Antitumor agents. Part 209: Pheophorbide-a derivatives as photo-Independent cytotoxic agents.

Article abstract of DOI:10.1016/S0968-0896(01)00305-4

A methanolic crude extract of the plant Garuga pinnata Roxb. (Burseraceae) showed promising cytotoxic activity against a panel of human tumor cell lines in vitro, including KB and its drug-resistant sublines (Ferguson et al. Cancer Res. 1988, 48, 5956). Pheophorbide-a and-b methyl esters (3,4) were isolated as active principles with broad photo-dependent cytotoxic activities in the micromolar range. These findings prompted SAR studies of known and novel pheophorbide-a derivatives as photo-dependent and photo-independent cytotoxic agents. The results showed that zinc-protoporphyrin IX (10), zinc 13(R)-hydroxypheophorbide-a methyl ester (22), and zinc chlorin-e6 trimethyl ester (13) possessed photo-independent cytotoxic activity. Compounds 13 and 22 were the most active cytotoxic agents of the series (mean ED(50) 4.6 +/- 1.0 microM and 5.7 +/- 0.7 microM, respectively) against KB cells incubated in the dark.

Full text of DOI:10.1016/S0968-0896(01)00305-4

Bioorganic & Medicinal Chemistry 10 (2002) 583–591
Antitumor Agents. Part 209: Pheophorbide-a Derivatives as
Photo-Independent Cytotoxic Agentsy
Prapai Wongsinkongman, Arnold Brossi, Hui-Kang Wang, Kenneth F. Bastow
and Kuo-Hsiung Lee*
Natural Products Laboratory, School of Pharmacy, University of North Carolina, Chapel Hill, NC 27599-7360, USA
Received 6 March 2001; accepted 28 August 2001
Abstract—A methanolic crude extract of the plant Garuga pinnata Roxb. (Burseraceae) showed promising cytotoxic activity against
a panel of human tumor cell lines in vitro, including KB and its drug-resistant sublines (Ferguson et al. Cancer Res. 1988, 48, 5956).
Pheophorbide-a and-b methyl esters (3,4) were isolated as active principles with broad photo-dependent cytotoxic activities in the
micromolar range. These findings prompted SAR studies of known and novel pheophorbide-a derivatives as photo-dependent and
photo-independent cytotoxic agents. The results showed that zinc-protoporphyrin IX (10), zinc 132(R)-hydroxypheophorbide-a
methyl ester (22), and zinc chlorin-e6 trimethyl ester (13) possessed photo-independent cytotoxic activity. Compounds 13 and 22
were the most active cytotoxic agents of the series (mean ED50 4.6ꢀ1.0 mM and 5.7ꢀ0.7 mM, respectively) against KB cells incu-
bated in the dark. # 2002 Elsevier Science Ltd. All rights reserved.
Introduction
ultimately traced to the need for short periods of photo-
irradiation (daily routine microscopic examination of
cultures). In order to extend these findings and establish
structure–activity relationships (SAR), a series of pheo-
phorbide-a derivatives was prepared and evaluated for
cytotoxic activities in vitro. When compared with com-
mercially available porphyrin-based compounds, such
as protoporphyrin IX (9) and chlorin-e6 (11), pheo-
phorbide-a derivatives showed greater photo-dependent
cytotoxic activities. Pheophorbide-based photo-
sensitizers represent good leads for the treatment of
cancer using photodynamic therapy, but their effective-
ness and utility is dependent on light penetration to the
tissue, which greatly restricts their clinical utility. How-
ever, in the present work, certain metal analogues of
pheophorbide-a and chlorin-e6 were found to exhibit
potent but essentially photo-independent cytotoxic
activity. Therefore, such compounds also represent
novel lead molecules with a broader potential for anti-
cancer drug development.
Pheophorbide-a derivatives, such as pheophytin-a and
pheophorbide-a methyl esters, inhibit replication of a
hepatoma tissue culture (HTC) cell line following light
irradiation.1 These pheophorbide-a derivatives are
structurally related to porphyrins and are of current
interest as photosensitizers for use in photodynamic
therapy (PDT), a modality developed to treat cancer
with a combination of light and photosensitizers.2 At
present, only photofrin II is approved by the USFDA to
treat esophageal and endobronchial non-small cell lung
cancers, as well as certain types of early-stage lung can-
cer.3,4 Photofrin II is a complex mixture of hematopor-
phyrin dimer, trimer, tetramer and pentamer, as well as
dehydration products with both ester and ether lin-
kages.2 Thus, potent, less complex photosensitizers
would be valuable for clinical evaluation and drug
development.
Our preliminary work led to the discovery of pheo-
phorbide-a and -b methyl esters from Garuga pinnata
(Burseraceae) as cytotoxic agents. However, the activity
was extremely variable, and reproducibility was
Results and Discussion
Natural products isolation and purification
yFor Part 208, see ref 5.
*Corresponding author. Tel.: +1-919-962-0066; fax: +1-919-966-
3893; e-mail: khlee@unc.edu
The structures of chlorophyll-a (1) extracted from spi-
nach, pheophorbide-a and -b methyl esters extracted
0968-0896/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(01)00305-4

584
P. Wongsinkongman et al. / Bioorg. Med. Chem. 10 (2002) 583–591
from Garuga pinnata Roxb. leaves (3,4), pheophorbide-a
derivatives purified from Clerodendrum calamitosum
Linn. leaves, and semi-synthetic derivatives (2, 3 and 5–8)
are shown in Scheme 1.5
abstracted from 3 at position 132 using DBU, followed
by hydroxylation with 10-camphorsulfonyloxaziridine
as described in Dolphin et al.7 (ꢁ)-(1R)-(10-Cam-
phorsulfonyl)oxaziridine gave 14 selectively (>99% ee),
while (ꢁ)-(1S)-(10-camphorsulfonyl)oxaziridine gave a
mixture of 8 and 14 (55% ee, data not shown).
Synthesis
Syntheses of chlorin-e6 trimethyl ester (12) and the zinc
derivative (13) of chlorin-e6 (11) are shown in Scheme 2.
