Synthesis, photophysical properties, and photodynamic activity of positional isomers of TFPP-glucose conjugates

Article abstract of DOI:10.1016/j.bmc.2018.02.031

The synthesis and characterization of a ‘complete set’ of positional isomers of tetrakis(perfluorophenyl)porphyrins (TFPP)-glucose conjugates (1OH, 2OH, 3OH, 4OH, and 6OH) are reported herein. The cellular uptake and photocytotoxicity of these conjugates were examined in order to investigate the influence of location of the TFPP moiety on the D-glucose molecule on the biological activity of the conjugates. An In vitro biological evaluation revealed that the certain of these isomers have a greater effect on cellular uptake and cytotoxicity than others. The TFPP-glucose conjugates 1OH, 3OH, and 4OH were found to exert exceptional photocytotoxicity in several types of cancer cells compared to 2OH and 6OH substituted isomers.

Full text of DOI:10.1016/j.bmc.2018.02.031

Accepted Manuscript
Synthesis, Photophysical Properties, and Photodynamic Activity of Positional
Isomers of TFPP-Glucose Conjugates
Arif Fadlan, Hiroki Tanimoto, Tatsuya Ito, Yusuke Aritomi, Maho Ueno,
Masaya Tokuda, Shiho Hirohara, Makoto Obata, Tsumoru Morimoto, Kiyomi
Kakiuchi
PII:
S0968-0896(17)32442-2
https://doi.org/10.1016/j.bmc.2018.02.031
BMC 14217
DOI:
Reference:
Bioorganic & Medicinal Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
18 December 2017
13 February 2018
18 February 2018
Please cite this article as: Fadlan, A., Tanimoto, H., Ito, T., Aritomi, Y., Ueno, M., Tokuda, M., Hirohara, S., Obata,
M., Morimoto, T., Kakiuchi, K., Synthesis, Photophysical Properties, and Photodynamic Activity of Positional
Isomers of TFPP-Glucose Conjugates, Bioorganic & Medicinal Chemistry (2018), doi: https://doi.org/10.1016/
j.bmc.2018.02.031
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Article
Synthesis, Photophysical Properties, and Photodynamic Activity of Positional Isomers
of TFPP-Glucose Conjugates
Arif Fadlan a, Hiroki Tanimoto a, Tatsuya Ito b, Yusuke Aritomi b, Maho Ueno b, Masaya
Tokuda b, Shiho Hirohara b,*, Makoto Obata c, Tsumoru Morimoto a, Kiyomi Kakiuchi a,*
aGraduate School of Materials Science, Nara Institute of Science and Technology,
Takayama 8916-5, Ikoma, Nara 630-0192, Japan
bDepartment of Chemical and Biological Engineering, National Institute of Technology,
Ube College, 2-14-1 Tokiwadai, Ube,Yamaguchi 755-8555, Japan
cInterdisciplinary Graduate School of Medicine and Engineering University of Yamanashi,
Kofu 400-8510, Japan
1

Corresponding Author footnote
* Address for correspondence:
Prof. Kiyomi Kakiuch (K. Kakiuchi), Graduate School of Materials Science, Nara Institute
of Science and Technology, Takayama 8916-5, Ikoma, Nara 630-0192, Japan. Tel: +81-
8836355609, Fax: +81-836217117
E-mail addresses: kakiuchi@ms.naist.jp
Prof. Shiho Hirohara (S. Hirohara), Department of Chemical and Biological Engineering,
National Institute of Technology, Ube College, 2-14-1 Tokiwadai, Ube, Yamaguchi 755-
8555, Japan. Tel: +81 836355609, Fax: +81 836355469.
E-mail addresses: hirohara@ube-k.ac.jp.
2

Keywords: Photodynamic therapy (PDT); positional isomers; glucosylated TFPP; cellular
uptake; photocytotoxicity
3

ABSTRACT
The synthesis and characterization of a ‘complete set’ of positional isomers of
tetrakis(perfluorophenyl)porphyrins (TFPP)-glucose conjugates (1OH, 2OH, 3OH, 4OH,
and 6OH) are reported herein. The cellular uptake and photocytotoxicity of these
conjugates were examined in order to investigate the influence of location of the TFPP
moiety on the D-glucose molecule on the biological activity of the conjugates. An In vitro
biological evaluation revealed that the certain of these isomers have a greater effect on
cellular uptake and cytotoxicity than others. The TFPP-glucose conjugates 1OH, 3OH, and
4OH were found to exert exceptional photocytotoxicity in several types of cancer cells
compared to 2OH and 6OH substituted isomers.
4

1. Introduction
Human life expectancy is increasing year by year thanks to the dramatic developments in
medical technology and the availability of advanced medical treatment.1 With such a social
situation, numerous people are seeking cancer treatment that has few side effects and is
non-invasive. Photodynamic therapy (PDT) represents such a treatment that satisfies these
requirements.2 PDT is a therapeutic method based on generating cytotoxic reactive oxygen
species (ROS) by the visible light irradiation of a tumor-containing photosensitizer (PS) for
selectively killing tumor tissue by apoptosis or necrosis. Porphyrin derivatives are the most
useful PS in PDT applications, because of their inherent ability to selectively accumulate in
tumors and for their ability to efficiently generate ROS.2,3 Until now, considerably efforts
have been devoted to developing high performance PS in terms of light absorbing ability in
a longer wavelength region (frequently referred to as the PDT-window), efficient ROS
generation, and high tumor selectivity.
In the development of drugs with improved tumor accumulation, one approach is to
conjugate the PS to an inactive site and transform it into an activated PDT-PSs by light or
enzymes, namely prodrug activation.3 Another simple approach involves the conjugation of
a PS with a biologically active element such as a peptide,4 an anti-tumor monoclonal
antibody,5 folic acid,6 or a sugar.7 Among these approaches, glycoconjugation, a chemical
modification using sugar, is a promising strategy and has attracted considerable attention.8
In addition to providing polar hydroxy groups and to specifically bind to receptors on the
cell surface,9 sugars are also known to have a key role in cancer,8a,10 immune dysfunction,11
congenital disorders,12 and infectious diseases.8b The potential of sugar is also supported by
5

