ones.
Ϫ
5
Photobinding to BSA. A solution of HTB (6
ϫ
10 M) and BSA
Ϫ
5
(4 ϫ 10 M) in PBS was allowed to incubate in the dark for 30
min. Then samples were irradiated under usual conditions (see
above) for various time periods. Controls included drug–protein
mixtures kept in the dark and BSA solutions irradiated for the same
periods of time. The drug was separated from the protein using a
Sephadex G-10 column equilibrated with PBS. The photobinding
was monitored by UV–Vis and fluorescence spectroscopy.
Figure 2. Photodegradation of HTB in PBS (1 mM) under aerated
and deaerated conditions.

Photochemistry and Photobiology, 2001, 73(5) 465
Figure 4. Intersection of normalized excitation (␭em
emission (␭exc 320 nm) spectra of HTB in PBS.
ϭ 425 nm) and
ϭ
Figure 3. Absorption spectra of HTB in MeOH and KOH/MeOH.
The emission spectrum of HTB in PBS showed a broad
maximum at 408 nm; the excitation spectrum was similar to
the long wavelength band of the absorption spectrum. From
the intersection of the normalized excitation and emission
reaction (19). The same GC pattern was observed when ben-
zophenone instead of HTB was used as standard triplet pho-
tosensitizer.
Finally, a solution of HTB (1 mM) and NPH (3.4 mM) in
KOH (1 mM)/MeOH was irradiated for 20 min under nitro-
Ϫ
1
spectra a singlet energy value (E0–0) of
ϳ338 kJ mol was
Ϫ
1
estimated, similar to the E0–0 334 kJ mol measured for
ϳ
gen (excitation at 355 nm), making sure that
Ն99% of the
the related SA in PBS under the same conditions (Fig. 4).
A summary of the photophysical data of HTB is given in
Table 1. The fluorescence quantum yield was 0.62 in PBS,
whereas the corresponding value for SA was only 0.31. Na-
proxen in acetonitrile (⌽flu 0.47) was used as a standard
light was absorbed by deprotonated HTB. Partial inhibition
of the photoreaction by NPH was clearly observed (62%
conversion vs 84% in the absence of NPH).
At this point inhibition of the HTB photodegradation by
quenchers such as oxygen, CHD and NPH had been dem-
onstrated, which supports involvement of the triplet state.
From the mechanistic point of view, formation of the CHD
dimers with the typical distribution of an energy transfer
reaction constitutes a further indirect evidence for the role
of triplet HTB. However, the role of the excited singlet re-
mained unclear, and on the other hand direct detection of
the short-lived transients involved in the photodegradation
mechanism was still missing. Therefore, photophysical stud-
ies on HTB were undertaken.
(16,17). The excited singlet states of both HTB and SA were
found to be quenched by molecular oxygen. The quenching
constants were calculated by means of time-resolved fluo-
rescence spectroscopy. Thus, the fluorescence lifetime (␶) of
HTB in PBS was 8.4 ns under N , 8.3 ns in aerated solutions
2
and 7.9 ns in oxygen-saturated solutions. Analogously, in
methanolic solutions the lifetimes were 7.9 ns under N , 5.7
2
ns in aerated solutions and 3.5 ns in oxygen-saturated so-
9
lutions. Based on these data, values of 5.9
ϫ
10 and 1.5
ϫ
10
Ϫ
1
Ϫ1
10
M
s
were calculated for the quenching constant by
oxygen in PBS or methanol, respectively. Under the same
conditions, the fluorescence lifetime of SA in PBS was 4.2
Photophysical properties
Fluorescence. The absorption spectrum of HTB in methanol
has a long wavelength maximum at 303 nm, extending into
the UVB and UVA regions. Under basic conditions (KOH/
MeOH), the absorption spectrum of deprotonated HTB is
shifted to the blue with a maximum at 300 nm. The spectrum
in PBS is similar to that obtained in KOH/MeOH (Fig. 3).
ns under N , 4.1 ns in aerated solutions and 4.0 ns in oxygen-
2
9
saturated solutions. From these data a value of 8.3
ϫ 10
Ϫ
1
Ϫ1
M
s
was calculated for the quenching constant by oxy-
gen, very similar to that calculated for HTB in the same
solvent (Table 1).