The synthetic Schemes affording pheophorbide-a deri-
vatives are shown in scheme 3 (see Methods in Experi-
mental). Pheophorbide-a (2), protoporphyrin IX (9),
and chlorin-e6 (11) were purchased from porphyrin
products (Logan, UT, USA). Zinc protoporphyrin IX
(10) was purchased from Aldrich (Milwaukee, WI,
USA).
Because zinc protoporphyrin IX (10) possessed photo-
independent cytotoxic activity (Tables 1 and 2), metal
coupling pheophorbide derivatives were prepared.
Metal chelation of free base analogues was highly
dependent on both metal salt and solvent. Zinc(II)
chloride did not generate zinc-coupled target com-
pounds, but zinc(II) acetate provided adequate yields
(20–30%). A non-polar sovent was preferable to a polar
solvent, perhaps due to less ionic interaction between
metal salt and solvent. Methanol was not a suitable
solvent for this chelation, while refluxing in toluene-
based solvent provided the desired products. With the
same procedure, nickel pyropheophorbide-a methyl
ester (23) was prepared from pyropheophorbide-a
Novel alkyl ester analogues of pheophorbide-a (17, 19–
21) were obtained in 30–40% yield using a method
described by Borovkov et al.6 To prepare 132(R)-
hydroxy pheophorbide-a methyl ester, a hydrogen was
Scheme 1. Structures of chlorophyll-a (1), pheophorbide derivatives (2–8), protoporphyrin IX (9) and zinc protoporphyrin IX (10).
Scheme 2. Synthesis of zinc chlorin-e6 trimethyl ester (13): (a) CH2N2; (b) Zn(OAc)2, toluene, reflux, 2 h.

P. Wongsinkongman et al. / Bioorg. Med. Chem. 10 (2002) 583–591
585
methyl ester (15) using nickel (II) acetate.
neously using a dark condition for cell culture) are given
in Table 2. 5, 10, 15, 20-Tetrakis(N-methylpyridinium-4-
yl)porphyrin (TMPyP) (24, Scheme 4) was also eval-
uated using both treatment conditions and was found to
be less cytotoxic than pheophorbide and chlorin deri-
vatives (Tables 1 and 2). TMPyP interacts with DNA
and, upon irradiation, causes DNA damage leading to
cell death.8 Because TMPyP was reported to produce
singlet oxygen, the compound could directly induce
DNA damage through electron transfer from guanine
base to the bound photo-excited drug.9 To determine
whether pheophorbide derivatives act by causing DNA
damage, DNA breaks were measured using a gel lysis
assay.
In vitro evaluation
Nakatini et al. reported previously that pheophytin and
its derivatives possessed photo-dependent cytotoxic
activities.1 We observed variable cytotoxic activity in
our bioassay-guided fractionation and isolation study
using Garuga pinnata Roxb. and further discovered that
normal monitoring of the morphology of treated cells
using light microscopy was sufficient to photo-activate
the active principles, which were identified as pheopho-
bide-a and -b methyl esters (3 and 4, respectively).
Therefore, in order to determine the SAR and photo-
dependence of pheophorbide derivatives, it was impor-
tant to compare cytotoxic activities between those using
a standard photo-irradiation step and those determined
in the dark. The photo-dependent cytotoxic activities of
compounds 1–24 are shown in Table 1 and correspond-
ing photo-independent activities (determined simulta-
Gel lysis assay
The induction of double-stranded (ds) cellular DNA
breaks by selected compounds was determined using a
semi-quantitative gel lysis assay (Fig. 1). Interestingly,
Scheme 3. Synthesis of pheophorbide-a derivatives (3, 14–22): (a) CH2N2; (b) DBU, ꢁ25 ꢂC, followed by (ꢁ)-(1R)-(10-camphorsulfonyl)oxazirdine,
ꢁ25 ꢂC; (c) Zn(OAc)2 or Ni(OAc)2, toluene, reflux, 2 h; (d) ROH/Boc2/DMAP, reflux, 1 h; (e) 2,4,6-collidine, reflux, 1.5 h; (f) NaBH4, TFA, reflux, 3
h; (g) Ni(OAc)2, toluene, reflux, 2 h.

586
P. Wongsinkongman et al. / Bioorg. Med. Chem. 10 (2002) 583–591
among the compounds tested, 13 and 22 caused ds-
DNA breaks under the dark treatment conditions. In
contrast, 8, 11 and 16 did not cause DNA breaks with-
out a photo-irradiation step (Fig. 2). Overall results for
compounds from photo-dependent and photo-indepen-
dent groups are given in Table 3. On the basis of the
results, DNA is a possible target of cytotoxic pheo-
phorbide compounds. The mechanism of action of
photo-independent compounds is currently under
investigation and will be reported elsewhere.
Structure–activity relationships
The in vitro cytotoxicity assay results (Tables 1 and 2)
led to the following SAR conclusions: (a) at C-7,1 an
aldehyde group (4) decreased cytotoxic activity at least
10–14-fold, compared with a methyl group (3); (b) at
C-131, a carbonyl group was necessary for cytotoxic
activity against KB cells (cf., 15 and 16); (c) compounds
with a carboxylic acid substituent at position 174 (2, 5,
Table 1. Photo-dependent cytotoxic activitiesa of pheophorbide-a,
chlorin-e6 and porphyrin analogues against human tumor cell lines
panel (KB, KB-VCR, KB-7d, SK-MEL-2), compared with human
embryonic lung cell line (HEL)
Compound
ED50 (mM)
KB
KB-VCR
KB-7d
SK-MEL-2
HEL
1
2
3
4
5
6
7
8
24.4ꢀ0.9 23.7ꢀ2.2 17.5ꢀ0.5
0.8ꢀ0.1 0.8ꢀ0.1 0.4ꢀ0.0
0.6ꢀ0.1 4.1ꢀ0.1 0.3ꢀ0.1
38.4ꢀ0.1 21.4ꢀ3.3
2.9ꢀ0.7
1.8ꢀ0.0
28.2ꢀ1.3
3.6ꢀ0.2
>16
1.3ꢀ0.3
0.8ꢀ0.2
>8
Figure 1. A representative result showing double-stranded (ds) cellular
DNA breaks detected using gel lysis assay. (1) KB (dark); (2) VP16 (20
mM); (3) KB (dark); (4) KB (irr); (5) 2 (8 mM, dark); (6) 2 (4 mM, irr);
(7) 5 (3 mM, irr); (8) 6 (15 mM, irr); (9) 3 (3 mM, irr); (10) 7 (2 mM, irr);
(11) 8 (2 mM, dark); (12) 8 (1 mM, irr); (13) 14 (3 mM, irr).
