Synergistic Photobactericidal Activity Based on Ultraviolet-A Irradiation and Ferulic Acid Derivatives

Article abstract of DOI:10.1111/php.12507

Ultraviolet-A (UV-A)-mediated bactericidal activity was enhanced by combined treatment with trans-ferulic acid (trans-FA, compound 1) or its derivatives. Derivative compounds 4 and 10 contain a phenyl group or an l-tyrosine HCl tert-butyl ester, respectiv

Full text of DOI:10.1111/php.12507

Photochemistry and Photobiology, 2015, 91: 1422–1428
Synergistic Photobactericidal Activity Based on Ultraviolet-A Irradiation
and Ferulic Acid Derivatives
1
2
1
Akihiro Shirai* , Masato Kajiura and Takeshi Omasa
1
Department of Biological Science and Technology, Biosystems Engineering, Institute of Technology and Science,
Tokushima University, Tokushima, Japan
Department of Biological Science and Technology, Faculty of Engineering, Tokushima University, Tokushima, Japan
2
Received 24 May 2015, accepted 29 July 2015, DOI: 10.1111/php.12507
ABSTRACT
0
Photosensitizers are initially transformed from the S ground
state to the S excited state by a pÀp* and/or a nÀp* transition
1
Ultraviolet-A (UV-A)-mediated bactericidal activity was
enhanced by combined treatment with trans-ferulic acid
by absorption of light at a particular wavelength. Eventually,
interaction between the excited state and molecular oxygen
results in the production of reactive oxygen species (ROS) via
electron or hydrogen transfer reactions and singlet oxygen via
energy transfer (7). Extensive research has revealed that the
antimicrobial activity of photosensitizers involves the generation
of ROS after light excitation (7).
(
trans-FA, compound 1) or its derivatives. Derivative com-
pounds 4 and 10 contain a phenyl group or an L-tyrosine
HCl tert-butyl ester, respectively, linked to the carboxyl
group of trans-FA. Of the three compounds, 10 exhibited the
highest synergistic activity in a photobactericidal assay based
on treating Escherichia coli with a derivative compound and
UV-A irradiation (wavelength 350–385 nm). Inactivation of
viable cells at a 4.9 J cm UV-A fluence increased from 1.90
to 5.19 logs in the presence of 10 (100 lM); a 4.95-log inacti-
vation was achieved with 10 (5 lM) and a 7.4 J cm UV-A
fluence. Addition of antioxidants significantly suppressed
photosynergistic bactericidal activity, suggesting that reactive
oxygen species (ROS) are involved in the combined bacterici-
dal mechanism. Flow cytometry revealed that combined
treatment with UV-A and compound 10, which showed the
highest photobactericidal activity, generates an excess of
oxidative radicals in bacterial cells. The bactericidal activity
of compound 10 may be due to electrostatic interaction
between the molecule’s cationic moiety and the cell surface,
followed by amplification of ROS generation in the cells.
Curcumin, which is composed of two feruloyl chromophores,
is an effective photosensitizer for the inactivation of Candida
albicans and Staphylococcus epidermidis (8,9). Chignell et al.
À2
(10) reported that the photoactivation of curcumin involves the
À2
formation of singlet oxygen and superoxide anion radicals. They
also reported that a curcumin half structure composed of one fer-
uloyl chromophore is also capable of producing ROS following
irradiation with UV-A light. We hypothesized that the bacterici-
dal efficacy of UV-A irradiation, which is dependent on the gen-
eration of ROS such as hydrogen peroxide and hydroxyl radicals
(11), could be improved by combined treatment with an appro-
priate photosensitizer composed of 3-(4-hydroxyphenyl)-2-prope-
noic acid, a close analog of FA. To explore the potential for
synergistically enhanced UV-A/photosensitizer-mediated bacteri-
cidal activity, we previously investigated the efficacy of com-
bined treatment with UV-A irradiation and a CA derivative,
phenyl 3-(4-hydroxyphenyl)-2-propenoate possessing part of the
chromophore of half-curcumin. As expected, the combination
treatment showed higher bactericidal activity than UV-A irradia-
tion alone (A. Shirai, unpublished), suggesting that the derivative
plays a role in synergistic ROS generation under UV-A irradia-
tion. The potential role of the derivative in synergistic ROS gen-
eration is supported by data indicating that the absorption spectra
of half-curcumin and the CA derivative in protic solvents are
very similar, with a maximum absorbance wavelength (kmax) of
INTRODUCTION
Ferulic acid (FA) is a naturally occurring phenolic acid found in
wheat, barley, potatoes, tomatoes and many vegetables (1,2). FA
usually exists as the trans-isomer. Absorption of ultraviolet (UV)
light by trans-FA produces a mixture of cis-trans isomers. For
example, in methyl alcohol, a portion of the trans-FA molecules
are converted to the cis-isomer upon exposure to UV-A at a
wavelength of 365 nm (3). Similarly, trans-p-coumaric acid
(
CA), which forms part of the basic structure of trans-FA, is also
3
35 nm for half-curcumin (10) and 313 nm for the CA deriva-
isomerized by exposure to UV-A in the range 320–400 nm (4).
Photoisomerization of the basic structure as a result of absorption
of light energy during UV irradiation can be initiated by the gen-
eration of phenoxy radicals followed by binding of a hydrogen
atom (5) and by excitation to the S
sition state (6). The photoactivation mechanism of the latter pho-
toisomerization reaction is similar to that of photosensitizers.
tive (our data).