that a radiolabeled D-glucose derivative, namely 2-[18F]fluoro-2-deoxy-D-glucose (18F-
FDG), is frequently used in cancer diagnosis by positron emission tomography (PET).8a,13
Many sugar-conjugated PSs have been synthesized and their photophysical properties,
cellular uptake behavior, and in vitro photocytotoxicity examined.7a,14 A few sugar-
conjugated PSs have also been tested in in vivo conditions.7a,14b,15 The findings indicate
sugar-conjugated PSs exhibit a high cell uptake and photocytotoxic effect by virtue of the
fact that the molecule contains a sugar moiety. Kessel and co-workers reported that not
only the type of substituent (namely, the biologically active element) conjugated to the
tetraphenylporphyrin sulfonic acid (TPPS), but also the substitution number and pattern
(i.e. mono-, bis cis-, bis trans-, tri-, and tetra-sulfonic acids) greatly influence the extent of
cellular uptake and the overall effect of the PS.16 In a previous study, we reported on the
synthesis of tetrakispentafluorophenyl porphyrin (TFPP) derivatives with glucose
oligomers with varying chain lengths and binding positions (namely, mono-, bis cis-, bis
trans-, tri-, and tetra-glucoses) and examined the PDT effect. The findings indicated that
the TFPP bis trans-glucose derivatives exhibited an almost 20-fold higher
photocytotoxicity compared with the other TFPPs.17
Lippard and co-workers recently investigated the effect of the position of substitution on D-
glucose on the biological activity of Pt complex-glucose conjugates (Glc-Pts) for an
anticancer agent.8c Although the synthesis of a complete set of positional isomers of a
glycoconjugate and a study of the influence the substitution position on its biological
activity has not been reported over the past two decades, they prepared all possible
positional isomers (C1α, C1β, C2, C3, C4, and C6) of a Glc-Pt complex studied their
6

effects, both in vitro and in vivo (Chart 1). Their findings indicate that the recognition and
efficiencies varied for different positional isomers of a glycoconjugate and that different
positions of substitution resulted in a different response for cellular uptake, cytotoxicity
profile, and the inhibition of the facilitative membrane-bound glucose transporter (GLUT1)
which is usually specifically overexpressed in various cancers. While the C3 isomer
showed the lowest efficiency in cancer cells, the C1α and C2 isomers accumulation the
most efficiently, compared to the others. Furthermore, the C2 isomer exhibited the highest
GLUT1-specific internalization in cancer cells.
Chart 1 Structure of Pt complex-glucose conjugates8c
In order to study the effect of D-glucose positional isomers on their chemical and biological
properties, we synthesized a complete set of positional isomers of D-glucose by attaching
TFPP to the 2-, 3-, 4-, and 6-positions by sulfide bonding, i.e., at the 2OH, 3OH, 4OH, and
6OH positions examined their chemical and biological properties, and compared their
activities with those for the 1OH isomer The findings indicate that the position of
substitution is an important factor in ROS generation, cellular uptake, GLUT-1 inhibition,
and cytotoxicity.
7

Scheme 1. Synthesis of 2-, 3-, 4-, and 6-Ac-S-AcGlc
2. Results and discussion
2.1. Chemistry
To systematic study the effect of D-glucose positional isomers, D-glucose derivatives that
were deoxythiolated at the 2-, 3-, 4-, and 6-positions were synthesized following previously
reported procedures, with some modifications. Starting with D-mannose, tetraacetyl
mannose 2-triflate (2-Tf-O-AcMan) was prepared through peracetylation followed by
rearrangement of the acetyl group through orthoester formation (details of their precursor
preparations are shown in Supporting Information).18 An acethylthio group was introduced
by an SN2 reaction to afford tetraacetyl 2-thio glucose (2-Ac-S-AcGlc) as the β-anomer.19
From commercially available diacetone D-glucose, the free 3-hydroxy group was
epimerized through oxidation and reduction20 followed by triflation21 and acetyl thiolation22
to give diacetone 3-deoxy-3-acetylthio D-glucose (Diacetone-3-deoxy-Ac-S-Glc), which
was transformed into the pyranose form to yield the tetraacetyl 3-thio glucose (3-Ac-S-
8

