A study of the stability of tri(glucosyloxyphenyl)chlorin, a sensitizer for photodynamic therapy, in human colon tumoural cells: A liquid chromatography and MALDI-TOF mass spectrometry analysis

Article abstract of DOI:10.1016/j.bmc.2004.04.022

Asymmetrical glycoconjugated tetrapyrrolic macrocycles are under study as efficient sensitizers for photodynamic therapy (PDT). In this context, tri(meta-O-β-glucopyranosyloxyphenyl)chlorin [TPC(m-O-Glu)₃] 2a/3a was found to be four times more

Full text of DOI:10.1016/j.bmc.2004.04.022

Bioorganic & Medicinal Chemistry 12 (2004) 3673–3682
A study of the stability of tri(glucosyloxyphenyl)chlorin, a sensitizer
for photodynamic therapy, in human colon tumoural cells: a liquid
chromatography and MALDI-TOF mass spectrometry analysis
I. Laville,a S. Pigaglio,a J.-C. Blais,b B. Loock,c Ph. Maillard,c D. S. Griersonc
and J. Blaisa,*
aLPBC, UMR CNRS 7033, Universitꢀe Pierre et Marie Curie, 4 place Jussieu, case 138, 75252 Paris Cedex 05, France
bLCSOB, UMR CNRS 7613, Universitꢀe Pierre et Marie Curie, 75252 Paris Cedex 05, France
cUMR 176 CNRS Institut Curie, Facultꢀe des Sciences, 91405 Orsay, France
Received 17 October 2003; accepted 15 April 2004
Available online 18 May 2004
Abstract—Asymmetrical glycoconjugated tetrapyrrolic macrocycles are under study as efficient sensitizers for photodynamic
therapy (PDT). In this context, tri(meta-O-b-glucopyranosyloxyphenyl)chlorin [TPC(m-O-Glu)3] 2a/3a was found to be four times
more photoactive in vitro than Foscanꢀ. In a further study of this interesting glycoconjugate, its metabolism by cellular glycosidases
in HT29 cells has to be explored. Cellular extracts of HT29 cells incubated with TPC(m-O-Glu)3 (24 h, 6 lM) were analyzed by
MALDI-TOF mass spectrometry and high performance liquid chromatography (HPLC). In MALDI-TOF mass spectra, the
presence of compounds distinct from TPC(m-O-Glu)3 (m=z 1151) were observed at m=z 989, 827 and 665 corresponding to the loss of
one, two or three glucose units (162 u) and were be ascribed to TPC(m-OH)(m-O-Glu)2 2/3b,b0,b00, TPC(m-OH)2(m-O-Glu) 2/3c,c0,c00
and TPC(m-OH)3 isomers 2d/3d, respectively. The porphyrins resulting from chlorin oxidation TPP(m-O-Glu)3 4a, TPP(m-OH)(m-
O-Glu)2 4b,b00, TPP(m-OH)2(m-O-Glu) 4c,c00 and TPP(m-OH)3 4d were also observed. The HPLC profile (kanal: ¼ 420 nm) showed
eight peaks consistent with mass spectra. The kinetics of deglucosylation was studied from HPLC profiles between 1 and 48 h
incubation. The concentration of triglucoconjugated and diglucoconjugated molecules was maximum around 3 and 8 h incubation,
respectively, whereas, totally deglucosylated species appeared only after incubation for more than 10 h. The fully deglycosylated
porphyrin TPP(m-OH)3 is the final metabolite, being observed at a concentration 15 times higher than that of the remaining TPC(m-
O-Glu)3 2a/3a. Compared to the photobiological activity of the parent molecule [TPC(m-O-Glu)3], a three times higher TPP(m-OH)3
concentration was necessary to observe a similar in vitro photoactivity.
ꢁ 2004 Elsevier Ltd. All rights reserved.
Abbreviations: DMEM, Dulbecco’s modified Eagle’s medium; DMSO, dimethyl sulfoxide; FCS, fetal calf serum; HPLC, high performance liquid
chromatography; HT29, human colorectal adenocarcinoma cells; MALDI-TOF, matrix assisted laser desorption ionization-time of flight; MTT,
3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide; PBS, phosphate buffer saline; PDT, photodynamic therapy; m-THPC, 5,10,15,20-
tetrakis(meta-hydroxyphenyl)chlorin; TPC(m-O-Glu)3, 5,10,15-tri(meta-O-b-
O-b- -glucopyranosyloxyphenyl)-20-phenyl-7,8-chlorin; TPC(m-OH)(m-O-Glu)2, 5,10-di(meta-O-b-
phenyl)-20-phenyl-2,3-chlorin, 10,15-di(meta-O-b- -glucopyranosyloxyphenyl)-5-(meta-hydroxyphenyl)-20-phenyl-2,3-chlorin, 5,10-di(meta-O-b-
glucopyranosyloxyphenyl)-15-(meta-hydroxyphenyl)-20-phenyl-7,8-chlorin, 5,15-di(meta-O-b- -glucopyranosyloxyphenyl)-10-(meta-hydroxyphenyl)-
20-phenyl-2,3-chlorin, 5,15-di(meta-O-b- -glucopyranosyloxyphenyl)-10-(meta-hydroxyphenyl)-20-phenyl-7,8-chlorin and 5,10-di(meta-O-b- -gluco-
pyranosyloxyphenyl)-15-(meta-hydroxyphenyl)-20-phenyl-17,18-chlorin; TPP(m-OH)(m-O-Glu)2, 5,10-di(meta-O-b- -glucopyranosyloxyphenyl)-15-
(meta-hydroxyphenyl)-20-phenyl porphyrin and 5,15-di(meta-O-b- -glucopyranosyloxyphenyl)-10-(meta-hydroxyphenyl)-20-phenyl porphyrin; TP-
C(m-OH)2(m-O-Glu), 5-(meta-O-b- -glucopyranosyloxyphenyl)-10,15-di(meta-hydroxyphenyl)-20-phenyl-2,3-chlorin 10-(meta-O-b- -glucopyrano-
syloxyphenyl)-5,15-di(meta-hydroxyphenyl)-20-phenyl-2,3-chlorin, 5-(meta-O-b- -glucopyranosyloxyphenyl)-10,15-di(meta-hydroxyphenyl)-20-phenyl-
7,8-chlorin and 10-(meta-O-b- -glucopyranosyloxyphenyl)-5,15-di(meta-hydroxyphenyl)-20-phenyl-7,8-chlorin; TPP(m-OH)2(m-O-Glu), 5-(meta-O-
b- -glucopyranosyloxyphenyl)-10,15-di(meta-hydroxyphenyl)-20-phenyl porphyrin and 10-(meta-O-b- -glucopyranosyloxyphenyl)-5,15-di(meta-
D
-glucopyranosyloxyphenyl)-20-phenyl-2,3-chlorin and 5,10,15-tri(meta-
D
D
-glucopyranosyloxyphenyl)-15-(meta-hydroxy-
D
D-
D
D
D
D
D
D
D
D
D
D
D
hydroxyphenyl)-20-phenyl porphyrin; TPC(m-OH)3, 5,10,15-tri(meta-hydroxyphenyl)-20-phenyl-2,3-chlorin and 5,10,15-tri(meta-hydroxyphenyl)-
20-phenyl-7,8-chlorin; TPP(m-OH)3, 5,10,15-tri(meta-hydroxyphenyl)-20-phenyl porphyrin.
Keywords: Photodynamic therapy; Glucoconjugated chlorins; Metabolization; Deglucosylation.
