Octachloro-fluorescein: Synthesis and photosensitizer performance evaluation

Article abstract of DOI:10.1016/j.dyepig.2019.107635

Fluorescence image-guided photodynamic therapy (PDT) receives great attention since it provides both the diagnostic and therapeutic information. Theoretically, fluorescence and the photodynamic performance (singlet oxygen) can be traced back to the same o

Full text of DOI:10.1016/j.dyepig.2019.107635

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
DyesandPigments170(2019)107635
Contents lists available at ScienceDirect
Dyes and Pigments
journal homepage: www.elsevier.com/locate/dyepig
Octachloro-fluorescein: Synthesis and photosensitizer performance
evaluation
Ying Wang, Hanjiao Chen, Chenghui Li, Peng Wu∗
Analytical & Testing Center, Sichuan University, Chengdu, 610064, China
A R T I C L E I N F O
A B S T R A C T
Keywords:
Fluorescence image-guided photodynamic therapy (PDT) receives great attention since it provides both the
diagnostic and therapeutic information. Theoretically, fluorescence and the photodynamic performance (singlet
oxygen) can be traced back to the same origin, namely the Jablonski energy diagram. Therefore, designing
photosensitizers with balanced fluorescence and singlet oxygen generation is essentially important. Heavy atom
effect (HAE) is an effective approach for designing highly efficient photosensitizers, but at the sacrifice of
fluorescence. Herein, through analysis of the well-known fluorophore fluorescein and the photosensitizer Rose
Bengal (RB), a new kind of photosensitizer, octachloro-fluorescein (OCF), was designed and synthesized.
Structurally, OCF was obtained through anchoring chlorine atoms (heavier than H but lighter than I) onto all the
existing open sites of fluorescein, resting in a 5-fold singlet oxygen quantum yield increase over fluorescein, but a
much higher fluorescence quantum yield (Φfl = 0.628) over RB. The photodynamic performance of OCF was
characterized both photophysically and photochemically. Besides, the image-guided PDT performance of OCF
was further evaluated with HeLa cells, demonstrating that elegant heavy atom substitution of fluorescein may
provide an effective way of balancing the fluorescence and singlet oxygen generation.
Photosensitizer
Singlet oxygen
Heavy atom effect
Image-guide PDT
Fluorescein
1. Introduction
attention since it unites the diagnostic and therapeutic effect in a single
agent, allowing for simultaneous diagnosis and treatment responses
Photosensitization or a photodynamic process refers to the light-
induced reaction, in which light, photosensitizer, and oxygen are the
three basic components [1,2]. During such a process (Scheme 1), the
photosensitizer (PS) first absorbs light for excitation to the excited
singlet state, followed by either emitting prompt fluorescence via de-
caying back to the ground state, or activating the triplet state through
intersystem crossing (ISC). Upon interaction of oxygen with the triplet
state, energy transfer from triplet state to ground state oxygen (3O2)
occurs, leading to the generation of singlet oxygen (1O2). The highly
active nature of singlet oxygen can be harvested for a variety of pho-
tosensitization or photodynamic applications, such as photodynamic
antimicrobial chemotherapy (PACT) [3,4], and photoinduced organic
synthesis [5]. Particularly, photodynamic therapy (PDT), relying on cell
ablation by photosensitized generation of singlet oxygen, has been
widely treatment of cancer [6–10] and other disease (e.g., Alzheimer's
disease [11]). To increase the singlet oxygen generation, anchoring
photosensitizers with heavy atoms, namely heavy atom effect (HAE), is
an effective approach, which promotes spin-orbit coupling and thus
increased ISC [12–14].
