Journal of Photochemistry & Photobiology, B: Biology 178 (2018) 201–210

Contents lists available at ScienceDirect

Journal of Photochemistry & Photobiology, B: Biology

journal homepage: www.elsevier.com/locate/jphotobiol

Development of photoprotective, antiphototoxic, and antiphotogenotoxic T

formulations of ocular drugs with fluoroquinolones
⁎
Anna Zgadzaja, , Joanna Kornackaa, Anita Jastrzębskaa, Andrzej Parzonkob, Sylwester Sommerc,
Grzegorz Nałęcz-Jaweckia
a
Department of Environmental Health Sciences, Medical University of Warsaw, Poland
b
Department of Pharmacognosy and Molecular Basis of Phytotherapy, Medical University of Warsaw, Poland
c
Institute of Nuclear Chemistry & Technology, Centre for Radiobiology & Biological Dosimetry, Warsaw, Poland

A R T I C L E I N F O A B S T R A C T

Keywords: The development of innovative solutions in photosafety of photolabile pharmaceutical products may help to
Fluoroquinolones reduce the adverse effects of these products, caused by light exposure. Providing new data in this area of study is
Phototoxicity particularly important in case of drugs applied topically on sensitive organs such as eyes. The main goal of this
P-coumaric acid research is to investigate whether two potential excipients, namely: p-coumaric acid and benzophenone-4, affect
Benzophenone-4
the photodegradation, phototoxicity and photogenotoxicity of water solutions of four fluoroquinolones: cipro-
floxacin, lomefloxacin, fleroxacin and clinafloxacin. We conducted a set of bioassays combined with the ap-
plication of high-performance liquid chromatography and mass spectrometry techniques. The significant re-
duction of phototoxic and photogenotoxic abilities was evaluated in mixtures with ciprofloxacin and p-coumaric
acid by using the umu test with Salmonella typhimurium TA1535/pSK1002, the methylthiazol tetrazolium re-
duction assay, and the micronucleus assay with the V79 cell line. In the bacterial assay the opposite effect was
observed for the formulation with lomefloxacin and p-coumaric acid. This may be explained by the significant
differences in the profile of the lomefloxacin photodegradation products. Further, the photoprotective and an-
tiphotomutagenic abilities of ciprofloxacin mixed with benzophenone-4 were assessed. Promising results ob-
tained in compositions with ciprofloxacin may be a basis for further research. Nevertheless, the increase in the
DNA damage potential in mixtures with p-coumaric acid and two other antibiotics shows the importance of the
safety evaluation of such innovative combinations.

1. Introduction
The development of innovative solutions in photosafety of photo-
labile pharmaceutical products may help to reduce the adverse effects
of these products, caused by light exposure. Providing new data in this
area is particularly important in case of drugs applied topically on
sensitive organs such as eyes. Eyes are a complex optical system ex-
posed to light from the surrounding environment. They are particularly
vulnerable to the ultraviolet (UV) light present in the range of sunlight
and generated by some artificial light sources; exposure to UV light may
result in corneal damage, cataract formation and the progress of age
related macular degeneration. Photosafety testing is particularly re-
levant for chemicals with a phototoxic and photogenotoxic potential
[1], such as fluoroquinolones (FQ). FQ are broad-spectrum antibiotics
often used as components of various pharmaceutical preparations, in-
cluding eye drops and ocular ointments [2,3]. Photolysis of these an-
tibiotics may lead not only to light induced side effects but also to the
reduction of their antibacterial activity. The effects may decrease the
effectiveness of the treatment and stimulate the development of more
resistant bacterial strains. There are many reports about FQ cytotoxicity
and genotoxicity in the presence of light in in vivo and in vitro tests
[4–10]. Among the ocular side effects of this group of antibiotics are
local irritation and burning, hyperemia, eyelid edema, lid margin
crusting, superficial punctate keratitis, corneal precipitation and per-
foration, blurred vision, and lacrimation [3]. Moreover there are re-
ports about the risk of reduction in the transparency of the eye lenses
due to the photopolymerization of α-cristallin and of damage to the lens
epithelial cells [11]. An analysis of FQ with respect to their photo-
reactivity not only as single compounds but also as components of
mixtures may provide new information in the photochemistry and new
possibilities to increase the photosafety of pharmaceutical formula-
tions. There are reports about the modification of FQ photoreactivity by
a combination with various reactive oxygen species scavengers. Ume-
zawa et al. [4] examined the influence of superoxide dismutase,

⁎
Corresponding author.
E-mail address: azgadzaj@wum.edu.pl (A. Zgadzaj).