* Corresponding author. Tel.: +33-1-44-27-75-46; fax: +33-1-44-27-75-60; e-mail: jblais@lpbc.jussieu.fr
0968-0896/$ - see front matter ꢁ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2004.04.022

3674
I. Laville et al. / Bioorg. Med. Chem. 12 (2004) 3673–3682
1. Introduction
in erythrocytes membranes,12 the b-glucosidase activity
in human plasma being about 20-fold lower than that of
a-glucosidases. Glycosidases promoted metabolism of
glycoconjugated porphyrins/chlorins can thus occur at a
number of levels after administration in vivo of these
agents. However, a certain specificity and stability can
be achieved. Monsigny et al.,13 showed that glycosyl
moieties of glycoconjugated drugs could be recognized
by membrane lectins, the drug being released in different
organelles (endoplasmic reticulum and Golgi apparatus)
that contain specific glycosidases. The resulting degly-
cosylation processes were shown to depend on the nat-
ure of the link between the drug and the glycosyl
moieties.
Photodynamic therapy (PDT) is an emerging treatment
modality of solid tumours based on the administration
of a light-activated drug (photosensitizer) followed by
the laser illumination of pathological areas. Upon sen-
sitizer photoactivation, the generation of reactive oxy-
gen species leads to irreversible destruction of the
treated tissues.1;2
Several important milestones in this field include the
regulatory approval given in 2000 for Visudyneꢀ (BPD-
MA) for the treatment of age-related macular degener-
ation (AMD), and for Foscanꢀ (mTHPC, 2002) as an
agent for palliative treatment of head and neck cancer.
However, for these, and other promising molecules un-
der study, the risk of long-term cutaneous photosensi-
tivity remains a significant adverse effect. Continued
research is thus necessary to develop new sensitizers with
high photoactivity and increased therapeutic index.
Concerning the porphyrin/chlorin component in such
tetrapyrrolic sensitizers, there was little data pertaining
to their metabolism.14;15 As Photofrinꢀ, the first com-
mercialized sensitizer, is a mixture of porphyrin deriva-
tives, HPLC analysis of tissue extracts is difficult to
interpret. The presence of a metabolite of Foscanꢀ
(mTHPC) was detected by Whelpton et al. in the rat
liver.16 In contrast, Cai et al. have studied the metabo-
lism of Foscanꢀ and found no evidence either in vitro
and in vivo (absence of light) that this agent is meta-
It has been demonstrated that the overall lipophilicity of
the sensitizer plays an important role in PDT efficacy2;3
and in this context, glycoconjugation is undergoing
evaluation as an effective mean to modulate the amph-
iphilicity of tetrapyrrolic compounds. Of note is the
work by Pandey and co-workers4 who have recently
prepared a number of galactosylated chlorins and
shown that the position of the sugar motif has a strong
influence on photobiological activity.5 In our laborato-
ries, efforts have been focused on the preparation and
in vitro evaluation of the phototoxicity of a broad series
of neutral tri and tetraglycoconjugated tetrapyrrolic
macrocycles.6;7 It was found that, in general, the trigly-
coconjugated sensitizers exhibit higher photobiological
activity than either the parent tetrapyrroles or the
symmetrical tetraglycoconjugated derivatives.6 To look
at the origin of this effect in more detail, a recent study
was made of the cellular internalization and photody-
namic activity of new tri and tetraglucosylated conju-
gate of Foscanꢀ, tri(meta-O-b-glucopyranosyloxy-
phenyl)chlorin [TPC(m-O-Glu)3] and tetra(meta-O-b-
bolized.17 Visudyneꢀ,
a monoacid benzoporphyrin
derivative, was reported to be hydrolyzed to its diacid
derivative in vivo. However, the extent of formation of
this metabolite was less than 10% of the administered
dose.18 Similarly, no metabolites were detected in a
study of mono-L-aspartyl chlorin e6 metabolism in a
mouse mammary tumour model19 or for palladium
bacteriopheophorbide (Tookadꢀ), a new sensitizer in
clinical trial for the PDT treatment of prostate cancer,
and gallium-porphyrin (ATX-70).20;21
In the present study, we have explored the possible
cellular in vitro metabolization of TPC(m-O-Glu)3 2a/3a
(Fig. 1), a compound that exhibits interesting photo-
biological characteristics for PDT treatment,8 using
MALDI-TOF mass spectrometry and high performance
liquid chromatography (HPLC). We provide data to
demonstrate that the three sugar motifs in the chlorin
2a/3a undergo sequential hydrolysis, and that these
processes are interconnected to the oxidative metabo-
lism of the chlorin system to the corresponding por-
phyrins.
glucopyranosyloxyphenyl)chlorin
[TPC(m-O-Glu)4].8
The triglucoconjugated TPC(m-O-Glu)3 is four times
more photoactive in vitro than Foscanꢀ itself and could
be of potential interest in the PDT treatment of
tumours.
At present, little is known concerning the metaboliza-
tion of glycosylated porphyrins or chlorins either in vivo
or in cell based assays. In vivo metabolism is a partic-
ularly important issue when considering the use of
glycoconjugated compounds in PDT treatment, since
cleavage of the glycoside bond by glycosidases will result
in modifications of amphiphilic properties, biodistribu-
tion, blood clearance and drug–cell interactions.
Metabolism of these agents may in turn alter their
photobiological activity.
2. Results and discussion
2.1. Synthesis
The isomeric glucoconjugated chlorins TPC(m-O-Glu)3
2a and 3a (a 50/50 mixture of two inseparable isomers)22
were obtained by diimide reduction and deacetylation of
the pentaacetate derivative 1 of triglucosylated por-
phyrin [TPP(m-O-Glu)3] 4a (Scheme 1).8 In a similar
way, the unsubstituted chlorins TPC(m-OH)3 2d/3d were
obtained in two steps from TPP(m-OMe)3 5. Porphyrin
4d [TPP(m-OH)3] was obtained by direct treatment of 5
The level of glycosidase activity and expression has been
found to be organ dependent9 and to be higher in certain
tumoural tissues compared to normal tissues.10;11 Glu-
cosidases have also been identified in human plasma and

I. Laville et al. / Bioorg. Med. Chem. 12 (2004) 3673–3682
3675
Figure 1. Typical MALDI-TOF mass spectrum of a cellular extract of HT29 cells treated with 2a/3a (6 lM, 24 h incubation). The isotopic distri-
bution of each main signal indicates the presence of two species separated by 2 u (as an example: see insert in the case of peak at m=z 665).
Scheme 1. Synthesis and structure of glycoconjugated chlorins and porphyrins. Reagents and conditions: (i) (a) toluene-4-sulfonohydrazide,
anhydrous K2CO3, dry pyridine, 100 ꢂC; (b) ortho-chloranil, ethyl acetate; (ii) MeONa/MeOH, 20 ꢂC.
with BBr3. As determined by NMR, the mixture of tri-
glycosylated chlorins 2a/3a contains a small quantity
(<10%) of the triglucosylated porphyrin 4a as a con-
taminant.22

3676
I. Laville et al. / Bioorg. Med. Chem. 12 (2004) 3673–3682
The possible transformation of the glucoconjugated
chlorins 2a/3a by cellular enzymes has been investigated
using HT29 human adenocarcinoma colon cells.
