Carboxylate modified benzylidene cyclopentanone dyes for one- and two-photon excited photodynamic therapy

Article abstract of DOI:10.1016/j.jphotochem.2011.06.002

Benzylidene cyclopentanone dyes have high photosensitization activity and large two-photon absorption (TPA) cross-section in near infrared region. To explore their application potentials in two-photon excited photodynamic therapy (TPE-PDT), a series of carboxylate modified benzylidene cyclopentanone dyes were synthesized by varying number of carboxylate groups. Their linear and nonlinear photophysical properties, reactive oxygen yields and in vitro PDT activities were investigated systematically. The results showed that the water solubility of these dyes was significantly enhanced after introducing carboxylate groups. All dyes exhibited large TPA cross-section around 1000-1400 GM at 840 nm. In addition, they could generate both superoxide anion radical and singlet oxygen species under illumination. However, only the dye Y1 modified by one carboxylate group presented strong PDT activity to human rectal cancer 1116 cells, which indicated that moderate water-lipid amphipathy was a very important factor for PDT dyes. Finally, Y1 was proved to have large potential in TPE-PDT by in vitro experiments.

Full text of DOI:10.1016/j.jphotochem.2011.06.002

Journal of Photochemistry and Photobiology A: Chemistry 222 (2011) 228–235
Contents lists available at ScienceDirect
Journal of Photochemistry and Photobiology A:
Chemistry
journal homepage: www.elsevier.com/locate/jphotochem
Carboxylate modified benzylidene cyclopentanone dyes for one- and two-photon
excited photodynamic therapy
Wei Yanga,c, Qianli Zoua,c, Yang Zhoub, Yuxia Zhaoa,∗, Naiyan Huangb, Ying Gub,∗, Feipeng Wua,∗
a Technical Institute of Physics and Chemistry, Chinese Academy of Sciences, Beijing 100190, PR China
b Department of Laser Medicine, Chinese PLA General Hospital, Beijing 100853, PR China
c Graduate School of Chinese Academy of Sciences, Beijing 100049, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Benzylidene cyclopentanone dyes have high photosensitization activity and large two-photon absorption
(TPA) cross-section in near infrared region. To explore their application potentials in two-photon excited
photodynamic therapy (TPE-PDT), a series of carboxylate modified benzylidene cyclopentanone dyes
were synthesized by varying number of carboxylate groups. Their linear and nonlinear photophysical
properties, reactive oxygen yields and in vitro PDT activities were investigated systematically. The results
showed that the water solubility of these dyes was significantly enhanced after introducing carboxylate
groups. All dyes exhibited large TPA cross-section around 1000–1400 GM at 840 nm. In addition, they
could generate both superoxide anion radical and singlet oxygen species under illumination. However,
only the dye Y1 modified by one carboxylate group presented strong PDT activity to human rectal cancer
1116 cells, which indicated that moderate water–lipid amphipathy was a very important factor for PDT
dyes. Finally, Y1 was proved to have large potential in TPE-PDT by in vitro experiments.
Received 3 February 2011
Received in revised form 30 May 2011
Accepted 2 June 2011
Available online 12 June 2011
Keywords:
Two-photon absorption
Photodynamic therapy
Benzylidene cyclopentanone
Water–lipid amphipathy
© 2011 Elsevier B.V. All rights reserved.
1. Introduction
In fact, water–lipid amphiphilic photosensitizers are more photo-
dynamically active than hydrophobic or hydrophilic ones because
Two-photon excited photodynamic therapy (TPE-PDT) is a novel
phototherapy technology [1–3], which attracts much attention
in recent years due to its advantages of deep biological tissue
penetration and highly selective targeting damage. Compared to
conventional PDT [4–6], in which photosensitizing drugs generate
cytotoxic reactive oxygen species (ROS), such as superoxide anion
radical (O2−•) or singlet oxygen (1O2), through absorbing one pho-
ton under visible illumination, TPE-PDT achieves photocytotoxicity
through absorbing two photons of long wavelength simultane-
ously. Therefore a large two-photon absorption cross-section (ꢀ)
is an important prerequisite for TPE-PDT photosensitizers.
Conventional clinical PDT drugs were used in early TPE-PDT
studies. However, their ꢀ values were too small to be suitable
clinically, for example, the ꢀ value of Photofrin was only 7.4 GM
[3] at 850 nm and the ꢀ value of Verteprofin was 51 GM [7]
at 900 nm (1 GM = 10−50 cm4 s molecule−1 photon−1). In the last
decade, a great number of compounds with large ꢀ values were
reported. However, only a few of them could be used as photo-
sensitizer directly for TPE-PDT [8]. The main reason is most of
reported compounds are hydrophobic, having no biocompatibility.
hydrophility is necessary when photosensitizers transport in the
aqueous biological fluids (such as blood, intercellular substance
or cytoplasm) and lipophility is required when photosensitizers
diffuse into phospholipid membranes or cellular organs. Thus a
combination with hydrophilic and lipophilic properties is impor-
tant for a PDT photosensitizer [9,10].
Recently various strategies have been employed to increase bio-
compatibility of photosensitizers with large two-photon absorp-
tion (TPA) cross-section [11–16]. Among them, modifications of
hydrophobic photosensitizers with water-soluble polyethylene
glycol or carboxylate groups have been proven to be efficient. How-
ever, till now, most of these studies have been focused on porphyrin
derivatives [17]. The reason for this preference should be due to the
successful application of hematoporphyrin derivatives as clinical
PDT drugs in last decades together with natural porphyins fre-
quently occurring in living matter. Less attention has been paid on
other photosensitizers. Usually, the specificity of a photosensitizer
on cancer tissues is considered a prerequisite for PDT application.