11.3ꢀ0.3
>8
>8
0.6ꢀ0.0 0.6ꢀ0.1 0.5ꢀ0.1
3.0ꢀ0.1 3.3ꢀ0.2 1.5ꢀ0.1
0.4ꢀ0.1 1.8ꢀ0.2 0.4ꢀ0.1
0.2ꢀ0.0 1.6ꢀ0.2 0.3ꢀ0.02
0.9ꢀ0.2
4.0ꢀ0.5
0.6ꢀ0.1
0.5ꢀ0.1
4.4ꢀ0.3
NDa
0.8ꢀ0.2
NDa
0.7ꢀ0.0
0.6ꢀ0.2
24.8ꢀ3.5
9
>18
3.7ꢀ0.9 13.7ꢀ0.9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
VP16
14.7ꢀ1.4 12.8ꢀ0.9 12.6ꢀ0.9
16.4ꢀ1.3 16.8ꢀ1.3 15.4ꢀ0.8
0.5ꢀ0.0 3.7ꢀ0.2 0.6ꢀ0.1
2.4ꢀ0.1 3.3ꢀ0.2 2.1ꢀ0.1
0.5ꢀ0.1 2.1ꢀ0.2 0.4ꢀ0.0
0.8ꢀ0.1 10.9ꢀ1.7 0.8ꢀ0.1
3.4ꢀ0.4 9.1ꢀ0.1 7.9ꢀ0.2
14.0ꢀ1.2 12.0ꢀ0.5
33.5ꢀ0.8
1.7ꢀ0.3
1.7ꢀ0.1
1.0ꢀ0.2
2.3ꢀ0.5
3.3ꢀ0.5
0.3ꢀ0.1
1.6ꢀ0.2
>3
21.3ꢀ0.1 28.0ꢀ0.2
>29
>16
5.1ꢀ0.1
12.7ꢀ0.3
10.8ꢀ0.3
22.8ꢀ0.1 14.4ꢀ0.3
>16
>16
>15
>15
>14
29.8ꢀ0.5
6.7ꢀ0.1
>16
5.9ꢀ0.1
13.0ꢀ0.3
13.9ꢀ0.3
6.0ꢀ0.5
24.1ꢀ3.2 14.6ꢀ1.0
26.3ꢀ0.3
2.5ꢀ0.4
NDb
>14
1.7ꢀ0.1
NDb
1.8ꢀ0.0 3.1ꢀ0.1 1.9ꢀ0.1
87.7ꢀ3.3 87.7ꢀ2.9 54.6ꢀ3.5
>100
>100
>15
>100
NDb
>100
NDb
>100
NDb
4.0ꢀ0.0
aDetermined using a sulforhodamine B-staining assay of Rubinstein et
al.20 Cultured cells were observed at daily intervals using a Nikon ꢃꢁ1
inverted microscope (40ꢃmagnification) irradiated by microscopic
light for 2–3 min each day over 3 days before processing. Results are
mean concentration (mM) of compounds that inhibit 50% growth.
Standard deviation values were calculated from results of independent
triplicate assays. VP-16 was included as a control.
Figure 2. Double-stranded (ds) DNA breaks using gel lysis assay of
photo-dependent cytotoxic agents (8, 11, 16) and photo-independent
cytotoxic agents (13, 22). Open bars represent compounds evaluated
using the dark culture condition, and filled bars represent activities
under a matched photo-irradiated condition. Compounds were tested
as described in the experimental section and results are the mean
ꢀSEM from two independent experiments.
bND, not determined.
a
Table 2. Cytotoxic activities of pheophorbide, chlorin and porphyrin analogues in the absence of light against human tumor cell lines
Compound
ED50 (mM)
KB
KB-VCR
KB-7d
SK-MEL-2
HEL
1
21.6ꢀ1.0
17.9ꢀ0.8
4.6ꢀ1.0
5.7ꢀ0.7
69.5ꢀ3.3
41.1ꢀ1.6
26.1ꢀ1.1
19.6ꢀ1.4
5.1ꢀ1.3
5.8ꢀ0.7
69.5ꢀ2.5
34.5ꢀ1.3
18.7ꢀ1.2
15.6ꢀ1.9
5.0ꢀ0.7
5.1ꢀ0.1
56.3ꢀ2.1
>50
NDb
15.8ꢀ1.8
4.7ꢀ0.7
6.1ꢀ0.4
NDb
12.8ꢀ0.0
10.4ꢀ1.6
NDb
10
13
22
23
24
4.4ꢀ0.7
NDb
>50
NDb
aDetermined using a sulforhodamine B-staining assay of Rubinstein et al.12 Plates were covered by aluminum foil and incubated for 3 days and
cultures were observed using a Nikon ꢃꢁ1 inverted microscope after 3 days immediately before processing. Results are mean concentration (mM) of
compounds that inhibit 50% growth. Standard deviation values were calculated from results of independent triplicate assays. Compounds 2–9, 11,
12, and 14–21 were inactive at 40 mg/mL in the absence of light; therefore, they were excluded from this table.
bND, not determined.