Extraction was performed in the dark on treated cells
with 2a/3a and cellular extracts were then analyzed by
MALDI-TOF mass spectrometry and HPLC.
bolization is not affected by the meta- or para-position
of sugar units (spectra not shown). However, one cannot
exclude the metabolic pathway could be affected by the
chemical structure of the sensitizer.
The metabolic degradation of glucoconjugated com-
pounds have to be considered whatever the cell line used
depending on the glucosidase activity. As an example,
similar degradation pattern of the TPC(m-O-Glu)3 2a/3a
has also been observed with the loss of one, two and
three glucose units in cellular extracts of treated cells
(6 lM, 24 h incubation) in the case of murine melanoma
cells (B16).
2.2. Mass spectrometry analysis of cellular extracts
A typical MALDI-TOF mass spectrum of a cellular
extract of HT29 treated with TPC(m-O-Glu)3 2a/3a
(6 lM, 24 h incubation) recorded prior to any chro-
matographic separation is shown in Figure 1. In addi-
tion to the protonated molecule for the two isomeric
triglucosylated chlorins 2a/3a (m=z 1151), intense signals
at m=z 989, 827 and 665 were observed. These signals
corresponded to the loss of one, two or three glucose
units (162 u) from these chlorins, and were assigned to
TPC(m-OH)(m-O-Glu)2 (six isomers; 2/3b,b0,b00),
TPC(m-OH)2(m-O-Glu) (six isomers; 2/3c,c0,c00) and
TPC(m-OH)3 (two isomers; 2d/3d).
2.3. HPLC analysis
Chromatographic analysis of cellular extracts of treated
cells was conducted in order to elucidate the different
steps of the deglucosylation process. The HPLC profile
[absorption detection (k ¼ 420 nm)] of a cellular extract
of HT29 cells (Fig. 2) incubated with 2a/3a (24 h incu-
bation) was consistent with the mass spectra displayed
with eight peaks or group of peaks. Chromatograms
obtained with a dual absorption (k ¼ 420 nm) or fluo-
rescence detection (kexc ¼ 420 nm, kem ¼ 650 nm) were
similar in both cases indicating the eight absorbing
species are fluorescent (data not shown).
The isotopic distribution observed for these signals (see
for example insert Fig. 1) was different from that ex-
pected for a single species, and clearly indicated the
presence of the corresponding porphyrin derivatives,
TPP(m-O-Glu)3 4a (m=z 1149), TPP(m-OH)(m-O-Glu)2
4bb00 (m=z 987), TPP(m-OH)2(m-O-Glu) 4c,c00 (m=z 825),
and TPP(m-OH)3 4d (m=z 663).
MALDI-TOF mass spectra of HPLC fractions allowed
to directly ascribed fraction 1 to 2a/3a (m=z 1151),
fraction 2 to TPC(m-OH)(m-O-Glu)2 (six isomers) (m=z
989) and fraction 3 to 4a (m=z 1149) (see Table 1). Peaks
7 and 8 could be assigned by using model compounds.
MALDI-TOF mass spectra and HPLC characteristics of
2d/3d (m=z 665, 29 min retention time) and 4d (m=z 663,
37 min retention time) clearly indicated that peaks 7 and
8 corresponded to the totally deglucosylated chlorin and
porphyrin, respectively (see Table 1).
As a control, 2a/3a was kept in culture medium (6 lM)
in the dark for 24 h at 37 ꢂC, extracted, and then dried
under vacuum. In the mass spectrum of the extracted
control sample the protonated molecule at m=z 1151 was
observed. Compound 4a was also detected correspond-
ing to the porphyrin contaminant8 with an abundance
(approx. 7%), which was lower than that measured in all
cases for the cellular extracts.
It can thus been assumed that the loss of sugars only
results from an enzymatic reaction between cellular b-
glucosidase and the sensitizer, the tetrapyrrolic ring
aromatization implicating an other enzymatic pathway.
Activity of intact cell b-glucosidase has been evaluated
to 6.9( 1) · 10ꢀ8 nmol/min/cell in HT29. As a compari-
son, mannosidase activity was two times lower
[3.5( 0.13) · 10ꢀ8 nmol/min/cell]. In as far as enzymatic
activities are specific of the cell line and sugar moieties,
deglycosylation processes may be expected to be highly
dependent on the chemical structure of glycoconjugated
compounds.
It is interesting to note that glucosidic bond metaboli-
zation processes do not depend upon the saturated or
unsaturated character of the tetrapyrrolic ring. Besides
the molecular ion at m=z 1149, the mass spectrum of a
cellular extract of HT29 treated with 4a (6 lM, 24 h
incubation), also displayed three peaks at m=z 987, 825
and 663, corresponding to the loss of one, two and three
sugar units. An identical degradation pattern was found
for cells treated with the analogous para-substituted
compound, TPP(p-O-Glu)3 indicating the cellular meta-
Figure 2. Representative HPLC profile (absorption detection
k ¼ 420 nm) of a cellular extract of HT29 cells treated with 2a/3a
(6 lM, 24 h incubation). Fraction numbers 1 to 8 correspond to the fn
values reported in Table 1.

I. Laville et al. / Bioorg. Med. Chem. 12 (2004) 3673–3682
3677
Table 1. Fraction number (fR), corresponding retention times (tR) and
1.2
1.0
0.8
0.6
0.4
0.2
0.0
m=z values obtained from HPLC and MALDI-TOF mass spectrome-
try analysis of cellular extracts of HT29 treated with TPC(m-O-Glu)3
2a/3a (6 lM, 24 h incubation)
fn
tR (min)
m=z
Compound
1
2
3
4
5
6
7
8
5.1–5.3
7.8–8.2
13
1151
989
1149
987
827
825
665
663
2a/3a: TPC(m-O-Glu)3
2/3b,b0,b00: TPC(m-OH)(m-O-Glu)2
4a: TPP(m-O-Glu)3
15.2
16
4b,b00: TPP(m-OH)(m-O-Glu)2
2/3c,c0,c00: TPC(m-OH)2(m-O-Glu)
4c,c00: TPP(m-OH)2(m-O-Glu)
2d/3d: TPC(m-OH)3
0
10
20
30
40
50
60
(A)
Incubation (hours)
19
29.5
37.8
1.4
1.2
1
4d: TPP(m-OH)3
0.8
0.6
0.4
0.2
0
Assignment of peaks 4, 5 and 6 could not be done from
MALDI-TOF data. However, on the basis of respective
retention times and polarity considerations, peak 4 was
assigned to diglucoconjugated porphyrin isomers, and
peaks 5 and 6 to monoglucosylated chlorin and por-
phyrin isomers, respectively (Table 1). Occurrence of
deglucosylation processes results in the existence of
isomeric forms that could explain the structure observed
for peak 2 and the poor resolution of some other peaks.
0
10
20
30
40
50
60
(B)
Incubation (hours)
Figure 3. Time course of metabolites determined from HPLC analysis
of cellular extract of HT29 cells treated with 2a/3a (6 lM, incubation
times going from 1 to 48 h). (A) Chlorin derivatives: N 2a/3a, r 2/
3b,b0,b00, · 2/3c,c0,c00. (B) Porphyrin derivatives: N 4a, r 4b,b00, · 4c,c00,
The HPLC analysis wavelength of 420 nm corresponded
to optimized conditions to observe both chlorin and
porphyrin derivatives. Nevertheless, HPLC profile does
not reflect the concentration of each species since maxi-
mum extinction coefficients are different and highly
dependent on the solvent environment. The concentra-
tion of metabolites isolated from the cellular extracts
was determined from calibration curves performed in
the case of peaks 1, 2, 3, 7 and 8 using the corresponding
model compounds. Pure mono and diglucoconjugated
derivatives were not available.