Till now, most of PDT studies show that the selective tumor uptake
of a sensitizer is probably not mainly due to the special proper-
ties of the sensitizer, but rather to differences in the physiologies
between tumors and normal tissues [18,19]. The special retention
of a sensitizer in the malignant tissue is a consequence of different
kinetics of the sensitizer removal from malignant and healthy tis-
sues. The removal from healthy tissue is faster. The concentration
∗
Corresponding authors. Tel.: +86 10 82543569; fax: +86 10 82543491.
E-mail addresses: yuxia.zhao@mail.ipc.ac.cn (Y. Zhao), guyinglaser@sina.com
(Y. Gu), fpwu@mail.ipc.ac.cn (F. Wu).
1010-6030/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jphotochem.2011.06.002

W. Yang et al. / Journal of Photochemistry and Photobiology A: Chemistry 222 (2011) 228–235
229
O
O
CN
COONa
HCl
OHC
N
OHC
N
NaOH
N
N
NaOH
Y3
S1
COONa
COONa
O
CH2CH3
(i)
N
(ii) NaOH
CH2CH3
DEM
O
O
COONa
N
N
Y2
COONa
N
N
COONa
Y1
(i) Acrylic acid
(ii) NaOH
COONa
O
DEM
OHC
NH
COONa
NaOH
N
N
S2
H
S3
O
NaOH
O
O
Acrylic acid
NaOH
NaOOC
COONa
HN
NH
N
N
Y4
S4
COONa
COONa
COONa
COONa
Scheme 1.
difference depends on the nature of the sensitizer. This result brings
more vitality for exploring new photosensitizers. In our previous
work, benzylidene cyclopentanone dyes were found having large
ꢀ values (400–3300 GM around 800 nm, which are comparatively
high based on their simple molecular structures) and high pho-
tosensitizing efficiencies in two-photon polymerization [20–22],
indicating they could have potential in TPE-PDT. However, these
dyes were all hydrophobic. In this work, four carboxylate mod-
ified benzylidene cyclopentanone dyes Y1–Y4 were synthesized
(Scheme 1). Their solubility, photophysical properties and PDT
activities were investigated systematically.
Science & Technology Co. Ltd. Cell counting Kit-8 (CCK-8) was from
Beyotime Institute of Biotechnology. Fetal bovine serum (FBS) was
obtained from Hangzhou Sijiqing Co. Ltd. Other A.R. grade reagents
were all from Beijing Beihua Co. Ltd., and used after purification by
common methods. Sodium 3-(4-formylphenylamino)-propanoate
(S2) and 2-(4-(diethylamino)-benzylidene)-cyclopentanone (DEA)
were prepared according to literature [23,24]. Human rectal cancer
1116 cells (HRC-1116 cells, from Department of Laser Medicine,
Chinese PLA General Hospital) were grown in culture medium of
DMEM supplemented with 10% FBS at 37 ◦C in a humidified atmo-
sphere of 5% CO2 in air. Cells were suspended at a concentration of
1 × 105 cells/mL.
2. Experimental
2.1.1. Sodium 3-[(4-formylphenyl) (methyl)-amino]-
propanoate (S1)
2.1. Materials
Sodium hydroxide (140 mmol, 5.6 g) and 3-((4-formylphenyl)-
(methyl)-amino)-propanenitrile (47.8 mmol, 9 g) were dissolved in
200 ml water. The reaction mixture was refluxed for 4 h, then cooled
and filtrated. A dilute HCl solution was added dropwise into the fil-
trate under stirring. Precipitate was collected, washed with water
(3 × 150 mL), and neutralized with a dilute NaOH solution. The neu-
tralized solution was dried under vacuum to afford 10.5 g of yellow
solid, yield 96%. 1H NMR (400 MHz, D2O) ı (ppm): 2.44 (t, J = 7.16 Hz,
2H), 3.03 (s, 3H), 3.71 (t, J = 7.36 Hz, 2H), 6.81 (d, J = 9.04 Hz, 2H), 7.72
(d, J = 9.08 Hz, 2H), 9.43(s, 1H).
5,5-Dimethyl-1-pyrroline-N-oxide
(DMPO),
2,2,6,6-tetra-
methyl-4-piperidone (TEMP) and 9,10-dimethylanthracene
(DMA) were purchased from Sigma–Aldrich Chemical Company.
Cyclopentanone was purchased from Tianjin Jinke Institute of Fine
Chemical. 3-((4-Formylphenyl)-(methyl)-amino)-propanenitrile
was obtained from Wenzhou Longsheng Chemical Co. Ltd. Dul-
becco’s modified Eagle’s medium (DMEM, low glucose, containing
100 unit/mL penicillin, 100 mg/mL streptomycin) and phosphate
buffered saline (PBS, pH = 7.4) were purchased from Beijing Solarbio

230
W. Yang et al. / Journal of Photochemistry and Photobiology A: Chemistry 222 (2011) 228–235
2.1.2. 2-[4-[(2-Sodiumcarboxylate-ethyl)(methyl)-amino]-
benzyliden]-5-[4-(diethylamino)-benzyliden]-
cyclopentanone (Y1)
4H), 2.43 (t, J = 7.4 Hz, 8H); HRMS-ESI: m/z Calcd. for C31H35N2O9
[M−4Na+5H]+: 579.2347, found: 579.2337.
DEA (11.2 mmol, 2.72 g) and S1 (10 mmol, 2.28 g) were mixed in
methanol (70 mL) with 0.09 g NaOH as catalyst. The reaction mix-
ture was refluxed under stirring for 24 h. After cooling, the solvent
was removed by rotary evaporation. The residue was dissolved
in water and neutralized with a dilute HCl solution. Precipitate
was collected, washed with water (3 × 50 mL), and further purified
2.2. Method
UV–Vis absorption spectra were measured on a Hitachi U-
3900 spectrophotometer with a concentration of 10−5 mol/L.