In this study, the half-curcumin analog trans-FA and several
trans-FA derivatives were investigated with respect to their
photobactericidal efficacy in combined treatment with UV-A
irradiation (wavelength range of 350–385 nm produced using a
light-emitting diode [LED] source). A distinct property of the
parental trans-FA compound is that its UV absorption properties
are almost identical to those of the CA derivative described
above. Our data suggest that trans-FA may serve as a potent
1
state through the pÀp* tran-
*Corresponding author email: a.shirai@tokushima-u.ac.jp (Akihiro Shirai)
© 2015 The American Society of Photobiology
1
422

Photochemistry and Photobiology, 2015, 91 1423
enhancer of UV-A-mediated sterilization. Chemically modified
derivatives of trans-FA show promise as photosynergistic com-
pounds. To completely inactivate bacteria using UV-A irradiation
250
2
00
alone, a high fluence rate is required; for example, a fluence of
À2
3
15 J cm
for a 75-min incubation is required for complete
inactivation of Escherichia coli (11). A number of researchers
have investigated the efficacy of combining UV irradiation with
photosensitizers such as riboflavin (12) or antimicrobial agents
such as quaternary pyridinium salt (13). However, there have
been (to the best of our knowledge) few reports regarding
improved UV-A-mediated bactericidal activity triggered in the
presence of trans-FA. In this report, we demonstrate enhanced
bactericidal efficacy through combined treatment with UV-A irra-
diation and trans-FA or several of its derivatives. Our results
show that the photobactericidal activity is UV-A fluence depen-
dent. The high bactericidal activity associated with the combined
treatment is also dependent on enhanced generation of ROS.
150
1
00
50
0
250 300 350 400 450 500 550 600
Wavelength (nm)
Figure 1. Emission spectrum of UV-A-LED as used in this study, which
exhibits
4
a
spectral maximum at 365 nm. The fluence rate was
À2
.09 mW cm at a distance of 65 mm between the illumination source
and the UV meter. LED, light-emitting diode.
MATERIALS AND METHODS
Chemistry. All chemicals used for compound synthesis were reagent
grade commercially available materials and were used without further
purification. Reagents were purchased from Tokyo Chemical Industry
Co., Ltd., (Tokyo, Japan), Wako Pure Chemical Industries, Ltd. (Osaka,
Japan), Sigma-Aldrich Co., LLC (St. Louis, MO), Kanto Chemical Co.,
Inc. (Tokyo, Japan) and Merck KGaA (Darmstadt, Germany). For synthe-
ses requiring water-free conditions, solvents were dried using type-3A
molecular sieves. Granules of sodium hydroxide were added to pyridines
to ensure dryness. All reactions were monitored by thin-layer chromatog-
raphy (TLC) using normal-phase thin-layer plates (Merck silica gel
the propenoic group double bond. The purity of the synthesized com-
pounds, as demonstrated by TLC and HPLC, was 97% or higher. These
results confirmed that the synthesized trans-FA derivatives were the
intended compounds.
Measurement of UV and fluorescence spectra. Solutions of com-
pounds 1, 4 and 10 were prepared at a concentration of 50 lM with 0.4%
(v/v) DMSO. UV spectra were measured using a 1-cm path-length
cuvette and a U-3300 spectrophotometer (Hitachi Ltd., Tokyo, Japan).
Fluorescence spectra at an excitation of 340 nm were measured using a
96-well culture plate with black wells (Corning Inc., Corning, NY) and a
microplate reader (Infinite M200; Tecan Japan Co., Ltd, Kanagawa,
Japan). UV and fluorescence spectra are shown in Figure S2.
60 F254; thickness, 0.25 mm). Nuclear magnetic resonance (NMR) spec-
tra were recorded using a JEM-EX 400 spectrometer (JOEL, Tokyo,
Japan), with tetramethylsilane as an internal standard. The purity of com-
pounds was assessed by TLC and by determination of melting point
using a micromelting apparatus (Thermo Fisher Scientific, Waltham,
MA). Prior to mass analysis, compound purity was also confirmed by
HPLC with UV detection using a Hitachi L-2000 HPLC system (Hitachi
High-Technologies Corporation, Tokyo, Japan) equipped with a COS-
MOSIL column (Cholester, 4.6 mm 9 150 mm; Nacalai Tesque, Inc.,
Kyoto, Japan). Mass spectra were recorded on an Acquity UPLC-LCT
Premier liquid chromatography–mass spectrometry (LC-MS) system (Ni-
hon Waters K.K., Tokyo, Japan). The crude compounds were purified by
flash column chromatography on silica gel (silica gel 120, mesh 70-230
spherical; Nacalai Tesque, Inc.). Except for synthesis reactions performed
in aqueous solution, all reactions were conducted under a stream of nitro-
gen.
UV-A source and irradiation. The UV light source was an LED
device, as previously described (13); the device is constructed with 27
LED elements (NCSU033A; Nichia Corp., Anan, Japan) arranged in a
three-column-by-nine-row array on a basal plate of 100 mm 9 300 mm.
The LED has a radiation angle of about 120° as the full width at half
maximum. The UV LED was irradiated downward to the top of a
48-well culture plate (AGC Tecno Glass Co., Ltd., Tokyo, Japan). The
irradiation distance between the LED device and the surface of bacterial
suspensions was set at 65 mm. The spectrum and fluence rate were
measured at the distance (65 mm) (Fig. 1). The fluence rate over the
wavelength range 350–385 nm (peak wavelength, 365 nm) emitted by
À2
this device was 4.09 mW cm . The total fluence for a 10-, 20- and
30-min irradiation, calculated based on fluence rate and exposure time,
À2
Synthesis. As shown in Figure S1, two trans-FA derivatives (com-
pounds 4 and 10) were synthesized by modifying previously described
methods (14–18).
was 2.5, 4.9 and 7.4 J cm , respectively.
