The 2-aminoglucosamide motif improves cellular uptake and photodynamic activity of tetraphenylporphyrin

Article abstract of DOI:10.1016/j.ejmech.2005.04.007

Several strategies have been proposed to improve the efficiency of photosensitizers used in photodynamic therapy (PDT). In this context, the synthesis of mono- (1) and di-glucosylated (2) porphyrins, and mono-glucosylated chlorin (3) was performed. HT29:h

Full text of DOI:10.1016/j.ejmech.2005.04.007

European Journal of Medicinal Chemistry 40 (2005) 1111–1122
www.elsevier.com/locate/ejmech
Original article
The 2-aminoglucosamide motif improves cellular uptake
and photodynamic activity of tetraphenylporphyrin
Benoît Di Stasio a,b, Céline Frochot a, Dominique Dumas c, Pascale Even a, Jurriaan Zwier d,
Audrey Müller a, Jacques Didelon b, François Guillemin b, Marie-Laure Viriot a,
Muriel Barberi-Heyob b,*
a DCPR-GRAPP, UMR 7630 CNRS-INPL, Groupe ENSIC, 1, rue Grandville, 54000 Nancy, France
b Centre Alexis Vautrin, CRAN-UMR 7039 CNRS-UHP-INPL, Avenue de Bourgogne, Brabois, 54511 Vandœuvre-les-Nancy, France
c LEMTA, Equipe Mécanique et Ingénierie Cellulaire et Tissulaire-Service Imagerie Cellulaire, UMR 7563 CNRS-INPL-UHP, Faculté de Médecine,
BP 184, 54505 Vandœuvre-les-Nancy, France
d University of Amsterdam, Institute of Molecular Chemistry, Organic Photonic Materials Group, Nieuwe Achtergracht 129, 1018 WS, The Netherlands
Received 8 December 2004; received in revised form 17 March 2005; accepted 27 April 2005
Available online 15 June 2005
Abstract
Several strategies have been proposed to improve the efficiency of photosensitizers used in photodynamic therapy (PDT). In this context,
the synthesis of mono- (1) and di-glucosylated (2) porphyrins, and mono-glucosylated chlorin (3) was performed. HT29 human adenocarci-
noma cells were significantly more sensitive to asymmetric and less hydrophobic glucosylated photosensitizers-mediated PDT (1, 3), com-
pared to tetraphenylporphyrin (TPP). The lowest photosensitivity observed for TPP was consistent with the lowest uptake. Moreover, the most
pronounced photodynamic activity measured for 3 was in relation with the improvement of cellular uptake, the singlet oxygen quantum yield
and the high extinction coefficient value at 650 nm compared to porphyrins. Cellular localization analysis showed that 1 and 3 accumulated
mainly inside the endoplasmic reticulum.
© 2005 Elsevier SAS. All rights reserved.
Keywords: Photodynamic therapy; Glucosylation; Porphyrin; Chlorin; Targeting; In vitro
1. Introduction
ophthalmic diseases [1]. PDT is based on the use of photo-
sensitizers, which are preferentially taken up and/or retained
by diseased tissues. Upon photo-activation with visible light
at the appropriate wavelength, the generation of cytotoxic spe-
cies, such as reactive singlet oxygen, leads to irreversible
destruction of the treated tissues [2]. Normal cells, however,
are also able to accumulate photosensitizers and be damaged
by them, so that prolonged skin photosensitization, light-
sensitivity of the eye and other side effects have proved to be
severe limitations of PDT.
The ideal drug delivery system should enable the selective
accumulation of the photosensitizer within the diseased tis-
sue and the delivery of therapeutic concentrations of photo-
sensitizer to the target site with little or no uptake by non-
target cells. The carrier must also be able to incorporate the
photosensitizer without loss or alteration of its activity. In
view of the high probability of repetitive dosing schedules,
the system must also be biodegradable and have little or no
immunogenicity [3]. Another reason for using targeting or
Photodynamic therapy (PDT) is a promising treatment for
a variety of oncological, cardiovascular, dermatological, and
Abbreviations: ATP, adenosine triphosphate; Boc, tert-butoxycarbonyl;
BSA, bovine serum albumin; CDCl3, chloroform-d; CH2Cl2, dichloro-
methane; DCC, dicyclohexylcarbodiimide; DDQ, 2,3-dichloro 5,6-dicyano-
1,4-benzoquinone; DiOC6, 3,3-dihexyloxacarbocyanine iodide; DMAP,
4-dimethylaminopyridine; DMF, dimethyl formamide; DMSO, dimethyl sul-
foxide; EDCI, 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochlo-
ride; EDTA, ethylene diamine tetra acetic acid; EtOH, ethanol; HCl, hydro-
chloric acid; BtOH, 1-hydroxybenzotriazole; I.D., internal diameter; KQ,
quenching constant; L.D.50, light doses yielding 50% growth inhibition;
MeOH, methanol; MgSO4, magnesium sulfate; MTT, 3-(4,5-dimethylthiazol-
2-yl)-2,5-diphenyltetrazolium bromide; NaHCO3, sodium bicarbonate;
NaOH, sodium hydroxide; NHS, N-hydroxysuccinimide; NMR, nuclear
magnetic resonance spectroscopy; PBS, phosphate buffered saline; P1COOH,
4-carboxyphenylporphyrin; PDT, photodynamic therapy; RP-HPLC, rever-
sed phase high performance liquid chromatography; TFA, trifluoroacetic acid;
THF, tetrahydrofuran; TMS, tetramethylsilane; TPP, tetraphenylporphyrin.
* Corresponding author. Tel.: +33 3 83 59 83 06; fax: +33 3 83 44 60 71.
E-mail address: m.barberi@nancy.fnclcc.fr (M. Barberi-Heyob).
0223-5234/$ - see front matter © 2005 Elsevier SAS. All rights reserved.
doi:10.1016/j.ejmech.2005.04.007