One-photon fluorescence spectra were performed on a Hitachi
F-4500 fluorescence spectrophotometer with a concentration of
10−5 mol/L. 1H NMR spectra were obtained on Bruker DPX400 spec-
trometer. Mass spectra were carried out on Bruker Apex IV FTMS.
Fluorescence quantum yields (˚f) were measured in dilute solu-
tions using Rhodamine 6G in ethanol as a standard (˚f = 0.95) [25].
Electron paramagnetic resonance (EPR) spectra were moni-
tored on Bruker ESP-300E. DMPO and TEMP were employed as
trapping agents for superoxide anion radical and singlet oxygen,
respectively. Samples were dissolved in oxygen-saturateddimethyl
sulfoxide (DMSO) with a concentration of 10−4 mol/L for Y1–Y3 and
2 × 10−5 mol/L for Y4. After the addition of spin-trapping reagent,
the oxygen-saturated solution was irradiated with a 532 nm laser.
Quantum yield of singlet oxygen (˚ꢁ) was measured by monitor-
ing the time-resolved laser photolysis experiment [26]. A 532 nm
diode laser was adopted as light source. 9,10-Dimethylanthracene
(DMA) was used as trapping agent for singlet oxygen, and Rose Ben-
gal was used as a standard (˚ꢁ = 0.76 in methanol, 0.162 in DMSO,
1.0 in octanol) [27].
TPA cross-section (ꢀ) values of four compounds in methanol
solution (2 × 10−4 mol/L) were determined using the two-photon
excited fluorescence (TPEF) technique with femtosecond laser
pulses. The excitation light source used was a mode-locked
Tsunami Ti:sapphire laser (720–880 nm, 80 MHz, <130 fs). Rho-
damine B in methanol solution (10−4 mol/L) was used as Ref. [28].
Lipid–water partition coefficient (PC) which showed relative
solubility in octanol versus PBS was obtained by the stir-flask
method [29]. Photosensitizers were diluted to a concentration of
20 ␮mol/L in 5 mL PBS. 5 mL n-octanol was added. The mixture of
the two solvents were stirred gently for 24 h, then centrifuged at
4000 rpm for 10 min to separate the phases. The lipid–water PC val-
ues were calculated by the ratio of photosensitizer concentration
in n-octanol to that in PBS solution.
Before in vitro experiments, Y1–Y4 were dissolved in PBS to
make a series of concentration. In dark cytotoxicity test, flat-bottom
96-well plates were divided into two groups: experimental group
(E) and reference group (R). Experimental group was seeded with
100 ␮L HRC-1116 cell suspension per well, while reference group
was filled with 100 ␮L culture medium. One well was kept as blank
in each group. After 24 h incubation, 10 ␮L PBS was added into blank
wells (B). The rest of experimental wells and reference wells (X)
were injected with 10 ␮L PBS solutions of dyes with proper con-
centration to make the final concentrations of dyes are 2, 5, 10, 25,
50 ␮g/mL.
Cell viability was measured using CCK-8 assay [30]. After 4 h
incubation with dyes, 10 ␮L CCK-8 was added into each experi-
mental well and 10 ␮L PBS was added into each reference well.
Cells were continuously incubated for another 2 h. The absorbance
of each well at 450 nm was determined using a microplate reader.
Cell viability was calculated by the following equation:
by column chromatography on silica gel (VCHCl : VMeOH = 100 : 3).
3
The eluted product was neutralized with dilute NaOH solution, and
dried under vacuum to afford red solid Y1 (1.73 g, yield 40%). 1H
NMR (400 MHz, DMSO-d6) ı (ppm): 1.12 (t, J = 7.04 Hz, 6H), 2.07 (t,
J = 7.36 Hz, 2H), 2.95 (s, 3H), 2.99 (s, 4H), 3.4 (q, J = 7 Hz, 4H), 3.54
(t, J = 7.12 Hz, 2H), 6.74 (d, J = 8.92 Hz, 4H), 7.27 (s, 2H), 7.48 (dd,
J1 = 6.52 Hz, J2 = 2.52 Hz, 4H). HRMS-ESI: m/z Calcd. for C27H33N2O3
[M−Na++2H+]+: 433.2491, found: 433.2477.
2.1.3. 2-[4-[Bis-(2-sodiumcarboxylate-ethyl)-amino]-
benzyliden]-5-[4-(diethylamino)-benzyliden]-
cyclopentanone (Y2)
DEA (5.7 mmol, 1.4 g) and S2 (5 mmol, 1.07 g) were dissolved in
ethanol (32 mL) and heated to reflux under nitrogen. 0.05 g NaOH
was added as catalyst. After 6 h of reflux, the reaction mixture
was cooled and filtrated. The solid was collected and washed with
ethanol (3 × 30 mL). Intermediate S3 was obtained with a yield of
86%. 1H NMR (400 MHz, D2O) ı (ppm): 0.93 (t, J = 7.04 Hz, 6H),
2.36 (t, J = 7.12 Hz, 2H), 2.43 (s, 4H), 3.03 (q, J = 6.96 Hz, 4H), 3.22
(t, J = 7.04 Hz, 2H), 6.39 (d, J = 8.32 Hz, 2H), 6.56 (d, J = 8.36 Hz, 2H),
7.09 (d, J = 8.6 Hz, 2H), 7.17 (t, J = 8.24 Hz, 4H).
S3 (4.3 mmol, 1.9 g) was dissolved in hydrochloric acid (1 mol/L,
7 mL). 34 mL acrylic acid was added. The reaction mixture was
stirred at 80 ◦C for 8.5 h under nitrogen, then cooled and filtrated.
The filtrate was added into water (200 mL) and stirred at 5 ◦C for 3 h.