https://doi.org/10.1016/j.jphotobiol.2017.11.011
Received 15 August 2017; Received in revised form 6 November 2017; Accepted 7 November 2017
Available online 09 November 2017
1011-1344/ © 2017 Elsevier B.V. All rights reserved.
A. Zgadzaj et al.
catalase, sodium azide and 1,4-diazabicyclo-[2,2,2]-octane on the
photodynamic calf thymus DNA strand-breaking activity of several FQ.
The obtained results differed depending on the structure of the irra-
diated antibiotic. Bulera et al. [8] co-incubated the CHO cells with
clinafloxacin and several antioxidants which reduced the hydroxyl ra-
dical formation but inhibited the photogenotoxicity only to a limited
extent. These reports show the variety of results obtained by examining
FQ photoreactivity in the presence of antioxidants. The reactive oxygen
species scavengers may also be used as ingredients of ocular formula-
tions. There are examples of the photoprotective potential of similar
compositions. One of them was eye drops with p-coumaric acid that
successfully reduced the photoirritation in rabbit eyes [12]. Moreover,
these compounds may exhibit antimutagenic activity, as with bacterial
assays [13]. These potential ocular excipients have never been in-
vestigated in combination with FQ. We supposed that an analysis of
photodegradation, phototoxicity and photogenotoxicity of the above-
mentioned composition would provide interesting data because of the
previous reports about the photoprotective and antimutagenic potential
of p-coumaric acid. Another way to reduce the degradation of photo-
labile drugs applied topically is to design innovative pharmaceutical
formulations with chemical light-absorbers. As reported by Ioele et al.
[14], the addition of sunscreens such as octisilate and octyl methox-
ycinnamate successfully inhibited the diclofenac photodegradation in
gel formulations. Similar modifications were considered in some com-
positions of ocular preparations. A liquid sunscreen applied topically
may block the harmful effect of UV irradiation and protect the sensitive
surface of an eye. Cejka et al. [15,16] confirmed the photoprotective
abilities of actinoquinol as an ocular light absorber through in vivo tests.
They investigated the changes in corneal optics, hydration, and im-
munohistochemical conditions. Nowadays, eye drops with this in-
gredient and hyaluronic acid are available in the market registered as a
medical device in Europe. Kek et al. [17] compared the efficacy of seven
water- and oil- soluble compounds in eight possible ocular formulations
applied directly on the surface of an eye ex vivo. They used an ocular
spectrometer system to evaluate the changes in the transmission of UV
radiation through the anterior eye. Significant increases in the ab-
sorption of the UV spectrum were detected in seven of the eight studied
formulations, demonstrating their potential as topical ocular sunsc-
reens. One of these compounds was benzophenon-4 (sulisobenzone).
Except our latest work about ofloxacin in ointments [18], there are only
few reports of a simultaneous evaluation of the photodegradation,
photogenotoxicity, and phototoxicity of FQ in various pharmaceutical
formulations with different excipients. We investigated four variants of
ointments and obtained a significant reduction of ofloxacin photolysis
in the ointment with bisoctrizole. Hubicka et al. [19] assessed the
photodegradation of ciprofloxacin, moxifloxacin, norfloxacin and
ofloxacin in the presence of excipients from tablets; however, they did
not perform cytotoxicity or genotoxicity tests.
The main goal of this work was to investigate whether two potential
excipients, namely, p-coumaric acid and benzophenone-4, could affect
the photodegradation, phototoxicity and photogenotoxicity of water
solutions of the four selected FQ. For this purpose, we conducted a set
of bioassays combined with the application of high-performance liquid
chromatography (HPLC) and mass spectrometry techniques.
2. Materials and Methods
2.1. Chemicals
Ciprofloxacin (CP) (CAS No. 85721-33-1), fleroxacin (FR) (CAS No.
79660-72-3), clinafloxacin (CL) (CAS No. 105956-97-6) and benzo-
phenone-4 (BP-4) (CAS No. 4065-45-6) were purchased from Fluka.
Lomefloxacin (LM) (CAS No. 98-79-52-8), p-coumaric acid (p-CA) (CAS
No. 501-98-4), ethyl methanosulfonate (EMS) (CAS No. 62-50-0), 4-
nitroquinoline N-oxide (4-NQO) (CAS No. 56-57-5) and 2-aminoan-
thracene (2-AA) (CAS No. 613-13-8) were purchased from Sigma.
202
Journal of Photochemistry & Photobiology, B: Biology 178 (2018) 201–210
Acetic acid (CAS No. 64-19-7), neutral red (NR) (CAS No. 553-24-3)
and ethanol (CAS No. 64-17-5) were purchased from POCh S.A.
Acetonitrile (CAS No. 75-05-8) and methanol (CAS No. 67-56-1) were
purchased from Merck. Trifluoroacetic acid (CAS No. 76-05-1) was
purchased from J.T. Baker. Vectashield mounting medium with 40,6-
diamidino-2-phenylindole (DAPI, CAS No. 28718-90-3) was purchased
from Vector Laboratories. Sodium lauryl sulfate (SLS) (CAS No. 151-21-
3) was purchased from BDH Chemicals.
2.2. Biological Cultures
Salmonella typhimurium (S. typhimurium) TA1535/pSK1002 was
purchased from Deutsche Sammlung von Mikroorganismen Und
Zellkulturen GmbH in Germany. The Chinese hamster lung fibroblasts
cell line V79 (ATCC CCL-93™) was purchased from American Type
Culture Collection.
2.3. Preparation and Irradiation of Tested Solutions
The four FQ investigated in this study were CP, LM, FR and CL. They
were selected based on the differences in their chemical structure, i.e.,
the increasing number of halogen atoms in the molecule connected to
the increasing photolability of these compounds. All the compounds
were water soluble. Two (CP and LM) of them were selected because of
their application in ocular formulations. All the solutions were prepared
in water or PBS. Solvents and the concentration range differed de-
pending on the bioassay. Because of the very low range of tested con-
centrations in the bacterial assay, deionised water was chosen as a
solvent to avoid the influence of other ions in the solution on the ki-
netics of the photodegradation process. Further, deionised water is a
preferred negative control in the umu test according to the ISO protocol.
For the neutral red uptake assay (NRU), the micronucleus test, and the
methylthiazol tetrazolium reduction assay (MTT), all FQ were dissolved
in PBS because of the relatively high sensitivity of mammalian cells to
the changes in the medium osmolarity. In the bioassays, we could
choose a relatively high range of FQ concentration; therefore, the in-
fluence of additional ions on the kinetics of the photodegradation
process was acceptable. The concentrations of BP-4 and p-CA used in
the mixtures with FQ were selected based on their preliminary toxicity
and phototoxicity evaluation. If the range of nontoxic concentrations
was wide, we selected the lowest concentration that exhibited a suffi-
cient photoprotective effect in combination with at least one of the
drugs. We also considered the available literature data [12,17]. BP-4
had good water solubility, and p-CA was diluted in water with ethanol
(1:1). All the solutions were freshly prepared before every bioassay.
The samples were irradiated in the sunlight simulator SUNTEST CPS
+ (Atlas) with a 1500 W xenon lamp and maximum air cooling. The
lamp emitted light in the UV–Vis (300–800 nm) wavelength range, and
the radiation intensity was 58 mW/cm2. The temperature inside the
SUNTEST chamber had no effect on the FQ degradation at least during
this experiment. The samples were irradiated in plugged quartz tubes
just before each assay. The irradiation time differed among the com-
pounds depending on their photolability. For the umu test, the solutions
were exposed to light for 30 min (CP) or 5 min (LM, FR, and CL), which
was sufficient for their total photodegradation without the excipients.
For the NRU and MTT assay, because of the relatively high range of the