[15–17]. From the photophysical point of view, photosensitizers pro-
vides intrinsic chances for designing image-guided PDT. On one hand,
the fluorescence from the photosensitizers allows sensitive imaging-
based operations. On the other hand, singlet oxygen generated from
photosensitization can be explored for effective therapy. Therefore, the
key factor to achieve effective image-guided PDT is to design photo-
sensitizers capable of simultaneous bright fluorescence and good singlet
oxygen generation performance. Although heavy atom effect is effective
in promoting singlet oxygen generation, it is at the sacrifice of fluor-
escence energy, namely quenching of fluorescence [18–21]. Therefore,
designing photosensitizers with balanced fluorescence and singlet
oxygen generation is essentially important for potential image-guided
PDT.
Fluorescein (Scheme 1B), a classical xanthene dye with high fluor-
escence quantum yield and excellent water solubility, is well-known for
its widespread use as fluorescent detection reagents and probes for
bioimaging applications [22–24]. Some derivatives of fluorescein are
also the most widely used fluorescent derivatization reagents in cy-
tology and immunohistochemistry. Normally, the singlet oxygen
quantum yield of fluorescein is typically very low (ΦΔ = 0.03), which is
For modern cancer therapy, image-guided PDT receives great
∗ Corresponding author.
E-mail address: wupeng@scu.edu.cn (P. Wu).
https://doi.org/10.1016/j.dyepig.2019.107635
Received 17 May 2019; Received in revised form 7 June 2019; Accepted 7 June 2019
Availableonline11June2019
0143-7208/©2019ElsevierLtd.Allrightsreserved.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Y. Wang, et al.
DyesandPigments170(2019)107635
Scheme 1. Image-guided photodynamic cancer therapy of OCF.
not suitable for PDT applications. Heavy atom substituted fluorescein
derivatives, such as tetrachlorotetraiodo-fluorescein (Rose Bengal, RB,
Scheme 1B), are well-known photosensitizers for a variety of photo-
dynamic applications [25]. Although featuring high singlet oxygen
quantum yield (ΦΔ = 0.75) [26], the fluorescence of RB (Φfl = 0.003) is
largely lower than that of fluorescein (Φfl = 0.848), which limited its
use in image-guided PDT.
Therefore in this work, we designed and synthesized a kind of
photosensitizer through substituting the iodine atoms in RB with
chlorine, namely 2′,4′,5′,7′-tetrachloro-4,5,6,7-tetrachlorofluorescein
(OCF). The photosensitization performance of OCF was evaluated
through detailed photophysical characterizations. By anchoring
chlorine atoms onto the xanthene ring of fluorescein, OCF displayed a
∼5-fold singlet oxygen quantum yield increase, but still possesses high
fluorescence quantum yield (Φfl = 0.628). Besides, the image-guided
PDT performance of OCF was further evaluated in vivo with HeLa cells,
demonstrating that elegant heavy atom substitution of fluorescein may
provide an effectively way of balancing the fluorescence and singlet
oxygen generation performance of fluorescein.
Scheme 2. Synthesis procedures for OCF.
dichlororesorcinol (compound 3, Fig. S2). Later, OCF (4) was obtained
through condensation between 2,4-dichlororesorcinol (3) and tetra-
chlorophthalic anhydride, which was further characterized with 1H
NMR (Fig. S3), 13C NMR (Fig. S4), MS (Fig. S5), and FT-IR (Fig. S6).
2. Results and discussion
2.1. Synthesis of OCF
Halo-fluorescein can be easily introduced into the ring of fluorescein
via one-step condensation reaction with resorcinol and phthalic anhy-
dride [27]. The detailed synthetic route for the new compound OCF
(compound 4) from 2,4-dichlororesorcinol (compound 3) and tetra-
chlorophthalic anhydride was given in Scheme 2. To prepare compound
3, 2,4-dihydroxybenzoic acid (compound 1) was taken as the precursor,
followed by chlorination with HCl and H2O2 to yield 3,5-dichloro-2,4-
dihydroxybenzoic acid (compound 2, Fig. S1). Subsequent elimination
of the carboxyl group in N,N-dimethylaniline yielded 2,4-
2.2. Photophysical characterizations
The photophysical properties of OCF, including the absorption,
fluorescence, and phosphorescence spectra, were first collected. For
conforming the heavy atom effect, the spectra of fluorescein and RB
were also included. As shown in Fig. 1, the absorption and fluorescence
spectra of these three compounds are generally similar in shape. The
maximum absorption and fluorescence emission of fluorescein locate at
491 nm and 514 nm, respectively. Upon chlorination of fluorescein to
2

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Y. Wang, et al.