Precipitate was collected, washed with water (30 mL). The prod-
uct was neutralized with dilute NaOH solution, dried to afford
0.72 g of dark red solid (yield 32%). 1H NMR (400 MHz, D2O) ı
(ppm): 1.08 (t, J = 7.12 Hz, 6H), 2.47 (t, J = 7.64 Hz, 4H), 2.57 (s, 4H),
3.22 (q, J = 6.96 Hz, 4H), 3.59(t, J = 7.72 Hz, 4H), 6.21 (d, J = 8.92 Hz,
2H), 6.77 (d, J = 8.88 Hz, 2H), 7.17 (s, 2H), 7.32 (d, J = 8.96 Hz, 2H),
7.41 (d, J = 8.88 Hz, 2H). HRMS-ESI: m/z Calcd. for C29H35N2O5
[M−2Na++3H+]+: 491.2546, found: 491.2546.
2.1.4. 2,5-Bis-[4-[(2-sodiumcarboxylate-ethyl)(methyl)-
amino]-benzylidene]-cyclopentanone (Y3)
S1 (40 mmol, 9.172 g) and cyclopentanone (20 mmol, 1.68 g)
were mixed in water (80 mL). 0.3 g NaOH was added as cata-
lyst. After 6 h of reflux, the reaction mixture was dispersed into
ethanol (600 mL) and stirred at 5 ◦C for 3 h. Precipitate was col-
lected, washed with ethanol (50 mL), and dried to afford 2.7 g of
dark red solid (yield 27%). 1H NMR (400 MHz, D2O) ı (ppm): 2.31
(t, J = 7.52 Hz, 4H), 2.43 (s, 4H), 2.73 (s, 6H), 3.45 (t, J = 7.68 Hz, 4H),
6.57 (d, J = 8.84 Hz, 4H), 7.09 (s, 2H), 7.25 (d, J = 8.8 Hz, 4H). HRMS-
ESI: m/z Calcd. for C27H31N2O5 [M−2Na++3H+]+: 463.2233, found:
463.2225.
2.1.5. 2, 5-Bis-[4-[bis-(2-sodiumcarboxylate-ethyl)-
amino]-benzylidene]-cyclopentanone (Y4)
ꢀ
ꢁ
AEX − ARX
AEB − ARB
Cell viability (%) =
× 100%
S4 and Y4 were synthesized by similar procedure described
above. S4 (yield 40.6%). 1H NMR (400 MHz, D2O) ı (ppm): 7.27
(d, J = 8.4 Hz, 4H), 7.12 (s, 2H), 6.57 (d, J = 8.4 Hz, 4H), 3.18 (t,
J = 7.0 Hz, 4H), 2.50 (s, 4H), 2.38 (t, J = 7.0 Hz, 4H). HRMS-ESI: m/z
Calcd. for [M−2Na+3H]+: 435.1925, found: 435.1915. Y4 (yield
78.4%). 1H NMR (400 MHz, D2O) ı (ppm): 7.46 (d, J = 8.8 Hz, 4H),
7.18 (s, 2H), 6.74 (d, J = 8.8 Hz, 4H), 3.54 (t, J = 7.4 Hz, 8H), 2.76 (s,
Here A stands for absorbance; subscription E and R represents
experimental and reference group; X is well with a given concen-
tration; B is blank well.
In photocytotoxicity test (one-photon PDT activity experiment),
the same operation is performed except that the final concentra-
tions of dyes are 0.25, 0.5, 1, 2, 5 ␮g/mL. Moreover, after 4 h of

W. Yang et al. / Journal of Photochemistry and Photobiology A: Chemistry 222 (2011) 228–235
231
incubation with dyes, cells were irradiated with a 532 nm diode
(a)
In THF
laser (20 mW/cm2) for 20 min. Then they were continuously cul-
tured in dark for another 24 h before cell viability was monitored
using CCK-8 assay.
1.0
0.8
0.6
0.4
0.2
0.0
In DMSO
In methanol
In PBS
In two-photon induced cytotoxicity experiment, HRC-1116 cells
were incubated with 2 ␮g/mL dye solution for 4 h, followed by
irradiation under a mode-locked femtosecond Ti:sapphire laser
(Tsunami Ti:sapphire, 800 nm, 80 MHz, 80 fs) through an objec-
tive (10×, NA = 0.4, Olympus). The sample was fixed on a xyz-step
motorized stage controlled by a computer. Before illumination, the
focused laser point was adjusted to the middle of a targeted cell.
The illumination intensity was 16 mW. Blank group was observed
under the same condition in the absence of photosensitizer.
3. Results and discussion
350
400
450
500
550
600
Wavelength (nm)
3.1. Synthetic strategy
(b)
In THF
In DMSO
In methanol
In PBS
Scheme 1 provides the synthesis route of target dyes Y1–Y4.
Sodium carboxylate groups were introduced to the aryl moi-
eties of a 2,5-bis[4-(diarylamino)-benzylidene]-cyclopentanone
chromophore to improve its hydrophility by base-catalyzed con-
densation of carboxylate modified aromatic aldehydes S1–S2
with cyclopentanone or DEA. S1 was synthesized by hydrolysis
reaction of the nitrile group in 3-((4-formylphenyl)-(methyl)-
amino)-propanenitrile with very high yield (96%). S2 was obtained
by addition reaction of 4-aminobenzaldehyde with acrylic acid [23].
Due to the strong electron-withdrawing ability of aldehyde group
at the para-position of amino group in 4-aminobenzaldehyde, the
Michael addition of 4-aminobenzaldehyde with acrylic acid gave
only the nomo carboxylate adduct S2. When the condensation reac-
tion of S2 with cyclopentanone or DEA finished, the nucleophilicity
of nitrogen in S3 and S4 increased. Thus, the further Michael addi-
tion reaction of S3 or S4 with acrylic acid provided target compound
Y2 and Y4.