AcGlc) as an α/β-anomeric mixture.23 For the tetraacetyl 4-thio glucose (4-Ac-S-AcGlc),
the synthesis was started from D-galactose, the most reliable and reproducible route. Global
benzoylation except for the axial 4-hydroxy group was successful, giving 4-hydroxyl
tetrabenzoyl glucose in moderate yield,24 which was transformed to the triflate derivative.25
Reaction with potassium thioacetate gave 4-Ac-S-Bz-Glc in moderate yield. Finally, the
fully 4-Ac-S-AcGlc was synthesized as an α/β-anomeric mixture, which was previously
prepared via a different route.26 The benzoyl groups of 4-Ac-S-Bz-Glc were replaced with
acetyl groups because the removal of benzoyl group was difficult at the stage of the last
step after the conjugation with TFPP. 6-Acetyl 6-thio (6-Ac-S-AcGlc)27 was prepared from
D-glucose by a one-pot regioselective tosylation/tetraacetylation28 and acetylthiolation was
accomplished similar to the procedure used for the synthesis of 2-Ac-S-AcGlc.19
Scheme 2. Synthesis of TFPP-glucose conjugates
9

SNAr reactions of TFPP with peracetylated thioglucose were carried out using a method
similar to that used to prepare the TFPP-glucose conjugate 1Ac (Scheme 2).17a All other
thioglucopyranoses, 2-Ac-S-AcGlc, 3-Ac-S-AcGlc, 4-Ac-S-AcGlc, and 6-Ac-S-AcGlc,
were less reactive than 1-Ac-S-AcGlc. The SNAr reaction of TFPP with 6-Ac-S-AcGlc
required an unusually high temperature of 40 ~ 50oC or a longer reaction time of 3 days or
more. This is because the other positions (C2, C3, C4, and C6) react more slowly than the
anomeric position,29 that is, it appears to be difficult to generate thiolate ions in these
compounds, and therefore, to bind with porphyrin. This was confirmed by the fact that the
thioglucopyranose derivatives (2-Ac-S-AcGlc, 3-Ac-S-AcGlc, 4-Ac-S-AcGlc, and 6-Ac-S-
AcGlc) were recovered from the reaction mixture after column purification. The reaction
mixtures were separated by conventional silica gel column chromatography to give the
peracetylated TFPP-glucose conjugates 2Ac, 3Ac, 4Ac, and 6Ac as deep purplish red
10

solids. These compounds were fully characterized by 1D (1H, 13C, and 19F) and 2D (COSY
and HSQC) NMR, UV-vis spectroscopy, MALDI-TOF high resolution mass spectrometry.
1H NMR spectra of 2Ac, 3Ac, 4Ac, and 6Ac were recorded in CDCl3. The β-pyrroles
afforded two peaks centered at 8.9 ppm due to asymmetric conjugation, which is similar to
that for 1Ac.17a The protons of acetyl groups and inner pyrrole protons were found around
2.3 ~ 2.1 ppm and −2.9 ppm, respectively. Glucose protons afforded heavily coupled peaks
in the range of 7 ~ 3 ppm. The peracetylated TFPP-glucose conjugates 2Ac and 6Ac were
obtained only β-anomeric form. In contrast, 3Ac and 4Ac were obtained as anomeric
mixtures. These compounds showed two doublet peaks that were assigned to the α-
anometic proton (3J = 4.0 Hz) at 6.4 ppm and the β-anomeric one (3J = 8.0 Hz) at 5.8
ppm.17b,30 The α- and β-isomeric ratios for these derivatives were determined to be 5:5 and
3:7 for 3Ac and 4Ac, respectively. Interestingly, this stereochemical difference resulted in
the chemical shift of the 2,6- and 3,5-F of glucose-substituted perfluorophenyl group,
which are separated from the anomeric position. Stereochemically pure 2Ac and 6Ac
showed two peaks corresponding to the 2,6- and 3,5 positions at −136.54 ~ −138.02 and
−131.85 ~ −133.77 ppm. In contrast, 3Ac and 4Ac afforded two set of two peaks whose
area ratios approximately coincided with that determined by 1H NMR spectroscopy.
Further separation of the anomeric mixture was not feasible or very inefficient at least.
Fortunately, this did not impact our goal because all acetyl groups were removed to give
anomeric mixtures of all these glucose conjugates, which existed in an equilibrium between
α- and β-anomers in solution and were subjected to further physiological and biological
investigations (vide infra). The part of crude acetylated compounds 2Ac, 3Ac, 4Ac and 6Ac
11

were used for the synthesis of deacetylated compounds. After chromatographic purification,
four TFPP-glucose conjugates 2OH, 3OH, 4OH, and 6OH were obtained in moderate
yields and acceptable purities. The structures of TFPP-glucose conjugates 2OH, 3OH,
1
4OH, and 6OH were confirmed by H, 19F, and 13C NMR spectroscopies with a 2D
1
technique such as COSY, HSQC and H-19F HOESY, UV-vis spectroscopy, Steady-state
luminescence spectroscopy, and MALDI-TOF high resolution mass spectrometry. All of
the TFPP-glucose conjugates were detected as [M]+ species in MALDI-TOF high
resolution mass spectrometry.
After hydrolysis by treatment with sodium methoxide and chromatographic purification, all
peaks assigned to the acetyl groups disappeared. 19F NMR spectra of 2OH, 3OH, 4OH,
and 6OH clearly indicated that remaining 4-F nucleus of the pentafluorophenyl groups
were conserved. Unlike peracetylated TFPP-glucose conjugates, all of the TFPP-glucose
conjugates showed two sets of two peaks conforming the 2,6- and 3,5-F nucleus of glucose-
conjugated perfluorophenyl group. This is reasonably explained by the fact that the α- and
β-pyranose forms are in equlibrium. The peak area ratio between the two sets was
approximately 4:6 which is in reasonable agreement with the value for D-glucose in water
(37.3:62.7).
The stock solution of all compounds 2OH, 3OH, 4OH, and 6OH were prepared using
qNMR data, and the physical properties (ROS-generation test and In vitro assay) were
evaluated using these stock solutions.
12