Determination of bactericidal activity. Bactericidal activity of com-
pounds 1, 4 and 10 (see Fig. 2 for compound structures) against E. coli
NITE Biological Resource Center (NBRC) 12713 was assessed by count-
ing the colony forming units (CFUs) appearing after treatment of a sus-
Details regarding synthesis procedures and physicochemical properties
e.g. appearance, yield, m.p., R and results of NMR and mass spectral
f
(
6
À1
analyses) are summarized in the Supporting Information. The identity of
pension of bacteria (2.0 9 10 cells mL ) prepared as described in a
previous report (13).
1
each synthesized compound was confirmed by H-NMR and MS. The
NMR spectra showed a trans (E) configuration, as demonstrated by the
coupling constants (ranging from 15.6 to 16.3 Hz) of the a-H and b-H of
An aliquot of bacterial suspension (0.1 mL) was added to each well
of a 48-well culture plate already containing 0.01 mL of test compound
Figure 2. Chemical structures of compounds 1, 4 and 10 used in the combination photosterilization process.

1
424 Akihiro Shirai et al.
(
10 mM in 80% DMSO) and 0.89 mL of sterile water per well. The plate
0
1
2
3
0
-1
-2
-3
-4
-5
-6
was then incubated for 10, 20 and 30 min (UV-A fluences of 2.5, 4.9
and 7.4 J cm ) at 30°C with UV-A irradiation or with no irradiation (in
À2
*
-
-
-
a dark room). After incubation, a 0.15 mL aliquot from each well was
diluted 10-fold with SCDLP broth (Nihon Pharmaceutical Co., Ltd.,
Tokyo, Japan) followed by serial 10-fold dilutions with 0.8% (w/v) phys-
iological saline containing 0.7% (w/w) Tween 80. Each serial dilution
was plated on an SCDLP agar (Nihon Pharmaceutical Co., Ltd.) plate,
which was incubated at 37°C for 48 h. The number of CFUs was then
determined and used to calculate bactericidal activity, expressed as the
log survival ratio according to the following equation: log survival
ratio = log (N
before bactericidal treatment and N
treatment for time t.
In order to investigate the effect of ROS on bacterial inactivation in
combined treatment with compounds 1, 4 and 10 and UV-A irradiation,
catalase (from bovine liver; Wako Pure Chemical Industries, Ltd.) or
N-acetylcysteine (NAC; Tokyo Chemical Industry Co., Ltd.) was dis-
solved in Milli-Q water and then added to the bacterial suspension in a
**
-4
-5
(A)
(B)
*
*
-6
(
)
0
1
2
3
4
5
6
7
8
1 2 4
0 3 6 7 8
5
t
/N
0
), where N
0
represents the number of initial CFUs
represents the number of CFUs after
2
2
Total fluence (J/cm )
Total fluence (J/cm )
t
0
0
-1
-2
-3
-4
-5
-6
*
-1
-2
*
*
-3
**
4
8-well culture plate. The bacterial suspension was preincubated at 37°C
-4
-5
-6
for 30 min before the addition of test compound and UV-A irradiation.
Substrates for catalase or NAC were added to the bacterial suspensions at
final concentrations of 1 kU per well or 1 mM, respectively. As a control
for catalase, albumin (from bovine serum, Cohn fraction V, pH 7.0;
Wako Pure Chemical Industries, Ltd.) was added to the bacterial suspen-
sions at a final concentration equal to the weight of catalase added. Pre-
liminary experiments (data not shown) demonstrated that incubation with
the respective substrates (1 h, 37°C, in the dark) had no significant effect
on viability.
Flow cytometry analysis of intracellular ROS production. The relative
level of intracellular ROS production was determined using flow cytome-
try according to a previously described method (13). Hydroxyphenyl flu-
orescein (measurement wavelength,
Sekisui Medical Co., Ltd., Tokyo, Japan) was used as a specific fluores-
cence probe for hydroxyl radicals. Specific probe-labeled E. coli was pre-
**
(C)
(D)
*
*
**
(
)
(
)
0
1
2
3
4
5
6
7
8
1 2 4
0 3 6 7 8
5
2
2
Total fluence (J/cm )
Total fluence (J/cm )
Figure 3. UV-A fluence-dependent change in inactivation of Escherichia
coli after treatment: no additive (A), compound 1 (B), 4 (C) or 10 (D),
with UV-A irradiation (unshaded) or 30-min incubation without UV-A
shaded). Test compound was added at a concentration of 100 lM. Cells
were suspended at an initial density of 10 CFUs mL . Data are pre-
(
6
À1
sented as mean Æ SD (n = 3). Significant differences (*P < 0.05 and
k
ex = 490 nm,
kem = 515 nm;
*
*P < 0.01) are indicated relative to viability after irradiation treatment
at each UV-A fluence without additive. Samples for which survival was
10 CFUs mL (lower limit of detection) are noted in parentheses.
À1
<
pared by gently shaking (75 rpm)
a
suspension of bacteria
2.0 9 10 cells mL ) in D-PBS(À) (Nacalai Tesque, Inc.) containing
0 lM specific fluorescent probe for 80 min at 37°C in the dark. The
7
À1
(
3
À2
compound and a UV-A fluence of 7.4 J cm , representing a
6-log reduction in survival (noted by data points enclosed in
parentheses in the figures). In the presence of compound 10, the
cells were then collected by centrifugation (6570 g for 3 min at 4°C),
washed three times with D-PBS(À), adjusted to an optical density at
6
60 nm of 0.02 (UV-1700 spectrophotometer; Shimadzu Corp., Kyoto,
Japan), and distributed at 495 lL per well in a 48-well culture plate.
