Bioactivation of carbamate-based 20(S)-camptothecin prodrugs

Article abstract of DOI:10.1016/j.bmc.2004.01.039

Two new prodrugs of CPT were synthesized, based on carbamate linkages between the 20-hydroxy group of CPT and a linker designed to be enzymatically removed by either Penicillin-G-Amidase or catalytic antibody 38C2. Cell growth inhibition assays showed an

Full text of DOI:10.1016/j.bmc.2004.01.039

Bioorganic & Medicinal Chemistry 12 (2004) 1859–1866
Bioactivation of carbamate-based 20(S)-camptothecin prodrugs
Neta Pessah,a Mika Reznik,b Marina Shamis,a Ferda Yantiri,c Hong Xin,c
Katherine Bowdish,c Noam Shomron,b Gil Astb and Doron Shabata,*
aSchool of Chemistry, Faculty of Exact Sciences, Tel-Aviv University, Tel Aviv 69978, Israel
bDepartment of Human Genetics and Molecular Medicine, Tel Aviv University Medical School,
Tel-Aviv University, Tel Aviv 69978, Israel
cAlexion Antibody Technologies, Inc., 3985 Sorrento Valley Blvd., Suite A, San-Diego, CA 92121, USA
Received 2 September 2003; revised 18 December 2003; accepted 27 January 2004
Abstract—Two new prodrugs of CPT were synthesized, based on carbamate linkages between the 20-hydroxy group of CPT and a
linker designed to be enzymatically removed by either Penicillin-G-Amidase or catalytic antibody 38C2. Cell growth inhibition
assays showed an up-to-2250-fold difference in toxicity between the prodrugs and the active drug. Asignificant increase in toxicity
was observed upon incubation of the enzyme or the catalytic antibody with the corresponding prodrug The described derivatives of
CPT further our knowledge in the design of prodrugs for use in selective approaches for targeted chemotherapy.
# 2004 Elsevier Ltd. All rights reserved.
1. Introduction
The antitumor activity of 20(S)-camptothecin (CPT), a
pentacyclic plant alkaloid, was recognized 20 years ago.
It was first isolated from the Asian tree Camptotheca
acuminata by Wall et al. in 1966.4 CPT exerts its anti-
tumor activity mainly through inhibition of topoisome-
rase I.5 This enzyme, which is found in mammals, binds
preferentially to double-stranded DNA, cleaving one
strand and forming an enzyme-DNAcovalent bond
between a tyrosine residue and the 30 phosphate of the
cleaved DNA. Drug-induced accumulation of topoi-
somerase I-DNAcomplexes was identified as an essen-
tial step, ultimately leading to cell death by apoptosis.6
Chemotherapy remains the major systemic treatment of
malignant diseases. However, it is not very effective
against tumors, especially once they have metastasized,
mainly because of insufficient drug concentrations
inside the tumors, systemic toxicity, development of
resistance, and lack of selectivity for tumor cells over
normal cells. The clinical efficacy of chemotherapy can
be improved by selective delivery of the available drugs
to malignant cells, thus reducing toxicity, and permit-
ting much higher drug doses and more frequent treat-
ments.1,2 One approach to overcome these drawbacks is
the development of relatively non-toxic anticancer
agents, in a prodrug form, specifically activated in or
near the tumor tissue. The ideal prodrug should be
stable in vivo, far less toxic than its parent form, and
activated specifically in or within the microenvironment
of the tumor cells. However, at this point, a major
problem with the prodrug strategy in vivo is the enzy-
matic moiety: High background reactions by endogen-
ous enzymes result in loss of selectivity.3
The most common problem associated with CPT is the
stability of the 20-hydroxy lactone, which is easily
hydrolyzed at neutral pH to provide the inactive car-
boxylate. It was shown that both the lactone ring stabi-
lity and the 20-hydroxy group of CPT are critical for its
antitumor activity. It was found that masking the 20-
hydroxy group of CPT, results in reduced activity of the
drug and increased stability of the lactone ring.7 The 20-
hydroxyl generates an intramolecular hydrogen bond
with the carbonyl moiety of the lactone, which accel-
erates the hydrolysis of the otherwise stable lactone
ring. Therefore, masking this hydroxyl by a chemical
linker that can be selectively removed, is a convenient
approach for generation of a prodrug from CPT.8
The masking linker serves two purposes: (1) removal of
the intramolecular hydrogen bond, resulting with a
Keywords: Prodrug Activation; Selective Chemotherapy; Enzymes;
ADEPT.
* Corresponding author. Tel.: +972-3640-8340; fax: +972-3640-9293;
e-mail: chdoron@post.tau.ac.il
0968-0896/$ - see front matter # 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2004.01.039