P. Wongsinkongman et al. / Bioorg. Med. Chem. 10 (2002) 583–591
587
Table 3. Intracellular DNA damage a induced by pheophorbide, chlorin, and porphyrin derivatives
Compound
Intensity of double stranded-DNA breaks (fold over control)
Irradiated
Dark
2
3
4
5
6
7
8
96.8ꢀ14.0
256.0ꢀ147.0
85.3ꢀ47.7
NDb
NDb
4.1ꢀ1.1
90.1ꢀ88.7
NDb
94.4ꢀ5.20
NDb
687.0ꢀ201.0
259.0ꢀ168.0
341.0ꢀ68.0
72.1ꢀ16.0
NDb
12.6ꢀ6.3
10
11
12
13
14
15
16
19
21
22
VP-16
NDb
0
238.0ꢀ48.5
366.0ꢀ58.0
444.0ꢀ107.2
142.0ꢀ45.4
558.0ꢀ147.0
241.0ꢀ90.1
68.9ꢀ15.0
NDb
171
NDb
NDb
0
NDb
NDb
272.0ꢀ74.6
100.0ꢀ22.9
259.2ꢀ2.8
90.0ꢀ17.2
aIntracellular DNA breaks were studied using the gel-plug method as described in Bastow et al.22 Standard error values, where indicated, were
calculated from results of independent duplicate assays. The semi-quantitative values were the fold of increased intensity in DNA damage compared
to control. The numeric values were determined with densitometry of ethidium bromide-stained agarose gels using STORM and ImageQuant 5.0
software. Weakly cytotoxic compounds were excluded from this assay (ED50>10 mg/mL). VP-16, a topoisomerase II inhibitor that induced double-
stranded protein-linked DNA breaks, was included as a positive control in certain experiments.
bND, not determined.
nickel analogue of pyropheophorbide-a methyl ester
(23) was inactive both in the dark or in the presence of
light.
Conclusions
Generally, pheophorbide-a derivatives with longer alkyl
esters at C-174 were only weakly cytotoxic compared to
the methyl ester analogue (3); however, the butyl ester
analogue (19) was more active than ethyl (18), hexyl
(20), octyl (21), or benzyl (17) ester analogues under
photo-irradiation (Table 1). The free-base analogues (9,
12, 14, 15) were photo-dependent cytotoxic agents (no
activity under the dark culture condition at 40 mg/mL),
while the zinc-chelated analogues (10, 13, 22) displayed
improved photo-independent cytotoxic activities in the
micromolar range (Table 2). Interestingly, cytotoxic
activities of zinc chelates compared to their free-base
analogues (cf., 12 and 13; 14 and 22) decreased around
5–10-fold under photo-irradiation, but their activities
increased more than 40-fold in the dark. In contrast, the
nickel analogue (23) was inactive under either drug
treatment condition.
Scheme 4. Structure of 5, 10, 15, 20-tetrakis(N-methylpyridinium-4-
yl)porphyrin (24).
6) were not recognized as p-glycoprotein substrates; (d)
at C-174, a methyl ester group (3) was preferred to other
alkyl esters (17–21) for KB cytotoxicity; (e) replacing a
hydrogen at C-132 (3) with a hydroxy (8) or a methoxy
group (7) improved cytotoxic activity (2–3-fold) against
KB cells; (f) the stereochemistry of the hydroxy group at
C-132 of pheophorbide-a methyl ester was not an
important determinant of cytotoxic activity in resistant
cell lines (cf., 8 and 14); (g) the free bases (9, 12, 14, 15)
possessed photo-dependent cytotoxic activity; (h)
although metal chelated compounds (10, 13, 22, 23)
were 4–5-fold less cytotoxic than the corresponding free
bases (9, 12, 14, 15) in the presence of light, the action
of the metal analogues was largely light-independent;
(i) the metal chelated complexes (10, 23) were poorer
substrates of p-glycoprotein, except 13 and 22; (j) the
Theoretically, singlet oxygen is the major cytotoxic spe-
cies during photodynamic therapy, although evidence is
lacking for its production in vivo. The photodynamic
action of photosensitizers is initiated by photon
absorption, which induces singlet oxygen production
and subsequently causes DNA, RNA, or protein
damage via oxidative stress. Molecular oxygen plays a
key role in molecular damage, resulting in vascular col-
lapse, tissue degradation, and cell death.2 It has been
shown that certain metal groups may affect singlet oxy-
gen formation.3 Therefore, nickel may also prevent such

588
P. Wongsinkongman et al. / Bioorg. Med. Chem. 10 (2002) 583–591
production, while zinc may potentiate production,
causing oxidative stress without requiring photo-irra-
diation. Although pheophorbide-a can generate singlet
oxygen after photo-irradiation,10 the metal chelated
compounds may have other novel mechanisms of
action, including DNA interaction leading to cellular
ds-DNA breaks as reported herein. Additional
mechanistic studies are needed to evaluate and further
develop photo-independent pheophorbide-a derivatives
as anticancer drug candidates.
23H, 25H-phorbine-21-carboxylic acid methyl ester (3).
C36H38N4O5. Compound 3 was obtained semi-syntheti-
cally and was also isolated as a natural product. For
synthesis, a solution of pheophorbide-a [200 mg; Por-
phyrin Products (Logan, UT, USA)] in Et2O and
MeOH (1:1; 50 mL) was treated with diazomethane at
0 ꢂC with stirring for 15 min. The solvent was evapo-
rated and the residue was purified by silica gel column
chromatography, using CHCl3/hexane/EtOAc (1:3:2) as
the eluent. Recrystallization of 3 from CH2Cl2/MeOH
provided the desired compound as dark blue crystals
(190 mg; 95% yield). For phytochemical isolation, dried
leaves (4 kg) of Garuga pinnata Roxb. were treated as
above to give an 80% MeOH extract (330 g). This
extract was partitioned between H2O and CHCl3. The
CHCl3 extract was partitioned between hexane (10 g
extract) and 90% MeOH (114 g extract). From the lat-
ter extract (114 g), 3 was purified by silica gel column
chromatography, using CHCl3/hexane/EtOAc (1:3:2) as
the eluent, yielding 200 mg (0.005% yield from dried
leaves). Mp 238 ꢂC (lit.12 241 ꢂC, lit.13 228 ꢂC); 1H
NMR1,7,12 and UV–vis data14 were consistent with
literature values; FAB-MS: m/z 607 (M+H)+.