Table 1 UV-Vis and Photoluminescence Spectral Data of TFPP-glucose conjugates 1OH,
2OH, 3OH, 4OH, and 6OH in DMSO at 25°C.a
UV-vis
max / nm (ε × 10−4 / M−1 cm−1)
Q bands
Luminescence
λ
b
f>500 nm
λ
max / nmc
Soret
412.5
505.5
(2.09)
537.5
(0.07)
582.0
(0.61)
1OH
2OH
3OH
4OH
6OH
0.03
0.03
0.03
0.03
0.04
637.5, 702.5
637.5, 702.5
637.5, 703.0
637.5, 703.0
637.5, 702.5
(29.3)
412.5
(28.9)
506.0
(1.93)
537.5
(0.03)
582.0
(0.55)
413.0
(30.4)
505.5
(2.13)
537.5
(0.06)
581.0
(0.63)
413.0
(30.2)
506.0
(2.01)
537.5
(0.02)
582.0
(0.57)
412.5
(35.1)
505.5
(2.52)
536.5
(0.06)
581.0
(0.71)
a
[1OH] = 5.00 µM, [2OH] = 4.44 µM, [3OH] = 4.48 µM, [4OH] = 4.05 µM, [6OH] =
4.19 µM. b Oscillator strength in the range above 500 nm estimated as 4.32 × 10−9 ∫ε(v)dv. c
The excitation wavelength was adjusted to the maximum absorption wavelength of the
Soret band.
All TFPP-glucose conjugates 2OH, 3OH, 4OH, and 6OH exerted typical phyllo-type UV-
vis spectra with one intense peak (Soret band) at approximately 413 nm and three vibronic
peaks (Q bands) around 505 to 648 nm in DMSO (Table 1 and vide infra). The TFPP-
glucose conjugates 2OH, 3OH, 4OH, and 6OH afforded broad peaks in UV-vis spectra
recorded in PBS containing 1 vol% DMSO (Figure 1 and Table 2), and resemble to the
13

spectrum of 1OH that was reported previously.17a These results suggested that the TFPP-
glucose conjugates form aggregates similar to 1OH in PBS. In addition, the oscillator
strength in the region over 500 nm (f>500nm), which is an important measure for evaluating
the probability of photoexicitation, was nearly the same for of the TFPP-glucose conjugates.
15
: 1OH
: 2OH
: 3OH
10
: 4OH
: 6OH
5
ε
0
350
450
550
650
750
λ
/ nm
Figure 1. UV-vis spectra of TFPP-glucose conjugates 1OH, 2OH, 3OH, 4OH, and 6OH
in PBS Solution containing 1% DMSO at 25°C.
14

Table 2 UV-Vis and Photoluminescence Spectral Data of TFPP-glucose conjugates 1OH,
2OH, 3OH, 4OH, and 6OH in PBS containing 1% DMSO at 25°C.a
UV-vis
max / nm (ε × 10−4 / M−1 cm−1)
Q bands
Luminescence
λ
b
f>500 nm
λ
max / nmc
Soret
421.5
(9.1)
509.5
(1.77)
557.0
(0.23)
586.5
(0.60)
1OH
2OH
3OH
4OH
6OH
0.06
0.06
0.05
0.06
0.06
645.0, 709.0
417.5
(12.2)
509.0
(2.04)
556.5
(0.23)
586.0
(0.69)
646.5, 662.0,
702.5
421.5
(8.90)
509.0
(1.67)
557.0
(0.23)
586.0
(0.58)
645.5, 663.0,
703.0
421.5
(10.5)
510.0
(1.90)
557.0
(0.21)
586.5
(0.61)
646.0, 662.5,
703.0
421.0
(11.2)
510.0
(2.06)
557.0
(0.25)
586.0
(0.70)
645.5, 702.5
a
[1OH] = 5.00 µM, [2OH] = 4.44 µM, [3OH] = 4.48 µM, [4OH] = 4.05 µM, [6OH] =
4.19 µM. b Oscillator strength in the range above 500 nm estimated as 4.32 × 10−9 ∫ε(v)dv. c
The excitation wavelength was adjusted to the maximum absorption wavelength of the
Soret band.
PDT efficacy crucially depends on the generation of reactive oxygen species (ROS) by the
photoirradiation of photosensitizers located in tumor tissue.2,3a There are two types of
photochemical reactions, namely hydrogen or electron transfer generating superoxide
anions, hydroxyl radicals (•OH) and other (types I reaction) and energy transfer producing
1
singlet oxygen (1O2) (type II reaction). In general, O2 is known to be the major cytotoxic
15

species in PDT applications because of its relatively long lifetime. In recent years, it was
reported that hematoporphyrin (HP) and imidazole conjugated porphyrin not only generate
singlet oxygen, but also hydroxyl radicals.17 Therefore, both the type and efficiency of ROS
generation are important issues for successful photosensitizers in PDT applications. The
relative quantum yield for 1O2 generation (Φ∆) was determined by means 1O2 luminescence
measurements at 1270 nm in O2-saturated MeOH and normalized to the values for 1OH.
On the other hand, the relative quantum yields (Φ•OH) of •OH were evaluated by the rate
constant for the increment of the initial rate of HPF-fluorescence intensity31 and normalized
to the value for hematoporphyrin (HP). The relative Φ∆ and Φ•OH values for the TFPP-
glucose conjugates are listed in Table 3. No differences in the Φ∆ values was found in
methanol. However, similar to 1OH,17b TFPP-glucose conjugates 2OH, 3OH, 4OH, and
6OH showed higher •OH generation than HP. The Φ•OH values for the TFPP-glucose
conjugates increased in the order of 2OH < 1OH < 4OH < 3OH << 6OH. Hence, the Φ•OH
values for TFPP-glucose conjugates were dependent, not only on the S-glycosylation
pattern, but also the specific of D-glucose.
16