1112
B. Di Stasio et al. / European Journal of Medicinal Chemistry 40 (2005) 1111–1122
carriers is to provide an environment where the photosensi-
tizer can be administered in a monomeric form. Indeed, due
to their chemical composition, most of the photosensitizers
tend to aggregate in aqueous media as a result of the propen-
sity of the hydrophobic skeleton to avoid contact with water
molecules [4]. This state is one of the determining factors,
which can hinder the efficacy of the drug in vivo by increas-
ing its bioavailability and limiting its capacity to absorb light
[5]. An approach developed by several groups is to modulate
the amphiphilicity of the photosensitizer [6]. Structural modi-
fications induced by glycoconjugation of the tetrapyrrole sys-
tem appear as an effective mean to create the equilibrium
between hydrophilicity and hydrophobicity. Moser et al. mea-
sured a 50–100-fold higher photocytotoxicity for diglucoa-
mido porphyrinoids compared to non glucosylated ana-
logues [7]. The position of sugar substituent is crucial to photo
biological activity [8]. Whether glycoconjugation is obvi-
ously a good means to introduce such a balance between
hydrophilicity and hydrophobicity [9], the nature of the sugar
residues seems to take a significant part in the photosensitiz-
ing properties and remains to be elucidated. This should be
undertaken in the light of the knowledge of the various lec-
tines occurring at the surface of the cell membrane and their
implication in glycoconjugated dyes internalization. Py-
ropheophorbide 2-deoxyglucosamide was evaluated in a 9L
glioma rat model as a new photosensitizer targeting glucose
transporters (GLUT). Fluorescence imaging studies demon-
strated that this conjugate was trapped in tumor via the
GLUT/hexokinase pathway [10]. If O- or S-glycosylated por-
phyrins have already been synthesized, no report has appeared
concerning aminoglycosamide tetraphenyl of porphyrins and
chlorin. In comparison with O-glycosylated photosensitiz-
ers, CONH– glycosyl bonds should resist endogenous hy-
drolysis catalyzed by glycosidases [11].
active compounds have been characterized by photophysical
properties (molar extinction coefficient, fluorescence and sin-
glet oxygen quantum yields). Cellular internalization pro-
cess, cellular localization and photocytotoxic properties of
1–3 have been compared to those of tetraphenylporphyrin
(TPP) in HT29 human adenocarcinoma cells.
2. Results and discussion
2.1. Synthesis, photophysical and chemical characteristics
2.1.1. Synthesis
The synthesis of novel conjugates 1 and 3 is shown in
Scheme 1, the synthesis of 2 in Scheme 2. The synthesis of
the O-acetylated glucosamine has already been described [12].
Synthesis of 1 was carried out in CH2Cl2/DMF by cou-
pling the O-acetylated glucosamine with 5-(4-carboxy-
phenyl)-10,15,20-triphenyl-21,23-porphyrin (4), mediated by
1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochlo-
ride (EDCI) at 0 °C. The compound was purified by chroma-
tography and the yield of pure product was about 65%. The
compound 4 was obtained according to the Little method [13]
by condensation of pyrrole (4 eq.) with benzaldehyde (3 eq.)
and carboxybenzaldehyde (1 eq.) in propionic acid.After puri-
fication by SiO2 and C18 chromatography the yield was 7%.
The 5-(4-carboxyphenyl)-5,10,15-triphenyl-chlorin (5) was
prepared by the diimide reduction of the corresponding por-
phyrin (yield: 70%) [14]. EDCI has been also used to couple
O-acetylated glucosamine to the monocarboxyphenyl chlo-
rin 5 to afford the monoglucosylated chlorin 3. The glyco-
conjugated benzaldehyde 7 was obtained by amidation of the
carboxybenzaldehyde using two equivalents of O-acetylated
glucosamine and the coupling reagents DCC-BtOH in DMF-
CH2Cl2. The deprotection of the aldehyde function (partial
imine formation was observed) was performed with hydro-
chloric acid (HCl) 5 N and the product obtained after chro-
matography column with a yield of 87%.
In this context, glycosamide porphyrins and the corre-
sponding chlorin have been synthesized in order to study the
influence of structural modifications induced by symmetric
or asymmetric glycoconjugation on photophysical properties
and photosensitivity. Synthesis of mono- and di-glucosylated
porphyrins were performed, 1 and 2, respectively, and the
corresponding mono-glucosylated chlorin 3. The photo-
The trans-biglucosylated porphyrin has been synthesized
using an intermediate step that is to say a glucosylated dipyr-
romethane. This dipyrromethane is stable in the purified form
Scheme 1
. Synthesis of 5-[4-(1,3,4,6-tetra-O-acetyl-2-amido-2-desoxy-b-D-glucopyranose phenyl)]-10,15,20-triphenyl porphyrin (1) and 5-[4-(1,3,4,6-tetra-
O-acetyl-2-amido-2-desoxy-b-D-glucopyranose phenyl)]-10,15,20-triphenyl chlorin (3).

B. Di Stasio et al. / European Journal of Medicinal Chemistry 40 (2005) 1111–1122
1113
Scheme 2
.
Synthesis of 5,15-bis[4-(1,3,4,6-tetra-O-acetyl-2-amido-2-desoxy-b-D-glucopyranose phenyl)]-10,20-biphenyl porphyrin (2).
in absence of light and air and the reaction of the meso-
substituted dipyrromethane with the benzaldehyde under the
conditions of the two-step one flask porphyrin synthesis
affords a direct route to trans-substituted meso-porphyrin [15].
The synthesis was carried out by condensation of the acetylg-
lycocongugate benzaldehyde with a large excess of pyrrole
and a catalytic amount of trifluoroacetic acid (TFA). The yield
is 85%. The dipyrromethane could be then condensed with
the benzaldehyde to afford the trans-substituted porphyrin 2.
The purification was done by chromatography and the pure
product was obtained with 7% yield, which is acceptable for
this kind of synthesis. The structures of the purified conju-
recognition on photodynamic activity. The in vitro photocy-
totoxic results of tetracarbohydrated porphyrins indicated that
the glucose moiety protected with acetyl groups specifically
increased the incorporation of the drug into the cell [17].
2.1.2. Photophysical characteristics
The absorption spectra of photo-active compounds were
typical of porphyrin derivatives with a Soret band of about
417 nm with high extinction coefficient (e, Table 1) and four
Q bands at ca. 515, 550, 590 and 650 nm (Fig. 1). The pres-
ence of amino glucosylated groups leads to a low decrease of
extinction coefficients in the Soret band (Table 1). As ex-
pected, 3 displayed a high extinction coefficient value at
650 nm compared to porphyrins (Table 1).
The fluorescence quantum yield values were calculated by
steady state comparative method using TPP as a reference
[18]. The fluorescence quantum yield values were compa-
rable (Table 1). Singlet oxygen formation quantum yields val-
ues were calculated by direct measurement of the infrared
luminescence (1270 nm), using TPP as a reference. 3 dis-
played the highest U 102 value (Table 1) and it was the only
chlorin tested in this study. This result is in good agreement
with Mikata et al. who demonstrated that the sugar-linked
chlorins were more effective than TPP tetrasulfonic acid
1
gates were corroborated by H nuclear magnetic resonance
spectroscopy (NMR), UV–Vis analysis and mass spectra
(MALDI-TOFMS, data not shown).
We chose to study glucosamide conjugates of 4 in which
the b-glucose is protected. Indeed, we previously demon-
strated that when glucosylated fluorescent tracers had
hydroxyl-protected sugar with acetyl groups, they did not
cause toxicity in comparison with those linked with free
hydroxyl sugars [16]. Moreover, Mikata et al. [17] reported
the in vitro photocytotoxicity of four families of hydroxyl-
protected and unprotected sugar-linked TPPs, in order to
clarify the effect of amphiphilic character and carbohydrate
Table 1
Molar extinction coefficients (e), fluorescence emission maximum wavelength (kmax), fluorescence (Uf) and singlet oxygen formation (U 1O2) quantum yields,
and triplet state lifetimes (s)
Photosensitizer
kmax
e a
Soret band
117
k (nm)
e
kmax
Q bands
Uf
U 1O2
s1
s2
1
415
420
415
417
650
646
650
650
1.2
3.9
6.9
3.4
652/720
654/718
652/720
654/718
0.09
0.1
0.55
0.33
0.73
0.55
1.15 0.03 35
n.d. n.d.
6
2
149
3
58
0.16
0.11
1.03 0.03 70 17
1.45 0.08 66 12
TPP
556
s1: Triplet state life time in aerated solution (µs).
s2: Triplet state life time in argon saturated solution (µs).
n.d.: Not determined.
a (103 l mol–1 cm–1).