Experimental
General experimental procedures
All melting points were determined on a Fisher–Johns
1
melting point apparatus and are uncorrected. H NMR
spectra were obtained using a Varian 300 MHz NMR
spectrometer with TMS as the internal standard and
CDCl3 as solvent. All chemical shifts are reported in
ppm. FAB-MS spectral analyses were determined on a
VG Analytical VG-70E spectrometer, using chloroform
as a solvent. Analytical thin-layer chromatography
(TLC) was carried out on Merck precoated aluminum
silica gel sheets (Kieselgel 60 F-254). Preparative TLC
was carried out on 1 mm Analtech precoated silica gel
plates (Kieselgel 60-F254). The solvent systems used are
described under syntheses of each compound. The UV–
vis spectrum was obtained on a Beckman DU-600 UV–
vis spectrophotometer, using chloroform as a solvent.
For the cytotoxicity assay, a THERMOmaxTM micro-
plate reader (Molecular Devices Inc., Menlo Park, CA,
USA) was used to measure the absorbance. Compounds
were obtained as amorphous, green powders unless
indicated otherwise.
Pheophorbide-b methyl ester or 14-ethyl-13-formyl-3-(2-
methoxy-carbonyl-ethyl)-4, 8, 18-tetramethyl-20-oxo-9-
vinyl-23H, 25H-phorbine-21-carboxylic acid methyl ester
(4). C36H36N4O6. From a 90% methanolic extract of
Garuga pinnata Roxb. leaves, 4 was purified by silica gel
column chromatography, eluted with CHCl3/hexane/
EtOAc (1:3:2), yielding 66 mg of yellow powder
(0.002% yield from dried leaves). Mp 218 ꢂC (lit.15
270 ꢂC); 1H NMR data were consistent with those in the
literature;1,16 UV–vis (lmax, e) (CHCl3): 439 (10,665);
528 (1434), 600 (1108), 655 (2728); FAB-MS: m/z
621(M+H)+.
Chlorophyll-a (1). C55H72N4O5Mg. Chlorophyll-a pur-
chased from Aldrich was found to contain at least three
major constituents (TLC in CHCl3/hexane/EtOAc
1:3:1), therefore it was isolated from fresh spinach
leaves to ensure sufficient pure compound for biological
testing. Fresh leaves (1 kg) were dried at 60 ꢂC overnight
providing dried leaves (40 g) as starting material (4% w/w),
which were macerated in 80% MeOH for 3 days. The
extract was evaporated under vacuum providing 3.5 g
of dark green residue (8.8% yield, dried weight basis).
This crude extract was purified using repetitive silica
column chromatography (2.5ꢃ25 cm), eluting sequen-
tially with CHCl3/hexane (4:1), CHCl3/hexane/MeOH
(9:1:0.1), and CHCl3/MeOH (95:5), and MeOH by col-
lecting 30-mL fractions. Chlorophyll-a was obtained by
using a mixture of CHCl3/hexane (4:1) as the eluent (5
mg; 0.012% yield from dried leaves). 1H NMR data
were consistent with those in the literature;11 UV–vis
(lmax, e) (CHCl3): 412 (18,309); 536 (6368); 611 (2925);
668 (8418); FAB-MS: m/z 894 (M+ H)+.
132(S)-Methoxy pheophorbide-a (5); 132(S)-hydroxy
pheophorbide-a (6); 132(S)-methoxy pheophorbide-a
methyl ester (7); 132(S)-hydroxy pheophorbide-a methyl
ester (8). The method to isolate and modify these com-
pounds is described in Cheng et al.5
Protoporphyrin IX (9). Red powder purchased from
Porphyrin Products (Logan, UT, USA).
Zinc protoporphyrin IX (10). Red powder purchased
from Aldrich (Milwaukee, WI, USA).
Chlorin-e6 (11). Brownish-green powder purchased
from Porphyrin Products (Logan, UT, USA).
Chlorin-e6 trimethyl ester (12). C37H42N4O6. A solution
of 11 (50 mg) was methylated using diazomethane at 0 ꢂC
for 15 min, providing 48 mg dark green powder (96%
yield). Mp 205–207 ꢂC (lit.4,17 207–208 ꢂC); 1H NMR
data were consistent with literature values;4,7 UV–vis
(lmax, e) (CHCl3): 403 (91,207); 501 (8942); 531 (4322);
609 (4089); 665 (28,071); FAB-MS: m/z 639 (M+H)+.
Pheophorbide-a (2). Dark brown powder purchased
from Porphyrin Products (Logan, UT, USA).
Zinc chlorin-e6 trimethyl ester (13). C37H40N4O6Zn.
Compound 12 (20 mg) was refluxed with zinc(II) acetate
in toluene for 2 h, washed with water and dried over
Pheophorbide-a methyl ester or 14-ethyl-3-(2-methoxy-
carbonyl-ethyl)-4, 8, 13, 18-tetramethyl-20-oxo-9-vinyl-

P. Wongsinkongman et al. / Bioorg. Med. Chem. 10 (2002) 583–591
589
ꢂ
1
sodium sulfate, providing 6 mg dark green powder
(30% yield). Mp 130 ꢂC; 1H NMR (300 MHz, CDCl3): d
9.53 (1H, s), 9.48 (1H, s), 8.54 (1H, s), 8.05 (1H, dd,
J=18, 2 Hz), 6.18 (1H, dd, J=17, 2 Hz), 6.03 (1H,
dd,J=17, 2 Hz), 5.20 (2H, q), 4.35 (1H, m), 4.29 (3H, s),
4.20 (1H, m), 3.80 (3H, s), 3.70 (2H, q), 3.58 (3H, s),
3.44 (3H, s), 3.33 (3H, s), 3.28 (3H, s), 2.55 (2H, m), 2.25
(2H, m), 1.80 (3H, m), 1.70 (3H, m); UV–vis (lmax, e)
(CHCl3): 413 (186,320); 512 (7842); 596 (12,635); 639
(73,915); FAB-MS: m/z 702.4 (M+H)+.