1.0
0.8
0.6
0.4
0.2
0.0
500
550
600
650
700
750
Wavelength (nm)
Fig. 1. Normalized absorption and emission spectra of Y1 in different media.
3.2. Linear photophysical properties
quantum yields could be obtained for these dyes in this solvent.
Because octanol is a lipid-like solvent, the action of dyes in cells
could be more similar to that of dyes in octanol compared to other
solvents.
The UV–Vis absorption spectra as well as emission spectra of dye
Y1 in THF, DMSO, methanol and PBS are shown in Fig. 1. The spectra
of other three dyes Y2–Y4 are quite similar to that of Y1. Due to the
small difference of the electron donating abilities of their terminal
substitutes, only small variations in the absorption and emission
maxima of four dyes can be found. The linear photophysical prop-
erties of four dyes are all listed in Table 1. Y4 is not soluble in THF, so
no data is available. Moreover, the molar extinction coefficients of
Y2–Y3 in THF are also unavailable because their solubility in THF is
very poor. The red-shift of the maximum absorption band and the
enlarged Stokes shift (ꢂꢃ) with increasing solvent polarity (from
THF to DMSO) confirm the absorption is due to a ␲–␲* transition
and the emission is mainly from an intermolecular charge transfer
state [31,32]. From DMSO (a polar aprotic solvent) to methanol (a
less polar but protic solvent compared to DMSO), a further red-shift
of the absorption band and a large increase of Stokes shift should be
due to the hydrogen bond interaction between dyes and methanol.
In addition, the sharp decreases of singlet oxygen and fluorescence
quantum yields in methanol compared to data in DMSO are also
relative to the hydrogen bond interaction. The further decrease
of fluorescence quantum yield in PBS versus in methanol is in
agreement with solvent-polarity-induced nonradiative decay [33].
However, the solution viscosity shows an opposite effect. Octanol
which possesses relative high viscosity sharply reduces radiation-
less transition [34], leading to exceptional high singlet oxygen and
fluorescence quantum yields of Y1. It is a pity that other three dyes
cannot be dissolved in octanol, no singlet oxygen and fluorescence
3.3. Photosensitization activities
PDT is based on the dye-sensitized photooxidation of biologi-
cal matter in the target tissue. There is much evidence to suggest
that reactive oxygen species (ROS) are the major damaging species
in PDT [19]. As typical ROS, superoxide anion radical (O2−•) and
singlet oxygen (1O2) are generated through Type I and Type II mech-
anisms [35,36] respectively. Both of them are based on excited
triplet state of photosensitizer. The generation of ROS can be mon-
itored by EPR spectra [37]. A typical EPR signal (inset in Fig. 2a) of
−•
DMPO–O2 adduct with hyperfine couple constants (˛N = 12.7 G,
˛
H = 10.3 G, ˛ꢄ H = 1.3−G•) was trapped in DMSO solutions of four
ˇ
dyes plus DMPO as O2 trapping reagent under irradiation with a
532 nm laser. Using TEMP as 1O2 trapping reagent, the EPR spec-
trum (inset in Fig. 2b) of the oxidation product TEMPO with triplet
peaks in equal intensity (˛N = 15.9 G) was also observed in DMSO
solutions of four dyes under identical irradiation.
The time dependent of EPR signal intensity are illustrated in
Fig. 2. Due to the poor solubility of Y4 in DMSO, its concentration
(2 × 10−5 mol/L) is only one fifth of the concentration (10−4 mol/L)
of other three dyes, leading to weak EPR signal intensity. For
Y1–Y3, considering their different extinction coefficient at 532 nm

232
W. Yang et al. / Journal of Photochemistry and Photobiology A: Chemistry 222 (2011) 228–235
Table 1
Linear and nonlinear photophysical properties of Y1–Y4 in different media.
abs
Solvent
ꢅ
(nm)
ꢅflmax (nm)
εmax (104 M−1 cm−1
)
ꢂꢃ (cm−1
)
˚
f
˚
ꢁ
ꢀmax (GM)
max
Y1
THF
460
479
487
491
512
461
479
492
516
454
473
488
507
473
489
513
513
565
575
614
646
511
565
617
645
511
561
613
644
558
613
644
6.22
7.48
7.42
6.23
3.71
2246
3178
3142
4080
4051
2123
3178
4118
3876
2457
3316
4179
4196
3221
4137
3965
DMSO
Octanol
Methanol
PBS
0.184
0.358
0.024
0.002
0.067
0.201
0.016
1026
1401
Y2
Y3
Y4
THF
DMSO
Methanol
PBS
8.99
7.41
6.21
0.187
0.032
0.003
0.056
0.021
THF
DMSO
Methanol
PBS
DMSO
Methanol
PBS
6.26
5.28
4.78
6.94
4.89
5.27
0.154
0.024
0.002
0.146
0.050
0.005
0.058
0.019
1053
1167
0.053
0.042
abs
abs
ꢅ
is one-photon absorption band maximum; εmax is molar extinction coefficient at ꢅ
; ꢅflmax is fluorescence band maximum; ˚f is fluorescence quantum yield; ˚ꢁ is
max
max
singlet oxygen quantum yield; ꢀmax is the maximum TPA cross section obtained within 720–870 nm.
(Y2 > Y1 > Y3), their photosensitizing efficiency are very similar. The
results of EPR spectra demonstrate that all four dyes can effec-
tively generate ROS through both Type I and Type II mechanisms.
Although the 1O2 quantum yields (˚ꢁ) of four dyes (listed in
Table 1) are low in polar solvent such as DMSO and methanol, the
˚
of Y1 in lipid-like octanol as 0.201 is acceptable. Considering
ꢁ
target points of PDT are normally lipoidic cellular membrane or
cellular organs (such as mitochondria and lysosome), the ˚ꢁ in
octanol is more useful.