À2
decrease in viable cell count at a UV-A fluence of 4.9 J cm
Each well was supplemented with 5 lL of 80% DMSO with or without
À2
represented a 5.19-log reduction in survival (Fig. 3D), whereas
reductions of 1.90, 3.58 and 3.42 logs were observed at the same
UV-A fluence for samples treated with UV-A alone, compound
1
0 mM test compound (1, 4 or 10) and irradiated with 7.4 J cm of
UV-A at 30°C. The resulting suspensions were treated using the same
method described in a previous report (13) and analyzed using a FACS
Verse flow cytometer (Becton Dickinson, Franklin Lakes, NJ). The level
of ROS generated by oxidative processes was expressed in arbitrary units
1
and compound 4, respectively. In the absence of UV-A irradia-
tion, the three test compounds exhibited minimal bactericidal
activity at 100 lM, yielding 0.78-log or lower reductions in sur-
vival for a 30-min treatment (the incubation time necessary to
provide a 7.4 J cm UV-A fluence).
The concentration of compound 10 lethal to 10 CFUs mL
of E. coli was investigated at a UV-A fluence of 7.4 J cm
(
a.u.) normalized to the mean fluorescence value of control samples incu-
bated in the dark for 30 min without test compound.
Statistical analysis. All experiments were performed as three indepen-
dent procedures, and results are presented as the mean and standard devi-
ation. Inferential analysis was performed using a two-tailed, unpaired
Student’s t-test. Ps < 0.05 were considered to be indicative of signifi-
cance.
À2
6
À1
À2
(
Fig. 4). A concentration of 10 lM or higher was required to
À1
reduce the viable cell count to <10 CFUs mL . A concentration
of 5 lM produced a 4.95-log reduction in survival, which was a
significantly greater reduction in survival than observed with
UV-A treatment alone (P < 0.05).
RESULTS
Bactericidal activity
Bactericidal activity against E. coli was investigated by treating
6
À1
bacteria at an initial density of 10 CFUs mL with a combina-
tion of trans-FA 1 or its derivatives (4 and 10) and UV-A irradi-
ation. Test compound was added at a concentration of 100 lM.
Irradiation with UV-A (7.4 J cm ) in the absence of test com-
pound resulted in a 3.60-log reduction in the viable cell count
Effect of ROS on inactivation of Escherichia coli
When the test compounds were present, we predicted that hydro-
gen peroxide would be produced upon UV-A irradiation, thus
enhancing the efficacy of photosterilization. The role of hydrogen
peroxide was investigated in E. coli treated with a combination
of compound 1, 4 or 10 (100 lM) and UV-A irradiation.
Figure 5 shows that the addition of catalase at 1 kU per well sig-
nificantly attenuated the bactericidal effect in samples treated
with compound 1 (change from 4.01- to 0.61-log reduction in
À2
(
Fig. 3A). As shown in Fig. 3(B,C), irradiation in the presence
of compound 1 or 4 resulted in a significant decrease in viable
cell count (P < 0.05 and 0.01) compared with irradiation alone
À2
at various fluences (2.5, 4.9 and 7.4 J cm ). The viable count
À1
was reduced to <10 CFUs mL in combined treatment with test

Photochemistry and Photobiology, 2015, 91 1425
1
1
1
1
.8
.6
.4
.2
0
1
2
3
*
*
-
-
-
-
-
-
*
1
.0
4
5
6
0.8
*
0
Time (min) 30
30
0
30
0
30
0
30
30
30
30
*
*
**
(
)
(
)
Total fluence
0
7.4
7.4
7.4
7.4
2
(
J/cm )
0
10
20
30
40
50
Compound
–
1
4
10
–
1
4
10
Compound 10 (µM)
Figure 6. Changes in intracellular fluorescence level in Escherichia coli
after UV-A irradiation with or without compound 1, 4 or 10 (100 lM).
The fluorescence level for each treatment is expressed in arbitrary units
(a.u.) of mean fluorescence as determined by flow cytometry analysis and
compared to the value for control samples incubated without test com-
pound in the dark (30-min incubation). Data are presented as mean Æ SD
Figure 4. Dose-dependent effect of compound 10 on inactivation of
À2
Escherichia coli treated with a total fluence of 7.4 J cm . Cells were
6
À1
suspended at an initial density of 10 CFUs mL . Data are presented as
mean Æ SD (n = 3). Significant differences (*P < 0.05 and **P < 0.01)
are indicated relative to viability after irradiation treatment with the UV-
A fluence without additive.
(n = 3). *P < 0.05, **P < 0.01.
activity by 1.21 logs (P < 0.01), compared with a 3.07-log inac-
tivation without NAC. Attenuation of bactericidal activity by the
addition of NAC was not observed with combined treatment with
compounds (1 and 4) and UV-A irradiation (4.9 J cm ) (data
not shown).