1114
B. Di Stasio et al. / European Journal of Medicinal Chemistry 40 (2005) 1111–1122
the glucosylated porphyrins and the singlet oxygen forma-
tion [20]. In our case, only the number of sugars changes and
the few compounds tested does not allow us to discuss any
correlation. 1 and 3 triplet state lifetimes were measured by
laser flash photolysis in the presence of aerated solution and
in argon-saturated solutions. Typical triplet–triplet absorp-
tion spectrum is shown on Fig. 2. The triplet–singlet differ-
ence spectra show a maximum absorption in the blue spectral
region at 450 nm. The first peak sometimes found in the por-
phyrin triplet–triplet absorption spectrum [21] located in the
near-UV region (330–350 nm) can not be distinguished from
the noise. In the presence of oxygen, the decays follow a first-
order dependence (data not shown) and this is essentially the
same under nitrogen. The triplet states of these compounds
are efficiently quenched by oxygen (Table 1).
Fig. 1. Absorption and emission spectra of the glucosylated compounds and
2.1.3. Chemical characteristics
TPP in ethanol (normalized). For emission spectra, excitation wavelength
was performed at 417 nm.
The amphiphilicity of dyes is a characteristic, which may
be decisive for photosensitizing activity since this parameter
may influence their ability to cross membrane as well as their
localization within the cell. High performance liquid chroma-
tography (HPLC) analysis allows us to compare the hydro-
phobic characteristic between the different photo-active
compounds. This was done by reverse phase liquid chroma-
tography. The retention times for TPP, 2 and 1 were
35.60 0.04, 26.3 0.03 and 23.3 0.05 min, respectively.
which is known as a photosensitizer that produce singlet oxy-
gen efficiently [19]. Singlet oxygen formation quantum yields
seem to be influenced by the number of glucosylated group,
illustrating that the changes in the structure could influence
the triplet quantum yield and/or the efficiency of energy trans-
fer to molecular oxygen (Table 1). Oulmi et al. have found a
linear dependence between the number of the aryl groups of
Fig. 2
106 l mol–1 s–1, respectively.
.
Transient absorption spectrum (kexc, 515 nm and increment, 50 ns) (c) of 1 [A] and 3 [B] in DMSO with quenching constants (KQ), 98 × 104 and 111 ×

B. Di Stasio et al. / European Journal of Medicinal Chemistry 40 (2005) 1111–1122
1115
Both isomers were found for 3: 18.1 0.02 and
18.9 0.02 min. Actually, isomeric 2,3- and 7,8 mono-
glucosylated chlorins were obtained as an inseparable mix-
ture. These two isomers are different by the position of the
reduced double bond in the macrocycle. TPP was the most
hydrophobic compound and conversely, 3 was more hydro-
philic than mono- and di-glucosylated porphyrins. The
increase of acetylated D-glucosamine group number around
the macrocycle confers a decrease of hydrophobic character
to the compounds, confirmed by the values of retention times.
Aggregation and hydrophobicity are conclusive factors, which
can influence photosensitizer uptake [9]. For in vitro quanti-
tative structure-activity relationship studies, it was noted that
amphiphilic compounds exhibit greater photosensitizing abil-
ity than symmetrically hydrophobic or hydrophilic com-
pounds [22]. This was most clearly demonstrated by studies
showing that disulfonated phthalocyanines and TPPs with the
two sulfonate groups adjacent to each other were more active
than derivatives with two sulfonate groups on opposite sides
of the chromophore [23].
higher hydrophobicity (TPP) corresponding to lower uptake.
For mono- or di-glucosylated compounds (1 and 2), it is not
so clear as position of sugar substituent could also influence
cellular uptake and photobiological activity. Improvement of
cellular uptake for the glucosylated compounds was partly
related to the increase of their hydrophilicity. Nevertheless,
gross cellular uptake was not in itself the only determinant of
PDT activity, but localization to a sensitive subcellular target
may be important. In this current study, fluorescence micros-
copy was used to examine the subcellular localization pat-
terns of glucosylated compounds.
2.3. Antiproliferative assay
The 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium
bromide (MTT) test was used to evaluate the cytotoxicity of
TPP, 1–3 in the dark for concentrations varying from 0.5 to
30.0 µM (Fig. 4). Experiments, yielding surviving cell frac-
tions higher than 85%, demonstrated that incubation of
HT29 cells with the above TPP concentrations for 24 h,
induced no cytotoxicity in the absence of light exposure. Fig. 4
demonstrates that 24-h incubation with mono-glucosylated
compounds (1 and 3) induced no cytotoxicity for concentra-
tions lower than 10 µM. However, 2 revealed more cytotoxic
effects than mono-glucosylated compounds (Fig. 4).
2.2. Cellular uptake kinetics
The HT29 cellular uptake of TPP, 1–3, as a function of
incubation times, was examined for a non-cytotoxic photo-
sensitizer concentration (10 µM, Fig. 3). During the first 3 h,
the photosensitizers accumulated similarly in HT29 cells with
no significant difference in the mean normalized fluores-
cence intensities of the cell suspension. From 15-h incuba-
tion, the accumulation of glucosylated photosensitizers was
significantly higher (P < 0.05) than the cellular uptake of TPP
(Fig. 3). The cellular uptake of 1 after a 24-h exposure was
on average about 11.5-fold higher than the concentration of
TPP and the accumulation of 2 and 3 compounds reached
5.0- and 6.4-fold more, respectively. The intracellular kinetic
experiments of glucosylated compounds demonstrated that a
steady state was reached at about 20 h. It may be noted that
accumulation can be related to the dye retention time values,
2.4. Photocytotoxicity
The cells were incubated with the photo-active com-
pounds at 10–6 M and irradiated by red light as we previously
described [24], except that HT29 cells were exposed to the
dye for 24 h at 37 °C. The intracellular uptake kinetics dem-
onstrated that a steady state was reached at about 20-h expo-
sure (Fig. 3). Because one aim of this work was to improve
photosensitization to red light, which is a light wavelength
able to enter deeply in living tissues, we compared the activ-
ity of glucosylated compounds and of TPP following red light
irradiation (k = 650 nm, fluence 4.54 mW cm–2). The light
Fig. 3. Cellular uptake time course as determined from the cellular fluores-
cence of HT29 intensities with [lozenge] 1, [triangle] 2, [square] 3, and [cir-
cle] TPP at 10 µM. Cellular fluorescence intensities were measured at 650 nm.
For each experiment, the fluorescence intensity of each sample was norma-
lized to cell concentration.
These data represent the mean values from three or four independent expe-
riments. Error bars are standard deviations.
Fig. 4. Survival of HT29 cells treated with [black] 1, [wide shading] 3, [fine
shading] 2 for 24-h incubation without irradiation. Concentration depen-
dence of cytotoxicity for glucosylated compounds. Experiments were per-
formed by MTT test. Data represent the mean values from three independent
experiments. Error bars are standard deviations.