using butyl alcohol (30% yield). Mp 110 C; H NMR
(300 MHz, CDCl3): d 9.47 (1H, s), 9.33 (1H, s), 8.49
(1H, s), 7.99 (1H, dd, J=18, 2 Hz), 6.30 (1H, dd,
J=11.7, 2 Hz), 6.27 (1H, s), 6.19 (1H, dd, J=11.7, 2
Hz), 4.50 (1H, dq), 4.21 (1H, m), 3.90 (3H, s), 3.80 (3H,
s), 3.62 (2H, q), 3.34 (3H, s), 3.18 (3H, s), 3.09 (2H, m),
2.58 (2H, m), 2.29 (2H, m), 1.79 (3H, t), 1.71 (3H, t), 1.50
(4H, m), 1.10 (3H, t), 0.81 (1H, s), 1.70 (1H, s); UV–vis
(lmax, e) (CHCl3): 413 (84,438); 507 (9787), 538 (8904),
610 (7695), 668 (38,210); FAB-MS: m/z 649 (M+H)+.
132(R)-Hydroxy pheophorbide-a methyl ester (14).
C36H38N4O6. Compound 14 was prepared by a litera-
ture procedure,7 providing 80 mg brownish-green pow-
Pheophorbide-a hexyl ester (20). C41H48N4O5. Com-
pound 20 was prepared as described above for 17, but
using hexyl alcohol (30% yield). Mp 175 ꢂC; H NMR
1
der (80% yield). Mp 233 ꢂC (lit.17 >300 ꢂC); H NMR
(300 MHz, CDCl3): d 9.53 (1H, s), 9.39 (1H, s), 8.56
(1H, s), 7.9 (1H, dd, J=18, 2 Hz), 6.30 (1H, dd, J=12,
1.3 Hz), 6.24 (1H, s), 6.19 (1H, dd, J=11.7, 1.3 Hz),
4.47 (1H, m), 4.21 (1H, m), 3.88 (3H, s), 3.82 (2H, q),
3.66 (3H, s), 3.40 (2H, m), 3.38 (3H, s), 3.24 (3H, s), 2.58
(2H, m), 2.40 (2H, m), 1.80 (3H, t), 1.71 (3H, d), 1.54
(4H, m), 1.34 (4H, m), 1.10 (3H, t), 0.81 (1H, s), 1.70
(1H, s); UV–vis (lmax, e) (CHCl3): 413 (84,438); 507
(9787), 538 (8904), 610 (7695), 668 (38,210); FAB-MS:
m/z 677.6 (M+H)+.
1
data were consistent with literature values;1 UV–vis
(lmax, e) (CHCl3): 415 (48,498); 506 (5363); 536
(4220); 611 (3668); 669 (23,066); FAB-MS: m/z 623
(M+H)+.
Pyropheophorbide-a methyl ester (15). C34H36N4O3.
Compound 15 was prepared by a literature procedure,7
providing 27 mg brown powder (90% yield). Mp 233 ꢂC
(lit.17 217–219 ꢂC); 1H NMR and UV–vis data were
consistent with literature values;17 FAB-MS: m/z 549
(M+H)+.
Pheophorbide-a octyl ester (21). C43H52N4O5. Com-
pound 21 was prepared as described above for 17, but
131-Deoxypyropheophorbide-a
methyl
ester
(16).
using octyl alcohol (30% yield). H NMR (300 MHz,
1
C34H38N4O2. Compound 16 was prepared by a litera-
ture procedure,17 providing 12 mg brown powder (60%
yield). Mp 170 ꢂC [lit.18 180–182 ꢂC]; 1H NMR data
CDCl3): d 9.51 (1H, s), 9.38 (1H, s), 8.55 (1H, s), 7.90
(1H, dd, J=18, 2 Hz), 6.25 (1H, s), 6.20 (1H, dd,
J=11.7, 1.3 Hz), 6.19 (1H, dd, J=11.7, 1.3 Hz), 4.50
(1H, m), 4.10 (1H, dq, J=7, 1.8 Hz), 3.90 (3H, s), 3.74 (3H,
s), 3.62 (2H, q), 3.45 (2H, m), 3.34 (3H, s), 3.18 (3H, s),
2.58 (2H, m), 2.29 (2H, m), 1.79 (3H, d), 1.71 (3H, t), 1.70
(12H, m), 0.90 (3H, m), 0.81 (1H, s), 1.70 (1H, s); UV–vis
(lmax, e) (CHCl3): 413 (84,438); 507 (9787), 538 (8904), 610
(7695), 668 (38,210); FAB-MS: m/z 705 (M+H)+.
were consistent with literature values;1,18 UV–vis (lmax
,
e) (CHCl3): 403 (81,423); 503 (8,821), 593 (3,158), 648
(18,184); FAB-MS: m/z 536 (M+H)+.
Pheophorbide-a benzyl ester (17). C42H42N4O5. Com-
pound 2 (10 mg) was stirred with di-tert-butyl dicarbo-
nate (1 equiv, 0.16 mmol) and
2
mg of
dimethylaminopyridine for 10 min. Then, benzyl alco-
hol (1 equiv, 0.16 mmol) was added and stirring con-
tinued for 1 h. The product was purified using silica gel
column chromatography, eluting with CHCl3/hexane/
EtOAc (1:3:2), providing 3 mg brown powder (30%
Zinc 132(R)-hydroxy pheophorbide-a methyl ester (22).