4d. The maximum concentration was 2.8 0.03 · 10ꢀ6 lmol/lg of
ꢁ
proteins in the case of 2a/3a and 3.6 0.03 · 10ꢀ6 lmol/lg of proteins
in the case of 4d. Errors bars, SE (n ¼ 3).
Table 2. Ratio of TPP(m-O-Glu)3/TPC(m-O-Glu)3 (4a/2a–3a) and
TPP(m-OH)3/TPC(m-O-Glu)3 (4d/2a–3a) concentration as determined
from HPLC profiles
0 h
3 h
8 h
14 h
24 h
48 h
0.07
15
4a/2a–3a
4d/2a–3a
0.07
0
0.12
0
0.33
0.12
0.18
1.2
0.06
4.36
2.4. Kinetics of TPC(m-O-gluOH)3 deglucosylation
Figure 3A and B show the time course of A=Amax ratio
where A is the peak area obtained for a given incubation
time t and Amax the peak area corresponding to the
maximum value observed for each product isolated from
the cellular extracts.
to that of 2a/3a (4a/2a–3a) increased during the first 8 h
from 7% to 33% then falling down to 6% at 24 h. As
mentioned before this value corresponds to the initial
amount of contaminant porphyrin. The time course of
the relative concentration of triglucoconjugated por-
phyrin to chlorin (Table 2) clearly shows that 2a/3a
chlorin oxidation does result from a cellular process.
The concentration of 2a/3a first sharply increased
reaching a maximum value [2.8( 0.03) · 10ꢀ6 lmol/lg
of proteins] at 3 h incubation. That concentration
decreased with increasing incubation time––lowered by
60% at 8 h––and remained quite constant between 8 and
14 h incubation. This suggests a possible competition
between cellular uptake and metabolization processes
the latter becoming predominant for long incubation
times. The concentration of 2a/3a determined at 24 h
was 0.55( 0.07) · 10ꢀ6 lmol/lg of proteins and only
0.24( 0.05) · 10ꢀ6 lmol/lg of proteins at 48 h incuba-
tion.
Both species corresponding to peaks 4, 5 and 6 (Fig. 3A
and B) exhibited a maximum concentration around 8 h
decreasing at longer times.
The two compounds resulting from the loss of three
glucose units, appeared only after incubation for more
than 10 h. The determination of the time course of 2d/3d
abundance was somewhat hazardous since peak 7 was
poorly resolved with a low intensity. As shown in Figure
3B, 4d concentration slowly increased with incubation
time. Concentrations of 2.4( 0.8) · 10ꢀ6 lmol/lg of
proteins and 3.6( 0.8) · 10ꢀ6 lmol/lg of proteins were
measured at 24 and 48 h, respectively. These concen-
trations were much higher than that of the parent mol-
ecule 2a/3a (Table 2).
Kinetics of formation of the triglucoconjugated por-
phyrin 4a (peak 2) (Fig. 3B) also shows a maximum
concentration at 3 h incubation that subsequently de-
creased for longer times. The ratio of 4a concentration

3678
I. Laville et al. / Bioorg. Med. Chem. 12 (2004) 3673–3682
140
120
100
80
a
b
60
40
c
d
20
0
0
0.5
1
1.5
2
2.5
Concentration (µM)
Figure 5. Concentration dependence of toxicity in HT29 (3 h incuba-
tion) expressed as cell survival fraction in the dark and exposed to
514 nm light with a light fluence of 5 J/cm2: 2a/3a (a) and (c), 4d (b) and
(d). Experiments have been triplicated. Errors bars, SE.
2.5. Phototoxicity of the final metabolite
Figure 4. HPLC profile (absorption detection, k 420 nm) of a cellular
extract of HT29 treated with 2a/3a (6 lM): (A) 3 h incubation, (B) cells
washed at 3 h incubation and kept in free drug medium up to
24 h. Fraction numbers 1 to 8 correspond to the fn values reported in
Table 1.
In as far as the 4d concentration becomes much higher
than the maximal concentration of the initial com-
pound, the question arises about the consequences of
metabolization processes on the sensitizer photoactivity
for long incubation times. The phototoxicity of 4d in
HT29 cells has been determined and compared to that of
2a/3a. Cells were incubated for 3 h in order to minimize
metabolization of the parent molecule. In both cases the
cytotoxicity was found negligible for a dose going from
0.5 to 2 lM (Fig. 5). The phototoxicity (514 nm, 5 J/cm2)
of 2a/3a and 4d was found quite similar (v2 test, NS,
a ¼ 2:5%) with a LD50 value around 0.7 lM (dose cor-
responding to a survival fraction of 50%). However, the
phototoxicity must be related to the respective amount
of internalized drug. As determined by cellular ex-
traction, 4d concentration was found to be
5.9( 0.8) · 10ꢀ6 lmol/lg of proteins whereas it was
lower [2.5( 0.8) · 10ꢀ6 lmol/lg of proteins] for 2a/3a. A
2–3 higher concentration of deglucosylated porphyrin is
thus necessary to observe a similar activity to that of the
initial glucoconjugated chlorin.
It is interesting to note that the ratio of 4a to 2a/3a
concentration increased with time (Table 2). In addition
the comparison of the HPLC profile of a cellular extract
obtained at 3 h incubation (Fig. 4A) and that of an ex-
tract obtained with cells first incubated for 3 h, then
washed and kept in the dark in free drug medium up to
24 h (Fig. 4B) showed that 4d becomes the most abun-
dant species in the cellular extract. These findings clearly
show that the deglucosylated porphyrin 4d observed
only resulted from a metabolism process and can thus be
assumed to be the final metabolite of the triglucosylated
chlorin 2a/3a. A possible metabolic pathway is illus-
trated in Scheme 2.
3. Conclusion
In conclusion, metabolism of the triglucoconjugated
chlorin, TPC(m-O-Glu)3, 2a/3a results in the formation
of partially or totally deglucosylated compounds, chlo-
rins by action of cellular glucosidases and porphyrins by
oxidation of chlorin macrocycles. The final metabolite
resulting from the loss of three glucose units and oxi-
dation of macrocycle exhibits a biological activity lower
than that of the initial compound. It may thus be as-
sumed that cellular metabolization of TPC(m-O-Glu)3
affects the global photoactivity of this compound. In vivo
metabolization studies are currently under investigation.
4. Experimental
Acetonitrile (HPLC grade), methanol (HPLC grade),
DMSO (analytical-reagent grade), trifluoroacetic acid
(>99% pure) were purchased from VWR (France).
Scheme 2. Possible metabolic pathway of glycoconjugated chlorin
2a/3a.

I. Laville et al. / Bioorg. Med. Chem. 12 (2004) 3673–3682
3679
4.1. Chemistry
mixture was stirred under argon for a further 10 min.
Then a BF3-etherate solution in methylene chloride
(0.3 mL, 0.5 M) was added and the mixture was stirred
under argon then after 2 h BF3-etherate solution
(0.3 mL, 0.5 M) was added again. The mixture was
stirred at room temperature overnight. o-Chloranil
(1.2 g, 4.87 · 10ꢀ3 mol) was added and the solution was
refluxed for 2 h and Et3N (1.5 mL) was added. Silica gel
(10 g) was added to the dark solution and the solvent
was evaporated. The absorbed crude products were
placed on the top of a silica gel chromatographic col-
umn. The porphyrin mixture was eluted with methylene
chloride/acetone (10/1, v/v). The fifth red fraction was
collected and purified again by preparative thin layer
silica gel chromatography eluted with methylene chlo-
ride/acetone (10/1, v/v). The titled pure triglycosylated
porphyrin was crystallized from methylene chloride/
heptane (405 mg, yield 15%).