1116
B. Di Stasio et al. / European Journal of Medicinal Chemistry 40 (2005) 1111–1122
with 3 (Fig. 6F). The diffuse cytoplasmic fluorescence
observed for cells treated with glucosylated photosensitizers
is consistent with previous results reported in literature [25].
Fig. 6 clearly shows the advantage of introduction of sugar
moiety in membrane permeability of sugar-linked photosen-
sitizer.
Confocal images of the double stained HT29 cells with 1
and DiOC6 endoplasmic reticulum marker are presented in
Fig. 7. The representative image of 1 is shown in red (Fig. 7B),
the organelle-specific dye in green (Fig. 7A), and the over-
lapped image in yellow (Fig. 7C). Subcellular localization
analysis with organelle probes in HT29 cells shows that 1
accumulated mainly inside the endoplasmic reticulum
(Fig. 7C). Fluorescence topographic profiles of DiOC6 and
of 1 checked and proved this localization (Fig. 8). The mito-
chondrial image showed no overlap with the corresponding 1
image. Moreover, accumulation and the staining pattern of
lysosomal probe Lucifer Yellow® and 1 were different, thus
indicating that very little 1 accumulated inside lysosomes (data
not shown).According to Laville et al., the tri-glucoconjugated
chlorin could be partly internalized via an active receptor-
mediated endocytosis mechanism and this higher phototox-
icity has been related to a greater mitochondrial affinity [25].
In our case, mitochondrial-localizing mitotracker® and 1 fluo-
rescence patterns were different, indicating a poor affinity of
1 for this organelle. There is strong evidence that photosen-
sitizers with an acute localization in mitochondria promote
the release of cytochrome c upon irradiation [26]. This loss
of cytochrome c can be lethal to cells either because of the
disruption of the mitochondrial respiratory chain with the
eventual reduction of cellular ATP levels, or through caspase
initiation with subsequent apoptotic cell death [27]. The
impact of PDT on the endoplasmic reticulum has not very
often been taken into account. Only a few studies have dem-
onstrated the importance of photochemical damage to endo-
plasmic reticulum for the inactivation of tumor cells in cul-
ture [28–31]. The endoplasmic reticulum is mainly localized
in the perinuclear area of the cytoplasm, and like Golgi appa-
ratus they also interact together in the case of newly synthe-
sized proteins. The endoplasmic reticulum is known to play a
central role in the biosynthesis, segregation, and transport of
proteins and lipids, as well as in the release of intracellular
stores of calcium [32].
Fig. 5. Measurement of PDT sensitivity for [lozenge] 1, [triangle] 2, [square]
3, and [circle] TPP in HT29 cells. Survival curves, obtained by MTT test,
were assessed for cells exposed to increasing doses of light from 1.0 to
20.0 J cm–2 (for details see Section 4.6). Cells were incubated with photo-
sensitizers at 10–6 M for 24 h before light treatment. Data represent the mean
values from three independent experiments. Error bars are standard devia-
tions.
doses yielding 50% growth inhibition (L.D.50) values of 3, 1,
2 and TPP were 4.4, 5.2, 11.1 and 30.2 J cm–2, respectively.
Thus, 3 and 1 were statistically significantly more efficient
6.9-fold and 5.8-fold, respectively, compared to TPP
(P < 0.05). Fig. 5 shows that TPP control photosensitizer dis-
played a low photocytotoxicity in our experimental condi-
tions. Conversely, survival measurements with the MTT test
demonstrated that using glucosylated compounds, the photo-
sensitivity was improved in comparison to TPP-mediated pho-
tosensitization (Fig. 5). The lowest uptake observed for TPP
is consistent with the lowest photosensitivity (L.D.50 = 30.2 J
cm–2) and the most pronounced photodynamic activity mea-
sured for 3 (L.D.50 = 4.4 J cm–2) is in relation with the
1
improvement of cellular uptake (6.4-fold), the high U 02
value (0.73) and the high extinction coefficient value at
650 nm compared to porphyrins.
The 2 symmetric glucosylated porphyrin exhibited a low
photocytotoxicity compared to 1. This observation confirms
conclusions previously described. It was found that com-
pounds bearing asymmetric mono-saccharide units are, in
general, more phototoxic than symmetrical compounds [9].
2.5. Visualization of cellular uptake and subcellular
localization
We can also add that using protected sugar motifs acts in
favor of endoplasmic reticulum localization.Actually, we pre-
viously demonstrated that coumarins linked with free glu-
cosamine presented an intracellular fluorescence distribution
in the cytoplasmic compartment whereas those protected with
acetylated glucosamine showed an accumulation inside the
endoplasmic reticulum [16].
Cellular uptake of TPP and 1 (1 µM) was also compared
using confocal fluorescence microscopy by incubating
HT29 cells for 24 h (Fig. 6). The cells displayed a pattern of
weak intracellular TPP fluorescence emission, indicative of
extensive intracellular aggregation (Fig. 6D). Moreover, this
fluorescence emission appeared more restricted to the cyto-
plasmic compartment with little or no detectable nuclear stain-
ing observed. With the same concentration employed with
TPP, the cells exhibited an intense and diffuse intracellular 1
fluorescence (Fig. 6E). Comparable staining was observed
3. Conclusions
Our results show that glycoconjugation is obviously an
interesting means to introduce such a balance between hydro-

B. Di Stasio et al. / European Journal of Medicinal Chemistry 40 (2005) 1111–1122
1117
Fig. 6
. Visualization of cellular uptake by confocal fluorescence microscopy for HT29 cells exposed for 24-h incubation to TPP (A, D); 1 (B, E) and 3 (C, F).
Light transmitted (left column) and the corresponding fluorescence images (right column). Scale bars are indicated on images.
phylicity and hydrophobicity, previous observations suggest-
ing the requirement of amphiphilicity for efficient photody-
namic activity [33]. Thus, the improvement of cellular uptake
achieved for glucosylated compounds, was partly related to
the increase of hydrophylicity. The cellular uptake of 1 was
on average about 11.5-fold higher than the concentration of
TPP and the accumulation of 3 reached 6.4-fold more. The
higher photodynamic efficiency has been related to this greater
cellular uptake. Survival measurements using photocyto-
toxicity assay demonstrated that HT29 cells were signifi-
cantly more sensitive to asymmetric mono-glucosylated
photosensitizer-mediated PDT. Visualization of cellular
uptake demonstrated that the cells displayed a pattern of weak
intracellular TPP fluorescence emission, associated with an
intense intracellular aggregation. Subcellular localization
analysis with organelle probes showed that 1 accumulated
mainly inside the endoplasmic reticulum. Thus, we have suc-
cessfully synthesized new glucosylated porphyrins and chlo-
rin, which could be considered as potential agents in cancer
phototherapy.
4. Experimental
4.1. Synthesis
Chemicals and solvents were purchased from Aldrich
(Saint Quentin Fallavier, France) and SdS (Peypin, France).
Reactions were monitored by thin-layer chromatography
(Kieselgel 60 F254, SdS). Chromatographic purification was
performed with silical gel (200–400 mesh).