1860
N. Pessah et al. / Bioorg. Med. Chem. 12 (2004) 1859–1866
stable lactone and (2) increasing solubility of CPT,
which is almost insoluble in aqueous and organic media
without the linker.
1B). Both prodrugs were purified by standard flash
chromatography on silica gel and characterized by
NMR, MS and HPLC chromatograms. The carbamate
prodrugs were obtained as a yellowish powder and
found to be soluble in most of the organic solvents. The
prodrugs were used for further studies by preparing
stock solutions in DMSO.
The low chemical reactivity of the 20-hydroxy group
results from a strong steric hindrance, due to its func-
tionality as a tertiary alcohol. Esterification of this
hydroxy group under standard conditions is the easiest
method for masking it. Therefore, most of the prepared
prodrugs of CPT were ester derivatives.9,10 The major
disadvantage of an ester prodrug, which needs to be
activated specifically, is its poor stability to hydrolysis in
physiological pH (exceptional example is of the CPT-
PEG ester, which was reported by Greenwald10). Fur-
thermore, endogenous enzymes with esterase activity
can also hydrolyze such an ester prodrug. The relative
instability of the ester CPT prodrug results in a small
difference of the toxicity levels between the prodrug and
the free drug; for example, CPT prodrugs with ester
linkage are only 10–20-fold less toxic than their parent
drug.11 Here we report on the synthesis and character-
ization of new CPT prodrugs that are masked through a
stable carbamate linkage and are more than 1000-fold
less toxic than the parental CPT in cell growth inhibi-
tion assays. This CPT prodrugs could be activated
by Escherichia coli penicillin-G amidase8,12 (PGA) or by
catalytic antibody 38C2,11,13,14 as was shown by in vitro
studies on three different cancer cell lines.
2.2. Prodrugs activation
The masking linker of prodrug 7 was designed for
removal by PGA. This enzyme catalyzes the amide
bond cleavage of phenylaceteamide 7 to generate inter-
mediate 8 (Fig. 2A), which goes through a 1,6-quinone-
methide like rearrangement, followed by decarboxylation
to generate intermediate 9. The latter is self-cyclized
spontaneously to release the free drug. Activation of the
prodrug by PGAwas performed by incubating the pro-
drug with the enzyme at 37 ꢀC in cell-culture-medium.
The conversion of the CPT prodrug to its parental drug
was monitored by HPLC as presented in Figure 3A. It is
clearly shown that upon incubation of prodrug 7 with
PGA, intermediates 8 and 9 immediately start to appear
along with small amounts of CPT. After a while, the
intermediates are decreased and converted completely
to CPT.
Similarly, the masking linker of prodrug 7a was
designed to be removed by catalytic antibody 38C2.
This aldolase antibody catalyzes a retro-aldol retro-
Michael cleavage reaction, followed by spontaneous
decarboxylation to generate intermediate 9. The latter
goes through a self-cyclization reaction to release the
free drug (Fig. 2B). The conversion of the CPT prodrug
7a to the parental drug was monitored by HPLC, and
the results are presented in figure 3B. It is shown that
upon incubation of antibody 38C2 with the prodrug
intermediate 9 begins to appear, along with small
amounts of CPT. After a while, both the prodrug and
the intermediate are decreased and converted to CPT.
When prodrug 7a is incubated in PBS 7.4 at 37 ꢀC only,
no background reaction is observed for at least a week.
2. Results
2.1. Synthesis
CPT prodrug 7 was synthesized by a coupling reaction
between compunds 3 and the mono-Boc-N,N0-ethylene-
diamine derivative of camptothecin 6 after removal of
the Boc protecting group (Fig. 1A). CPT prodrug 7a
was synthesized by direct reaction between linker 3a
(prepared in situ from compound 2a and trifluoroacetic
acid) and the p-nitrophenyl-carbonate derivative of
camptothecin 5 in the presence of triethylamine (Fig.
Table 1. Concentration of CPT, proCPT 7, and proCPT 7+PGA, which inhibited cell proliferation of cancer cells by 50% after 72 h of incubation
time
Cell line
Tissue
IC 50 (mM)
Drug
Prodrug
Prodrug+Enzyme PGA
293T
HeLa
HepG2
Embryonic kidney
Cervix carcinoma
Hepatocellular carcinoma
0.04
0.05
0.06
56.23
18.62
35.48
0.50
5.01
0.49
Table 2. Concentration of CPT, proCPT 7a, and proCPT 7a+antibody 38C2, which inhibited cell proliferation by 50% after 72 h of incubation
time
Cell line
Tissue
IC 50 (mM)
Drug
Prodrug
Prodrug +2.4 mM 38C2
Prodrug +24 mM 38C2
Du145
A375
MCF-7
BxPc-3
Prostate cancer
Malignant melanoma
Breast adenocarcinoma
Pancreatic adenocarcinoma
0.012
0.003
0.082
0.02
11.55
3.32
47.5
45
0.311
0.121
1.75
1.5
0.061
0.05
1.1
ND

