Purpurinimides as photosensitizers: Effect of the presence and position of the substituents in the in vivo photodynamic efficacy

Article abstract of DOI:10.1016/S0960-894X(00)00274-2

This study presents a novel approach for the regioselective synthesis of a series of alkyl ether analogues of purpurin-18-N-alkylimide. In the purpurinimide series, this is the first example which demonstrates that the presence and position of the substituents in the macrocycle makes a remarkable difference in the in vivo PDT efficacy. (C) 2000 Elsevier Science Ltd. All rights reserved.

Full text of DOI:10.1016/S0960-894X(00)00274-2

Bioorganic & Medicinal Chemistry Letters 10 (2000) 1463±1466
Purpurinimides as Photosensitizers: E€ect of the Presence and
position of the Substituents in the In Vivo Photodynamic Ecacy
a,y
a
a
Ankur Rungta, Gang Zheng, Joseph R. Missert,
a
a
a,b,
William R. Potter, Thomas J. Dougherty and Ravindra K. Pandey *
aPhotodynamic Therapy Center, Roswell Park Cancer Institute, Bu€alo, NY 14263, USA
bDepartment of Nuclear Medicine, Roswell Park Cancer Institute, Bu€alo, BY 14263, USA
Received 24 March 2000; accepted 1 May 2000
AbstractÐThis study presents a novel approach for the regioselective synthesis of a series of alkyl ether analogues of purpurin-18-N-
alkylimide. In the purpurinimide series, this is the ®rst example which demonstrates that the presence and position of the substituents
in the macrocycle makes a remarkable di€erence in the in vivo PDT ecacy. # 2000 Elsevier Science Ltd. All rights reserved.
For quite some time one of the main objectives of our
laboratory has been to establish structure±activity rela-
tionships in a variety of photosensitizers. This study is
Our next step was to investigate the e€ect of substituent
regiochemistry on PDT ecacy. In this e€ort, purpurin-
imides 2±5 were synthesized (Scheme 1) and their in vivo
ecacy under similar treatment conditions was com-
pared with compound 1. Compounds 2 and 4, in which
1
aimed to establish a generic structural requirement(s)
for e€ective photosensitizing agents. The ultimate goal
is to increase the tumor selectivity of a drug, so that the
undesirable side-e€ects associated with most of the
photosensitizers could be eliminated or reduced. In such
e€ort, we have previously reported our results obtained
from QSAR studies of a series of alkyl ether analogues of
0
the (1 -heptyloxyethyl) group was regioselectively intro-
duced at positions-8 and -20 of the macrocycle, were
prepared to investigate the e€ect of the same substituent
present at various positions of the ring system. In order
to investigate the e€ect of a primary versus secondary
2
0
pyropheophorbide a. These data indicated that overall
alkyl ether group in PDT ecacy, the (1 -heptylox-
lipophilicity of the molecule made a signi®cant di€erence
in PDT ecacy. We then extended this approach to the
purpurinimide series in which the lipophilic characteristic
of the molecule can be varied either by altering the length
of the imide-N-alkyl group or the alkyl ether substituent
regioselectively introduced at position-3 of the macro-
yethyl) substituent present at position-3 of purpur-
inimide 1 was replaced with an octyloxymethyl group
(compound 3) without altering the overall lipophilicity.
Finally, in our attempt to determine the importance of
the secondary heptyloxyethyl group in PDT ecacy,
compound 5 lacking alkyl ether substituent at position
3, but, possessing lipophilicity similar to purpurinimide
1 was also prepared.
3
cycle. Among the compounds evaluated so far, the 3-
devinyl-3(l -heptyloxy ethyl)-purpurin-18-N-hexylimide 1
0
was found to be quite e€ective (100% tumor cure at day
3
0 at a dose of 1.0 mmol/kg, treated with light (135J/
2
cm ), 24 h post injection of the drug). In the same series,
we have also shown that among the compounds with
similar lipophilicity, the photosensitizers containing the
methyl ester functionalities are much more e€ective
Results and Discussion
Purpurinimides 1±5 were synthesized by following the
reaction sequence shown in Scheme 1. In brief, methyl-
pheophorbide-a 6 was extracted from the algae Spir-
4
than those possessing amide substituents.
5
ulina paci®ca by following the literature procedure.
The ®ve member isocyclic ring present in 6 was then
converted into a fused six member cyclic anhydride sys-
tem 7, known as purpurin-18 methyl ester by following the
methodology developed in our laboratory. Reaction of
purpurin-18 with n-hexylamine produced 8 as a mixture of
two isomers, which were not separated into individual
*
8
Corresponding author. Tel.: +1-716-845-3203; fax: +1-716-845-
920; e-mail: rpandey@sc3103.med.bu€alo.edu
6
y
Research student (July 1998±December 1999) from Williamsville East
High School, East Amherst, NY 14051, USA.
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00274-2