À2
survival with catalase addition at a UV-A fluence of 4.9 J cm
;
P < 0.01). In contrast, exposure to an equivalent weight concen-
tration of nonenzymatic albumin had no effect on the bactericidal
activity (P < 0.01). In combined treatment with compound 4 at
the same UV-A fluence, the bactericidal activity was significantly
decreased in the presence of catalase, with the reduction in sur-
vival declining from 3.40 to 0.41 logs. Addition of albumin
instead of catalase partially restored the bactericidal activity
À2
UV-A-induced ROS generation
The relative level of intracellular ROS generated was estimated
(
1.08-log reduction in survival). The reduction in survival after
by flow cytometry analysis using E. coli cells treated with a
À2
combined treatment with UV-A irradiation (2.5 J cm ) and
compound 10 (3.07 logs) was almost the same as that observed
with catalase addition at 1 and even 5 kU per well (data not
shown). Addition of NAC at a concentration of 1 mM as a cell-
permeating hydroxyl radical quencher attenuated the bactericidal
À2
combination of 100 lM compound 1, 4 or 10 and a 7.4-J cm
UV-A fluence. As shown in Fig. 6, incubating bacteria for
3
0 min with compound 1 or 4 in the dark produced only a mini-
mal increase in fluorescence intensity compared with control
samples incubated without test compound. Under the same con-
ditions, incubating bacteria with compound 10 induced a 1.19-
fold increase in fluorescence intensity. No significant increase in
fluorescence intensity was observed in samples exposed to UV-A
in the absence or presence of compound 1 or 4. In contrast,
combined compound 10/UV-A-irradiation treatment induced a
5
*
*
4
3
2
1
**
*
*
*
*
1
.56-fold increase in fluorescence intensity as compared with the
control.
*
*
DISCUSSION
Photodynamic inactivation of microbes is based on combined
treatment with photosensitizers and irradiation with UV or visible
light. Curcumin, which is composed of two feruloyl chro-
mophores, is an example of a photosensitive compound. The
antimicrobial activity of curcumin against C. albicans and S. epi-
dermidis is greatly enhanced when combined with blue-light irra-
diation (8,9). The photobactericidal efficacy of compounds such
as curcumin appears to involve ROS generation through the reac-
tion of oxygen molecules with photosensitizers excited by
absorption of light energy (7,10). In an investigation of the prop-
erties of curcumin, Chignell et al. (10) reported that ROS such
as superoxide anion radicals and singlet oxygen are formed in
solutions irradiated with light at 420 nm. The authors also
reported that a half structure of curcumin, which is an analog of
0
Total fluence
2
4.9
–
4.9
+
4.9
–
4.9
–
4.9
+
4.9
–
2.5
–
2.5
–
(
J/cm )
1
kU/well Catalase
Albumin
–
–
–
–
+
–
–
–
–
–
+
–
–
–
–
+
1
mM NAC
Compound
1
4
10
Figure 5. Effect of ROS on inactivation of Escherichia coli in combined
treatment with 100 lM compound 1, 4 or 10 and a fluence of 2.5 or
UV-A. Cell suspensions (initial density, 10 CFUs mL )
were pretreated (30 min at 37°C) with catalase (1 kU per well), an equiv-
alent concentration of albumin or 1 mM NAC for compound 10 only.
Do xa yt ag e an r es p pe rc ei es es .n ted as mean Æ SD (n = 3). **P < 0.01. ROS, reactive
À2
6
À1
4
.9 J cm

1
426 Akihiro Shirai et al.
FA, produces singlet oxygen and superoxide anion radicals fol-
lowing irradiation with UV-A at 342 nm.
Combined treatment with a noncationic compound, in which the
NH
2
ÁHCl group of compound 10 is protected with tert-buty-
À2
Based on the abovementioned previous reports, we predicted
that photobactericidal efficacy could be enhanced by combining
trans-FA treatment with UV-A irradiation, because the base
structure of trans-FA is almost identical to that of half-curcumin.
We hypothesized that the UV spectra of compounds 4 and 10, in
which the carboxyl group in FA is directly substituted with a
phenyl group or with the phenyl group of L-Tyr, respectively,
would exhibit a red shift compared with the parental compound,
FA 1, due to an extended p-electron system. Indeed, the UV
loxycarbonyl (Boc) group, and a UV-A fluence of 2.5 J cm
(10-min incubation) resulted in a minimal reduction (0.34 logs)
in the viable cell count (data not shown). In contrast, the com-
bination treatment with compound 10 induced a 2.63-log reduc-
tion in survival. During the initial period of the irradiation
treatment, the bactericidal activity of compound 10 was signifi-
cantly higher than that of the noncationic compound modified
with a Boc group. Adsorption of the compound onto the cell
surface could promote initial photoinduced damage to the
outer membrane, similar to that produced by a cationic photo-
sensitizer (20).
k
max for compounds 4 and 10 shifted to 326 and 327 nm,
respectively, compared with a value of 314 nm for parental FA
. The red shift could enhance the synergistic bactericidal activ-
1
The concentration dependence of the UV-A-triggered bacteri-
cidal activity of compound 10 was also investigated. The bacteri-
ity with UV irradiation. Proanthocyanidin irradiated at 365 nm
generates hydroxyl and superoxide anion radicals more effec-
tively than when irradiated with visible light (19). This effect is
dependent on the wavelength at which the compound shows high
absorptivity. Compound 10, containing an L-Tyr HCl moiety in
addition to a phenyl group, exhibits cationic properties, which
may facilitate its interaction with bacterial cells via electrostatic
interaction with the anionic components of the cell surface, as
described in a previous report (20). As a consequence, photosen-
sitizers with cationic properties can be used to promote photoan-
timicrobial effects (20,21).