1118
B. Di Stasio et al. / European Journal of Medicinal Chemistry 40 (2005) 1111–1122
Fig. 7
.
Confocal fluorescence microscopy images of HT29 cells, double stained with 1 and an organelle probe. Dual staining confocal microscopy was perfor-
med for 1 emitting a red fluorescence signal (B). The cells were exposed to DiOC6 (A) staining the endoplasmic reticulum (green fluorescence) and the
overlapped image in yellow (C). The cells were exposed for 24-h incubation to 1 (1 µM), then exposed to DiOC6 (C).
Scale bars are indicated on images.
NMR spectra were recorded on aAvance Brucker 300 MHz
spectrometer. 1H NMR spectra are reported in units of d with
tetramethylsilane (TMS) resonance at 0 ppm as a reference.
MALDI sample preparation: a saturated ␣-cyano-4-
hydroxy-trans-cinnamic acid (CHCA) solution, in 50% aceto-
nitrile, 0.1% of trifluoroacetic acid, and 1 µl of sample were
spotted on the stainless steel MALDI targets. The molar ratio
of analyte to matrix was 1:104. The solvent was evaporated
prior to insertion in the source. Mass spectra were acquired
over the range 0–3000 Da.
The porphyrins and chlorin and the general procedure of
their synthesis are reported in Schemes 1 and 2.
5-(4-Carboxyphenyl)-10,15,20-triphenyl-21,23-porphy-
rin (4) was synthesized via condensation of benzaldehyde,
4-carboxybenzaldehyde and pyrrole under acid catalysis [34].
solution was then washed with 1 N HCl, NaHCO3 saturated
solution, distilled water and dried over magnesium sulfate
(MgSO4).
The compound was purified by SiO2 chromatography using
3:97 EtOH/CH2Cl2. The yield is 65%.
Rf 0.52 (SiO2, 4:96 EtOH/CH2Cl2), mp 184 °C, 1H NMR
(CDCl3) d: –2.70 (s, 2H, NH), 2.18 (s, 3H, CH3), 2.23 (s, 3H,
CH3), 2.24 (s, 3H, CH3), 2.32 (s, 3H, CH3), 4.00 (m, 1H,
Glc-H-5), 4.30, 4.42 (m, 2H, Glc-H-6, CH2OAc), 4.80 (m,
1H, Glc-H-2), 5.41 (m, 2H, Glc-H-3, Glc-H-4), 6.01 (d, 1H,
Glc-H-1, J = 8.7 Hz), 6.58 (s, 1H, CONH), 7.81 (m, 9H, Phe-
nyl), 8.12 (2H, d, ArH, J = 8.2 Hz), 8.26 (2H, ArH), 8.40 (m,
2H, d, ArH, J = 8.2 Hz), 8.82, 8.91 (m, 8H, pyrrole-H), MS
(MALDI-TOFMS) m/z: 988.35 Da.
4.1.2. Synthesis of the 4-(1,3,4,6-tetra-O-acetyl-2-amido-2-
desoxy-b-D-glucopyranose) benzaldehyde (6)
4.1.1. Synthesis of 5-[4-(1,3,4,6-tetra-O-acetyl-2-amido-2-
desoxy-b-D-glucopyranose phenyl)]-10,15,20-triphenyl
porphyrin (1)
To a solution of 2-amino-1,3,4,6,-tetra-O-acetyl-2-deoxy-
b-D-glucose (0.147 mmol, 51 mg) and 5-(4-carboxyphenyl)-
10,15,20-triphenyl-21,23-porphyrin (0.147 mmol, 100 mg)
in CH2Cl2 was added 1.1 eq. of 1-(3-dimethylaminopropyl)-
3-ethyl-carbodiimide (0.162 mmol, 31 mg) and catalytic
amount of DMAP (0.015 mmol, 1.8 mg). The mixture was
stirred for 3 h at 0 °C and 24 h at room temperature. The
A solution of 4-carboxybenzaldehyde (1 mmol, 0.150 g)
in 10 ml CH2Cl2 was cooled to 0 °C under argon atmosphere.
1-hydroxybenzotriazole (BtOH) (1.2 mmol, 0.162 g) in 4 ml
DMF and then DCC (1.2 mmol, 0.247 g) in 5 ml CH2Cl2 was
added drop by drop. The solution was stirred for 2 h at room
temperature. 2-Amino-1,3,4,6,-tetra-O-acetyl-2-deoxy-b-D-
glucose (2 mmol, 0.694 g) in 5 ml was added to the mixture
which was allowed to stir all night. The mixture was then fil-
tered and the solvent was removed under reduced pressure. The

B. Di Stasio et al. / European Journal of Medicinal Chemistry 40 (2005) 1111–1122
1119
Fig. 8
. Confocal fluorescence microscopy images of HT29 cells double stained with 1 and DiOC6 (A) and the corresponding image obtained by transmission
(B). Fluorescence topographic profiles of DiOC6 (D) correspond to the endoplasmic reticulum (D) and 1 (E) corresponds to mono-glucosylated porphyrin.
deprotection of the aldehyde function (partial imine formation
was observed) was realized with 3 ml HCl 5 N in 30 ml acetone.
The mixture was stirred for 15 min and acetone was removed
under reduced pressure. The product was dissolved in CH2Cl2,
the organic phase was washed with HCl 1 N, water and dried
(MgSO4). The solvent was removed under reduced pressure.
The product was obtained with 87% yield.
diluted with CH2Cl2 (50 ml) and washed with 0.1 N aq.
NaOH, water and dried with MgSO4. The solvent was
removed under reduced pressure and the unreacted pyrrole
by vacuum distillation at room temperature. The white prod-
uct is purified by chromatography on SiO2, using 4:96 EtOH/
CH2Cl2, the yield is 85%.
Rf 0.37 (SiO2, 4:96 EtOH/CH2Cl2), m.p. 250 °C, 1H NMR
(CDCl3) d: 1.92 (s, 3H, CH3), 1.99 (s, 3H, CH3), 2.00 (s, 3H,
CH3), 2.04 (s, 3H, CH3), 3.75 (m, 1H, Glc-H-5), 4.08, 4.23
(d, 2H, Glc-H-6 CH2OAc), 4.46 (m, 1H, Glc-H-2), 5.16 (m,
2H, Glc-H-3, Glc-H-4), 5.44 (s, 1H, CH), 5.72 (d, 1H, Glc-
H-1, J = 8.7 Hz), 5.81 (m, 2H, pyrrole), 6.06 (m, 2H, pyr-
role), 6.10 (s, 1H, CONH), 6.64 (s, 2H, pyrrole), 7.22 (d, 2H,
ArH, J = 8.1 Hz), 7.56 (d, 2H, ArH, J = 8.1 Hz), 7.89 (broad
band, 2H, NH pyrrole), MS.
Rf 0.66 (SiO2, 4:96 EtOH/CH2Cl2), m.p. 157 °C, 1H NMR
(CDCl3) d: 2.07 (s, 3H, CH3), 2.12 (s, 3H, CH3), 2.13 (s, 3H,
CH3), 2.21 (s, 3H, CH3), 3.50 (1H, m, Glc-H-2), 3.86 (1H,
m, Glc-H-5), 4.21, 4.30 (2H, dd, Glc-H-6), 5.28 (2H, m, Glc-
H-3, Glc-H-4), 5.84 (1H, d, Glc-H-1, J = 8.8 Hz), 6.35 (1H,
s, NHCO), 7.86, 7.97 (4H, d, Ar–H, J = 8.1 Hz), 10.1 (1H, s,
CHO).
4.1.3. Synthesis of meso-(4(1,3,4,6-tetra-O-acetyl-2-amido-
2-desoxy-b-D-glucopyranose) phenyl dipyrromethane (7)
A solution of the glycoconjugated benzaldehyde 6
(0.45 mmol, 200 mg) and pyrrole (18 mmol, 1.3 ml) was
degassed by bubbling with argon for 10 min. Trifluoroacetic
acid (0.045 mmol, 3.6 µl) was added. The solution was stirred
for 15 min more at room temperature. The mixture was then
4.1.4. Synthesis of 5,15-bis[4-(1,3,4,6-tetra-O-acetyl-2-
amido-2-desoxy-b-D-glucopyranose pheny)]-10,20-
biphenyl porphyrin (2)
A solution of meso-(4-(1,3,4,6-tetra-O-acetyl-2-amido-2-
desoxy-b-D-glucopyranose)) dipyrromethane (0.2 mmol,
119 mg) and benzaldehyde (0.2 mmol, 21 mg) in 20 ml