tested concentrations, the photolysis was slower; therefore, we ex-
tended the irradiation time to 15 min (FR), 30 min (LM and CL), and
90 min (CP). FQ concentrations before and after light exposure were
measured with HPLC.
2.4. Chromatographic Analysis
The drug concentrations during irradiation were monitored using a
Shimadzu HPLC system with LC-10AT pumps, CTO-10AS column oven,
and SPD-M10Avp photodiode-array detector. The column parameters
A. Zgadzaj et al.
were as follows: Luna (Phenomenex, Torrance, CA, U.S.) C18 (2)
column, 150 × 4.6 mm, particle size of 5 μm. The flow rate was 1 mL/
min. The injection volume was 20 μL. The mobile phases consisted of
HPLC grade water with 0.01% trifluoroacetic acid as eluent A and
acetonitrile with 0.01% trifluoroacetic acid as eluent B. The gradient (%
B) was as follows: 0 min, 20%; 2 min, 20%; 9 min, 90%; 10 min, 90%;
10.1 min, 20%; and 13 min, 20%. The quantitative analysis was per-
formed at 278 nm for CP, 288 nm for LM, 287 nm for FR, and 297 nm
for CL. The areas of all FQ peaks were converted to the concentrations
[mg/L] through the calibration curves.
Photodegradation products were detected using Agilent 1260 in-
finity HPLC coupled with a Hybrid Triple Quadrupole/Linear Ion Trap
mass spectrometer QTRAP® 4000 (AB SCIEX, Framingham, MA, USA).
Chromatographic separation was achieved with a Kinetex RP-18
column (100 mm, 4.6 mm, particle size = 2.6 μm) supplied by
Phenomenex (Torrance, CA, USA). The column was maintained at 40 °C
at a flow rate of 0.5 mL/min. The mobile phases consisted of HPLC
grade water with 0.2% formic acid as eluent A and acetonitrile with
0.2% formic acid as eluent B. The gradient (%B) was as follows: 0 min,
10%; 1 min, 10%; 25 min, 90%, 35 min, 90%; 36 min, 10%; and
45 min, 10%. The injection volume was 10 μL. The tandem mass
spectrometer was operated with electrospray ionization (ESI) in the
positive ionization mode using Full Scan and Multiple Reaction
Monitoring (MRM) modes. The target FQ was analyzed in the MRM
mode. Two transitions between the precursor ion and the most abun-
dant fragment ions were monitored. The full scan spectra were acquired
in the m/z range from 200 to 400 Da. The positive ionization (capillary
voltage = +5000 V) method was used.
2.5. Umu Test
The umu test is a biological assay for evaluating the genotoxicity of
water samples or chemical compounds. The test organism was S. ty-
phimurium TA1535/pSK1002. The assay was performed according to
the ISO 13829 protocol, in 96-well microplates with and without me-
tabolic activation as described earlier [18]. All the samples were tested
at least three times (on different days with freshly prepared and irra-
diated solutions) in a dilution series of six concentrations (four data
points for each option). The DNA damage of the tested sample in the
umu test was represented by the induction ratio (IR). Samples with IR
of > 1.5 were considered genotoxic. The sample toxicity was re-
presented as the reduction of the growth factor (GF). For negative
control, GF was equal to 1.0. If GF was smaller than 0.5, the sample was
considered toxic and the genotoxicity could not be evaluated.
2.6. NRU Assay
The NRU test was performed on the V79 mammalian cell line as a
preliminary cytotoxicity study before the micronucleus assay. The
quantitative estimation of viable cells was evaluated based on the
neutral red uptake in comparison to the negative control. The cells were
seeded in 96-well microplates (104 cells/100 μL) in the DMEM (Lonza)
culture medium (supplemented with 10% FBS, 0.01 M HEPES, 100 IU/
mL penicillin and 0.1 mg/mL streptomycin) and incubated for 24 h (5%
CO2, 37 °C, > 90% humidity). Then, each plate was examined under a
microscope to ensure that the cells formed a half-confluent monolayer.
The culture medium was replaced by the tested samples mixed with the
treatment medium with a reduced level of serum: in eight concentra-
tions in a two-fold dilution series for 24 h. The treatment medium was
removed; the cells were washed with PBS and treated with the neutral
red medium for 3 h. Then, the medium was discarded; the cells were
washed with PBS again and treated with an ethanol and acetic acid
solution. The amount of neutral red released from the cells was eval-
uated colorimetrically at 540 nm. As negative and positive controls,
PBS and SLS were used, respectively. The cytotoxicity of all the samples
was tested two times (four replicates for one concentration each time).
203
Journal of Photochemistry & Photobiology, B: Biology 178 (2018) 201–210
2.7. MTT Assay
The MTT test was performed on the V79 mammalian cell line in
addition to the NRU assay for the mixture of p-coumaric acid and CP in
the range of higher concentrations. The measurement of cell viability in
this test was based on the metabolic activity of the cells. Mitochondrial
succinate dehydrogenase reduced the water-soluble MTT salt in the
viable cells to a violet insoluble formazan that could be quantified
colorimetrically after dissolving it in alcohol. The cells were seeded and
treated with samples diluted in the treatment medium as described in
the NRU test. After exposition, the cells were washed and treated with
the medium containing the MTT, which penetrated into the living cells.
The medium with the dye was removed after 2 h and replaced with
isopropanol that dissolved and released the formazan from the cells.
The colour intensity that correlated with the viability of the cell culture
was evaluated with a microplate spectrophotometer at 570 nm. As ne-
gative and positive controls, PBS and Triton X-100 were used, respec-
tively. The cytotoxicity of all the samples was tested three times in a
dilution series of eight concentrations (four replicates for one con-
centration each time). The range of tested concentrations was from
88 μg/mL and 12 μg/mL to 1500 μg/mL and 200 μg/mL for CP and p-
CA, respectively. The samples were irradiated until the photodegrada-
tion of 50% CP in the formulation without excipients.
2.8. Micronucleus Assay
The micronucleus test was performed in vitro on Chinese hamster
lung fibroblasts (V79 cell line) as described earlier [18]. The cells in the
half-confluent monolayer were exposed for the irradiated or non-irra-
diated FQ and their mixtures for 6 h. Each sample was prepared in PBS
and mixed with the culture medium before the assay (1:3). Our pre-
liminary experiments showed that such modification did not affect both
the cell proliferation and the results of the assay. PBS and EMS were
used as the negative and positive controls, respectively. The micro-
nucleus frequencies were analyzed in at least 2000 binucleated cells per
sample and the controls, equally divided between the two replicates.
The assay was performed at least twice for all variants, each time with a
fresh cell culture. Glass slides stained with DAPI Vectashield were
counted with Metafer Slide Scanning Platform (MetaSystems).
2.9. Statistical Analysis
A statistical analysis of the present results was performed depending
on the presence or absence of the compliance with a normal distribu-
tion with a Student's t-test or Mann–Whitney U test; p < 0.05. For the
results of the micronucleus test, the two proportions z-test was used. In
the umu test, the significance of the differences between the obtained
results was calculated based on the disparity between the slope of the
linear correlation (FQ concentration and the IR) of the selected samples.
3. Results
3.1. Photodegradation
The most significant difference in the progress of photodegradation
in the solutions was observed between CP and the rest of the FQ. Under
the conditions tested, CP was the most photostable with t1/2 = 25 min
(R2 = 0.965) (Fig. 1). FR and CL were the most susceptible to photo-
degradation and exhibited similar photolability. LM was only slightly
more photostable than CL and FR, with t1/2 = 1.2 min (R2 = 0.997)