N. Pessah et al. / Bioorg. Med. Chem. 12 (2004) 1859–1866
1861
2.3. Cell growth inhibition
Another four human tumor cell lines were evaluated for
prodrug 7a: breast adenocarcinoma cell line MCF-7,
prostate cancer cell line Du 145, malignant melanoma
cell line A375, and primary pancreatic adenocarcinoma
cell line BxPc-3. Comparison of IC50 values demon-
strated that pro-CPT 7a was 500–2250-fold less toxic
than CPT (Table 2). To evaluate the efficacy of antibody
38C2-mediated pro-CPT activation, the activity of pro-
CPT 7a was monitored in the absence and presence of
catalytic antibody 38C2 (at two different concentra-
tions) with cultured human Du 145 prostate cancer cell
line (Fig. 4B). In this cell line, pro-CPT 7a demonstrated
around 1000-fold less toxicity than CPT. In the presence
of catalytic antibody 38C2, pro-CPT 7a was only 25-
fold and 5-fold less toxic than CPT, respectively (Table
The cytotoxicities of CPT and pro-CPTs were mon-
itored by the number of living cells in the presence of
drug and prodrug at a range of concentrations. Three
different human tumor cell lines were evaluated for
prodrug 7: embryonic kidney 293T, cervix carcinoma
HeLa, and hepatocellular carcinoma HepG2. Compar-
ison of IC50 values demonstrated that pro-CPT 7 was
372–1405-fold less toxic than CPT (Table 1). As
was shown previously, PGAcatalyzed the cleavage of the
relatively non-toxic pro-CPT 7 into active CPT in vitro
in three of these cell-based assays. Figure 4A shows
experimental plots measured for embryonal kidney
293T cell line.
Figure 1. (A) Synthesis of a campthothecin prodrug, designed for activation by penicillin-G amidase. (B) Synthesis of a campthothecin prodrug,
which can be activated by catalytic antibody 38C2.

1862
N. Pessah et al. / Bioorg. Med. Chem. 12 (2004) 1859–1866
Figure 2. (A) General description of the prodrug activation strategy by PGA catalyzed phenylaceteamide cleavage, followed by 1,6-quinone methide
rearrangement, spontaneous cyclization and release of the free drug. (B) Prodrug activation strategy by antibody 38C2 catalyzed the retro-aldol-
retro-Michael reaction sequence, followed by spontaneous decarboxylation, cyclization and release of the free drug.
Figure 3. (A) Monitoring of CPT release over time, upon incubation of prodrug 7 (500 mM) with penicillin-G amidase (5 mM) in PBS 7.4 at 37 ꢀC.
(B) Monitoring of CPT release over time, upon incubation of prodrug 7a (500 mM) with catalytic antibody 38C2 (50 mM) in PBS 7.4 at 37 ꢀC.