On the other hand, the influence of CHD and NPH on the
fluorescence of HTB was investigated. When increasing
amounts of saturated methanolic solutions of CHD or NPH
were added to a methanolic solution of HTB no significant
Table 1. Photophysical data of HTB and salicylic acid
HTB
Salicylic acid
PBS
MeOH
PBS
␾flu
␶
␶
␶
(N ) (ns)
(air) (ns)
0.27
0.62
0.31
4.2
4.1
2
7.9
5.7
3.5
1.5
8.4
8.3
7.9
5.9
(O ) (ns)
2
Ϫ
1
K
(O2) (M
4.0
q
1
Ϫ
1010
1010
1010
8.3 ϫ
s )
ϫ
ϫ
Scheme 2. Photobinding of HTB to the protein.

466 F. Bosc a´ et al.
9
Ϫ
1
Ϫ1
ϭ
8 ϫ 10 M s ), so it was tentatively assigned to the T–
T absorption of HTB. Similar spectra were obtained in
MeOH and KOH/MeOH solutions. In order to confirm the
nature of this transient, the effect of two more quenchers
(CHD and NPH) was studied. When increasing amounts of
the quenchers were added to a solution of HTB (1 mM) in
KOH (1 mM)/MeOH a quenching of the transient was ob-
served. Taking into account the lifetime of this transient, the
calculated value of the quenching rate constants were 4.7
ϫ
9
Ϫ
1
Ϫ
1
9
Ϫ
1
Ϫ1
10 M
s
(CHD) and 5.2
ϫ
10 M
s
(NPH).
The fact that CHD is able to quench the band at 480 nm
supported the assignment of this band to the triplet state of
HTB. Formation of the CHD dimers typical of an energy
transfer reaction (see above) was also in agreement with this
assignment.
Further, NPH was used to calculate the extinction coeffi-
cient of the triplet state of HTB. Thus, this extinction coef-
ficient in KOH/MeOH was estimated by monitoring the en-
ergy transfer reaction between HTB triplet and ground state
NPH:
3HTB
ϩ NPH →
HTB ϩ 3NPH
A deaerated KOH(1 mM)/MeOH solution of HTB (1 mM)
containing different amounts of NPH (from 0.8 to 1.4 mM)
was excited by a 355 nm laser pulse. An efficient quenching
of the triplet at 480 nm was observed, concomitantly with
the corresponding formation of the characteristic band as-
sociated with the triplet state of NPH at 415 nm (19). The
extinction coefficient (⑀) of the triplet state of HTB was then
calculated using equation:
OD( NPH[415 nm]) ϫ ⑀(3HTB[480 nm])
3
ϭ
OD( HTB[480 nm]) ϫ ⑀(3NPH[415 nm])
3
where OD are the optical densities of the HTB triplet state
(480 nm) at the beginning of the reaction and the NPH triplet
state (415 nm) at the end of the reaction (21). The extinction
3
Ϫ
3
Ϫ
1
Ϫ1
coefficient of NPH was taken to be 40 000 dm mol cm
22). This led to a value of 28 600 dm mol 1 cm for
( HTB[480 nm]), which was used to calculate the intersys-
Ϫ
3
Ϫ
Ϫ1
⑀
(
3
Figure 5. Transient absorption spectra (a) of a deaerated solution
of HTB in PBS measured at different times (0.3, 1, 2, 4 and 8 s)
after the pulse (266 nm); and (b) under air of a solution of HTB in
␮
tem crossing quantum yield (␾ISC) for HTB in deaerated
KOH/MeOH solutions using benzophenone (Ph CO) as stan-
2
PBS measured at different times (0.3, 0.5, 1, 2 and 4
pulse (266 nm).