3.4. Two-photon absorption cross section
5000
(a)
The two-photon excitation spectra of Y1–Y4 in methanol are
showed in Fig. 3. Compared to their linear absorption peaks
around 490 nm, the energy band of TPA is coincide in energy
with the vibronic shoulder of the 1PA. This suggests that it is
most probably vibronically enhanced S0 → S1 transition. Within
Y1
10 G
4500
4000
3500
3000
2500
2000
1500
1000
500
Y2
Y3
Y4
the whole measured range (720–870 nm), the maximum
ꢀ
(ꢀmax) of Y1–Y4 are 1026 GM, 1401 GM, 1053 GM and 1167 GM,
respectively. These data are quite comparable to that of 2,5-
bis-[4-(diethylamino)-benzylidene]-cyclopentanone (1032 GM in
N,N-dimethylformamide), indicating that the TPA property of
benzylidene cyclopentanone core has little influenced after modi-
fication with terminal hydrophilic group. Among these four dyes,
Y2 shows the largest ꢀ value due to the stronger electron-donating
capability of ethyl than that of methyl or –CH2CH2COONa group.
Considering their high TPA cross sections, these four dyes are
expected to have great application prospect on TPE-PDT.
0
0
50
100
150
200
250
Exposure Time (sec)
14000
12000
10000
8000
6000
4000
2000
0
(b)
Y1
1600
Y2
Y3
Y4
10 G
Y1
Y2
Y3
Y4
1400
1200
1000
800
600
400
200
0
σ
0
50
100
150
200
250
Exposure Time (sec)
700
720
740
760
780
800
820
840
860
880
−•
Fig. 2. EPR signal intensity of DMPO–O2 adduct (a) and TEMPO (b) as a function
of irradiation time (irradiated at 532 nm) for oxygen-saturated DMSO solution of
dyes ((1 × 10−4 mol/L for Y1–Y3 and 2 × 10−5 mol/L for Y4). Inset: the EPR spectra of
Wavelength (nm)
−•
Fig. 3. Two-photon excitation spectra of Y1–Y4 in methanol.
DMPO–O2 (a) and TEMPO (b).

W. Yang et al. / Journal of Photochemistry and Photobiology A: Chemistry 222 (2011) 228–235
233
Table 2
Solubility in PBS and octanol–water partition coefficient (PC) of Y1–Y4.
(a)
110
Y1
Y2
Y3
Y4
100
90
80
70
60
50
40
30
20
10
0
Compound
Solubility in PBS (mg/mL)
log PC
Y1
Y2
Y3
Y4
0.8
>5
>5
2.92
0.25
−0.65
>5
−0.79
3.5. Evaluation of amphiphilicity
It is clinically necessary that photodynamic drug should be
water–lipid amphiphilic in order to pass both hydrophilic and
lipophilic barriers in biological system. Normally, it is evaluated by
octanol–water partition coefficient (PC) [38]. In this work, differ-
ent number of carboxylate group was introduced to balance the
hydrophobicity of benzylidene cyclopentanone core. The results
showed that the water-solubility of dyes could be sharply improved
as increasing number of carboxylate group (Table 2). The solubil-
ity of Y1 in PBS is 0.8 mg/mL, while that of Y2–Y4 is over 5 mg/mL.
However, the polarity of carboxylate is so strong that the modified
target dye will lost its lipophilicity when more than one carboxylate
is linked. Y2–Y4 are almost insoluble in apolar solvents, whereas
Y1 with log PC of 2.92 still keeps good solubility in organic phase,
behaving as a water–lipid amphiphilic molecule.
0
10
20
30
40
50
Concentration (μg/mL)
(b)
120
Y1
Y2
Y3
Y4
100
80
60
40
20
0
3.6. One-photon PDT effect
A wide range of dye’s concentration was investigated in both
dark cytotoxicity and photocytotoxicity tests. Dark cytotoxicity test
was determined in darkness to exclude the influence of photody-
namic effect. Therefore, cell viability in dark cytotoxicity test could
be used to assess the toxicity of a dye indirectly. Photocytotoxicity
test was carried out under irradiation of a 532 nm laser, so cell via-
bility could evaluate the PDT effect of a dye. The results show (Fig. 4)
that cell viability decreases with increasing dye’s concentration.
Y2–Y4 are safely in dark even at concentration of 50 ␮g/mL (cell
viability over 70%), while evident dark cytotoxicity of Y1 appears
at concentration of 25 ␮g/mL. The phototoxicity of Y2–Y4 is lim-
ited, while Y1 achieves very clear phototoxicity at concentration of
0
1
2
3
4
5
Concentration (μg/mL)
Fig. 4. Dark cytotoxicity (a) and photocytotoxicity (b) of Y1–Y4 toward HEC-1116
cells using CCK-8 assay. The error bars denote standard deviation from three repli-
cates. Dyes were dissolved in PBS before injected into culture medium.
Fig. 5. Experimental group (a–c): Photographs of HRC-1116 cells incubated with 2 mg/mL PBS solution of Y1 before irradiation (a); after 1 min irradiation (b); 10 min after
irradiation (c). Blank group (d–f): Photographs of HRC-1116 cells before irradiation (d); after 10 min irradiation (e); 10 min after irradiation (f).

234
W. Yang et al. / Journal of Photochemistry and Photobiology A: Chemistry 222 (2011) 228–235
2 ␮g/mL. At concentration of 5 ␮g/mL, Y1 presents only 3.9% cell
viability under irradiation.
[7] M. Khurana, H.A. Collins, A. Karotki, H.L. Anderson, D.T. Cramb, B.C. Wilson,
Quantitative in vitro demonstration of two-photon photodynamic therapy
using photofrin and visudyne, Photochem. Photobiol. 83 (2007) 1441–1448.