À2
cidal activity of UV-A alone at a total fluence of 7.4 J cm was
significantly enhanced by combined treatment with compound 10
at 5 lM (3.60-log reduction in viability with UV-A treatment
alone versus 4.95-log reduction in combined treatment). Kumar
et al. (12) reported synergistic photoinactivation of E. coli using
50 lM riboflavin as a photosensitizer and broad-spectrum UV
À2
irradiation (265–370 nm) at a fluence of 10 J cm . Under these
conditions, they achieved 5.81-log inactivation, compared with a
5.30-log inactivation with UV irradiation alone (12). In the pre-
sent study, compound 10 showed potential synergistic activity at
a lower concentration and lower fluence than was used by
Kumar et al. (12) with riboflavin, suggesting that trans-FA ester
derivatives modified with a cationic moiety can enhance the effi-
cacy of UV-A-mediated photobactericidal activity. Further study
would demonstrate synergistic efficacy of compound 10 in com-
bination with UV-A irradiation for several microbes containing
Gram-positive bacteria and yeasts, such as Staphylococcus and
Candida.
Combining UV irradiation with one of the three compounds
(
1, 4 or 10), we describe in this study significantly accelerated
the photoinactivation effect of UV-A in a UV-A fluence-depen-
dent manner. In contrast, these compounds showed minimal bac-
tericidal activity in the dark. We also investigated the
modification effect of the carboxyl group in trans-FA on bacteri-
cidal activity. Compound 4, modified with a phenyl group,
exhibited synergistic bactericidal activity upon UV-A irradiation.
The high activity was, however, almost the same as that of par-
ental compound 1. The red shift in the UV spectra between com-
pounds 1 and 4 (kmax change of 12 nm) did not enhance the
photobactericidal activity.
On the basis of a report demonstrating the generation of
superoxide anion radicals via half-curcumin following irradiation
with UV-A at 342 nm (10), we hypothesized that ROS con-
tribute to the synergistic photobactericidal effects observed with
combined treatment with compound 1, 4 or 10 and UV-A irradi-
ation. Previous work demonstrated that inactivation of E. coli
induced by exposure to UV-A is attenuated by the addition of
catalase (quencher of hydrogen peroxide) to the treated bacterial
suspension (11). Indeed, in this study, inactivation of cells trea-
The three compounds isomerized upon trace UV-A irradiation
À2
at 0.14 J cm , as monitored in DMSO-d
6
by NMR analysis,
but photodegradation was not observed until a fluence of
À2
7
.4 J cm (data not shown). NMR analyses confirmed that the
photoreaction-induced cis-conversion of the three compounds
occurs at a ratio of approximate 50%. Phenyl molecules substi-
tuted with an a, b-unsaturated ketone moiety, such as cinnamic
acid and CA, are known to exhibit trans-to-cis conversion under
UV irradiation (4,22). The biological properties of cinnamic acid
À2
ted with UV-A alone at a total fluence of 7.4 J cm diminished
dramatically in the presence of catalase at 100 U per well
(change from 3.77- to 1.73-log reduction in survival with cata-
lase addition; P < 0.01), but the attenuation effect was not
observed upon the addition of an equivalent weight concentration
of nonenzymatic albumin instead of catalase (data not shown).
The role of hydrogen peroxide in reducing the viability of E. coli
treated with a combination of one of the three compounds
(
e.g. antimicrobial activity and inhibition of adenocarcinoma cell
invasiveness) are dependent on isomerization to the cis-isomer
22,23). The photobactericidal activity of compound 1, which
(
was comparable to that of compound 4, may be the result of a
synergistic bactericidal mechanism that involves not only ROS
generation (discussed below) but also cis-isomerization. This
possibility is supported by data demonstrating that cis-cinnamic
À2
(100 lM) and a 2.5 or 4.9 J cm fluence of UV-A was also
investigated. As shown in Fig. 3(B–D), the applied fluence of
À2
2.5 or 4.9 J cm yielded an ca 3-log reduction in survival for
À1
acid at a concentration as low as 2.5 lg mL can damage the
each test compound. In combination treatment with compound 1
or 4, the addition of catalase markedly attenuated the bactericidal
effect, and the attenuation depended on the use of catalase in
comparison with that of albumin. For compound 4, substituted
with a phenyl group, the resulting increase in hydrophobicity
may have interfered with the intrinsic bactericidal activity in the
presence of albumin. However, the change in inactivation
between catalase and albumin was notable. On the other hand,
outer membrane of bacteria (22).
Compared with the addition of compounds 1 and 4, com-
bined treatment with UV-A irradiation and compound 10, which
incorporates an L-Tyr HCl tert-butyl ester, resulted in a signifi-
cant 5.19-log decrease in cell viability at a UV-A fluence of
À2
4
1
.9 J cm . Structurally, compound 10 differs from compounds
and 4, in that 10 has a positive charge derived from L-Tyr.

Photochemistry and Photobiology, 2015, 91 1427
the addition of catalase with compound 10 had a minimal impact
on bactericidal activity. Significant suppression of bactericidal
activity was observed in the presence of the cell-permeating radi-
cal quencher NAC. Our results suggest that ROS such as hydro-
gen peroxide and bactericidal radicals (hydroxyl radicals, etc.)
are the primary mediators of the synergistic photobactericidal
activity observed with our combination procedure and that com-
pounds activated by UV-A irradiation induce the generation of
an excess of ROS in solution. A more direct assay of ROS gen-
eration (e.g. electron spin resonance analysis) would permit both
qualitative and quantitative analyses (19).