Table 3 Relative Quantum Yield of Singlet Oxygen (Φ∆) and Hydroxyl Radical (Φ•OH) of
TFPP-glucose conjugates 1OH, 2OH, 3OH, 4OH, and 6OH.
a
b
Φ∆
Φ•OH
25.0
20.5
36.9
28.4
61.2
1.0
1OH
2OH
3OH
4OH
6OH
HP
1.0
1.0
1.0
1.0
1.0
–
a
In O2-saturated methanol and normalized to the values for 1OH. The excitation
wavelength was 399.4 nm (1OH), 397.6 nm (2OH), 395.4 nm (3OH), 397.6 nm (4OH),
and 389.8 nm (6OH) (Abs = 0.2). b The Φ•OH value in O2-saturated PBS containing 1 vol%
DMSO and normalized to the values for HP.
The hydrophobicity parameter (Log P), the logarithmic scale of the partition coefficient
between n-octanol and water, is closely correlated to the penetration and uptake behavior of
drugs through the cellular membrane.32 The Log P values of the TFPP-glucose conjugates
1OH, 2OH, 3OH, 4OH, and 6OH were determined by reverse phase thin layer
chromatography (RP-TLC) (Table 4). Their hydrophobicity increased in the order of 1OH
< 3OH < 4OH < 6OH < 2OH.
17

Table 4 Rf values in RP-TLC and Hydrophobicity Parameter (Log P) of TFPP-glucose
conjugates 1OH, 2OH, 3OH, 4OH, and 6OH.
Rf
Log Pa
6.05
6.87
6.21
6.30
6.50
1OH
2OH
3OH
4OH
6OH
0.19
0.11
0.17
0.16
0.14
a Log P = 3.46 × Log (1/Rf) + 3.55.
2.2. Biology
The cellular uptake of these compounds was examined in stomach cancer, namely an RGK
gastric carcinoma mucosal cell line (RGK cells). Figure 2 shows the relative cellular uptake
of the TFPP-glucose conjugates 1OH, 2OH, 3OH, 4OH, and 6OH by RGK cells after
incubation with each compound (c = 2.5 µM) for 8 and 24h. The value was normalized to
that of 1OH for an 8h incubation. The amounts of these compounds that were taken up
increased with increasing incubation time from 8 to 24h. TFPP-glucose conjugate 4OH
showed a slightly higher uptake than the other conjugates.
18

2.5
2
1.5
1
0.5
0
1OH 2OH 3OH 4OH 6OH
Figure 2. Relative uptake amount of TFPP-glucose conjugates 1OH, 2OH, 3OH, 4OH,
and 6OH in RGK-1 cells for 8h (gray stick bars) and 24h (blue stick bars) co-incubation.
The concentration of the conjugates was 2.5 µM. Values are the mean standard deviation
of three replicate experiments.
Lippard and co-workers recently reported cellular uptake studies of platinum complexes
that were ligated to different positions on D-glucose (C1α, C1β, C2, C3, C4, and C6; Chart
1) using A2780, A549, and DU 145 cells after 7.5 ~ 72h co-incubation.8c They reported that
Glc-Pt C2 showed a high uptake in cancer cells that overexpress the membrane-bound
glucose transporter (GLUT1), resulting in a high anticancer activity. It is known that
GLUT1 is overexpressed in brain tumors and stomach cancer.35 Therefore, the cellular
uptake of TFPP-glucose conjugates 1OH, 2OH, 3OH, 4OH, and 6OH in the presence of
cytochalasin B, an inhibitor of GLUT1, was examined in RGK cells after a 8 and 24 h co-
incubation (Figure 3). The uptake of 2OH, 3OH, and 4OH by RGK cells appeared to be
slightly inhibited in the presence of cytochalasin B. This result suggests that the cellular
19

uptake of these conjugates is partly driven by GLUT1. However the effect of cytochalasin
B was too small, probably due to the strong hydrophobic nature of these conjugates, which
would result in enhanced non-specific interactions with cellular membrane.
2.5
(a)
2
1.5
1
0.5
0
1OH 2OH 3OH 4OH 6OH
2.5
(b)
2
1.5
1
0.5
0
1OH 2OH 3OH 4OH 6OH
Figure 3. Relative uptake amount of TFPP-glucose conjugates 1OH, 2OH, 3OH, 4OH,
and 6OH in RGK cells in the absence (red stick bars) and presence (gray stick bars) of
cytochalasin B (c = 50 µM), which is a GLUT1 inhibitor, for 8h (a) and 24h (b) co-
incubation. The concentration of the conjugates was 2.5 µM. Values are the mean
standard deviation of three replicate experiments.
20