DyesandPigments170(2019)107635
Table 1
Summary of the photophysical properties of fluorescein, OCF and RB (solvent:
H2O, pH = 9).
Fluorescein
OCF
RB
Absorption (nm)
491
9.95
514
537
5.49
553
549
8.57
566
ε (104 M−1 cm−1
)
Fluorescence (nm)
Stokes shift (nm)
τfl (ns)
23
16
17
4.58
0.848
633
4.15
0.628
680
0.08
0.003
732
Φfl
Phosphorescence (nm)
τphos (μs)
214.2
2.18
1.85
3.31
0.03
201.9
2.41
1.51
8.97
0.16
98.4
125
0.375
124
kdec × 108 (s−1
)
k
k
rad × 108 (s−1
)
nr × 107 (s−1
)
a
ΦΔ
0.75
a
For ΦΔ evaluation, a mixed solvent of D2O and CH3CN (VD2O: VCH3CN = 1:
15) was used.
Fig. 2. Fluorescence (A) and phosphorescence (B) lifetime of fluorescein, OCF,
and RB.
Fig. 1. The absorption, fluorescence and phosphorescence spectra of fluor-
escein, OCF, and RB (10 μM, H2O, pH = 9).
lifetime of fluorescein and OCF, it is difficult to be collected with the
typical multichannel scanning technique due to the weak phosphores-
cence and strong interference from the fluorescence signal. Therefore,
we employed the flash lamp for rough evaluation of the phosphores-
cence decay. As shown in Fig. 2B, the phosphorescence lifetime of OCF
and RB (20 μM, H2O) were also shortened as compared with that of
fluorescein. The lifetime changes agreed well with the characteristics of
heavy atom effect.
yield OCF, red shift in both absorption (537 nm) and fluorescence
emission (553 nm) occurred. For RB, the absorption and fluorescence
were even red-shifted to 549 and 566 nm, respectively. For phosphor-
escence measurements, the above solutions were first degassed with
nitrogen to remove dissolved oxygen, since phosphorescence is easily
quenched by oxygen. After removal of oxygen, the phosphorescence
emission (delay time of 50 μs) of fluorescein, OCF, and RB were ob-
served at 633 nm, 680 nm, and 732 nm, respectively. Such spectra red
shift is a clear sign of heavy atom effect together with the electron
withdrawing effect of chloride and iodine [14,27].
Next, the fluorescence and phosphorescence emission of fluorescein,
OCF, and RB were quantitatively compared. Generally, the apparent
signature of heavy atom effect is simultaneous fluorescence quenching
and phosphorescence enhancing [18]. As shown in Fig. 1, upon addi-
tion of chloride (OCF) and iodine (RB) to the skeleton of fluorescein,
apparent fluorescence quenching was observed. Meanwhile, phos-
phorescence intensity of OCF was ∼2-fold higher than fluorescein, but
still lower than that of RB. The fluorescence quantum yield (Φfl) of OCF
is 0.628 (Table 1, Fig. S8), which is only 1.35-fold lower than that of
fluorescein (0.848), ensuring the potential cell imaging ability of OCF.
But for RB, the Φfl is as low as ∼0.003, barely suitable for cell imaging
applications.
From the Jablonski energy diagram (Scheme 1A), it is known that
the occurrence fluorescence and phosphorescence is typically accom-
panied by a series of non-radiative transitions. On the basis of the above
photophysical parameters, the decay rate constants (kdec), radiative rate
constants (krad) and non-radiative rate constants (knr) of fluorescein,
OCF and RB were calculated. As pointed out in Table 1, when fluor-
escein was substituted with heavy halogen atoms, about ∼1-fold and
∼57-fold drop in kdec were obtained for OCF and RB, respectively.