1
464
A. Rungta et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1463±1466
Scheme 1. Synthesis of various purpurinimide analogues.
isomers and via intramolecular base catalyzed cyclization
a€orded purpurin-18-N-hexylimide 9 in 70% yield (based
on compound 7) exhibiting a long wavelength absorption
near 705 nm. Reaction of 9 with HBr/acetic acid and then
with 1-heptanol produced purpurinimide 1 (m/z calcd for
C H N O : 777.48; found: 778.2 (M+l)) in 72% yield.
In order to investigate the e€ect of a primary versus
secondary alkyl ether group in PDT ecacy, the 3-
vinyl-N-hexyl purpurinimide 9 was converted into the
corresponding formyl analogue 12 by reacting with
sodium periodate and osmium tetroxide as a catalyst in
70% yield. The completion of the reaction was mon-
itored by UV±visible spectroscopy (disappearance of a
peak at 705 nm and appearance of a new peak at 735
nm). Reaction of 12 with sodium borohydride (NaBH4)
in tetrahydrofuran (THF) produced the 3-devinyl-3-
hydroxymethyl-purpurin-18-N-hexyl-imide 13 in 78%
yield with long wavelength absorption at 700 nm (m/z
calcd for C H N O : 665.35; found: 666.50 (M+l)).
4
7
63
5
5
For the preparation of purpurinimide 2, the 3-vinyl
group in imide 9 was ®rst converted into an ethyl sub-
stituent on hydrogenation using Pd/C as a catalyst to
produce compound 10, which on reacting with osmium
7
tetraoxide/o-dichlorobenzene generated 8-vinyl analogue
11 (m/z calcd for C H N O : 677.37; found: 678.86
40 48 5 5
(M+l)). At the ®nal step, reaction of 11 with 30% HBr/
acetic acid and 1-heptanol gave the desired purpurin-
39
47
5
5
Further reaction of 13 with HBr/acetic acid and sub-
sequent treatment with 1-octanol produced the desired
3-(1-octyloxyethyl) derivative 3 in 60% yield (m/z calcd
imide 2 in 60% yield (m/z calcd for C47 H N O :
63
5
5
777.48; found: 778.50 (M+l)).