␮s) after the
dard. Excitation of benzophenone and HTB (OD
ϭ 0.07 at
355 nm) was carried out separately and ␾ISC was obtained
by application of the following equation:
quenching of the singlet state by CHD nor by NPH was
observed.
3
␾ISC(HTB) ϭ ␾ISC(Ph CO)
ϫ
OD( HTB[480 nm])
2
Finally, it appeared interesting to study the fluorescence
emission spectrum of HTA, the major photoproduct of HTB,
in aqueous solutions. This molecule still shows fluorescence
ϫ ⑀(3Ph CO[510 nm])/OD( Ph CO[510 nm])
3
2
2
ϫ ⑀(3HTB[480 nm])
(
(
20), so it could also be used in photobiological experiments
see below). Hence, the fluorescence spectrum of HTA in
where OD is the optical density of the benzophenone triplet
state and HTB triplet state at 510 and 480 nm, respectively.
Using the reported values (18) for benzophenone intersystem
PBS was recorded. Its maximum (425 nm) was different
from that of HTB (408 nm); thus, both compounds were
clearly distinguishable from each other by fluorescence. The
emission quantum yield of HTA was 0.58 in deaerated PBS
solutions.
Ϫ
3
Ϫ1
crossing quantum yield (ca 1) and
cm ), the ␾ISC estimated for HTB was 0.14.
⑀ (6500 dm mol
Ϫ
1
Photobinding to proteins
Laser flash photolysis. Figure 5 shows the transient ab-
sorption spectra of N -bubbled (a) and air-bubbled (b) so-
lutions of HTB in PBS measured at different times after laser
flash excitation at 266 nm. The band at 480 nm with a life-
At this point a detailed insight into the photochemistry and
the photophysics of the triflusal metabolite (HTB) had been
accomplished. The next step was to try to understand the
mechanism of photoallergy by studying the metabolite pho-
2
time of 1.5
␮s was quenched by oxygen (Fig. 5b) (K (O )
q 2

Photochemistry and Photobiology, 2001, 73(5) 467
sion shown by the protein fraction after filtration unequiv-
ocally demonstrated the covalent photobinding of HTB to
the protein.
CONCLUSIONS
HTB, the major metabolite of triflusal, has been found to be
photolabile. Steady-state photolysis indicates that the main
photodegradation pathway is the nucleophilic substitution of
fluoride by the solvent. Photophysical studies (fluorescence
and laser-flash photolysis) unequivocally support the in-
volvement of HTB triplet state in the photodegradation
mechanism.
The covalent photobinding of HTB to a model protein
such as BSA has been demonstrated using fluorescence spec-
troscopy. Nucleophilic groups of the protein (–OH, –SH and
Figure 6. Fluorescence spectra of mixtures HTB–BSA irradiated
and nonirradiated, before and after Sephadex filtration.
–
NH ) are likely responsible for the observed photobinding,
2
which is the first step in the onset of photoallergy.
tobinding to proteins. It is reasonable to think that if nucle-
ophilic solvents are able to attack the trifluoromethyl moiety
of HTB, nucleophilic groups of a protein (such as –OH,
Acknowledgements This work has been sponsored by the European
Union (BMH-4-97-2590). M.L.M. acknowledges Generalitat Val-
enciana for a grant (GV.DOC00-25-02) and M.C.C. thanks Funda-
ci o´ n Jose y Ana Royo for financial support.
–
SH, –NH ) could be able to attack HTB in the same way,
2
resulting in a covalent photobinding to the protein
Scheme 2). This would lead to the formation of a pho-
(
toantigen, which is the first step of a photoallergic reac-
tion.
REFERENCES
1
2
3
. Miranda, M. A. (1997) Phototoxicity of drugs. In In Vitro Meth-
ods in Pharmaceutical Research (Edited by J. V. Castell and
M. J. Gomez-Lechon), pp. 289–315. Academic Press, London.
. Gould, J. W., M. G. Mercurio and C. A. Elmets (1995) Cuta-
neous photosensitivity diseases induced by exogenous agents. J.