1120
B. Di Stasio et al. / European Journal of Medicinal Chemistry 40 (2005) 1111–1122
CH2Cl2 was purged with argon for 15 min, then trifluoroace-
tic acid (0.2 mmol, 15 µl) was added. The mixture was stirred
for 2 h at room temperature then DDQ (0.4 mmol, 91 mg)
was added. The mixture was stirred for 1 more h and the sol-
vent was evaporated. The compound was purified by SiO2
chromatography using 5:95 EtOH/CH2Cl2. The yield is 6%.
Rf 0.37 (SiO2, 4:96 EtOH/CH2Cl2), m.p. 186 °C, 1H NMR
(CDCl3) d: –2.76 (s, 2H, NH), 2.15 (s, 3H, CH3), 2.20 (s, 3H,
CH3), 2.22 (s, 3H, CH3), 2.28 (s, 3H, CH3), 3.80 (m, 1H,
Glc-H-5), 4.30, 4.41 (m, 2H, Glc-H-6, CH2OAc), 4.78 (m,
1H, Glc-H-2), 5.41 (m, 2H, Glc-H-3, Glc-H-4), 6.01 (d, 1H,
Glc-H-1, J = 8.7 Hz), 6.54 (s, 1H, CONH), 7.82 (m, 6H,
4 ArH), 8.12 (d, 4 ArH), 8.23 (d, 4 ArH), 8.32 (d, 4 ArH),
8.84, 8.90 (d, 8H, pyrrole-H, J = 4.6 Hz), MS (MALDI-
TOFMS) m/z: 1361.45 Da.
(m, 1H, CONH), 7.65–8.86 (m, 19H Phenyl, 6H pyrrole-H),
MS (MALDI-TOFMS) m/z: 990.37 Da.
4.2. Absorption and fluorescence
Absorption spectra were recorded on a Perkin–Elmer
(Lambda 2, Courtabœuf, France) UV–vis spectrophotom-
eter.
Fluorescence spectra were recorded on a SPEX Fluorolog-
2 spectrofluorimeter 1680 (JobinYvon, Longjumeau, France)
equipped with a thermostated cell compartment (25 °C), using
a 450 W Xenon lamp. Fluorescence quantum yields (Uf) were
determined using a TPP solution as fluorescence standard
(Uf = 0.11, in toluene, taking into account solvent refractive
index and absorption efficiencies).
4.1.5. Synthesis of 5-(4-carboxyphenyl)-10,15,20-
triphenyl-21,23-chlorin (5)
4.3. Determination of singlet oxygen quantum yield
(U 1O2)
C1COOH was synthesized as described by Whitlock et al.
[14]: 5-(4-carboxyphenyl)-10,15,20-triphenyl-21,23-porphyrin
(0.15 mmol, 100 mg), toluenesulfohydrazide (0.30 mmol,
81 mg) and potassium carbonate anhydrous (1.35 mmol,
186 mg) in 7 ml of dry pyridin were refluxed under argon.
Toluenesulfohydrazide in 0.2 ml of pyridin were added twice
at intervals of 2 h. After 6.5 h of reaction, the solution was
poured in a mixture of toluene and water (2:1) and stirred for
1 h under reflux at 90 °C. Organic phase was washed with
3 N HCl, water, saturated bicarbonate sodium solution and
dried over MgSO4. Visible spectrum of the product shows a
mixture of chlorin and bacteriochlorin.
Excitation occurred with a Xe-arc, the light was separated
in a SPEX 1680, 0.22 µm double monochromator. The detec-
tion at 1270 nm was done through a PTI S/N 1565 monochro-
mator, and the emission was monitored by a liquid nitrogen-
cooled Ge-detector model (EO-817L, North Coast Scientific
Co). The absorbance of the reference solution (TPP in CHCl3
UD = 0.55) [35] and the sample solution (at 515 nm) were
set equal (between 0.2 and 0.5) by dilution.
4.4. Determination of triplet state lifetime s
Laser flash photolysis was carried out at 295 K by using a
Coherent Infinity YAG laser equipped with an optical para-
metric oscillator (OPO). The pulse duration and energy were
5 ns and 4 mJ, respectively. The illuminated surface area was
0.1 cm2. Light from an EG and G FX504 low-pressure Xenon
flash lamp passing through the sample under 90° with respect
to the laser beam was used to probe the changes of the absorp-
tion spectra. The probe light was imaged using an Acton
Research Spectra pro 150 spectrograph onto a Princeton
Instruments ICCD-576-G/RB-EM gated intensified CCD
camera (gate width 5 ns). The compounds were irradiated at
515 nm, in the QIV band, because of the gap in the spectral
range of excitation of the Coherent Infinity YAG laser in the
Soret band.
To the solution in toluene was added portions of
tetrachloro-1,2-benzoquinone (0.08 mmol, 20 mg) at room
temperature to oxidize the bacteriochlorin. Then the organic
phase was washed with a solution of sodium hydroxide
(NaOH) (5%), concentrated phosphoric acid (68 wt.%) to
remove the residual porphyrin, water and saturated sodium
bicarbonate (NaHCO3) solution. The resulting solution con-
taining only the chlorin was evaporated and the solid recrys-
tallized in toluene (yield 72%).
Rf 0.38 (SiO2, 4:96 EtOH/CH2Cl2), m.p. 94 °C, 1H NMR
(DMSO-d6) d: –1.59, –1.51 (s, 2H, NH), 4.13 (m, 4H, CH2
chlorine), 7.46–8.30 (m, 6H pyrrole-H, 19 H phenyl). MS
(MALDI-TOFMS) m/z: 656. 07 Da.
4.1.6. 5-[4-(1,3,4,6-tetra-O-acetyl-2-amido-2-desoxy-b-D-
glucopyranose pheny)]-10,15,20-triphenyl chlorin (3)
A sample of 5 was treated identically as for 4, affording a
purple compound with 62% yield. Isomeric 2,3- and
7,8 mono-glucosylated chlorins showed two NH signals of
almost equal intensity at –1.50 and –1.41 ppm, indicating a
1:1 mixture. They were obtained as an inseparable mixture.
Rf 0.62 (SiO2, 4:96 EtOH/CH2Cl2), m.p. 200 °C, 1H NMR
(CDCl3) d: –1.50(s), –1.41(s) (2H, NH), 2.10–2.23 (m, 12H,
4 CH3), 3.93 (m, 1H, Glc-H-5), 4.17 (m, 2H, CH2 chlorine),
4.34, 4.38 (2H, Glc-H-6, CH2OAc), 4.70 (m, 1H, Glc-H-2),
5.35 (m, 2H, Glc-H-3, Glc-H-4), 5.94 (m, 1H, Glc-H-1), 6.46
4.5. HPLC conditions
An Alltech Appollo C18, 5 µm column (250 × 4.6 mm
internal diameter (I.D.),Alltech, Lokeren, Belgium) was used.
Fluorescence detection was optimized with excitation and
emission wavelengths of 414 and 650 nm, respectively.
Gradient elution conditions were adopted. The eluting sol-
vent was methanol (MeOH)/H2O (75:25, v/v) for 15 min, fol-
lowed by 100% methanol for the last 25 min. Room tempera-
ture and a flow-rate of 1.0 ml min–1 were maintained
throughout the analysis. A volume of 20 µl of the samples in