(Fig. 2). For CP, a correlation between the concentration and the pho-
tostability of the tested solution was observed; the photolysis was
slowed down with an increase in the concentration of the irradiated
drug. This finding is consistent with the literature data [21]. The pho-
toprotective effect of p-CA and BP-4 was observed in solutions with CP,
particularly in the lower range of concentrations irradiated for the umu
A. Zgadzaj et al.
test. In the presence of p-CA, the CP concentration decreased only by
10% after 60 min, while in the absence of p-CA, the CP level decreased
by 90% (Fig. 1). For BP-4, the effectiveness of photoprotection was
similar. For LM, p-CA caused an increase in t1/2 from 2.1 to 3.2 min
(R2 = 0.985). In contrast, in mixtures with FR and CL, we did not ob-
serve any significant changes in their photolability in comparison to the
solutions without the excipients.
The mass spectrometry analysis was performed with the CP and LM
solutions with and without p-CA. The irradiation of the FQ resulted in
the formation of new chromatographic peaks corresponding to the
transformation products generated during the photolysis process. The
analyzes of the full scan spectra revealed the peaks with an increasing
area with an increase in the irradiation time. Using the Analyst 1.6
software (MSD Sciex), we selected the most abundant product ions for
each precursor ion and the MRM acquisition methods were built au-
tomatically. For CP (m/z = 332), four degradation products were
monitored during the irradiation with m/z of 316 (−16 Da), 330
(− 2 Da), 346 (+14 Da), and 348 (+16 Da) (Fig. 1). The addition of p-
CA to the CP solution significantly reduced the formation of the pho-
todegradation products (Fig. 1). For LM (m/z = 352), the following
degradation products were monitored during the irradiation: m/z of
350 (−2 Da), 348 (− 4 Da), 336 (− 16 Da) and 332 (− 20 Da). Our
results indicated, that the deprotonation may occur at different sites of
the LM molecule, as several peaks can be seen in the chromatograms of
the photoproducts (Fig. 3). The relative peak areas of the photoproducts
clearly showed the different kinetics of the formation of different iso-
mers. Moreover, the presence of p-CA led to altered pathways of LM
degradation, as can be seen for two peaks MRM = 348/320 with re-
tention times of 4.1 and 4.7 min, respectively. This could result in the
different reactivity of the irradiated solutions.
204
Journal of Photochemistry & Photobiology, B: Biology 178 (2018) 201–210
Fig. 1. Relative peak area of the CP (MRM 332 → 314) and the
CP transformation products (MRM 316 → 298, MRM 330 →
231, MRM 346 → 328, MRM 348 → 330) in the photo-
degradation experiments carried out with CP alone (blue dia-
monds) and with p-CA (red squares). (For interpretation of the
references to colour in this figure legend, the reader is referred
to the web version of this article.)
3.2. Umu Test
The preliminary research confirmed the lack of toxicity, photo-
toxicity, and photogenotoxicity of p-CA and BP-4 in the concentrations
ranges tested in mixtures with antibiotics (results not shown). For all
FQ, the toxicity and genotoxicity exhibited dose dependence. The DNA
damage was evaluated only for concentrations that did not inhibit
bacterial growth below GF < 0.5. CP and CL were significantly more
toxic and genotoxic for S. typhimurium TA1535 than for LM and FR;
therefore, they were tested in the lower range of concentrations. All
non-irradiated solutions containing only FQ increased the IR, which
was consistent with the mechanism of action of this group of anti-
biotics. Non-irradiated mixtures with BP-4 did not show any differences
in toxicity or genotoxicity as compared to solutions without this ex-
cipient. But, the addition of p-CA caused significant changes in the umu
test results for solutions with CP, LM, and FR before the light exposition
(Fig. 4). The genotoxic potential of CP mixed with p-CA was lower than
that obtained for the solution containing only CP. It could be observed
in both cases: with and without metabolic activation. For non-irradiated
formulations with LM and FR the addition of p-CA induced the opposite
effect, the genotoxicity of the mixtures significantly increased as com-
pared to that in the solutions containing only antibiotics. This could
also be observed with and without metabolic activation. From the re-
sults for CL, we could not draw similar conclusions. The irradiation
reduced the concentration of FQ in all the tested solutions and in-
creased the amount of their photodegradation products which resulted
in the formation of mixtures of all these compounds. Therefore, the
differences in the IR between the irradiated and the non-irradiated
solutions with an equal amount of antibiotics resulted only from the
presence of the photodegradation products of the FQ. Because of the
A. Zgadzaj et al.
kinetics of the photodegradation process, the above finding was true
only for CP. The rest of the FQ degraded completely by the end of the
light exposition. Therefore, on the graphs (Fig. 4) with the IR results,
the x-axis represents the initial concentration of each antibiotic. BP-4
did not affect the genotoxic potential of any irradiated mixture, al-
though it slowed down the photodegradation of CP and LM. In contrast,
the addition of p-CA not only photoprotected some antibiotics in the
mixtures but also modified their genotoxicity after irradiation as com-
pared to the irradiated solutions without the excipients. The photo-
degradation products of CP irradiated with p-CA were significantly less
genotoxic than in the case without the antioxidant. The opposite effect
was observed in mixtures with LM and FR. The results for CL were not
unequivocal. Similar changes were obtained in both variants of the umu
test: with and without the metabolic activation. The S9 fraction did not
increase the IR of any solutions or mixtures irradiated or not, which
indicated the absence of possible precarcinogens in the tested samples
(results not shown). The metabolic activation reduced the IR values of
all the tested formulations, which might be explained by the protein
205
Journal of Photochemistry & Photobiology, B: Biology 178 (2018) 201–210
Fig. 2. Relative peak area of the LM (MRM 352 → 265) and
the LM transformation products (MRM 332 → 189, MRM
336 → 149, MRM 348 → 320 with retention times 4.1 min
and 4.7 min, MRM 350 → 332 with retentions time 8.7 min
and 12.2 min) in the photodegradation experiment carried
out with LM alone (blue diamonds) and in the presence of p-
CA (red squares). RT – retention time. (For interpretation of
the references to colour in this figure legend, the reader is
referred to the web version of this article.)
binding of the tested compounds and their photodegradation products.
Some of them may also be metabolized to less reactive compounds.
3.3. NRU Assay
The main task of the NRU assay in this project was to provide data
about the cytotoxicity of the selected FQ and excipients before and after
irradiation toward V79 cells as a preliminary study before the micro-
nucleus test. The comparable sensitivity of V79 and Balb/C 3T3 cells,
which are recommended for the evaluation of phototoxic potential,
were previously confirmed in our preliminary studies. Therefore, V79
fibroblasts were used in our in vitro assays. Three of the tested anti-
biotics (LM, FR, and CL) did not affect the cellular viability in the entire
range of the tested concentrations (from 250 mg/L to 2 mg/L) in
comparison with the negative control (Supplementary material). Only
CP at the highest concentration (250 mg/L) reduced the cellular sur-
vival rate to 70%. Irradiation of the tested FQ had no influence on their
cytotoxicity at any evaluated concentration. The cytotoxicity of the
A. Zgadzaj et al. Journal of Photochemistry & Photobiology, B: Biology 178 (2018) 201–210