[8] J.R. Starkey, A.K. Rebane, M.A. Drobizhev, F.-Q. Meng, A.-J. Gong, A. Elliott,
K. McInnerney, C.W. Spangler, New two-photon activated photodynamic
therapy sensitizers induce xenograft tumor regressions after near-IR laser
treatment through the body of the host mouse, Clin. Cancer Res. 14 (2008)
6564–6573.
[9] N. Cauchon, H.J. Tian, R. Langlois, Structure–photodynamic activity relation-
ships of substituted zinc trisulfophthalocyanines, Bioconjugate Chem. 16
(2005) 80–89.
[10] B.W. Henderson, D.A. Bellnier, W.R. Greco, A. Sharma, R.K. Pandey, L.A. Vaughan,
K.R. Weishaupt, T.J. Dougherty, An in vivo quantitative structure–activity
The strong PDT activity of Y1 should be due to its proper
water–lipid amphiphilicity because it is the only amphiphilic dye
among four dyes. For their lack of lipophilicity, Y2–Y4 cannot be
efficiently absorbed by tumor cells. The result confirms the great
importance of amphiphilicity for PDT dyes. Though the dynamic
range of Y1 is limited (2.5–25 ␮g/mL), it shows remarkable PDT
efficiency compare to other three dyes. If the photosensitizer’s dose
is properly chosen (for example at 5 ␮g/mL), Y1 is still a potential
candidate for PDT.
relationship for
a congeneric series of pyropheophorbide derivatives
as photosensitizers for photodynamic therapy, Cancer Res. 57 (1997)
4000–4007.
[11] H.A. Collins, M. Khurana, E.H. Moriyama, A. Mariampillai, E. Dahlstedt, M. Balaz,
M.K. Kuimova, D. Phillips, M. Drobizhev, A. Rebane, B.C. Wilson, H.L. Ander-
son, Blood-vessel closure using photosensitizers engineered for two-photon
excitation, Nat. Photonics 2 (2008) 420–424.
[12] E. Dahlstedt, H.A. Collins, M. Balaz, M.K. Kuimova, M. Khurana, B.C. Wilson, D.
Phillips, H.L. Anderson, One- and two-photon activated phototoxicity of conju-
gated porphyrin dimers with high two-photon absorption cross sections, Org.
Biomol. Chem. 7 (2009) 897–904.
[13] K. Ogawa, Y. Kobuke, Design of two-photon absorbing materials for molecu-
lar optical memory and photodynamic therapy, Org. Biomol. Chem. 7 (2009)
2241–2246.
[14] S.C. Boca, M. Four, A. Bonne, B. van der Sanden, S. Astilean, P.L. Baldeck,
G. Lemercier, An ethylene-glycol decorated ruthenium(II) complex for two-
photon photodynamic therapy, Chem. Commun. 30 (2009) 4590–4592.
[15] D. Belfield, M.V. Bondar, F.E. Hernandez, A.E. Masunov, I.A. Mikhailov, A.R.
Morales, O.V. Przhonska, S. Yao, Two-photon absorption properties of new
fluorene-based singlet oxygen photosensitizers, J. Phys. Chem. C 113 (2009)
4706–4711.
[16] M. Velusamy, J.Y. Shen, J.T. Lin, Y.C. Lin, C.C. Hsieh, C.H. Lai, M.L. Ho, Y.C. Chen, P.T.
Chou, J.K. Hsiao, A new series of quadrupolar type two-photon absorption chro-
mophores bearing 11, 12-dibutoxydibenzo[a,c]-phenazine bridged amines:
Their applications in two-photon fluorescence imaging and two-photon pho-
todynamic therapy, Adv. Funct. Mater. 19 (2009) 2388–2397.
[17] K. Ogawa, Y. Kobuke, Recent advances in two-photon photodynamic therapy,
Anticancer Agents Med. Chem. 8 (2008) 269–279.
[18] J. Moan, Q. Peng, An outline of the hundred-year history of PDT, Anticancer Res.
23 (2003) 3591–3600.
[19] B.W. Henderson, T.J. Dougherty, How does photodynamic therapy work, Pho-
tochem. Photobiol. 55 (1992) 145–147.
[20] J. Wu, Y.X. Zhao, X. Li, M.Q. Shi, F.P. Wu, X.Y. Fang, Multibranched benzylidene
cyclopentanone dyes with large two-photon absorption cross-sections, New J.
Chem. 30 (2006) 1098–1103.
[21] J.Q. Xue, Y.X. Zhao, J. Wu, F.P. Wu, Novel benzylidene cyclopentanone dyes
for two-photon photopolymerization, J. Photochem. Photobiol. A: Chem. 195
(2008) 261–266.
[22] J. Wu, M.Q. Shi, Y.X. Zhao, F.P. Wu, Two-photon absorption property
and photopolymerization sensitizing efficiency of asymmetrical benzylidene
cyclopentanone dyes, Dyes Pigments 76 (2008) 690–695.
[23] X.J. Wan, Y.X. Zhao, J.Q. Xue, F.P. Wu, X.Y. Fang, Water-soluble benzylidene
cyclopentanone dye for two-photon photopolymerization, J. Photochem. Pho-
tobiol. A: Chem. 202 (2009) 74–79.
[24] T. Wang, F.P. Wu, M.Q. Shi, Syntheses of novel asymmetric cyclopentanone
dyes and measurement of two-photon absorption cross-section, Chem. Res.
Chin. Univ. 19 (2003) 470–473.
[25] R.F. Kubin, A.N. Fletcher, Fluorescence quantum yields of some rhodamine dyes,
J. Lumin. 27 (1982) 455–462.
3.7. Two-photon induced cytotoxic effect
A single cell in the circular region (Fig. 5) was exposed to a
focused 800 nm laser beam. A targeted cell incubated with Y1 in
experimental group was irradiated for 1 min. The cellular edema
formed soon after treatment (Fig. 5b) and became more visible
10 min later (Fig. 5c), indicating a clear photo-damage of a tumor
cell. In addition, cells out of laser beam exhibited no change dur-
ing the entire process. In blank group, a targeted cell which was
treated without photosensitizers showed no significant change
after 10 min irradiation (Fig. 5e and f). It was confirmed that both
laser and photosensitizer were indispensable to exert a cytotoxic
effect on tumor cells. The results prove that the Y1 has great poten-
tial in TPE-PDT.