Considering the attenuation of killing efficacy observed with
NAC, unlike catalase, phototreatment with compound 10
resulted in the production of oxidative radicals in bacterial
cells. In the present work, we used flow cytometry to assess
the generation of ROS in bacterial cells after exposure to UV-
A in both the absence and presence of test compound (1, 4 or
SUPPORTING INFORMATION
Additional Supporting Information may be found in the online
version of this article:
Figure S1. Scheme for the synthesis of trans-FA derivatives
4
and 10: (i) acetic anhydride, pyridine, 4°C to rt, 5 h, 93% (2);
ii) oxalyl dichloride, phenol crystal, TEA, DMF, DCM, rt, 2 h,
8% (3); (iii) 35% hydrazine hydrate, THF, rt, 30 min, quant.
4); (iv) acetic anhydride, pyridine, 4°C to rt, 4 h; (v) acetylated-
, tert-BuOH, DCC, DMAP, DCM, 4°C to rt, 18 h, 79% (6);
vi) 35% hydrazine hydrate, AN, rt, 30 min, 63% (7); (vii) 2,
WSCI, DMAP, AN, rt, 22 h, 65% (8); (viii) 10% TFA in DCM,
°C to rt, 11 h, 70% (9); (ix) 35% hydrazine hydrate, AN, 4°C
(
8
(
5
(
4
to rt, 50 min; (x) sat. NH Cl aq., 75% (10).
4
Figure S2. UV absorption (A) and fluorescence (B) spectra of
compounds 1 (solid line), 4 (dotted line) and 10 (dashed line) at
a concentration of 50 lM.
Data S1. Details of compound synthesis for O-Acetyl-(E)-fer-
ulic acid (2), Phenyl O-acetyl-(E)-ferulate (3), Phenyl-(E)-ferulate
1
0). A fluorescent probe that responds only to strong oxidative
species such as hydroxyl radicals and peroxynitrite was used in
these analyses. After combination treatment with compound 10
and UV-A irradiation, a 1.56-fold increase in the fluorescence
intensity demonstrated the generation of an excess of strong
oxidative species within the bacterial cells, which corresponded
to the strongest bactericidal activity observed in combination
photoinactivation experiments. In contrast, combined treatment
with UV-A irradiation and noncationic compounds (1 and 4)
resulted in minimal elevation in fluorescence intensity. Only a
small increase in fluorescence intensity (1.19-fold) was
observed upon treatment with compound 10 in the dark, sug-
gesting that electrostatic interaction with the cell surface does
induce some degree of change in the cell membrane. Treatment
with sublethal concentrations of a disinfectant possessing a
cationic property has also been shown to induce ROS genera-
tion in bacteria (24). Therefore, the mechanism of the bacterici-
dal effect of combined UV-A irradiation in the presence of
compound 10 may involve electrostatic interaction with the bac-
terial cell surface due to the positive charge of compound 10,
followed by in situ amplification of strong oxidative radical
generation induced by photoactivation of the compound by
UV-A light. The oxidative damage caused by the intracellular
ROS would then effectively kill the cell. We suggest that inter-
action with the bacterial surface is important for amplifying
intracellular ROS generation, because the nearly identical UV
and fluorescence spectra of compounds 4 and 10 suggest that
the degree of UV-A-mediated photoactivation is also nearly
identical.
(
4), O-Acetyl-N-a-Boc-L-tyrosine t-butyl ester (6), N-a-Boc-L-ty-
rosine t-butyl ester (7), O-(O-Acetyl-(E)-ferulate)-N-a-Boc-L-ty-
rosine t-butyl ester (8), O-(O-Acetyl-(E)-ferulate)-L-tyrosine
trifluoroacetic acid t-butyl ester (9) and O-((E)-Ferulate)-L-ty-
rosine hydrochloride t-butyl ester (10).
REFERENCES
1
2
3
. Maillard, M.-N. and C. Berset (1995) Evolution of antioxidant activ-
ity during kilning: role of insoluble bound phenolic acids of barley
and malt. J. Agric. Food Chem. 43, 1789–1793.
. Mattila, P. and J. Hellstrom (2007) Phenolic acids in potatoes, veg-
etables, and some of their products. J. Food Compost. Anal. 20,
€
1
52–160.
. Hartley, R. D. and E. C. Jones (1975) Effect of ultraviolet light
on substituted cinnamic acids and the estimation of their cis
and trans isomers by gas chromatography. J. Chromatogr. 107,
213–218.
4. Kort, R., H. Vonk, X. Xu, W. D. Hoff, W. Crielaard and K. J.
Hellingwerf (1996) Evidence for trans-cis isomerization of the p-
coumaric acid chromophore as the photochemical basis of the photo-
cycle of photoactive yellow protein. FEBS Lett. 382, 73–78.
5
. Graf, E. (1992) Antioxidant potential of ferulic acid. Free Radic.
Biol. Med. 13, 435–448.
6
. Ryan, W. L., D. J. Gordon and D. H. Levy (2002) Gas-phase photo-
chemistry of the photoactive yellow protein chromophore trans-p-
coumaric acid. J. Am. Chem. Soc. 124, 6194–6201.
7
. Yin, R., T. Dai, P. Avci, A. E. S. Jorge, W. C. M. A. de Melo, D.
Vecchio, Y.-Y. Huang, A. Gupta and M. R. Hamblin (2013)
Light based anti-infectives: ultraviolet C irradiation, photodynamic
therapy, blue light, and beyond. Curr. Opin. Pharmacol. 13,
In conclusion, electrostatic interaction between compound 10
and the bacterial cell surface appears to play an important role in
the enhanced bactericidal activity observed in combined treat-
ment with compound 10 and UV-A irradiation. The electrostatic
interaction, mediated by the cationic L-Tyr HCl moiety in com-
pound 10, results in damage to the cell membrane and subse-
quent amplified generation of active radical species stronger than
hydrogen peroxide upon continued UV-A irradiation. The cis-
conversion induced by photoirradiation may also play a role in
damaging the cell by enhancing the reaction between ROS and
cell substrates.