XIC of +M R M (30 pairs): 350.122/332.100 D a ID : Lom eflx U V 350 1 from Sam ple 18 (L 3) of 2015-12-21C hinolony pC A U V.wiff (T urbo S p r a y ) M a x . 3 .8 e 6 c p s .

12.18
3.8e6

3.6e6

3.4e6

3.2e6

3.0e6

2.8e6

2.6e6

2.4e6

2.2e6
I n t e n s it y , c p s

2.0e6

1.8e6

1.6e6
11.66
1.4e6

1.2e6

1.0e6 8.68
8.0e5
8.28

6.0e5 9.22

4.0e5
8.11 12.88
2.0e5

0.0
7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 12.5 13.0 13.5 14.0
Tim e, m in

XIC of +M R M (30 pairs): 348.257/320.200 D a ID : Lom eflx U V 348 4 from Sam ple 18 (L 3) of 2015-12-21C hinolony pC A U V.wiff (T urbo S p r a y ) M a x . 7 .9 e 4 c p s .

4.06
7.9e4

7.5e4

7.0e4

6.5e4

6.0e4

5.5e4

5.0e4

4.5e4
I n t e n s it y , c p s

4.0e4

3.5e4

3.0e4

2.5e4

2.0e4
4.69 8.97
1.5e4

1.0e4

5000.0

0.0
3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0
Tim e, m in

Fig. 3. The extracted ion chromatogram of the LM photoproducts with the MRM: 350/332 on the top picture and 348/320 on the bottom picture before (black line) and after irradiation
(red line). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