Am. Acad. Dermatol. 33, 551–573.
. Ophaswongse, S. and H. Maibach (1993) Topical nonsteroidal
anti-inflammatory drugs: allergic and photoallergic contact der-
matitis and phototoxicity. Contact Dermatitis 29, 57–64.
. Epstein, J. H. (1983) Phototoxicity and photoallergy in man. J.
Am. Acad. Dermatol. 8, 141–147.
A usual test for photobinding involves monitoring of the
UV–Vis absorption changes upon irradiation of a drug–pro-
tein mixture and subsequent isolation of the protein by gel-
filtration chromatography (23,24). This technique allows the
detection of the presence of the drug chromophore. How-
ever, if the drug has a fluorescent chromophore this fact
could be used to test photobinding with a higher sensitivity
(
25).
4
5
Ϫ
Ϫ
Solutions of HTB (6
ϫ
10 5 M) and BSA (4 10 5 M)
ϫ
. Miranda, M. A., J. V. Castell, Gomez-Lechon, M. J., D. Her-
nandez and L. A. Martinez (1995) Photobinding of drugs to cells
as an indicator of potential photoallergy. Toxicol. In Vitro 9,
were irradiated under usual conditions for 20 min. The ab-
sorption spectra of the HTB–BSA mixture before and after
irradiation, prior to and after Sephadex filtration were re-
corded. Although some changes in the UV–Vis spectra were
observed, it was not enough to draw a clear conclusion, since
the drug absorption maximum is relatively close to the pro-
tein absorption bands. But, fluorescence measurements pro-
vided clear evidence for drug–protein photobinding (see Fig.
4
99–503.
6. Miranda, M. A., J. V. Castell, D. Hernandez, M. J. Gomez-
Lechon, F. Bosca, I. M. Morera and Z. Sarabia (1998) Drug
photosensitized protein modification: identification of the reac-
tive sites and elucidation of the reaction mechanism with tia-
profenic acid/albumin as a model system. Chem. Res. Toxicol.
1
1, 172–177.
6
). The emission maximum of the fluorescence spectrum of
7. Miranda, M. A., J. V. Castell, D. Hernandez, M. J. Gomez-
Lechon, I. M. Morera and Z. Sarabia (1999) An in vitro ap-
proach to drug photoallergy: use of drug-directed antibodies to
asses photobinding of non-steroidal anti-inflamatories to skin
cells. Toxicol. In Vitro 13, 701–75.
8. Miranda, M. A., J. V. Castell, Z. Sarabia, D. Hernandez, I. Puer-
tes, I. M. Morera and M. J. Gomez-Lechon (1997) Mechanism
of photosensitization by drugs. Involvement of tyrosines in the
photomodification of proteins mediated by tiaprofenic acid in
vitro. Toxicol. In Vitro 11, 653–659.
mixtures before and after irradiation was somewhat differ-
ent. This is due to the fluorescence of HTB (␭max 408 nm)
before irradiation and photoproduct-like chromophore (␭max
4
25 nm) after irradiation.
An important difference was observed between the fluo-
rescence of nonirradiated HTB–BSA mixtures prior to and
after Sephadex filtration. The fluorescence observed for the
mixture before Sephadex treatment is due to the metabolite,
which practically disappears when the sample is filtered
through Sephadex due to the complete removal of HTB from
the mixture. The same was observed when HTA was used
instead of HTB in this control experiment (data not shown).
This rules out the possibility that HTB and/or its photoprod-
uct, HTA, are tightly bound to BSA and may not be removed
by passage through Sephadex.
9
. Serrano, G., A. Aliaga and I. Planells (1987) Photosensitivity
associated with triflusal. Photodermatology 4, 103–105.
1
1
0. Gardeazabal, J., X. Eizaguirre, M. J. Calderon and J. L. Diaz
Perez (1990) Fotodermatitis liquenoide inducida por triflusal
GEF 1, 6.