A. Rungta et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1463±1466
1465
for C H N 0 : 778.06; found: 779.2). The synthesis of
4
(log P 10.83), mesopurpurin-18 methyl ester 7 was reacted
with N-decylamine. The intermediate amide 14 was con-
verted into the desired N-decylimide 5 in 60% yield in the
same manner as discussed for the preparation of the rela-
ted imide analogue 8. The reaction sequence followed for
the preparation of the desired purpurinimide is illustrated
in Scheme 3. The structure was con®rmed by NMR (which
clearly indicated the absence of the vinyl resonances and
presence of an additional ethyl group), and mass spectro-
metry analyses (m/z calcd for C H O N : 720.00; found:
7
63 5 5
compound 4 was achieved in a sequence of reaction
from purpurinimide 9 in good yield (m/z calcd for
C H O N; 792.08; found: 793.4 (M+1)).
4
8
65
5
The structures 1±4 were further con®rmed by NMR
spectrometry. Some important features observed in the
NMR spectra are shown in Figure 1. In all the alkyl ether
analogues, the vinyl(� CH=CH ) resonances generally
2
observed at d 7.84 (dd), d 6.27(d) and 6.16(d) were absent
44
57
4
5
(
Figs. 1A±D). In both 3- and 8-(l-heptyloxymethyl) ana-
721.10).
logues 1 and 4, the presence of a quartet at d 5.80 ppm
for � CH(O-heptyl), integrating for one proton, con®rmed
the presence of the desired products (Figs 1A and B). As
expected, in purpurin imides containing either a hydoxy-
methyl group or a hexyloxymethyl substituent at posi-
tion-3 the macrocycle showed singlet at d 5.82 ppm for
In Vivo Photosensitizing Ecacy
The photosensitizers at various doses were injected
intravenously to 5±7 weeks old mice (6 mice/group,
transplanted subcutaneously with RIF tumor into the
axilla). Mice were restrained in aluminum holders and each
CH -OR (R=H or � octyl) integrating for two protons
2
(
Fig. 1c). In the NMR spectrum 4, the presence of only
2
two down®eld singlets at the meso-region (9.12 and 9.35
ppm) indicated that the position adjacent to the reduced
ring is occupied with a heptyl oxymethyl group (the
methylene resonances appeared as doublets instead of a
singlet as observed for the 3-octyloxy-methyl analogue
tumor was illuminated with 135J/cm light from a laser
tuned at the longest wavelength absorption maximum of
the photosensitizers. The percentage of the tumor
regrowth was recorded daily.
3
action of the -CH O-heptyl protons with the 18-H pro-
ton of ring D (trans-reduced)). In order to prepare
purpurinimide 5, lacking an alkyl ether substituent,
but, having lipophilicity similar to photosensitizer 1
) at d 5.5 ppm and 5.2 ppm (Fig. 1D) due to the inter-
In an initial experiment, groups of mice bearing RIF
2
tumors were treated with laser light (75 mw/cm , 135J/
2
2
cm , 702 nm) after the administration of purpurinimide
1 at three variable doses (4.0, 1.0 and 0.4 mmol/kg) and
the best tumor response (100% tumor cure at day 30)
was obtained at a dose of 1.0 mmol/kg. The treatment
wavelength (702 nm) was selected by in vivo re¯ectance
8
spectroscopy. The sensitizing ecacy of other photo-
senitizers was then compared at the same dose and
treatment conditions. The biological results are summar-
ized in Table 1. As can be seen, among the compounds
tested, the 3-devinyl-3 (1-heptyloxymethyl)-purpurin-18-
N-hexylimide 1 appeared to be most e€ective, and pro-
duced 100% tumor cure at day 30. To our surprise, the
positional isomer 2, in which the heptyloxyethyl group
was substituted at position-8 of the macrocycle, under
similar treatment conditions did not produce any PDT
activity (day 1±2: tumor response 0%). Replacement of
the secondary heptyl ether side-chain at position-3 of
purpurinimide 1 with a primary alkyl group (purpur-
inimide 3) produced a remarkable decrease in antitumor
activity. Photosensitizers 4, containing an alkyl ether
side-chain at the meso-position was found to be toxic at
a dose of 1 mmol/kg (3 out of 5 mice died after treating
with light but the surviving mice were tumor free at day
Figure 1.
Table 1. Preliminary in vivo antitumor activity of certain purpurinimidesa
Compound
Log P
value
Drug dose
mmol/kg
Light dose
2
In vivo
absorption
Tumor response
(% tumor cure)
(J/cm )
Day 1±2
100
Day 7
100
Day 14
Day 21
100
Day 30
100
1
2
3
4
5
10.83
10.83
10.80
10.78
10.62
1.0
1.0
1.0
1.0
1.0
135
135
135
135
135
702 nm
702 nm
702 nm
702 nm
702 nm
100
No response
No response
50
b
50
100
50
50
50
Tumour regrowth on day 7
a4±6 mm diameter RIF tumors (6 mice/group) were exposed to 75 MW/cm2 for 30 mm from a tunable dye laser tuned to the maximun red
absorption peak at 24 h post injection.
b
At a dose of 1.0 mmol/kg, there was 50% mortality after light exposure. Further studies at lower doses are currently in progress.