7
31–762.
8
. Dovigo, L. N., A. C. Pavarina, A. P. D. Ribeiro, I. L. Brunetti, C.
A. de. S. Costa, D. P. Jacomassi, V. S. Bagnato and C. Kurachi
(2011) Investigation of the photodynamic effects of curcumin against
Candida albicans. Photochem. Photobiol. 87, 895–903.
9
. Hegge, A. B., E. Bruzellb, S. Kristensena and H. H. Tønnesena
(
2012) Photoinactivation of Staphylococcus epidermidis biofilms and
suspensions by the hydrophobic photosensitizer curcumin-effect of
selected nanocarrier: studies on curcumin and curcuminoides XLVII.
Eur. J. Pharm. Sci. 47, 65–74.
1
1
0. Chignell, C. F., P. Bilski, K. J. Reszka, A. G. Motten, R. H. Sik and
T. A. Dahl (1994) Spectral and photochemical properties of cur-
cumin. Photochem. Photobiol. 59, 295–302.
1. Hamamoto, A., M. Mori, A. Takahashi, M. Nakano, N. Wakikawa,
M. Akutagawa, T. Ikehara, Y. Nakaya and Y. Kinouchi (2007) New
water disinfection system using UVA light-emitting diodes. J. Appl.
Microbiol. 103, 2291–2298.
Acknowledgements—This work was supported by a Grant-in-Aid for
Young Scientists (B) from the Japan Society for the Promotion of
Science (25870476).

1
428 Akihiro Shirai et al.
1
2. Kumar, V., O. Lockerble, S. D. Kell, P. H. Ruane, M. S. Platz, C. B.
Martin, J.-L. Ravanat, J. Cadet and R. P. Goodrich (2004) Riboflavin
and UV-light based pathogen reduction: extent and consequence of
DNA damage at the molecular level. Photochem. Photobiol. 80, 15–21.
3. Shirai, A., M. Aihara, A. Takahashi, H. Maseda and T. Omasa
19. Nakamura, K., M. Shirato, H. Ikai, T. Kanno, K. Sasaki, M. Kohno
and Y. Niwano (2013) Photo-irradiation of proanthocyanidin as a
new disinfection technique via reactive oxygen species formation.
PLoS ONE 8, e60053.
20. Segalla, A., C. D. Borsarelli, S. E. Braslavsky, J. D. Spikes, G. Ron-
cucci, D. Dei, G. Chiti, G. Jori and E. Reddia (2002) Photophysical,
photochemical and antibacterial photosensitizing properties of a
novel octacationic Zn(II)-phthalocyanine. Photochem. Photobiol. Sci.
1, 641–648.
21. Shrestha, A. and A. Kishen (2012) Polycationic chitosan-conjugated
photosensitizer for antibacterial photodynamic therapy. Photochem.
Photobiol. 88, 577–583.
22. Chen, Y.-L., S.-T. Huang, F.-M. Sun, Y.-L. Chiang, C.-J. Chiang,
C.-M. Tsai and C.-J. Weng (2011) Transformation of cinnamic acid
from trans- to cis-form raises a notable bactericidal and synergistic
activity against multiple-drug resistant Mycobacterium tuberculosis.
Eur. J. Pharm. Sci. 43, 188–194.
23. Tsai, C.-M., F.-M. Sun, Y.-L. Chen, C.-L. Hsu, G.-C. Yen and C.-J.
Weng (2013) Molecular mechanism depressing PMA-induced inva-
sive behaviors in human lung adenocarcinoma cells by cis- and
trans-cinnamic acid. Eur. J. Pharm. Sci. 48, 494–501.
24. Nakata, K., T. Tsuchido and Y. Matsumura (2010) Antimicrobial
cationic surfactant, cetyltrimethylammonium bromide, induces super-
oxide stress in Escherichia coli cells. J. Appl. Microbiol. 110, 568–
579.
1
(
2014) Synergistic antimicrobial activity based on the combined use
of a gemini-quaternary ammonium compound and ultraviolet-A light.
J. Photochem. Photobiol. B 130, 226–233.
1
4. Erg €u n, B. Cß ., T. Cß oban, F. K. Onurdag and E. Banoglu (2011) Syn-
thesis, antioxidant and antimicrobial evaluation of simple aromatic
esters of ferulic acid. Arch. Pharm. Res. 34, 1251–1261.
5. Fan, Q., H. Jiang, E.-D. Yuan, J.-X. Zhang, Z.-X. Ning, S.-J. Qi and
Q.-Y. Wei (2012) Tyrosinase inhibitory effects and antioxidative
activities of novel cinnamoyl amides with amino acid ester moiety.
Food Chem. 134, 1081–1087.
1
1
1
1
6. Nakamura, K., E. Nomura and H. Taniguchi (2001) Chemical prop-
erties of acyl groups on phenolic hydroxyl group of tyrosine. Pept.
Sci. 38, 43–46.
7. Nakamura, K., T. Nakajima, T. Aoyama, S. Okitsu and M. Koyama
(
2014) One-pot esterification and amidation of phenolic acids. Tetra-
hedron 70, 8097–8107.
8. Narasimhan, B., D. Belsare, D. Pharande, V. Mourya and A. Dhake
(
2004) Esters, amides and substituted derivatives of cinnamic acid:
synthesis, antimicrobial activity and QSAR investigations. Eur. J.
Med. Chem. 39, 827–834.