4. Conclusion
Through introducing carboxylate group, four water soluble ben-
zylidene cyclopentanone dyes Y1–Y4 were synthesized. Among
them, only the dye Y1 modified by one carboxylate group presented
proper water–lipid amphopathy, other three dyes had good water
solubility but lost lipophilicity. The results of EPR spectra demon-
strated that all four dyes could effectively generate ROS through
both Type I and Type II mechanisms. Their maximum ꢀ were
all above 1000 GM at 840 nm. However, only Y1 showed strong
one- and two-photon excited PDT activity to human rectal cancer
1116 cells. It was proven that introduction of carboxylate group
was a successful way to improve biocompatability of benzylidene
cyclopentanone dye but the modified degree must be controlled
within a proper range to keep its water–lipid amphipathy, which
was a very important factor to maintain the PDT activity of dyes.
Acknowledgment
[26] Z.J. Diwu, J.W. Lown, Photosensitization by anticancer agents. 12. Perylene
quinonoid pigments, a novel type of singlet oxygen sensitizer, J. Photochem.
Photobiol. A: Chem. 64 (1992) 273–287.
This work was supported by the National Natural Science Foun-
dation of China (No. 60978057).
[27] R.W. Redmond, J.N. Gamlin, A compilation of singlet oxygen yields from bio-
logically relevant molecules, Photchem. Photobiol. 70 (1999) 391–475.
[28] C. Xu, W.W. Webb, Measurement of two-photon excitation cross sections of
molecular fluorophores with data from 690 to 1050 nm, J. Opt. Soc. Am. B 13
(1996) 481–491.
[29] J. De Bruijn, F. Busser, W. Seinen, J. Hermens, Determination of octanol/water
partition coefficients for hydrophobic organic chemicals with the ‘slow-
stirring’ method, Environ. Toxicol. Chem. 8 (1989) 499–512.
[30] Y.H. Kim, J.H. Park, M. Lee, Y.H. Kim, T.G. Park, S.W. Kim, Polyethylenimine
with acid-labile linkages as a biodegradable gene carrier, J. Control. Release
103 (2005) 209–219.
[31] K.A. Zachariasse, S.I. Druzhinin, W. Bosch, R. Machinek, Intramolecular
charge transfer with the planarized 4-aminobenzonitrile 1-tert-butyl-6-
cyano-1,2,3,4-tetrahydroquinoline (NTC6), J. Am. Chem. Soc. 126 (2004)
1705–1715.
[32] T.A. Fayed, Intramolecular charge transfer and photoisomerization of 2-(p-
dimethylaminostyryl)benzoxazole: a new fluorescent probe, J. Photochem.
Photobiol. A: Chem. 121 (1999) 17–25.
References
[1] W.G. Fisher, W.P. Partridge Jr., C. Dees, E.A. Wachter, Simultaneous two-photon
activation of type-I photodynamic therapy agents, Photchem. Photobiol. 66
(1997) 141–155.
[2] K. Ogawa, H. Hasegawa, Y. Inaba, Y. Kobuke, H. Inouye, Y. Kanemitsu, E. Kohno,
T. Hirano, S. Ogura, I. Okura, Water-soluble bis(imidazolylporphyrin) self-
assemblies with large two-photon absorption cross sections as potential agents
for photodynamic therapy, J. Med. Chem. 49 (2006) 2276–2283.
[3] A. Karotki, M. Khurana, J.R. Lepock, B.C. Wilson, Simultaneous two-photon exci-
tation of photofrin in relation to photodynamic therapy, Photochem. Photobiol.
82 (2006) 443–452.
[4] I.J. MacDonald, T.J. Dougherty, Basic principles of photodynamic therapy, J.
Porphyr. Phthalocya. 5 (2001) 105–129.
[5] E.S. Nyman, P.H. Hynninen, Research advances in the use of tetrapyrrolic pho-
tosensitizers for photodynamic therapy, J. Photochem. Photobiol. B: Biol. 73
(2004) 1–28.
[33] C. Cornelissen-Gude, W. Rettig, R. Lapouyade, Photophysical properties of
squaraine derivatives: evidence for charge separation, J. Phys. Chem. A 101
(1997) 9673–9677.
[6] S.B. Brown, E.A. Brown, I. Walker, The present and future role of photodynamic
therapy in cancer treatment, Lancet Oncol. 5 (2004) 497–508.

W. Yang et al. / Journal of Photochemistry and Photobiology A: Chemistry 222 (2011) 228–235
235
[34] S. Sharafy, K.A. Muszkat, Viscosity dependence of fluorescence quantum yields,
J. Am. Chem. Soc. 93 (1971) 4119–4125.
[37] Y.Z. Hu, L.J. Jiang•,− Generation of semiquinone radical anion and reactive
and •OH) during the photosensitization of
a water-
oxygen (1O2, O2
[35] M. Ochsner, Photophysical and photobiological processes in the photodynamic
therapy of tumours, J. Photochem. Photobiol. B: Biol. 39 (1997) 1–18.
[36] C.S. Foote, Type I and type II mechanisms of photodynamic action, ACS Symp.
Ser. 339 (1987) 22–38.
soluble perylenequinone derivative, J. Photochem. Photobiol. B: Biol. 33 (1996)
51–59.
[38] A. Leo, C. Hansch, D. Elkins, Partition coefficients and their uses, Chem. Rev. 71
(1971) 525–616.