206
A. Zgadzaj et al.
excipients was tested in combinations with the abovementioned com-
pounds and separately. BP-4 alone did not affect the cellular viability
from 250 mg/L to 2 mg/L before and after the light exposition. But, p-
CA at concentrations above 31.5 mg/L reduced the cellular survival rate
by 20%. Irradiation affected the cytotoxicity of the p-CA solution.
Samples after light exposition were slightly less cytotoxic for V79 cells
(results not shown). A juxtaposition of results obtained for mixtures and
plain solutions revealed the differences between their effects on the cell
culture. In all the cases combining antibiotics with excipients slightly
decreased the cellular viability in comparison with solutions of single
compounds. For LM, FR, and CL mixed with BP-4 the non-cytotoxic
proportions (no significant differences in the number of living cells
between the sample and the negative control) were as follows: 125 mg/
L of FQ with 125 mg/L of BP4. For CP, the proportion was 62.5 mg/L of
the antibiotic and the excipient. The highest non-cytotoxic concentra-
tion of p-CA in the mixtures was 15 mg/L combined with 62.5 mg/L of
207
Journal of Photochemistry & Photobiology, B: Biology 178 (2018) 201–210
Fig. 4. The results of the umu test without me-
tabolic activation before and after the light ex-
position (+ UV): IR is represented on the y-axis
and the FQ concentrations are on the x-axis [mg/
L]. The black dots and solid line illustrate the
antibiotic solution, the white dots with dashed
line illustrate the mixtures with p-CA (2:1 with
LM, FR and CL and 20:1 with CP), the grey dots
with grey line illustrate the mixtures with BP-4
(1:1 with LM, FR and CL and 10:1 with CP). In
the umu test that is a bacterial assay the dilution
ranges of all antibiotics had to be low enough so
as not to inhibit the growth of the S. typhimurium
TA1535/pSK1002. However, the concentrations
ranges evaluated in this assay was similar to the
possible drug accumulation in the ocular tissues
after eye drops application [20].
CP or 125 mg/L of three other antimicrobial drugs. Based on these re-
sults, the optimal proportions of the chemical compounds in the mix-
tures for the micronucleus assay were as follows: 62.5 mg/L of the
antibiotic and 62.5 mg/L of BP4 or 15 mg/L of p-CA. All concentrations
were obtained in the cell culture medium after a four-fold dilution of
the mixtures prepared in PBS. All the combinations were irradiated in
the undiluted form (250 mg/L of FQ + 250 mg/L of BP-4 or 60 mg/L of
p-CA).
3.4. MTT Assay
The range of drug concentrations allowed us to calculate the IC50
ratio for all the tested samples. The cytotoxicity of CP mixed with p-CA
(IC50 = 157 mg/L) did not differ significantly from that of the clear
antibiotic solution (IC50 = 138 mg/L) stored in the dark. The irradia-
tion of all the tested samples increased their cytotoxicity. But, the IC50
A. Zgadzaj et al.
ratio juxtaposition of the CP clear solution (IC50 = 81 mg/L) with CP
and p-CA mixture (IC50 = 104 mg/L) revealed the photoprotective
effect of the antioxidant. Moreover, the dose response relationship was
clearly visible in all the tested mixtures and solutions before and after
the light exposition (Fig. 5).
3.5. Micronucleus Test
Two of the tested antibiotics (LM and FR) did not change the
number of micronuclei or the proliferation index significantly before
and after the irradiation, although FQ in the light-exposed samples were
almost completely photodegraded. The addition of excipients had no
influence on the obtained results in the case of these drugs (results not
shown). The CP solution did not increase the chromosomal loss or
chromosomal breakage before irradiation in comparison with the ne-
gative control (p < 0.01). However, after 60 min and 90 min of light
exposition the number of micronuclei in the V79 cells significantly in-
creased (Fig. 6). The addition of p-CA and BP-4 visibly reduced these
changes in samples with CP after 60 min of irradiation, while no DNA-
protective effect of the excipients was observed in samples irradiated
30 min longer. Moreover, CP samples after 60 min and 90 min of irra-
diation reduced the cell proliferation to 50% and 35% in comparison
with the negative control and the dark control. This effect was sig-
nificantly inhibited by both the tested excipients in the samples after
60 min and 90 min of irradiation. One of the tested antibiotics (CL)
exhibited strong clastogenic activity even before the light exposition.
This result was slightly reduced in solutions with p-CA stored in the
dark but was not observed in any samples after the irradiation (results
not shown here).
4. Discussion
The increasing photolability of all the tested FQ (CP → LM →
FR = CL) correlated with the increasing number of their halogen
atoms. The results of the mass spectrometry analysis showed that the
transformations of CP occurred mainly through hydroxylation (→348),
defluorination and hydroxylation (→346) and oxidative breakdown of
the piperazine ring (→ 316 and → 330). The same most abundant MS/
MS product ions of the transformation products of CP were also iden-
tified by Haddad and Kummerer [22]. Ions detected in our samples of
the irradiated LM were also described by Budai et al. [23]. LM in water
degraded photochemically to seven compounds with m/z values of 350,
348, 336, 332, 316, 308 and 288. According to the authors, their MS/
MS data indicated the possibility of the loss of 2 and 4 hydrogens
leading to the formation of very reactive one and two double bounds in
the piperazine ring. The next two photoproducts were probably formed
by subsequent demethylation (→ 336) and defluorination (→332).
The genotoxic abilities of the CP photolytic decomposition products
208
Journal of Photochemistry & Photobiology, B: Biology 178 (2018) 201–210
Fig. 5. The results of MTT assay with the V79 cell line.
The y-axis represents the cell viability, while the x-axis
represents the initial concentration of CP. +UV describes
the preirradiated solutions.
described in this work are consistent with other published results [24].
Itoh et al. [25] observed DNA single strand-breaks in an in vitro comet
assay with WTK-1 cells. These results increased the concentration and
irradiation time dependence. Garcia-Kaufer and Haddad [26] found a
statistically significant increase in the genotoxic potential of CP samples
after 32 min of irradiation of human liver cells. Moreover, the MN
frequency declined for samples irradiated longer, which is similar to
our results. In contrast, in [26], CP before irradiation induced a sig-
nificant formation of micronucleated cells with the highest induction at
10 μM. No analogical observations were described in our project. The
main explanation for differences between our results and this work
could be the use of two various cell lines: a normal one and cancer cells.
Our cells were Chinese hamster lung fibroblasts while Garcia-Kaufer
et al. [26] used human liver tumour cell line HepG2. Other significant
differences were the longer incubation period (24 h, while in our ex-
periments it was 6 h) and lower initial concentration of ciprofloxacin
(0.4 μg/mL, while in our experiment it was 62.5 μg/mL).
The lack of antiphotomutagenic effect of p-CA in samples irradiated
for 90 min before micronucleus assay could not be clearly explained on
the basis of our results. The mass spectrometry analysis was performed
for CP and p-CA mixture irradiated up to 60 min: p-CA significantly
decelerated the photodegradation process of CP and the formation of
photoproducts that could reveal a genotoxic activity. It could be pos-
sible that after 90 min of irradiation the amount of that compounds in
mixtures increased similar to samples irradiated for 60 min without
excipients and their genotoxic effects became visible.
Differences in the observed photogenotoxic potential of LM may
vary on the selected irradiation and bioassay protocols. Marrot et al.
[10] described a significant damage of DNA in the comet assay with
cultured human cells after exposition to 5 μM of this antibiotic and
15 min of UVA irradiation, while in a test with S. cerevisiae D7, no
photogenotoxicity of LM was observed. Perruca et al. [27] described LM
as the most efficient molecule in hypoxia. LM underwent a heterolysis
of the C8eF bond to yield the corresponding triplet aryl cation that has
been shown to react with DNA nucleotides, particularly with purine
bases. According to Soldevila et al. [28] 70% of the DNA cleavage in the
test with supercoiled circular DNA was observed for LM (100 μM) after
a light dose of 240 mJ/cm2. This value was remarkably higher than the
15% obtained for monohalogenated CP.
An increase in the number of cells with chromosomal loss or chro-
mosomal breakage in the micronucleus assay after exposition to CL was
also observed in other studies. Snyder et al. [24], after 3 h of exposition
for non-irradiated clinafloxacin (40 μg/mL and 60 μg/mL), observed an
increased number of micronuclei (8.8% and 14%, respectively) in the
V79 cells. Williams and Brunnemann [29] explained this effect by CL's
ability to inhibit the eukaryotic topoisomerase II enzyme. They noted
similar results for LM in contrast to our research.
The effectiveness of p-CA as an antimutagenic compound was also
A. Zgadzaj et al. Journal of Photochemistry & Photobiology, B: Biology 178 (2018) 201–210