N. Pessah et al. / Bioorg. Med. Chem. 12 (2004) 1859–1866
1863
2). This shows clearly that catalytic antibody 38C2
cleaves pro-CPT into active CPT in vitro in cell culture.
incubation at 37 ꢀC in PBS-7.4, the prodrugs remain
unchanged with no background decomposition (data
not shown). We further evaluated the stability of pro-
drug 7 by incubating it with the extracted cells-super-
natant and with 100% mouse serum. In both cases, the
prodrug remained unchanged for 72 hr, meaning that
there is no natural enzyme in the cells and in the serum
that is capable of cleaving the masking linker in a man-
ner similar to PGA.
We also evaluated the minimum concentration of PGA
required to activate prodrug 7. Embryonic kidney 293T
cells were incubated with constant concentration of
prodrug 7 and different concentrations of PGA( Fig. 5).
Activation was clearly observed with an enzyme con-
centration as low as 0.001 mg/mL.
The CPT prodrugs were found to have increased solu-
bility in aqueous media with values of 2-fold for pro-
drug 7 and 3-fold for prodrug 7a (measurements were
performed by preparing saturated solutions of the com-
pounds in water and reading their concentration by
HPLC with a UV detector). As was indicated above,
by masking the 20-hydroxy group, it is possible to
3. Discussion
Prodrugs 7 and 7a are new reported carbamate deriva-
tives between a masking enzymatic substrate and the
20-hydroxy group of CPT. This carbamate linkage is
relatively stable at physiological pH. After 7 days of
Figure 4. (A) Cytotoxicity of prodrug 7 with PGAon human embryonal kidney 293T cell line. The cells were incubated with the drug and prodrug,
with or without 0.005 mg/mL PGA. The viability of the cells was monitored 72 hr after drug/prodrug addition by XTT cell proliferation assay and
detected by the absorbance at 470 nm, (&) CPT, (*) pro-CPT, (~) pro-CPT+ 1 mM PGA. (B) Cytotoxicity of drugs to Du 145 human prostate
cancer cells. Du 145 cells were exposed to drug and pro-drug in the absence and presence of catalytic antibody 38C2. (&) CPT, (*) pro-CPT, (~)
pro-CPT+ 2.4 mM 38C2, (!) pro-CPT+ 24 mM 38C2.