The photocytotoxicity of the TFPP-glucose conjugates 1OH, 2OH, 3OH, 4OH, and 6OH
were evaluated in Human cervical cell line (HeLa cells) at a concentration of 1.0 µM in the
dark and by irradiation. The cells were incubated with photosensitizers for 24h, and then
photoirradiated using a 100 W halogen lamp (λ> 500 nm) at a light dose of 0 J cm−2 (dark)
or 16 J cm−2. cell survival (%) at 24h post-photoirradiation was evaluated in terms of the
mitochondrial activity of NADH dehydrogenase using the WST-8® reagent from Cell
Counting Kit-8 and the value was normalized to the survival rate in the absence of
photosensitizer, namely vehicle. None of the conjugates showed any cytotoxicity in the
dark (vide infra). In contrast, these conjugates, except for 2OH, showed potent
photocytotoxicity under photoirradiation (Figure 4).
140
120
100
80
60
40
20
0
1OH 2OH 3OH 4OH 6OH
Figure 4. Photocytotoxicity of TFPP-glucose conjugates 1OH, 2OH, 3OH, 4OH, and
6OH in HeLa cells. The concentration of the conjugate was 1.0 µM. The light dose was 16
J cm−2 from a 100 W halogen lamp (λ > 500 nm). Values are the mean standard deviation
of six replicate experiments.
21

140
120
100
80
: 1OH
: 3OH
: 6OH
: 2OH
: 4OH
60
40
20
0
0.0 0.2 0.4 0.6 0.8 1.0
c/ µM
Figure 5. Plot of the cell survival rate (%) of HeLa cells treated with TFPP-glucose
conjugates 1OH, 2OH, 3OH, 4OH, and 6OH as a function of the concentration. The light
dose was 16 J cm−2 from a 100 W halogen lamp (λ > 500 nm). Incubation time with the
conjugates was 24 h. Values are the mean standard deviation of six replicate experiments.
The photocytotoxicity of TFPP-glucose conjugates 1OH, 2OH, 3OH, 4OH, and 6OH in
HeLa cells was examined in the concentration range from 0.1 to 1.0 µM (Figure 5). The
drug concentration needed to induce 50% cell death (EC50) values for the TFPP-glucose
conjugates 1OH, 3OH, and 4OH were somewhat lower than the values for 2OH and 6OH,
indicating a slightly higher photocytotoxic effect. In previous studies of TFPPs having S-
glucopyranosyl groups with possible patterns (mono-, cis-bis-, trans-bis-, tris-, and tetrakis-
substituted), we found that the photocytotoxic effect of these compounds is correlated with
the amount taken up by cells.17 In contrast, no correlation between cellular uptake and
22

photocytotoxicity was found for these mono-glucose conjugates. This can be attributed to
the high hydrophobicity of TFPP-glucose conjugates. PDT has been applied to the
treatment of early lung cancer, early esophageal cancer, uterine cancer and breast cancer,
and more recently it was approved as a treatment for early gastric cancer and brain cancer.36
Therefore, the photocytotoxicity of TFPP-glucose conjugates 1OH, 2OH, 3OH, 4OH and
6OH was examined in brain tumors, namely the human glioblastoma cell line (U251), and
RGK cells under the same conditions as for HeLa cells. Figure 6 shows cell survival (%) as
a function of concentration. The effect of positional isomers in both cells was found to be
similar to that in HeLa cells.
23

140
120
100
80
: 1OH
: 2OH
: 3OH
: 4OH
: 6OH
(a)
60
40
20
0
0.0 0.2 0.4 0.6 0.8 1.0
140
120
100
80
(b)
60
40
20
0
0.0 0.2 0.4 0.6 0.8 1.0
c/ µM
Figure 6. Plot of the cell survival rate (%) of U251 cells (a) and RGK cells (b) treated with
TFPP-glucose conjugates 1OH, 2OH, 3OH, 4OH, and 6OH as a function of the
concentration. The light dose was 16 J cm−2 from a 100 W halogen lamp (λ > 500 nm).
Incubation time with the conjugates was 24 h. Values are the mean standard deviation of
six replicate experiments.
24

3. Conclusion
We successfully synthesized five possible positional isomers of TFPP-glucose conjugates
1OH, 2OH, 3OH, 4OH, and 6OH. The TFPP-glucose conjugates 1OH, 3OH, and 4OH
exerted a higher photocytotoxicity in several cancer cells, such as HeLa cells, RGK cells
and U251 cells, compared to the 2OH and 6OH isomers. Hence, the location of
substitution on the D-glucose molecule influence the photocytotoxicity of the TFPP-glucose
conjugates. In addition, the cellular uptake of these conjugates in RGK cells suggested that
the cellular internalization of 2OH, 3OH, and 4OH was partly facilitated by the GLUT1
pathway. Accordingly, the specificity of these TFPP-glucose conjugate isomers would be
expected to be reduced by reducing their hydrophobicity.
4. Experimental
4.1. Materials
All chemicals were of analytical grade. Dulbecco’s modified Eagle’s medium (DMEM)
was purchased from Nissui pharamaceutical Co. Ltd. (Tokyo, Japan). Fetal calf serum
(FCS) was purchased from Hyclone Laboratories, Inc. (Logan, UT, USA). Antibiotic-
Antimyconic was purchased from Life Technologies Japan Ltd. (Tokyo, Japan). DMEM
nutrient mixture F-12 HAM (F-12 Ham/DMEM, 1/1) and tetraphenylporphyrin
tetrasulfonic acid (TPPS) and 1,3-diphenylisobenzofuran (DPBF) were purchased from
Sigma-Aldrich
Japan.
Cytochalasin
B,
hematoporphyrin
(HP),
3,5-
bis(trifluoromethyl)benzoic acid, Dulbecco’s phosphate buffered saline (PBS) and dimethyl
sulfoxide (DMSO) were purchased from Wako Pure Chemical Industries, Ltd. (Osaka,
25