Fig. 6. The results of the micronucleus test with the V79 cells. NK – negative control. The stars illustrate statistically significant differences (p < 0.01) between the NK (black stars) or
differences between the CP solutions and mixtures of the CP with p-CA and benzophenone-4 (BP4) (white stars). The x-axis represents the time of irradiation.

evaluated with a bacterial assay by Ferguson et al. [13]. In the Ames
test on S. typhimurium TA98 and TA102 strains, p-CA showed the ability
to reduce the genotoxic effect of three mutagens: bleomycin, hydrogen
peroxide and 2-amino-3-methylimidazo[4,5-f]quinoline. The anti-
mutagenic effect was more strongly marked in a pre-incubation pro-
tocol. Phenolic acids have strong antioxidant activity because of a
phenolic ring and an organic carboxylic acid function. The free radical
scavenging of these compounds depends on the number of hydroxyl
groups and their position in the aromatic ring. Therefore, hydroxylated
cinnamates are more effective antioxidants than their benzoate
counterparts [30]. Hydroxycinnamic acids have several hydrogen
atoms available for abstraction and can remain stable through the de-
localization of electrons across the conjugated ring and sidechains.
These hydrogen atoms may form hydrogen bonds with nitrogen atoms
such as 2-amino-3-methylimidazo[4,5-f]quinoline, reducing their affi-
nity to the DNA [13]. Further, the photoprotective abilities of p-CA
were confirmed as well by Lodovici et al. [12]. But, the anti-
photomutagenic effect of this compound in mixtures with widely used
photolabile ocular antibiotics was assessed for the first time. Promising
results obtained in compositions with CP may be a basis for further