C36H36N4O6Zn. Compound 14 (20 mg) was refluxed
with zinc(II) acetate (1 equiv) in toluene for 2 h, washed
with water and the organic solvent was dried over
anhydrous sodium sulfate. Silica gel column chromato-
graphy was used for purification, eluting with CH2Cl2/
hexane/EtOAc ꢂ(2:2:1) to obtain 6 mg product (30%
yield). Mp 118 ꢂC; H NMR (300 MHz, CDCl3): d 9.54
1
(1H, s), 9.40 (1H, s), 8.58 (1H, s), 8.0 (1H, dd, J=18, 2
Hz), 7.30 (5H, m), 6.30 (1H, dd, J=17, 1.3 Hz), 6.27
(1H, s), 6.19 (1H, dd, J=11.7, 1.3 Hz), 5.30 (2H, s), 4.48
(1H, dq, J=1.8, 7, 7, 1.8 Hz), 4.25 (1H, m), 3.90 (3H, s),
3.70 (3H, s), 3.67 (2H, q), 3.40 (3H, s), 3.22 (3H, s), 2.58
(2H, m), 2.29 (2H, m), 1.79 (3H, d), 1.71 (3H, t), 0.81
(1H, br s), 1.70 (1H, s); UV–vis (lmax, e) (CHCl3): 413
(84,438); 507 (9787), 538 (8904), 610 (7695), 668
(38,210); FAB-MS: m/z 684 (M+H)+.
1
yield). Mp 114 C; H NMR (300 MHz, CDCl3): d 9.76
(1H, s), 9.55 (1H, s), 8.70 (1H, s), 8.01 (1H, dd, J=18, 2
Hz), 6.34 (1H, dd, J=12, 2 Hz), 6.18 (1H, dd, J=12, 2
Hz), 5.30 (1H, s), 4.45 (1H, m), 4.25 (1H, m), 3.89 (3H,
s), 3.75 (2H, q), 3.53 (3H, s), 3.44 (3H, s), 3.27 (3H, s), 3.18
(3H, s), 2.30 (4H, m), 1.79 (3H, t), 1.71 (3H, t); UV–vis
(lmax, e) (CHCl3): 416 (141,592); 521 (7644); 568 (13,948);
609 (21,221); 657 (71,358); FAB-MS: m/z 684 (M+H)+.
Pheophorbide-a ethyl ester (18). C37H40N4O5. Com-
pound 18 was prepared as described above for 17, but
Nickel
pyropheophorbide-a
methyl
ester
(23).
C34H34N4O3Ni. Compound 15 (30 mg) was refluxed
with nickel(II) acetate in toluene for 2 h. The product
was washed with water and dried over anhydrous
sodium sulfate. The desired product was purified using
silica gel column chromatography, eluting with CH2Cl2/
hexane/EtOAc (2:2:1), providing 9 mg (30% yield). Mp
using ethyl alcohol (30% yield). Mp 155 ꢂC; H NMR
1
data were consistent with literature values;19 UV–vis
(lmax, e) (CHCl3): 413 (84,438); 507 (9787), 538 (8904),
610 (7695), 668 (38,210); FAB-MS: m/z 622 (M+H)+.
ꢂ
1
Pheophorbide-a butyl ester (19). C39H44N4O5. Com-
pound 19 was prepared as described above for 17, but
171–172 C; H NMR (300 MHz, CDCl3): d 9.39 (1H,
s), 9.18 (1H, s), 8.29 (1H, s), 7.90 (1H, dd, J=18, 2 Hz),

590
P. Wongsinkongman et al. / Bioorg. Med. Chem. 10 (2002) 583–591
6.30 (1H, dd, J=12, 2 Hz), 6.10 (1H, dd, J=12, 2 Hz),
4.94 (2H, m), 4.48 (1H, dq, J=7 Hz, 1.8 Hz), 4.25 (1H,
m), 3.74 (2H, m), 3.70 (3H, s), 3.60 (3H, s), 3.29 (3H, s),
3.23 (3H, s), 2.58 (2H, m), 2.29 (2H, m), 1.79 (3H, d),
1.73 (3H, t); UV–vis (lmax, e) (CHCl3): 420 (59,552); 651
(53,959); FAB-MS: m/z 605 (M+H)+.
H4 system format]. After pouring a 55 ꢂC gel solution
containing 0.8% w/vagarose in TBE and allowing it to
gel, electrophoresis was carried out at 3.5 V/cm (mea-
sured between electrodes) at 4 ꢂC for 16 h without buffer
recirculation. The gel were stained with 1 mg/mL ethi-
dium bromide in water, treated overnight at 25 ꢂC with
RNAase A (1 mg/mL), and then photographed under
ultraviolet illumination. A representative gel is shown in
Figure 1. Semi-quantitative analysis of double-stranded
DNA breaks was done by densitometry using a
STORM phosphorimager and ImageQuant 5.0 software
(Molecular Dynamics Inc., Sunnyvale, CA, USA).
In vitro cytotoxicity assay. The in vitro cytotoxicity
assay followed the procedure described by Rubinstein et
al.20 Test compounds were prepared in DMSO and
stored at ꢁ80 ꢂC to prevent the degradation and the loss
of biological activity that was found to occur upon sto-
rage at ꢁ20 ꢂC. The final solvent concentration in cell
culture was less than 2% v/v of DMSO, a concentration
without effect on cell replication. The human tumor cell
line panel included human oral epidermoid carcinoma
(KB, ATCC # CCL17), multi-drug resistant (MDR) KB
sub-line (KB-VCR), multi-drug resistance associated
protein (MRP) KB sub-line (KB-7d), and malignant
melanoma (SK-MEL-2, ATCC # HTB68). Human
embryo lung fibroblast (HEL, ATCC # CCL137) was
also included as an example of a non-tumorgenic
human cell line. Drug-resistant cell lines were a gener-
ous gift of Dr. Y. C. Cheng (Yale University) and were
described in detail.21 Cells were incubated in RPMI-
1640 with 100 mg/mL kanamycin and 10% v/v fetal
bovine serum in a 5% CO2 humidified incubator at
37 ꢂC. Initial seeding densities varied between cell lines
in order to ensure a final absorbance reading in the
range 1–2.5 A562 units. Cells were plated and treated
with compounds for 24 h, photo-irradiated using a 25W
incandescent lamp at a distance of 30 cm for 5 min and
repeating exposure the next day. For treatment in dark,
plates were covered by aluminum foil during incuba-
tion. After incubation for 3 days in total, the ED50
value, the concentration that reduced the absorbance by
50%, was interpolated from dose–response graphs
plotted using PrismTM (Graph Pad Inc., San Diego, CA,
USA). Each test was performed in triplicate and
experimental variation was assessed using at least three
independent experiments.