Japan). Hydroxyphenyl fluorescein (HPF) was purchased from Goryo Chemical, Inc.
(Tokyo, Japan). 5,10,15,20-Tetrakispentfluorophenyl porphyrin (TFPP) was prepared
36
according to the previous literature.
5-[4-(2,3,4,6-Tetra-O-acetyl-1-thio- -D-
glucopyranos-1-S-yl)-2,3,5,6-tetrafluoro-phenyl)]-10,15,20-tris(2,3,4,5,6-
pentafluorophenyl)porphyrin (1Ac) and 5-[4-(1-thio-D-glucospyranos-1-S-yl)-2,3,5,6-
tetrafluorophenyl)]-10,15,20-tris(pentafluorophenyl)porphyrin (1OH) were prepared
17a
according to our previous papers
Stock solutions of photosensitizers were prepared by
weighing the dried photosensitizers and dissolving them in DMSO, and kept in freezer
(‒30°C) until use.
4.2. General Methods
Reverse-phase thin layer chromatography (RP-TLC) was carried out using R-18F254s
(Merck Japan Ltd., Tokyo, Japan). The purities of peracetyrated TFPP-Glucose conjugates
2Ac, 3Ac, 4Ac and 6Ac were determined to be >99% on an HPLC system (Jasco PU-2086
Plus Intelligent pump, Jasco Co., Ltd., Tokyo, Japan) equipped with a UV-vis detector
(UV-2075 Plus, Jasco Co., Ltd.) and a silica gel column (COSMOSIL 5SL-II packed
column, 4.6 mmφ × 150 mm, Nacalai Tesque, Inc., Kyoto, Japan) using a mixture of
CH2Cl2 and ethyl acetate (9/1, v/v) at 30°C. The purities of TFPP-Glucose conjugates
2OH, 3OH, 4OH and 6OH were determined to be >97% on an HPLC system (Jasco PU-
2086 Plus Intelligent pump) equipped with a UV-vis detector (UV-2075 Plus) and a silica
gel column (COSMOSIL 5SL-II packed column, 4.6 mmφ × 150 mm, Nacalai Tesque,
Inc.) using a mixture of CH2Cl2 and CH3OH (9/1, v/v) at 30°C. Mass spectra were recorded
26

using JEOL spiral TOF JMS-S3000 (MALDI-TOF) (JEOL Ltd., Tokyo, Japan). UV-vis
spectra were recorded on UV-2550 spectrophotometer (Shimadzu Co., Kyoto, Japan) and
V-570 spectrophotometer (JASCO Co., Ltd., Tokyo, Japan). Steady-state fluorescence (FL)
spectra were recorded on FP-6300 spectrofluorometer (JASCO Co., Ltd., Tokyo, Japan).
NMR spectra were recorded using an AVANCE III HD (500 MHz; Bruker Biospin K.K.,
Yokohama, Japan) and a NMR-DD2 500PS (500 MHz; Agilent Technologies, CA, USA).
Luminescence spectra of singlet oxygen sensitized by each compound's solution was
recorded using a spectrometer (Jobin Yvon SPEX fluorolog3, HORIBA, Ltd., Kyoto,
Japan) equipped with a photomultiplier (NIR-PMT R5509-72, Hamamatsu Photonics K.K.,
Shizuoka, Japan) cooled to 193 K. The absorbance and the fluorescence intensity of each
well were determined using plate readers (Multiscan JX and Fluoroskan Ascent, Thermo
Fisher Scientific Co., Yokohama, Japan).
4.3.1. 1,3,4,6-Tetra-O-acetyl-2-S-acetyl-2-thio-β-D-glucopyranose (2-Ac-S-AcGlc, CCDC
1581347)18-19
A 0.05 M solution of 2-Tf-O-AcMan (501 mg, 1 mmol) in dimethylformamide (DMF, 21
mL) was treated with potassium thioacetate (KSAc, 1.2 g, 10 mmol) at room temperature
under nitrogen atmosphere. The reaction mixture was stirred for 2h and was partitioned
between diethyl ether (Et2O) and water. The organic phase was dried, filtered, and
concentrated in vacuo. Purification by column chromatography on silica with hexane : ethyl
acetate (EtOAc) = 3 : 1 as the eluant followed by recrystallization from Et2O gave 2-Ac-S-
AcGlc as colorless solids (354 mg, 84%). Rf value 0.46 (hexane : EtOAc = 1 : 1). m.p.
27