209
A. Zgadzaj et al.
research. Nevertheless, the increase in the DNA damage potential in
mixtures with p-CA and LM and FR in our bacterial bioassay shows the
importance of the safety evaluation of such innovative combinations. As
our MS/MS study suggests, this observation in mixtures with LM may
be connected with the altered pathways of LM degradation.
The photoprotective effect of BP-4 that we observed is consistent
with BP-4′s ability as a photoprotective water-soluble excipient de-
scribed by Yang et al. [31] in pharmaceutical formulations with arbutin
and hydroquinone used for bleaching hyperpigmented lesions.
The next step of the research on formulations tested in this work
could be the evaluation of their effectiveness and safety in the realistic
exposure conditions. Moreover, other irradiation schemes then de-
scribed in this paper could provide a valuable date on the effectiveness
of tested combinations, like preincubation before irradiation or con-
comitant light exposure. The application of bioassays with the re-
constructed human cornea like epithelium or a set of in vivo tests would
allow to include such factors as ocular tolerance or a drug distribution
in the eye tissues.
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.jphotobiol.2017.11.011.
Funding sources
The project was supported by the Medical University of Warsaw,
Faculty of Pharmacy from the Grant for Young Scientists managed by
Anna Zgadzaj in 2016–2017 (FW14/PM2/16).
References
[1] EMA ICH Guideline S10, Guidance for photosafety evaluation of pharmaceutical,
STEP (2015) 5.
[2] P.C. Appelbaum, P.A. Hunter, The fluoroquinolone antibacterials: past, present and
future perspectives, Int. J. Antimicrob. Agents 16 (2000) 5–15.
[3] M. Thompson, Ocular toxicity of fluoroquinolones, Clin. Exp. Ophthalmol. 35
(2007) 566–577.
[4] N. Umezawa, K. Arakane, A. Ryu, S. Mashiko, M. Hirobe, T. Nagano, Participation
of reactive oxygen species in phototoxicity induced by quinolone antibacterial
agents, Arch. Biochem. Biophys. 342 (1997) 275–281.
[5] Y. Seto, R. Inoue, M. Ochi, G. Gandy, S. Yamada, S. Onoue, Combined use of in vitro
phototoxic assessment and cassette dosing pharmacokinetic study for phototoxicity
characterization of fluoroquinolones, AAPS J. 13 (2011) 482–492.
[6] T. Yamamoto, Y. Tsurumaki, M. Takei, M. Hosaka, Y. Oomori, In vitro method for
prediction of the phototoxic potentials of fluoroquinolones, Toxicol. in Vitro 15
(2001) 721–727.
[7] D. Trisciuoglio, E. Krasnowska, A. Maggi, R. Pozzi, T. Parasassi, O. Sapora,
Phototoxic effect of fluoroquinolones on two human cell lines, Toxicol. in Vitro 16
(2002) 449–456.
[8] S.J. Bulera, J.C. Theiss, T.A. Festerling, F.A. de la Iglesia, In vitro photogenotoxic
activity of clinafloxacin: a paradigm predicting photocarcinogenicity, Toxicol. Appl.
Pharmacol. 156 (1999) 222–230.
[9] G. Miolo, G. Viola, D. Vedaldi, F. Dall'Acqua, A. Fravolini, O. Tabarrini,
V. Cacchetti, In vitro phototoxic properties of new 6-desfluo and 6-fluoro-8-me-
thylquinolones, Toxicol. in Vitro 16 (2002) 683–693.
[10] L. Marrot, J.P. Belaidi, C. Chaubo, J.R. Meunier, P. Perez, C. Agapakis-Causse,
Fluoroquinolones as chemical tools to define a strategy for photogenotoxicity in
vitro assessment, Toxicol. in Vitro 15 (2001) 131–142.
[11] B. Zhao, C.F. Chignell, M. Rammal, F. Smith, M.G. Hamilton, U.P. Andley,
J.E. Roberts, Detection and prevention of ocular phototoxicity of ciprofloxacin and
other fluoroquinolone antibiotics, J. Photochem. Photobiol. 86 (2010) 798–805.
210
Journal of Photochemistry & Photobiology, B: Biology 178 (2018) 201–210
[12] M. Lodovici, S. Caldini, L. Morbidelli, V. Akpan, M. Ziche, P. Dolara, Protective
effect of 4-coumaric acid from UVB ray damage in the rabbit eye, Toxicology 255
(2009) 1–5.
[13] L.R. Ferguson, I.F. Lim, A.E. Pearson, J. Ralph, P.J. Harris, Bacterial antimutagen-
esis by hydroxycinnamic acids from plant cell walls, Mutat. Res. 542 (2003) 49–58.
[14] G. Ioele, M. De Luca, L. Tavano, G. Ragno, The difficulties for a photolabile drug in
topical formulations: the case of diclofenac, Int. J. Pharm. 25 (2014) 284–290.
[15] C. Cejka, J. Luyckx, T. Ardan, J. Platenik, J. Sirc, J. Michalek, J. Cejkova, The effect
of actinoquinol with hyaluronic acid in eye drops on the optical properties and
oxidative damage of the rabbit cornea irradiated with UVB rays, Photochem.
Photobiol. 86 (2010) 1294–1306.
[16] C. Cejka, J. Luyckx, J. Cejkova Central, Corneal thickness considered an index of
corneal hydration of the UVB irradiated rabbit cornea as influenced by UVB ab-
sorber, Physiol. Res. 61 (2012) 299-30.
[17] W.K. Kek, J. Miller, E. Rawson-Lax, C.G. Wilson, D. Uttamchandani, In situ mea-
surement of spectral changes in the anterior eye following application of ultraviolet-
absorbing compounds, Eur. J. Pharm. Biopharm. 75 (2010) 200–205.
[18] A. Zgadzaj, A. Skrzypczak, I. Welenc, A. Ługowska, A. Parzonko, E. Siedlecka,
S. Sommer, K. Sikorska, G. Nałęcz-Jawecki, Evaluation of photodegradation, pho-
totoxicity and photogenotoxicity of ofloxacin in ointments with sunscreens and in
solutions, J. Photochem. Photobiol. B 144 (2015) 76–84.
[19] U. Hubicka, P. Żmudzki, P. Talik, B. Żuromska-Witek, J. Krzek, Photodegradation
assessment of ciprofloxacin, moxifloxacin, norfloxacin and ofloxacin in the presence
of excipients from tablets by UPLC-MS/MS and DS, Chem. Cent. J. 7 (2013) 133.
[20] R. Solomon, E.D. Donnenfeld, H.D. Perry, R.W. Snyder, W.C. Nedrud, J. Stein,
A. Bloom, Penetration of topically applied gatifloxacin 0.3%, moxifloxacin 0.5%,
and ciprofloxacin 0.3% into the aqueous humor, Ophthalmology 112 (2005)
466–469.
[21] K. Tornianien, S. Tammilehto, V. Ulvi, The effect of pH, buffer type and drug
concentration on the photodegradation of ciprofloxacin, Int. J. Pharm. 132 (1996)
53–61.
[22] T. Haddad, K. Kummerer, Characterization of photo-transformation products of the
antibiotic drug ciprofloxacin with liquid chromatography–tandem mass spectro-
metry in combination with accurate mass determination using an LTQ-Orbitrap,
Chemosphere 115 (2014) 40–46.
[23] M. Budai, P. Gróf, A. Zimmer, K. Pápai, I. Klebovich, K. Ludányi, UV light induced
photodegradation of liposome encapsulated fluoroquinolones: an MS study, J.
Photochem. Photobiol. A Chem. 198 (2008) 268–273.
[24] R.D. Snyder, C.S. Cooper, Phototoxicity of fluoroquinolones in Chinese hamster V79
cells: dependency on active topoisomerase II, Photochem. Photobiol. 69 (1999)
288–293.
[25] T. Itoh, K. Mitsumori, S. Kawaguchi, Y.F. Sasaki, Genotoxic potential of quinolone
antimicrobials in the in vitro comet assay and micronucleus test, Mutat. Res. 603
(2006) 135–144.
[26] M. Garcia-Kaufer, T. Haddad, M. Bergheim, R. Gminski, P. Gupta, N. Mathur,
K. Kummerer, V. Mersch-Sundermann, Genotoxic effect of ciprofloxacin during
photolytic decomposition monitored by the in vitro micronucleus test (MNvit) in
HepG2 cells, Environ. Sci. Pollut. Res. J. 19 (2012) 1719–1727.
[27] P. Perucca, M. Savio, O. Cazzalini, R. Mocchi, C. Maccario, S. Sommatis, D. Ferraro,
R. Pizzala, L. Pretali, E. Fasani, A. Albini, L.A. Stivala, Structure–activity relation-
ship and role of oxygen in the potential antitumour activity of fluoroquinolones in
human epithelial cancer cells, J. Photochem. Photobiol. B Biol. 140 (2014) 57–68.
[28] S. Soldevila, M.C. Cuquerella, V. Lhiaubet-Vallet, R. Edge, F. Bosca, Seeking the
mechanism responsible or fluoroquinolone photomutagenicity: a pulse radiolysis,
steady-state, and laser flash photolysis study, Free Radic. Biol. Med. 67 (2014)
417–425.
[29] G.M. Williams, K.D. Brunnemann, D.J. Smart, D. Molina, A.M. Jefferey, J. Duan,
N. Krebsfaenger, A. Kampkoetter, G. Schmuck, Relationship of cellular topoi-
somerase IIα inhibition to cytotoxicity and published genotoxicity of fluor-
oquinolone antibiotics in V79 cells, Chem. Biol. Interact. 203 (2013) 386–390.
[30] P. Sunkireddy, S.N. Jha, J.N. Kawran, S.C. Yadav, Natural antioxidant biomolecules
promises future nanomedicine based therapy for cataract, Colloids Surf. B:
Biointerfaces 112 (2013) 554–562.
[31] Ch. Yang, N. Chang, Y. Chen, S. Lee, P. Lin, Ch. Lin, Comparative study on the
photostability of arbutin and deoxy arbutin: sensitivity to ultraviolet radiation and
enhanced photostability by the water-soluble sunscreen, benzophenone-4, Biosci.
Biotechnol. Biochem. 77 (2013) 1127–1130.