All chemicals used were of reagent grade and were
purchased from Aldrich or Fluka. Merck silica gel 60
(0.040–0.060 mm) was used for column chromatogra-
phy. Methylene chloride stabilized with ethanol (6 mL/
L) used for synthesis of 5,10,15-tri[3-O-(20,30,40,60-tetra-
acetyl-b-D-glucopyranosyl)phenyl]-20-phenyl porphy-
rin, was kept on CaCO3 overnight, filtered then distilled
ꢁ
on CaCO3 and stored on molecular sieve 4 A. Pyrrole
was distilled on CaH2. BF3-etherate was distilled.
Macherey–Nagel precoated plates (SIL G-200, 2 mm)
were used for preparative thin layer chromatography.
Elemental analysis were carried out by the ‘Service
1
Central de Microanalyse du CNRS’. H NMR spectra
were obtained in the indicated deuteriated solvents with
Brucker AM-300 instruments. Acidic impurities of
chloroform-d3 were removed with anhydrous K2CO3.
Chemical shift values were given in ppm relative to
TMS. Coupling constants were given in Hz. Optical
spectra were recorded using a Varian DMS 200 spec-
trometer.
Anal. Calcd for C86H84N4O30Æ9H2O, theoretical: C,
56.89; H, 5.66; N, 3.09. Found: C, 56.71; H, 4.52; N,
2.68. UV–vis spectrum: kmax, nm (e, L/mmol cm):
(CH2Cl2) 417 (410.1), 513.5 (18.6), 548 (8.1), 588 (6.9),
644 (4.7). 1H NMR (CDCl3): 8.88 (m, 8H, pyrrole), 8.21
(d, 2H, ortho-phenyl), 7.93 (m, 3H, para-phenoxy), 7.88
(br d, 3H, ortho0-phenoxy), 7.78 (d, 3H, para- and meta-
phenyl), 7.68 (m, 3H, meta-phenoxy), 7.44 (d, 3H, ortho-
phenoxy, J ¼ 7:4 Hz), 5.37 (m, 6H, HC1 and HC2 or
HC3 ‘ose’), 5.31 (m, 3H, HC3 or HC2 ‘ose’), 5.17 (m, 3H,
HC4 ‘ose’), 4.18 (m, 3H, HC6a ‘ose’), 4.04 (m, 3H, HC6b
‘ose’), 3.80 (m, 3H, HC5 ‘ose’), 2,09 (s, 9H, acetyl), 2.03
(s, 9H, acetyl), 1.98 (s, 9H, acetyl), 1.39 (s, 9H, acetyl),
)2.80 (s, 2H, NH). 13C NMR (CDCl3): 170.3 (C@O
acetyl), 170.2 (C@O acetyl), 169.3 (C@O acetyl), 155.3
(meta-phenoxy), 143.5 (meso-phenoxy), 141.8 (meso-
phenyl), 134.5 (ortho-phenyl), 131 (CH pyrrole), 129.8
(para-phenoxy), 127.8 (meta-phenoxy), 126.8 (meta-
phenyl), 123 (para-phenyl), 123–122.5 (br, ortho0-phen-
oxy), 120.6 (meso-C-phenyl), 119.3–119.1 (meso-C-
phenoxy), 116.6 (ortho-phenoxy), 99.3 (C1 ‘ose’), 72.7
(C2=3 ‘ose’), 72.2 (C5 ‘ose’), 71.2 (C3=2 ‘ose’), 68.2 (C4
‘ose’), 61.9 (C6 ‘ose’), 20.7 (CH3 acetyl), 20.6 (CH3
acetyl), 20.5 (CH3 acetyl), 19.9 (CH3 acetyl).
4.1.1. 3-O-(20,30,40,60-Tetraacetyl-b-
D-glucopyranosyloxy)-
benzaldehyde. A solution of 3-hydroxybenzaldehyde
(5.12 g, 42 · 10ꢀ3 mol) in methylene chloride (50 mL) was
vigorously stirred with an aqueous solution of sodium
hydroxide (5%, 70 mL) and tetrabutylammonium bro-
mide (2.26 g, 7 · 10ꢀ3 mol) at room temperature. To this
mixture, a solution of 2,3,4,6-tetraacetyl-a-D-glucopyr-
anosyl bromide (11.510 g, 28 · 10ꢀ3 mol) in methylene
chloride (20 mL) was added. The mixture was vigorously
stirred at room temperature for three days. After sepa-
ration, the organic layer was washed by aqueous sodium
hydroxide solution (5%, 2 · 20 mL) and water. The or-
ganic phase was dried over sodium sulfate, filtered and
concentrated under vacuum. The crude yellow oil was
purified by chromatography on silica gel eluted by a
mixture of ethyl acetate/heptane (1/1, v/v) affording the
pure titled product after crystallization from a mixture
of ethanol/water (2.710 g, yield 21%).
Anal. Calcd for C21H24O11, theoretical: C, 55.75; H,
5.35. Found: C, 55.65; H, 5.57.
1H NMR (CDCl3): 9.96 (s, 1H, CHO), 7.57 (d, 1H, H
phenyl, J ¼ 7:3 Hz), 7.50 (d, 1H, H phenyl), 7.44 (d, 1H,
H phenyl, J ¼ 7:5 Hz), 7.25 (m, 1H, H phenyl), 5.29 (d,
1H, HC1 ‘ose’, J ¼ 7:8 Hz), 5.30 (m, 1H, H ‘ose’), 5.15
(m, 2H, H ‘ose’), 4.21 (m, 2H, HC6 ‘ose’), 3.91 (m, 1H,
HC5 ‘ose’), 2.08 (s, 3H, acetyl), 2.04 (s, 6H, acetyl), 2.02
(s, 3H, acetyl).
4.1.3. 5,10,15-Tri(3-methoxyphenyl)-20-phenyl porphyrin
5. Freshly distilled pyrrole (2.75 mL, 4 · 10ꢀ2 mol),
benzaldehyde (1.01 mL, 10ꢀ2 mol) and meta-anisalde-
hyde (3.35 mL, 3 · 10ꢀ2 mol) are dissolved in propionic
acid (250 mL). The solution was refluxed for 1 h then
quickly cooled. The crude black solution was concen-
trated under vacuum then methanol (100 mL) was
added. The blue precipitate was filtered and washed with
cold methanol. The mixture of porphyrins was separated
by silica gel chromatography column eluted with a
mixture of methylene chloride/heptane (5/1, v/v). The
third red band was collected and purified again by silica
gel chromatography column eluted by toluene. The pure
blue crystals of 5,10,15-tri(3-methoxyphenyl)-20-phenyl
porphyrin was obtained after crystallization from
methylene chloride/methanol mixture (456 mg, yield
6.5%).