117–119°C. HRMS (ESI) : m/ z for C16H22NaO10S [M+Na]+ calcd 429.08314, found
429.08357 (error 0.43 mmu, 1.00 ppm). IR (NaCl, neat) νmax 1755, 1707, 1369, 1221, 1103,
1
1049 cm-1. H NMR (CDCl3, 500.16 MHz, Si(CH3)4 = 0 ppm): δ (ppm) = 5.83 (1H, d, 3J2
3
= 9.0 Hz, 1-GlcH), 5.31 (1H, dd, J2 = 11.5 Hz, J4 = 9.0 Hz, 3-GlcH), 5.10 (1H, dd, J5 =
3
3
3
10.0 Hz, J3 = 9.0 Hz, 4-GlcH), 4.32 (1H, dd, J6’ = 12.5 Hz, J5 = 4.5 Hz, 6-GlcH), 4.10
3
3
3
3
3
(1H, dd, J6 = 12.5 Hz, J5 = 1.5 Hz, 6’-GlcH), 3.85 (1H, ddd, J4 = 10.0 Hz, 3J6 = 4.5 Hz,
3
3
3J6’ = 1.5 Hz, 5-GlcH), 3.73 (1H, dd, J3 = 11.0 Hz, J1 = 9.5 Hz, 2-GlcH), 2.34 (3H, s,
SCOCH3), 2.09 (6H, s, 2COCH3), 2.017 (3H, s, COCH3), 2.015 (3H, s, COCH3). 13C NMR
(CDCl3, 125.77 MHz, CDCl3 = 77.0 ppm): δ (ppm) = 192.3 (SCOCH3), 170.7 (OCOCH3),
170.0 (OCOCH3), 169.5 (OCOCH3), 169.0 (OCOCH3), 92.0 (1-GlcC), 72.6 (3-GlcC), 71.0
(4-GlcC), 68.8 (5-GlcC), 61.5 (6-GlcC), 47.1 (2-GlcC), 30.7 (SCOCH3), 20.7 (OCOCH3),
20.6 (OCOCH3), 20.5 (OCOCH3). [α]3D0 +15.9 (c 1.00 in CHCl3).
4.3.2. 1,2,4,6-Tetra-O-acetyl-3-S-acetyl-3-thio-D-glucopyranose (3-Ac-S-AcGlc)20-23
A solution of Diacetone-3-deoxy-Ac-S-Glc (330 mg, 1 mmol) in CF3COOH/H2O (4:1, 1.4
mL) was stirred at room temperature under nitrogen atmosphere for 30 min. Concentration
of the mixture gave a colorless residue, which was dissolved in acetic anhydride (Ac2O, 1.9
mL) containing sodium acetate (NaOAc, 48 mg, 0.6 mmol). This mixture was heated at
reflux (10 min) before being poured into ice/water. Standard work-up (EtOAc) and column
chromatography on silica with hexane : EtOAc = 3 : 1 as eluant gave 3-Ac-S-AcGlc as a
yellowish oil (180 mg, 42%). Rf value 0.40 (hexane : EtOAc = 1 : 1). HRMS (ESI): m/ z for
C16H22NaO10S [M+Na]+ calcd 429.08314, found 429.08323 (error 0.09 mmu, 0.22 ppm).
28

1
IR (NaCl, neat) νmax 1753, 1702, 1373, 1221, 1068 cm-1. H NMR (CDCl3, 500.16 MHz,
Si(CH3)4 = 0 ppm): δ (ppm) = 5.72 (1H, d, 3J2 = 8.0 Hz, 1-GlcH), 5.12 (2H, m, 2-,6-GlcH),
3
4.26 (1H, m, J6’ = 12.5 Hz, 3J5 = 5.0 Hz, 4-GlcH), 4.11 (1H, m, 5-GlcH), 3.86 (2H, m, 3-
,6’-GlcH), 2.34 (3H, s, SCOCH3), 2.10 (3H, s, COCH3), 2.08 (3H, s, COCH3), 2.03 (6H, s,
2COCH3). 13C NMR (CDCl3, 125.77 MHz, CDCl3 = 77.0 ppm): δ (ppm) = 193.1
(SCOCH3), 170.7 (OCOCH3), 169.3 (OCOCH3), 169.2 (OCOCH3), 169.0 (OCOCH3), 93.1
(1-GlcC), 75.3 (2-GlcC), 69.1 (4-GlcC), 66.7 (5-GlcC), 61.7 (6-GlcC), 47.7 (3-GlcC), 30.6
(SCOCH3), 20.8 (OCOCH3), 20.7 (OCOCH3), 20.52 (OCOCH3), 20.51 (OCOCH3). [α]2D8
+17.8 (c 1.00 in CHCl3).
4.3.3. 1,2,3,6-Tetra-O-acetyl-4-S-acetyl-4-thio-D-glucopyranose (4-Ac-S-AcGlc, CCDC
1581348 for β- form)24-25
4-Ac-S-BzGlc (20 mg, 30 µmol) was debenzoylated in methanol (MeOH, 0.6 mL) with
sodium methoxide (MeONa, 0.1 M, 31 µL). After stirring at room temperature for 60 min,
the mixture was neutralized with acetic acid (AcOH) and concentrated. The resulted residue
was dissolved in anhydrous pyridine (83 µL), cooled to 0°C and Ac2O (24 µL, 26 µmol)
was added. After 30 min the reaction mixture was allowed to reach room temperature,
stirred for additional 5h, solvent was removed under reduced pressure and the residue was
dissolved in dichloromethane (CH2Cl2, 10 mL). The solution was washed with saturated
sodium hydrogen carbonate (NaHCO3) solution followed by 0.2 M hydrochloric acid (HCl)
aq. and water, dried, concentrated under reduced pressure, and purified by column
chromatography on silica with hexane : EtOAc = 5 : 1 to give the product 4-Ac-S-AcGlc as
29