Meanwhile, the radiative rate constants were also decreased (fluor-
escein > OCF > RB), accompanied by increase in nonradiative rate
constants. A 3-fold increase in knr of fluorescence after introduced Cl to
the fluorescein core was recorded, indicating that besides the regular
non-radiative transitions, ISC from excited singlet state to triplet state
may occur. In order to get more formation about the excited triplet
state, the nanosecond transient absorption (TA) spectra of fluorescein
and OCF (20 μM) were collected in N2-saturated solutions under the
excitation of 532 nm laser. As shown in Fig. S9, the apparent TA signal
OCF is stronger than that of fluorescein, which is in according with the
above increased knr of fluorescence. For RB, even higher knr of fluor-
escence and stronger TA signal were obtained.
To further confirm the heavy atom effect, the fluorescence and
phosphorescence decay behaviors of fluorescein, OCF, and RB (10 μM,
H2O) were compared. As shown in Fig. 2A, the fluorescence lifetime of
OCF (4.15 ns) is shortened as compared with that of fluorescein
(4.58 ns). While for RB, its fluorescence lifetime (∼0.08 ns) is already
close to the detection limit of our instrument. For phosphorescence
3

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Y. Wang, et al.
DyesandPigments170(2019)107635
oxidation were identified with ROS-specific scavengers, namely tryp-
tophan for 1O2, mannitol for •OH, catalase for H2O2, and superoxide
dismutase for O2•- [31–35]. As shown in Fig. 3D, only tryptophan could
inhibit the TMB oxidation, indicating that singlet oxygen was the major
ROS generated from photosensitization of OCF.
The singlet oxygen quantum yields (ΦΔ) of fluorescein and OCF
were evaluated with the 1O2 phosphorescence emission integrated area
with Rose Bengal (RB) as the standard photosensitizer (ΦΔ = 0.75)
(Figure S10 and Figure S11) [26]. As shown in Table 1, ΦΔ of fluor-
escein and OCF were 0.03 and 0.16, respectively. Considering of the
relatively high QY of both fluorescence (0.628) and singlet oxygen
(0.16), OCF may be potentially useful for image-guided PDT applica-
tions.
2.4. Evaluation of the image-guide PDT performance of OCF
On the basis of the improved singlet oxygen generation from OCF
(versus fluorescein), the PDT performance of OCF was first evaluated
with HeLa cervical cancer cells as the target. For comparison, fluor-
escein and RB were also included. In cell culture media, slight red-
shifted absorption and fluorescence emission of these photosensitizers
were observed, accompanied by insignificant fluorescence quenching
(Fig. S12). Next, the fluorescence imaging performance of OCF was
evaluated. For such purpose, each dye was incubated with HeLa cells
for 12 h and then the fluorescent signals were collected by confocal
laser scanning microscope (CLSM). As show in Fig. 4, in the FITC
channel, fluorescence from both fluorescein and OCF was observed,
with higher signal from fluorescein-stained cells. While in the TRITC
channel, higher fluorescence from OCF-stained cells over the fluor-
escein-stained ones was observed, probably because of the spectra red-
shift. In addition, under laser irradiation (0.5 mW/cm2) for 60 min, few
fluorescence photo-bleaching was observed, demonstrating good pho-
tostability compared with fluorescein and RB (Fig. S13). Therefore, OCF
can be both a fluorescent indicator and a singlet oxygen generator
moiety, which is potentially useful for imaging-guided PDT. For RB-
stained cells, no appreciable fluorescence was observed from neither
the FITC nor the TRITC channel, indicating that RB may be solely
functioned as a singlet oxygen generator.