4.1.2. 5,10,15-Tri[3-O-(20,30,40,60-tetraacetyl-b-
D-gluco-
pyranosyloxy)phenyl]-20-phenyl porphyrin 1. Freshly
distilled pyrrole (0.410 g, 6.1 · 10ꢀ3 mol) in methylene
chloride (70 mL), 3-O-(20,30,40,60-tetraacetyl-b-
D-gluco-
pyranosyloxy)benzaldehyde(2.26 g, 5 · 10ꢀ3 mol) and
freshly distilled benzaldehyde (0.177 g, 1.7 · 10ꢀ3) in
methylene chloride (70 mL) were added to methylene
chloride (750 mL) purged by argon during 30 min. The

3680
I. Laville et al. / Bioorg. Med. Chem. 12 (2004) 3673–3682
Anal. Calcd for C47H36N4O3, theoretical: C, 80.09; H,
5.15; N, 7.95. Found: C, 79.43; H, 5.17; N, 7.88. UV–vis
spectrum: kmax, nm (e, L/mmol cm): (CH2Cl2) 418.5
until the absorption peak at 735 nm (bacteriochlorin)
disappeared. The solution was washed with aqueous
solution of NaHSO3 (5%, 100 mL), water (100 mL) and
dried over anhydrous sodium sulfate. The filtered solu-
tion was concentrated under vacuum. The residue was
purified by column chromatography on silica gel eluted
by a mixture of methylene chloride/heptane (5/1, v/v).
The pure product (red band) was crystallized from a
mixture of methylene chloride/heptane (121 mg, yield
60%).
1
(502.9), 515 (21.7), 549 (8.9), 590 (7.3), 646.5 (5.9). H
NMR (CDCl3): 8.88 (s, 6H, pyrrole), 8.83 (d, 2H, pyr-
role, J ¼ 4:2 Hz), 8.21 (d, 2H, ortho-phenyl, J ¼ 6:1 Hz),
7.80 (m, 3H, meta- and para-phenyl), 7.80 (m, 6H, ortho-
phenoxy), 7.64 (t, 3H, meta-phenoxy, J ¼ 7:7 Hz), 7.32
(d, 3H, para-phenoxy, J ¼ 8 Hz), 3.98 (s, 9H, OCH3),
)2.79 (s, 2H, NH). 13C NMR (CDCl3): 157.7 (meta-
phenoxy), 143.2 (meso-phenoxy), 134.3 (ortho-phenyl),
130.8 (C-pyrrole), 127.3 (meta0-phenoxy), 120 (meso-C),
113.2 (para-phenoxy), 55.3 (OCH3).
Microanalysis calcd for C47H38N4O3: C, 79.86; H, 5.42;
N, 7.93. Found: C, 79.75; H, 5.26; N, 7.62. UV–visible
spectrum in CH2Cl2 kmax (e, L/mmol cm): 401 (178.9),
419.5 (233.9), 518.5 (18.4), 545 (12.9), 598 (8.5), 651.5
1
4.1.4. 5,10,15-Tri(3-hydroxyphenyl)-20-phenyl porphyrin
4d. 5,10,15-Tri(3-methoxyphenyl)-20-phenyl porphyrin
5 (100 mg, 0.14 · 10ꢀ3 mol) was dissolved in dry methyl-
ene chloride (25 mL) at )20 ꢂC under argon. BBr3
(0.604 mL, 6.3 · 10ꢀ3 mol) was slowly added. The green
solution was stirred at )20 ꢂC for 3 h then at room
temperature overnight. The green mixture was diluted in
cold water and neutralized by a saturated aqueous
solution of NaHCO3. The solution was extracted by
ethyl acetate. The organic phase was washed with water
(twice), dried with sodium sulfate, filtered then evapo-
rated. The pure porphyrin was obtained after crystalli-
zation from ethyl acetate/heptane (81 mg, yield 84%).
(37.7). H NMR (CDCl3): 8.61ꢂ, 8.56ꢂꢂ (d, 2H, pyrrole,
J ¼ 4:9 Hz), 8.46ꢂ, 8.40ꢂꢂ (mꢂ, dꢂꢂ, 2H, pyr-
role, Jꢂ ¼ 4:50 Hz), 8.23ꢂ, 8.17ꢂꢂ (d, 2H, pyrrole,
J ¼ 4:85 Hz), 8.10 (dd, 1H, ortho-phenyl, J ¼ 2 and
7.6 Hz), 7.87 (dd, 1H, para-phenyl, J ¼ 2:4 and 7 Hz),
7.69 (t, 6H, para and meta-phenyl and ortho-phenoxy,
J ¼ 7 Hz), 7.58 (q, 3H, phenoxy, J ¼ 7:5 Hz), 7.44 (m,
3H, para-phenoxy), 7.26 (dd, 2H, meta-phenoxy,
J ¼ 2:6 and 8.20 Hz), 7.20 (dd, 1H, meta-phenoxy,
J ¼ 2:5 and 8.3 Hz), 4.19ꢂ, 4.17ꢂꢂ (m, 2H, CH2 pyrrole),
3.94ꢂ, 3.93ꢂꢂ (s, 9H, OMe), )1.47 (s, 2H, NH).
4.1.6. 5,10,15-Tri(3-hydroxyphenyl)-20-phenyl 2,3-chlo-
rin and 5,10,15-tri(3-hydroxyphenyl)-20-phenyl 7,8-chlo-
rin 2a/3a. 5,10,15-Tri(3-methoxyphenyl)-20-phenyl 2,3-
chlorin and 5,10,15-tri(3-methoxyphenyl)-20-phenyl 7,8-
chlorin (112 mg, 0.16 · 10ꢀ3 mol) was dissolved in dry
methylene chloride (30 mL) at )20 ꢂC under argon. BBr3
(0.685 mL, 7.1 · 10ꢀ3 mol) was slowly added. The green
solution was stirred at )20 ꢂC for 3 h then at room
temperature overnight. The green mixture was diluted in
cold water and neutralized by a saturated aqueous
solution of NaHCO3. The solution was extracted by
ethyl acetate. The organic phase was washed with water
(twice), dried with sodium sulfate, filtered then evapo-
rated. The pure chlorin was obtained after crystalliza-
tion from ethyl acetate/heptane (100 mg, yield 95%).
Anal. Calcd for C44H30N4O3, theoretical: C, 79.74; H,
4.56; N, 8.45. Found: C, 79.32; H, 4.75; N, 8.25. UV–vis
spectrum: kmax, nm (e, L/mmol cm): (acetone) 415.5
(368.7), 512.5 (17.8), 546.5 (7.8), 589.5 (6.1), 645.5 (4.7).
1H NMR (acetone-d6): 8.97 (s, 4H, pyrrole), 8.96 (d, 2H,
pyrrole, J ¼ 8:1 Hz), 8.84 (d, 2H, J ¼ 4:7 Hz), 8.20 (dd,
2H, J ¼ 2 and 7.4 Hz), 7.72 (m, 9H, ortho-phenoxy,
meta-phenyl and para-phenyl), 7.57 (m, 3H, meta-
phenoxy, J ¼ 8 Hz), 7.31 (d, 3H, para-phenoxy,
J ¼ 8 Hz), )2.72 (s, 2H, NH). 13C NMR (acetone-d6):
143.7 (meso-phenoxy), 134.6 (ortho-phenyl), 131.5
(C-pyrrole), 128.2 (para-phenyl), 128.1 (meta0-phenoxy),
127.1 (ortho-phenoxy), 126.8 (meta-phenyl), 122.4
(ortho-phenoxy), 120.5 (meso-C), 115.4 (para-phenoxy).