1. Lee, A. Y., S. H. Yoo and K. H. Lee (1999) A case of photo-
allergic drug eruption caused by triflusal (Disgren). Photoder-
matol. Photoimmunol. Photomed. 15, 85–86.
2. Miranda-Romero, A. and L. Navarro (1992) Fotosensibilidad
por triflusal GEF 5, 31.
3. Messa, G. L., M. Franchi, A. Auteri and T. Diperri (1993) Ac-
tion of 2-acetoxy-trifluoromethylbenzoic acid (triflusal) on
platelet-function after repeated oral-administration in man- A
1
1
After irradiation the fluorescence emission at 425 nm re-
mained the same before and after filtration, indicating that
HTB photodegradation had occurred. The fluorescence emis-

468 F. Bosc a´ et al.
pharmacological clinical study. Int. J. Clin. Pharmacol. Res. 13,
19. Bosca, F., S. Encinas, P. F. Heelis and M. A. Miranda (1997)
Photophysical and photochemical characterization of a photo-
sensitizing drug: a combined steady state photolysis and laser
flash photolysis study on carprofen. Chem. Res. Toxicol. 10,
263–273.
1
4. Plaza, L., L. Lopez-Bescos, L. Martin-Jadraque, E. Alegr ´ı a, J.
M. Cruz-Fern a´ ndez, J. Velasco, J. A. Ruip e´ rez, F. Malpartida,
A. Artal, A. Cabades and A. F. Zurita (1993) Protective effect
of triflusal against acute myocardial infarction in patients with
unstable angina: results of a Spanish multicenter trial. Cardi-
ology 82, 388–398.
5. De la Cruz, J. P., J. M. Mata and F. S. de la Cuesta (1992)
Triflusal vs aspirin on the inhibition of human platelet and vas-
cular cyclooxygenase. Gen. Pharmacol. 23, 297–300.
6. Moore, D. E. and P. P. Chappuis (1988) A comparative study
of the photochemistry of the non-steroidal anti-inflammatory
drugs, naproxen, benoxaprofen and indomethacin. Photochem.
Photobiol. 47, 173–180.
8
20–827.
2
0. Xinhua, Q., L. J. Kirschenbaum and E. T. Borish (2000) Hy-
droxyterephthalate as a fluorescent probe for hydroxyl radicals:
application to hair melanin. Photochem. Photobiol. 71, 307–
3
13.
1
2
1. Bensasson, R. V., E. J. Land and T. G. Truscott (1983) Flash
Photolysis and Pulse Radiolysis. Pergamon Press, Oxford.
2. Carmichael, I. and G. L. Hug (1986) Triplet–triplet absorption
spectra of organic molecules in condensed phases. J. Phys.
Chem. Ref. Data 15, 1–250.
2
1
2
3. Lovell, W. W. (1993) A scheme for in vitro screening of sub-
stances for photoallergic potential. Toxicol. In Vitro 7, 95–102.
4. Barratt, M. D. and K. R. Brown (1985) Photochemical binding
of photoallergens to human serum albumin: a simple in vitro
method for screening potential photoallergens. Toxicol. Lett. 24,
1
7. Martinez, L. J. and J. C. Scaiano (1998) Characterization of the
transient intermediates generated from the photoexcitation of
nabumetone: a comparison with naproxen. Photochem. Photo-
biol. 68, 646–651.
2
1
–6.
1
8. Hauptschein, M., E. A. Nodiff and A. J. Saggiomo (1954) Tri-
fluoromethyl derivatives of hydroxybenzoic acids and related
compounds. J. Am. Chem. Soc. 76, 1051–1054.
25. Moser, J., F. Bosca, W. W. Lovell, J. V. Castell, M. A. Miranda
and A. Hye (2000) Photobinding of carprofen to protein. J. Pho-
tochem. Photobiol. B: Biol. 58, 13–19.