B. Di Stasio et al. / European Journal of Medicinal Chemistry 40 (2005) 1111–1122
1121
methanol was injected into the column. The internal standard
stock solution of rhodamine 1 mg ml–1 was also prepared in
methanol.
medium was removed, cells were washed three times with
cold PBS and fresh RPMI was added. Each concentration
was tested in sextuplicate. As previously described [24] cell
survival was measured 24 h after by MTT test; 50 µl of 0.5%
MTT solution was then added in each well and incubated for
3 h at 37 °C to allow MTT metabolism. The formazan crys-
tals were dissolved by adding 50 µl per well of 25% sodium
dodecylsulfate solution and vigorous pipetting was per-
formed. Absorbance was measured at 540 nm using a Multi-
skan MCC/340 plate reader (Labsystem, Cergy-Pontoise,
France). Results were expressed as relative absorbance to
untreated controls. Absorbance values for wells containing
medium alone were subtracted from the results of test wells.
4.6. General procedure for in vitro experiments
4.6.1. Cell culture conditions
Human colorectal adenocarcinoma cells (HT29) were
grown in 75 cm2 plastic tissue culture flasks in RPMI
1640 medium supplemented with 9% heat inactivated fetal
calf serum, penicillin (100 i.u. ml–1), streptomycin
(100 µg ml–1) in an atmosphere of 37 °C and 5% CO2 atmo-
sphere. Cells were subcultured by dispersal with 0.25%
trypsin and seeded 5 × 104 cells per ml.
4.6.4. Photocytotoxicity assay
Cell survival was measured 24 h after photosensitization
using MTT assay. HT29 cells were seeded at the initial den-
sity of 1.104 cells per ml in 96-well microtitration plates.
Forty-eight hours after plating, cells were exposed to photo-
active compounds at 10–6 M. After 24-h incubation at 37 °C,
the medium was removed and cells were then washed three
times with cold PBS and fresh RPMI was added before cell
irradiation. Results are given as the percentage of the result
obtained with the control cultures exposed to photosensitizer
alone. Light doses, yielding 50% growth inhibition (L.D.50),
were calculated using medium effect algorithm [37], and
expressed as mean values of three independent experiments
performed during different weeks. We used one plate for each
dose of light from 1.0 to 20.0 J cm–2. In order to perform the
experiments exactly in the same conditions for the four pho-
tosensitizers (TPP, 1–3), we divided each plate into four parts
corresponding to each compound. The dark toxicity of pho-
tosensitizers was assessed separately following a similar pro-
cedure.
4.6.2. Photosensitizers uptake
Cells (1.104 cells per ml) were inoculated in 25 cm2 plas-
tic tissue culture flasks and incubated for 72 h for proper
attachment to the substratum. The HT29 cells were then
exposed to photosensitizers (TPP, 1, 3 and 2 at 10–5 M) in
RPMI supplemented with 9% of bovine serum albumin (BSA)
from 3 to 24 h. All stock solutions were prepared by dissolv-
ing 10–3 mol l–1 photosensitizer in DMSO and after diluted
in RPMI in order to achieve dye concentrations mentioned.
No cytotoxic effect of DMSO was checked. After incubation
with the photosensitizers, the cells were washed twice in phos-
phate buffered saline (PBS), harvested by enzymatic deseg-
regation [trypsin–ethylene diamine tetra acetic acid (EDTA)].
After, the cells were washed in ice-cold PBS using centrifu-
gation (200 × g) and the fluorescence from the drug-loaded
cells was measured in 3.0 ml ethanol (EtOH) solution [36].
The fluorescence intensity of each sample was normalized to
cell concentration, to molar extinction coefficient values at
417 nm and to fluorescence quantum yield assessed for each
compound, respectively). Spectra were collected for wave-
lengths ranging from 600 to 750 nm (the band pass of both
excitation and emission slits were 2.5 nm, in the case of drug-
loaded cells 2.5 and 5.0 nm, respectively; the excitation wave-
length was 417 nm; photomultiplier voltage was 770 V.
For all measurements, front surface accessories (Perkin–
Elmer L225 9051) have been used, providing small angle
(45°) front surface excitation geometry. All absorption spec-
tra were recorded with a Perkin–Elmer Lambda Bio spectro-
photometer, using a 10 mm quartz cuvette.
Irradiation was carried out at 650 nm, using a dye laser
(Spectra Physics 375B) pumped with an Argon laser (Spec-
tra Physics 2020, Les Ulis, France). The output power was
700 mW. The light spot was 14 cm in diameter, providing the
fluence of 4.54 mW cm–2. During irradiation, the tempera-
ture never exceeded 24 2 °C. This temperature did not influ-
ence cell viability.
4.6.5. Treatment of cells prior to microspectrofluorimetry
and confocal laser scanning microscopy
Dual staining confocal microscopy experiments were per-
formed for 1. Cells were plated at 103 cells per well in Labtek-
II® 4-chambered coverslips (Dutscher, Brumath, France).
After 48-h attachment at 37 °C, the cells were then incubated
with 1 (10 µM) and specific organelle markers, according to
the following protocol. Before observation, cells were then
washed three times with RPMI medium and imaged using
confocal laser scanning microscopy.
The organelles staining was investigated using the proce-
dure adapted from the experimental protocol previously
reported [38]. Endoplasmic reticulum was stained by incu-
bating cells for 30 s at room temperature with 0.25 µg ml–1
4.6.3. Cytotoxicity assays
Cell survival was measured using 3-(4,5-dimethylthiazol-
2-yl)-2,5-diphenyltetrazolium (MTT) assay. Before perform-
ing growth inhibition assays, we examined the linearity of
the MTT assay with increasing number of HT29 cells plated
between 103 and 5.106 cells per ml and found quite satisfac-
tory results (r > 0.995). Briefly, cells were seeded at the ini-
tial density of 1.104 cells per ml in 96-well microtitration
plates. Forty-eight hours after plating, cells were exposed for
24 h to photosensitizers (TPP, 1–3) for concentrations vary-
ing from 0.5 to 30 µM. After 24-h incubation at 37 °C, the