Acknowledgements
The authors acknowledge N. Soonthorncharoennon
(Mahidol University, Bangkok, Thailand) for providing
Garuga pinnata Roxb., and H. H. Cheng (Taipei Medi-
cal University, Taipei, Taiwan) for compounds 2, 5–8.
We thank S.L. Morris-Natschke (University of North
Carolina at Chapel Hill, NC, USA) for her kind manu-
script review. This investigation was supported in part
by grant CA17625 from the National Cancer Institute
(K.-H.L.) and a faculty development grant from the
Pharmacy Foundation of NC (K.F.B.). This work was
performed by P.W. in partial fulfillment of the require-
ments for the degree of Doctor of Philosophy. P.W.
acknowledges a Royal Thai government scholarship
through her study.
References and Notes
1. Nakatini, Y.; Ourisson, G.; Beck, J. P. Chem. Pharm. Bull.
1981, 29, 2261.
2. Pass, H. I. J. Nat. Canc. Instit. 1993, 85, 443.
3. Lier, J. E.; Ali, H. Chem. Rev. 1999, 99, 2379.
4. Pandey, R. K.; Bellnier, D. A.; Smith, K. M.; Dougherty,
T. J. Photochem. Photobiol. 1991, 53, 65.
5. Cheng, H. H.; Wang, H. K.; Junko, I.; Bastow, K. F.;
Tachibana, Y.; Nakanishi, Y.; Luo, T. Y.; Lee, K. H. J. Nat.
Prod. 2001, 64, 915.
6. Borovkov, V. V.; Gribkov, A. A.; Kozyrev, A. N.; Brandis,
A. S.; Ishida, A.; Sakata, Y. Bull. Chem. Soc. (Japan) 1992,
65, 1533.
7. Ma, L.; Dolphin, D. J. Org. Chem. 1996, 61, 2501.
8. Villanueva, A.; Juarranz, A.; Diaz, V.; Gomez, J.; Canete,
M. Anticancer Drug Des. 1992, 7, 297.
9. Fiel, R. J. J. Biomol. Struct. Dyn. 1989, 6, 1259.
10. Kuwabara, M.; Yamamoto, T.; Inanami, O.; Sato, F.
Photochem. Photobiol. 1989, 49, 37.
11. Dai, R.; Shoemaker, R.; Farrens, D.; Han, M. J.; Kim,
C. S.; Song, P. J. Nat. Prod. 1992, 55, 1241.
12. Ma, L.; Dolphin, D. Tetrahedron: Asymmetry 1995, 6,
313.
13. Smith, K. M.; Goff, D. A.; Simpson, D. J. J. Am. Chem.
Soc. 1985, 107, 4946.
14. Ma, L.; Dolphin, D. J. Org. Chem. 1996, 61, 2501.
15. Scheer, H.; Katz, J. J. J. Amer. Chem. Soc. 1978, 100, 561.
16. Stoll, A.; Wiedemann, E. Helv. Chim. Acta 1933, 16,
183.
17. Pandey, R. K.; Sumlin, A. B.; Constantine, S.; Aoudia,
M.; Potter, W. R.; Bellnier, D. A.; Handerson, B. W.; Rod-
gers, M. A.; Smith, K. M.; Dougherty, T. J. Photochem. Pho-
tobiol. 1996, 64, 194.
Gel lysis assay
The procedure used KB cells and was based on the
method of Bastow et al.22 Cells were plated at 2ꢃ104/
cm2 overnight. After discarding the medium, test com-
pounds (concentration at 5-fold of ED50 values) were
added in fresh medium and incubation was continued
overnight. Cultures were exposed to visible light using a
25 W incandescent lamp at a distance of 30 cm for 5
min. After 3 h of incubation, cells were scraped into ice-
cold PBS and pelleted by centrifugation (5ꢃ100g, 5
min). Washed cell pellets were resuspended in 80 mL of a
37 ꢂC gel solution (1% w/vlow-melting point agarose
gel in PBS), cast in a mold and refrigerated for 5 min to
form 6ꢃ7ꢃ2 mm agarose plugs. Plugs of cells were
incubated in lysis-digestion buffer (0.4 M EDTA, 0.01 M
Tris–HCl, pH 8.0, 1% w/v N-lauryl-sarcosine and 100
mg/mL proteinase K) at 43 ꢂC overnight. The treated
plugs were equilibrated in TBE buffer (10 min, ambient
temperature) and placed against a gel comb in a hor-
izontal gel cast tray [BRL Inc. (Bethesda, MD, USA),

P. Wongsinkongman et al. / Bioorg. Med. Chem. 10 (2002) 583–591
591
18. Pandey, R.; Constantine, S.; Tsuchida, T.; Zheng, G.; Med-
forth, C. J.; Aoudia, M.; Kozyrev, A. N.; Rodgers, M. A. J.; Kato,
H.; Smith, K. M.; Dougherty, T. J. J. Med. Chem. 1997, 40, 2770.
19. Nakamura, Y.; Murakami, A.; Koshimizu, K.; Ohigashi,
H. Biosci. Biotech. Biochem. 1996, 60, 1028.
R. M.; Tosini, S.; Skehan, P.; Scudiero, P. A.; Monks, A.;
Boyd, M. R. J. Natl. Cancer Inst. 1990, 82, 1113.
21. Ferguson, P. J.; Fisher, M. H.; Stephenson, J.; Li, D.;
Zhou, B.; Cheng, Y. Canc. Res. 1988, 48, 5956.
22. Bastow, K. F.; Wang, H. K.; Cheng, Y.; Lee, K. H.
Bioorg. Med. Chem. 1997, 5, 1481.
20. Rubinstein, L. V.; Shoemaker, R. H.; Paull, K. D.; Simo,