1
466
A. Rungta et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1463±1466
0
ethyl) group at position-8 (Fig. 2) and the corresponding
on day 30). Purpurinimide 2, containing a (1 -heptyloxy-
0
3
-(1 -heptyloxyethyl) analogue 1 (Fig. 3) had similar
tumor uptake, but produced a remarkable di€erence in
PDT ecacy (Table 1). These results indicate that high
tumor uptake of the photosensitizer is not the only cri-
teria for its ecacy. For a compound to be e€ective it is
likely that it binds or localizes into the more sensitive
site(s) of the tumor cells and these studies are under
investigation.
Figure 2. Open circles are for skin and closed circles are for tumor
uptake (mice transplanted with RIF tumor) of isomer 1, at a dose of
Acknowledgements
5
.0 mmol/kg, 24 h post injection.
The authors thank the National Institutes of Health
(
CA 55791) for the ®nancial support. Partial support by
the shared resources of the Roswell Park Cancer Centre
Support Grant (P30CA16056) is also acknowledged.
The help rendered by Dr. Shiv Dutta in analyzing new
samples by mass spectrometry analyses is also
acknowledged.
References and Notes
1
. Pandey, R. K.; Zheng, G. Porphyrins as Photosensitizers in
Photodynamic Therapy, The Porphyrin Handbook; Kadish, K.
M.; Smith, K. M.; Guilard, R., Eds.; Academic Press, San
Diego, 2000; Vol. 6.
2. (a) Pandey, R. K.; Sumlin, A. B.; Constantine, S.; Aoudia,
M.; Potter, W. R.; Beilnier, D. A.; Henderson, B. W.; Rod-
gers, M. A.; Smith, K. M.; Dougherty, T. J. Photochem. Pho-
tobiol. 1996, 64, 194. (b) Henderson, B. W.; Bellnier, D. A.;
Greco, W. R.; Sharma, A.; Pandey, R. K.; Vaughan, L. A.;
Weishaupt, K. R.; Dougherty, T. J. Cancer Res. 1997, 57,
Figure 3. Open circles are for skin and closed circles are for tumor
uptake (mice transplanted with RIF tumor) of isomer 2, at a dose of
5
.0 mmol/kg, 24 h post injection.
30). These results suggest that at this particular dose,
either the purpurinimide 4 is too toxic or is quite potent.
Further studies at lower drug doses are currently in pro-
gress. Compared to imide 1, photosensitizer 5, in which
the alkyl ether group at position-3 was replaced by an
ethyl substituent, and the N-hexyl chain is substituted
with a 10-unit saturated carbon chain to keep a similar
lipophilicity produced limited PDT ecacy.
4
3
000.
. Zheng, G. PhD thesis, Roswell Park Graduate Division
(
SUNY, Bu€alo), December 1999.
4. Zheng, G.; Potter, W. R.; Sumlin, A.; Dougherty, T. J.;
Pandey, K. K. Bioorg. Med. Chem. Lett. 2000, 10, 123.
5. (a) Smith, K. M.; Go€, D. A.; Simpson, D. J. J. Am. Chem.
Soc., 1985, 107, 4941. (b) Pandey, R. K.; Beilnier, D. A.; Smith,
K. M.; Dougherty, T. J. Photochem. Photobiol. 1991, 53, 65.
6
. Kozyrev, A. N.; Zheng, G.; Lazarou, E.; Dougherty, T. J.;
Smith, K. M.; Pandey, R. K. Tetrahedron Lett. 1997, 38,
3356.
. (a) Zheng, G.; Dougherty, T. J.; Pandey, R. K. J. Org.
The tumor versus muscle uptake of the purpurinimide
analogues and their in vivo shifts were determined by in
vivo re¯ectance spectroscopy. These compounds pro-
duced higher tumor uptake than the surrounding muscle,
and were retained in tumors for at least 24 h post injec-
tion. However, among all the analogues, purpurinimide
3
7
Chem. 1999, 64, 3751. (b) Zheng, G.; Kozyrev, A. N.;
Dougherty, T. J.; Smith, K. M.; Pandey, R. K. Chem. Lett.
1996, 739.
1 containing a secondary heptyloxyethyl group at posi-
tion-3 was found to be most e€ective (100% tumor cure
8. Pandey, R. K.; Potter, W. R.; Meunier, I.; Sumlin, A. B.;
Smith, K. M. Photochem. Photobiol. 1995, 62, 764.