Microanalysis calcd for C44H32N4O3ÆH2O: C, 77.40; H,
5.02; N, 8.21. Found: C, 77.23; H, 4.91; N, 8.04. UV–
visible spectrum in acetone kmax (e, L/mmol cm): 404
(107.1), 416 (121), 516 (10.2), 542 (7), 597 (4.3), 650.5
(24.4). 1H NMR (acetone-d6): 8.55ꢂ, 8.47ꢂꢂ (d, 2H,
pyrrole, J ¼ 4:5 Hz), 8.53 (s, 3H, OH), 8.33ꢂ, 8.23ꢂꢂ (d,
2H, pyrrole, J ¼ 4:5 Hz), 8.17ꢂ, 8.07ꢂꢂ (d, 2H, pyrrole,
J ¼ 4:6 Hz), 7.98 (m, 1H, ortho-phenyl), 7.79 (m, 1H,
para-phenyl), 7.59 (m, 3H, para and meta-phenyl), 7.44
(m, 6H, ortho-phenoxy), 7.26 (m, 3H, para-phenoxy),
7.09 (dd, 3H, meta-phenoxy, J ¼ 8 and 20 Hz), 4.10ꢂ,
4.07ꢂꢂ (m, 2H, CH2 pyrrole), )1.60 (s, 2H, NH).
4.1.5. 5,10,15-Tri(3-methoxyphenyl)-20-phenyl 2,3-chlo-
rin and 5,10,15-tri(3-methoxyphenyl)-20-phenyl 7,8-chlo-
rin. 5,10,15-Tri(3-methoxyphenyl)-20-phenyl porphyrin
5 (200 mg, 0.284 mmol) and anhydrous K2CO3 (265 mg)
were added to dry pyridine (13 mL) under argon. Tol-
uene-4-sulfonohydrazide (86 mg, 0.46 mmol) was then
added and the mixture was heated under argon at 100–
105 ꢂC for 24 h. Further quantities of toluene-4-sul-
fonohydrazide (86 mg in 0.5 mL of dry pyridine) were
added after 2, 4, 6 and 8 h. After cooling, the crude
mixture was treated with ethyl acetate (90 mL), water
(45 mL) and heated, at 100 ꢂC, for 1 h. After cooling, the
organic phase was separated and washed with HCl (2 M,
100 mL), water (75 mL) and saturated water solution of
NaHCO3. The presence of chlorin and bacteriochlorin
was controlled by UV–visible spectroscopy (bands at
651 and 738 nm, respectively). ortho-Chloranil (196 mg)
was slowly added to the stirred organic solution at 25 ꢂC
4.2. Cell culture and sensitizer incubation
Human colorectal adenocarcinoma cells (HT29) were
allowed to grow to confluence in Dulbecco’s modified
medium (DMEM) supplemented with 10% FCS, gluta-

I. Laville et al. / Bioorg. Med. Chem. 12 (2004) 3673–3682
3681
mine and antibiotics. All reagents were obtained from
4.6. Phototoxicity assay
Bio-Media (France). Cells were subcultured by dispersal
with 0.25% trypsin and seeded (106 cells/mL). Incuba-
tions were performed in the dark. Stock solutions of 2a/
3a (5 mg/mL) were prepared in DMSO and dilutions
were performed in culture medium in the presence of 2%
of fetal calf serum (FCS).
The phototoxicity of TPC(m-O-Glu)3 2a/3a was deter-
mined in HT29 cells following the detailed procedure in
Ref. 8. Briefly, cells were seeded into 96-well plates at
105 cells/mL and allowed to grow in an incubator (5%
CO2, 37 ꢂC, humidified atmosphere, Jouan, France). On
the day of experiment, fresh DMEM with 2% FCS
containing the photosensitizer at a final concentration
between 0.5 and 2 lM were added in each well. Cells
were incubated for 3 h and then washed before fresh
complete medium was added. Irradiations were per-
formed at 514 nm (5 J/cm2) with an Arþ laser (Spectra
Physics-2020) under sterile conditions. Cell viability was
measured 24 h later by determination of mitochondrial
activity using the MTT assay according to Mossmann.24
Optical densities of microplates were determined at 570
and 640 nm using a Labsystemꢀ.
4.3. Cell extraction
Treated cells were washed three times with cold PBS and
then removed using a cell scraper. After sonication,
EtOAc (3 mL) was added to 600 lL of cell suspension in
PBS. The resulting solution was left in the dark for
30 min with sustained shaking and then centrifuged at
1700g for 10 min. The supernatant was removed and
evaporated to dryness for further analysis. The sensitizer
was recovered in greater than 90% yield. As determined
by mass spectrometry, this procedure does not promote
any deglucosylation of our glycoconjugated chlorins
(ratio of 2a/3a to 4a remains the same). The corre-
sponding protein concentration was determined by
Lowry’s method23 using a protein assay (Kit P56 56,
Sigma Aldrich––France).
4.7. Intact cell enzymatic assay
Glucosidase activity in HT29 cells was determined
according to Sawkar et al.25 Cells were allowed to grow
for 24 h into 24-well plates (2 · 105 cells/well). Cellular
medium was replaced by 150 lL of NaOAc buffer (pH 4)
and the enzymatic reaction was started by the addition
of 100 lL of 4-methylumbelliferyl glucoside (5 mM).
One hour later, the reaction was stopped and the cells
were lysed by the addition of 2 mL of glycine buffer
(pH 10). The concentration of released 4-methylum-
belliferone (4-MU) was measured by fluorescence
spectroscopy (kexc: 365 nm, kem: 445 nm). Enzyme
concentration was determined using a calibration curve
established for 4-MU in glycine buffer.
4.4. Mass spectrometry analysis
Positive ion MALDI-TOF mass spectra were recorded
using a PerSeptive Biosystems Voyager Elite (Framing-
ham––USA) time-of-flight mass spectrometer equipped
with a 337 nm nitrogen laser (VSL 337ND). It was
operated in the reflectron delayed extraction mode
(acceleration voltage: 20 kV, percent grid voltage: 60%,
extraction delay: 125 ns) and the laser fluence was ad-
justed with a variable-beam attenuator. Ions were de-
tected by a dual channel plate detector. The mass
spectrum presented in Figure 1 represents an average
over 256 consecutive laser shots (3 Hz repetition rate).
For sample preparation, the matrix, a-cyano-4-hy-
droxycinnamic acid (HCCA), was dissolved at a con-
centration of 0.1 M in methanol. Ten microliters of
matrix solution were added to the evaporated superna-
tant of cellular extract and 1 lL deposited onto a pol-
ished gold sample stage and allowed to dry in air.
Note: ꢂ2,3-chlorin or 7,8-chlorin isomer, ꢂꢂ7,8-chlorin or
2,3-chlorin isomer. The ratio of isomers is 1/1.
Acknowledgements
This work has been supported by the French Associa-
tion for Cancer Research (Association pour la Recher-
che contre le Cancer, ARC), grant no 5334.
4.5. HPLC separation
References and notes
HPLC separations were achieved using a Spheri 5RP18
column (220 mmꢃ4.6 mm; 5 lm particle diameter). The
mobile phase was 0.1% aqueous trifluoroacetic acid
(TFA)/acetonitrile (30/70, v/v) pumped at a flow rate of
0.8 mL/min (LC-10AS pump, Shimadzu, Japan). Dried
extracts were diluted in 250 lL of mobile phase from
which 100 lL aliquots were injected. A Shimadzu SPD-
6AV UV–visible spectrophotometric detector set at
420 nm was used along with a fluorescence detector (FR-
551) set at 420 and 650 nm, for the excitation and
emission wavelengths, respectively. The HPLC device
was equipped with a CR5A chromatopac integrator
(Shimadzu, Japan).