1122
B. Di Stasio et al. / European Journal of Medicinal Chemistry 40 (2005) 1111–1122
[10] M. Zhang, Z. Zhang, D. Blessington, H. Li, T.M. Busch, V. Madrak,
J. Miles, B. Chance, J.D. Glickson, G. Zheng, Bioconjug. Chem. 14
(2003) 709–714.
[11] in: J. Defaye, J. Gelas (Eds.), Studies in Natural Products Chemistry,
Elsevier, Amsterdam, 1991.
[12] M. Bergmann, L. Zervas, Ber. 65B (1932) 1201–1205.
[13] R.G. Little, J.Z. Anton, P.A. Loach, J. Heterocyclic Chem. 12 (1975)
343–349.
[14] H.W. Whitlock, R. Hanauer, M.Y. Oester, B.K. Bower, J. Am. Chem.
Soc. 91 (1969) 7485–7489.
[15] C.H. Lee, J.S. Lindsey, Tetrahedron 50 (1994) 11427–11440.
[16] C. Geoffroy, M.C. Carré, F. Baros, S. Muller, D. Dumas, J.F. Stoltz,
P. Van Hoppen, F.C. De Schryver, M.L. Viriot, in: Proceedings of the
Third Conference on Fluorescence Microscopy and Fluorescent
Probes, 1999, pp. 137–143.
3,3-dihexyloxacarbocyanine iodide (DiOC6). DiOC6 was pur-
chased from Molecular Probes, Leiden, The Netherlands.
Mitochondria were identified after staining the cells for 30 s
at 37 °C with 0.25 µmol l–1 Mitotracker Green® (Molecular
Probes). For lysosomes identification, HT29 cells were incu-
bated with 1 for 24 h, and overnight with 1 µg ml–1 Lucifer
yellow, LY® (Sigma-Aldrich, France). The cells were then
washed 3 times in a dye free media before imaging in RPMI.
The cells double stained with 1 and organelle probes were
examined with a confocal laser scanning microscope (Leica
TCS SP2-AOBS, Germany) equipped with an X63, numeri-
cal aperture 1.3 oil immersion objective (Leica).A pinhole of
60.85 µm was used and each image recorded contained 512 ×
512 pixels. An argon laser was used as excitation light at
458 nm for all organelle probes and a helium/neon laser at
633 nm for 1. Fluorescence of the organelles probes was
detected on channel 1 with a 505–545 nm band pass emis-
sion filter. Channel 2 was used to detect the red fluorescence
of 1 with 640–660 nm band pass emission filter. The fluores-
cence images were displayed in green and red “false” color
output and electronically combined to visualize dual local-
ization in yellow. Controls (cells stained only with 1 or
organelle probes) were conducted in parallel to optimize the
staining protocol and the detecting parameters of confocal
laser scanning microscopy.
[17] Y. Mikata,Y. Onchi, K. Tabata, S.I. Ogura, I. Okura, H. Ono, S.Yano,
Tetrahedron Lett. 39 (1998) 4505–4508.
[18] J.N. Demas, G.A. Crosby, J. Phys. Chem. 75 (1971) 991–1024.
[19] Y. Mikata, Y. Onchi, K. Shibata, T. Kakuchi, H. Ono, S.I. Ogura,
I. Okura, S. Yano, Bioorg. Med. Chem. Lett. 8 (1998) 3543–3548.
[20] D. Oulmi, P. Maillard, C. Vever-Bizet, M. Momenteau, D. Brault,
Photochem. Photobiol. 67 (1998) 511–518.
[21] E. Reddi, G. Jori, M.A.J. Rodgers, J.D. Spikes, Photochem. Photo-
biol. 38 (1983) 639–645.
[22] R. Boyle, D. Dolphin, Photochem. Photobiol. 64 (1996) 469–485.
[23] D. Kessel, P. Thompson, K. Saatio, D. Nantwi, Photochem. Photobiol.
45 (1987) 787–790.
[24] M. Barberi-Heyob, P.O. Vedrine, J.L. Merlin, R. Millon, J. Abecassis,
M.F. Poupon, F. Guillemin, Int. J. Oncol. 24 (2004) 951–958.
[25] I. Laville, T. Figueiredo, B. Loock, S. Pigaglio, P. Maillard, D.S. Gri-
erson, D. Carrez, A. Croisy, J. Blais, Bioorg. Med. Chem. 11 (2003)
1643–1652.
[26] L.Y. Xue, S.M. Chiu, N.L. Oleinick, Exp. Cell Res. 263 (2001)
145–155.
4.7. Statistical analysis
[27] C.M. Yow, J.Y. Chen, N.K. Mak, N.H. Cheung, A.W. Leung, Cancer
Lett. 157 (2000) 123–113.
[28] P. Morlière, J.C. Mazière, R. Santus, C.D. Smith, M.R. Prinsep,
C.C. Stobbe, M.C. Fenning, J.L. Golberg, J.D. Chapman, Cancer Res.
58 (1998) 3571–3578.
Mann and Whitney U-test was used to test for the signifi-
cant level between independent variables. The level of sig-
nificance was set to P < 0.05.
[29] J.Y. Matroule, G. Bonizzi, P. Morlière, N. Paillous, R. Santus,
V. Bours, J. Piette, J. Biol. Chem. 274 (1999) 2988–3000.
[30] C. Fabris, G. Valduga, G. Miotto, L. Borsetto, G. Jori, S. Garbisa,
E. Reddi, Cancer Res. 61 (2001) 7495–7500.
References
[31] M.H. Teiten, L. Bezdetnaya, P. Morlière, R. Santus, F. Guillemin, Br.
J. Cancer 88 (2003) 146–152.
[32] M. Terasaki, J. Song, J.R. Wong, M.J. Weiss, L.B. Chen, Cell 38
(1984) 101–108.
[33] A.S. Sobolev, D.A. Jans, A.A. Rosenkranz, Prog. Biophys. Mol. Biol.
73 (2000) 51–90.
[34] C.B. Hansen, G.J. Hoogers, W.J. Drenth, J. Mol. Catal. 79 (1993)
153–163.
[35] R. Schmidt, E. Afshari, J. Phys. Chem. 94 (1990) 4377–4378.
[36] C. Frochot, B. Di Stasio, M. Barberi-Heyob, M.C. Carré, J.M. Zwier,
F. Guillemin, M.L. Viriot, Oftalmologia 56 (2003) 62–66.
[37] T.C. Chou, P. Talalay, in: K.R. Harrap, T.A. Connors (Eds.), New
Avenues in Development Cancer Chemotherapy, Academic Press,
New York, 1987, pp. 37–64.
[1] J.G. Levy, M. Obochi, Photochem. Photobiol. 64 (1996) 737–739.
[2] T.J. Dougherty, Photochem. Photobiol. 45 (1987) 879–889.
[3] G. Poste, G.L. Nicolson, Biomembranes 11 (1983) 341–364.
[4] I. Rosenthal, Photochem. Photobiol. 53 (1991) 859–870.
[5] U. Isele, K. Schieweck, R. Kessler, P. van Hoogevest, H.G. Capraro, J.
Pharm. Sci. 84 (1995) 166–173.
[6] I.J. MacDonald, T. Dougherty, J. Porphyrins and Phthalocyanines 5
(2001) 105–129.
[7] J.G. Moser, F.P. Montforts, D. Kusch, A. Vervoorts, D. Kirsch,
M. Berghahn, N. Akgün, A. Rueck, S. Andrees, B. Wagner, Proc.
SPIE, Int. Soc. Opt. Eng. 2924 (1996) 22–30.
[8] G. Zheng, A. Graham, M. Shibata, J. Missert, A.R. Oseroff,
T.J. Dougherty, R.K. Pandey, J. Org. Chem. 66 (2001) 8709–8716.
[9] M. Momenteau, P. Maillard, M.A. de Bélinay, D. Carrez, A. Croisy, J.
Biomed. Opt. 4 (1999) 298–318.
[38] J.L. Merlin, H. Gautier, M. Barberi-Heyob, M.H. Teiten, F. Guillemin,
Int. J. Oncol. 22 (2003) 733–739.