Fig. 3. Photochemical characterization of singlet oxygen: (A) phosphorescence
emission spectra of singlet oxygen generated from photosensitization of fluor-
escein, OCF, and RB (10 μM, in the mixed solution, VD2O: VCH3CN = 1: 15); (B)
EPR spectra singlet oxygen generated from photosensitization of fluorescein,
OCF, and RB (10 μM, green LED, 2 min) in the presence of TEMP (a specific spin
trap for 1O2, final concentration is 100 mM); (C) time-dependent absorption
spectra of TMB under LED (3 W) radiation in the presence of OCF (20 μM); and
(D) identification of specific ROS generated from photosensitization of OCF
with scavengers. (For interpretation of the references to color in this figure
legend, the reader is referred to the Web version of this article.)
The PDT performance of these photosensitizers was investigated
with a standard Cell Counting Kit-8 (CCK-8) assay. As shown in Fig. 5A,
after incubating HeLa cells with different concentrations of fluorescein/
OCF/RB for 24 h, the cell viability were higher than 88%, demon-
strating a good biocompatibility and low dark-toxicity of these dyes.
Upon LED-irradiation (520 nm, 3 mW/cm2) for 60 min, cell survival
rates decreased in the order of RB > OCF > fluorescein (Fig. 5B),
which is in good accordance with the singlet oxygen quantum yield of
these dyes. Although the 1O2 QY of OCF (ΦΔ = 0.16) was lower than
that of RB (ΦΔ = 0.75), the half maximal inhibitory concentration
(IC50) of OCF (62.1 μM) was only 3.5-fold higher than that of RB
(17.8 μM), implying good PDT performance.
The photo-toxicity of OCF was further confirmed by the live/dead
cells staining with the Caclein-AM/propidium iodide (PI) kit. After
staining, live and dead cells will exhibit green and red fluorescence,
respectively. As shown in Fig. 5C, no red fluorescence can be seen from
fluorescein-stained cells, while the green fluorescence decreased
sharply in the order of fluorescein > OCF > RB, which agreed well
with the above cell viability tests. Therefore, the PDT performance of
OCF locates between those of fluorescein and RB.
2.3. Photochemical characterizations
Molecule photosensitizers, typically working in type-II photo-
sensitization, can produce highly active singlet oxygen for photo-
dynamic applications. To monitor the photosensitized production of
singlet oxygen, the characterized weak phosphorescence from decay of
singlet oxygen [1O2 (1Δg) → O2 (3Σg−)] is the most representative
evidence. As shown in Fig. 3A, under the excitation at 536 nm, obvious
singlet oxygen phosphorescence at ∼1275 nm was collected from OCF
and RB in a D2O–CH3CN mixed solvent (v/v = 1: 15), but not fluor-
escein. To further confirm such generation trend, electron paramagnetic
resonance (EPR) was further explored for identification of the gener-
ated 1O2, with 4-hydroxy-2,2,6,6-tetramethylpiperidine as the trapping
agent. As shown in Fig. 3B, the EPR spectrum of OCF and RB displayed
the characteristic peak of singlet oxygen, namely 1: 1: 1 typical triplet
signal of 2,2,6,6-tetramethylpiperidine-noxyl (TEMPO). In agreement
with the above phosphorescence characterization, there is also no ap-
preciable EPR signal of singlet oxygen from fluorescein, confirming the
low singlet oxygen quantum yield of fluorescein.
Next, the specific reactive oxygen species (ROS) generated form
photosensitization of OCF was identified. First, we selected TMB
(3,3′,5,5′-tetramethylbenzidine) as the substrate for characterization of
the photosensitized oxidation capacity of OCF [28–30], which can be
turned from colorless to blue upon oxidation. As shown in Fig. 3C, OCF
could catalyze the oxidation TMB within 5 min under light emitting
diode (LED) (λ = 520 nm, 3 W). The specific ROS responsible for TMB
3. Conclusion
In summary, a new kind of photosensitizer octachloro-fluorescein
with balanced fluorescence and singlet oxygen generation, was de-
signed and synthesized through substituting the open sites of fluor-
escein with eight chlorine atoms. Compared with fluorescein, OCF
displayed a ∼5-fold singlet oxygen quantum yield increase, due to the
4