1. Hsi, R. A.; Rosenthal, D. I.; Glatstein, E. Photodynamic
therapy in the treatment of cancer. Drugs 1999, 57, 725–
734.
2. Mac Donald, I. J.; Dougherty, T. J. Basic principles of
photodynamic therapy. J. Porphyrins Phtalocyanines 2001,
5, 105–129.
3. Boyle, R. W.; Dolphin, D. Structure and biodistribution
relationships of photodynamic sensitizers. Photochem.
Photobiol. 1996, 64, 469–485.
4. Zheng, G.; Potter, W. R.; Camacho, S. H.; Missert, J.;
Wang, G.; Bellnier, D. A.; Hendersson, B. W.; Rodgers,
M. A. J.; Dougherty, T. J.; Pandey, R. K. Synthesis,
photophysical properties, tumor uptake and preliminary

3682
I. Laville et al. / Bioorg. Med. Chem. 12 (2004) 3673–3682
in vivo photosensitizing efficacy of a homologous series of
3-(10-alkyloxy)ethyl-3-divinylpurpurin-18-N-alkylimides with
variable lipophilicity. J. Med. Chem. 2001, 44, 1540–1559.
15. Campbell, G. A.; Bartels, K. E.; Arnold, C.; Healey, T.;
Cowell, R. L.; Lucroy, M. D.; Ronn, A. M. Tissue levels,
histologic changes and plasma pharmacokinetics of meta-
tetra(hydroxyphenyl)chlorin (mTHPC) in the cat. Lasers
Med. Sci. 2002, 17, 79–85.
16. Whelpton, R.; Michael-Titus, A. T.; Basra, S. S.; Grahn,
M. Distribution of temoporfin, a new photosensitizer for
the photodynamic therapy of cancer, in a murine tumor
model. Photochem. Photobiol. 1995, 4, 397–401.
5. Zheng, G.; Graham, A.; Shibata, M.; Missert, J.; Oseroff,
A. R.; Dougherty, T. J.; Pandey, R. K. Synthesis of b-
galactose-conjugated chlorins derived by enyne methath-
esis as galectin-specific photosensitizers for photodynamic
therapy. J. Org. Chem. 2001, 66, 8709–8716.
ꢀ
6. Momenteau, M.; Maillard, Ph.; de Belinay, M.-A.; Carrez,
D.; Croisy, A. Tetrapyrrolic glycosylated macrocycles for
an application in PDT. J. Biomed. Opt. 1999, 4, 298–318.
7. Momenteau, M.; Oulmi, D.; Maillard, Ph.; Croisy, A.
In vitro photobiological of a new series of photosensitiz-
ers: the glycoconjugated porphyrins. SPIE Photodyn.
Therapy Cancer 1994, 2325, 13–23.
17. Cai, H.; Wang, Q.; Luo, J.; Lim, C. K. Study of
temoporfin metabolism by HPLC and electrospray mass
spectrometry. Biomed. Chromatogr. 1999, 13, 354–359.
18. Houle, J. M.; Strong, A. Clinical pharmacokinetics of
verteporfin. J. Clin. Pharmacol. 2002, 42, 547–557.
19. Ferrario, A.; Kessel, D.; Gomer, C. J. Metabolic proper-
8. Laville, I.; Figueiredo, T.; Loock, B.; Pigaglio, S.; Mail-
lard, Ph.; Grierson, D. S.; Carrez, D.; Croisy, A.; Blais, J.
Synthesis, cellular internalization and photodynamic
activity of glucoconjugated derivatives of tri and tetra-
(meta-hydroxyphenyl)chlorins. Bioorg. Med. Chem. 2003,
11, 1643–1652.
9. Sperker, B.; Backman, J. T.; Kroemer, H. K. The role of
b-glucuronidase in drug disposition and drug targeting in
humans. Clin. Pharmacokinet. 1997, 33, 18–31.
ties and photosensitizing responsiveness of mono-L-aspart-
yl chlorin e6 in a mouse tumor model. Cancer Res. 1992,
52, 2890–2893.
20. Experts report, Birmingham, 2002, Feb. 22th.
21. Sasaki, K.; Yumita, N.; Nishigaki, R.; Sakata, I.; Naka-
jima, S.; Umemura, S. I. Pharmacokinetic study of a
gallium-porphyrin photo- and sono-sensitizer, ATX-70, in
tumor-bearing mice. Jpn. J. Cancer Res. 2001, 92, 989–
995.
10. Gil-Martin, E.; Rodriguez-Berrocal, J.; Paez de la Cadena,
M.; Fernandez-Briera, A. Alterations of glycosidases in
human colonic adenocarcinoma. Clin. Biochem. 1997, 30,
17–25.
11. Tiby, V.; Lombardo, A.; Goi, G.; Tettamanti, G.; Gaini,
S. M.; Canal, N. Study of some lysosomal glycohydrolases
in tumors of the nervous system. Res. Commun. Chem.
Pathol. Pharmacol. 1984, 44, 149–161.
12. Goi, G.; Bairati, C.; Segalini, G.; Burlina, A. B.; Massac-
cesi, L.; Lovagnini, A.; Lombaro, A. Alterations in the
activity of several glycohydrolases in red blood cell
membrane from type 2 diabetes mellitus patients. Meta-
bolism 1999, 48, 817–821.
22. The formation of isomeric chlorins and incomplete
reduction is a shortcoming inherent to the synthetic
method described by Bonnett et al. used to effect porphy-
rin to chlorin transformation. Bonnett, R.; White, S. G.;
Winfiel, U.-J.; Berembaum, M. C. Hydroporphyrins of the
meso-tetra(hydroxyphenyl)porphyrin series as tumor
photosensitizers. Biochem. J. 1989, 261, 277–280; Bon-
nett, R.; Nizhnik, A. N.; White, S. G.; Berembaum, M. C.
Porphyrin sensitizers in tumour phototherapy. Novel
sensitizers of the chlorin and bacteriochlorin class with
amphiphilic properties. J. Photochem. Photobiol. B: Biol.
1990, 6, 29–37.
23. Peterson, G. L. A simplification of the protein assay
method of Lowry et al. which is more generally applicable.
Anal. Biochem. 1977, 83, 346–356.
24. Mosmann, T. Rapid colorimetric assay for cellular growth
and survival: application to proliferation and cytotoxicity
assays. J. Immunol. Meth. 1983, 65, 55–63.
13. Monsigny, M.; Roche, A. C.; Kieda, C.; Midoux, P.;
ꢀ
Obrenovitch, A. Characterization and biological implica-
tions of membrane lectins in tumor, lymphoid and
myeloid cells sugar. Biochimie 1989, 70, 1633–1649.
14. Rovers, J. P.; Saarnak, A. E.; de Jode, M.; Sterenborg, H.
J.; Terpstra, O. T.; Grahn, M. F. Biodistribution and
bioactivity of tetra-pegylated meta-tetra(hydroxyphen-
yl)chlorin compared to native meta-tetra(hydroxyphen-
25. Sawkar, A. R.; Cheng, W. C.; Beutler, E.; Wong, C. H.;
Balch, W. E.; Kelly, J. W. Chemical chaperones increase
the cellular activity of N370S b-glucosidase: a therapeutic
strategy for Gaucher disease. Proc. Natl. Acad. Sci. U.S.A.
2002, 99, 15428–15433.
yl)chlorin in a rat liver tumor model. Photochem.
Photobiol. 2000, 71, 211–217.