1864
N. Pessah et al. / Bioorg. Med. Chem. 12 (2004) 1859–1866
In conclusion, we have designed and synthesized two
new camptothecin prodrugs that are suitable for activa-
tion with the enzyme PGAand catalytic antibody 38C2.
The prodrugs are based on a carbamate linkage between
the 20-hydroxy group of CPT and an amine masking
linker. These are the first reported carbamate prodrugs
of 20(S)-CPT, and they exhibit a more than 1000-fold
difference between the drug and the prodrug activities.
The masking molecule is constructed of two carbamate-
functional groups, which make the prodrug more
hydrophilic and, therefore, more soluble in water. Fur-
ther studies toward faster cyclization linkers may lead to
more practical carbamate CPT prodrugs.
Figure 5. The effect of different concentrations of PGAon the cyto-
toxicity level of prodrug 7. Human embryonal kidney 293T cells were
incubated with 0.5 mM prodrug (IC50 values were calculated as shown
in Fig. 4A) and the indicated concentration of PGA. Cell density was
measured as in Figure 4A.
4. Experimental
4.1. General
prevent the internal hydrogen bond of the hydroxyl
with the neighbored oxygen of the carbonyl, thus leav-
ing the hydroxyl available to a generate hydrogen bond
with water molecules.
All reactions requiring anhydrous conditions were per-
formed in oven-dried glassware under an Ar or N2
atmosphere. Chemicals and solvents were either puriss
p.A. or purified by standard techniques. THF was dis-
tilled from sodium-benzophenone. Thin layer chroma-
Although there is a slow activation rate of the prodrugs,
cytotoxicity studies performed with several cancer cell
lines clearly demonstrated the significantly reduced
toxicity of the pro-CPTs, when compared with their
parent drug. The best results for prodrug 7 were
obtained with human embryonic kidney cell line 293T,
in which the pro-CPT was 1405-folds less toxic than
CPT (Fig. 4A). When the studies were performed in the
presence of PGA, the toxicity of the pro-CPT increased
100-fold, compared with a control experiment in the
absence of PGA. The 10-fold cytotoxicity difference
between the prodrug+PGAand the drug alone is
derived from the slow cyclization of intermediate 9 to
free CPT (Fig. 2). Therefore, by increasing the rate of
cyclization, it is possible to design an improved pro-
drug. Addition of substitutions on the N,N0-dimethyl-
ethylenediamine spacer should speed up the desired
cyclization, by constraining the conformation of the
five-member ring. Recent studies in our lab have con-
firmed this assumption. The best results for prodrug 7a
were obtained for primary pancreatic adenocarcinoma
cell line BxPc-3, in which the pro-CPT was 2250 fold
less toxic than CPT. When the studies were performed
in the presence of catalytic antibody 38C2, the toxicity
of the pro-CPT increased 27–189-fold.
tography (TLC): silicagel plates Merck 60 F254
;
compounds were visualized by irradiation with UV light
and/or by treatment with a solution of 25 g phospho-
.
molybdic acid, 10 g Ce(SO4)2 H2O, 60 mL concd.
H2SO4 and 940 mL H2O, followed by heating and/or by
staining with a solution of 12 g 2,4-dinitrophenylhy-
drazine in 60 mL concd. H2SO4; 80 mL H2O and 200
mL 95% EtOH, followed by heating and/or immersing
into an iodine bath (30 g I2, 2 g Ki, in 400 mL EtOH/H2O
1:1) and warming. Flash chromatography (FC): silica gel
Merck 60 (particle size 0.040–0.063 mm), eluent given in
parentheses. 1H NMR: Bruker AMX 200. The chemical
shifts are expressed in d relative to TMS (d=0 ppm) and
the coupling constants J in Hz. The spectra were recor-
ded in CDCl3 as solvent at room temperature unless
stated otherwise. HR-MS: liquid secondary ionization
(LSI-MS): VG ZAB-ZSE with 3-nitrobenzyl-alcohol
matrix. All general reagents, including salts and sol-
vents, were purchased from Aldrich (Milwaukee, MN).
4.2. Synthesis of camptothecin prodrug 7
Linker 3 was prepared by the following method: Phe-
nylacetic acid was dissolved in thionyl chloride and the
reaction mixture was refluxed for 1 h. The excess of
thionyl chloride was removed under reduced pressure,
yielding phenylacetyl chloride. The product (3.14 g, 0.02
mol) was then dissolved in dry THF and added drop-
wise to a solution of 4-aminobenzylalcohol (1 g, 8.12
mmol) and DMAP (2.47 g, 0.02 mole) at 0ꢀc. The mixture
was stirred overnight and then washed with dichloro-
methane and NaOH 1M. The organic layer was dried over
magnesium sulfate. The solvent was removed under
reduced pressure, and the product was purified by column
chromatography on silica gel (ethylacetate/hexanes
40:60), yielding compound 1 (2.17 g, 74.5%). Compound
1 (2 g, 5.6mmol) was heated to 80ꢀc with ethylene-dia-
mine for 1.5 h and then washed with HCl 1M and Ethyl
acetate to give pure compound 2 (1.207 g, 89%).
Recently a new approach for selective chemotherapy
has been developed (Polymer Directed Enzyme Prodrug
Therapy15ꢁ17 (PDEPT)). This method is a two-step
antitumor approach in which both the prodrug and the
enzyme are targeted to the tumor site with a polymer
molecule. In the first step, a polymer-prodrug conjugate
is administered and trapped in tumor tissues through
the EPR (enhanced permeability and retention) effect. A
conjugate of an HPMA-copolymer and catalytic anti-
body 38C2 has already been reported.18 In this chemo-
therapeutic approach, the relatively slow release of the
free drug may well be advantageous in vivo. It will pro-
long the duration of the drug delivery, already targeted
to the tumor tissue by the carrier HPMA-copolymer
molecule.

N. Pessah et al. / Bioorg. Med. Chem. 12 (2004) 1859–1866
1865
1H NMR (200MHz,MeOD):d=7.52 (2H, d, J=8.5); 7.3
(2H, d, J=8.5); 4.37 (2H, s); 3.49 (2H, s).
stirred for 15 min, and the DMF was removed under
reduced pressure. The remaining residue was purified by
thin-layer chromatography on silica gel (ethylacetate/
methanol 98:2) yielding prodrug 7 (45.1 mg, 68%).
The p-nitrophenyl-carbonate of compound 2 was pre-
pared by the following procedure. Compound 2 (1.2g,
4.97 mmol), triethylamine (0.706 mL, 9.95 mmol) and
DMAP (cat. amount) were dissolved in dry THF. p-
Nitrophenyl-chloroformate (2 g, 9.95 mmol) was added,
and the mixture was stirred for 1.5 h under Argon. The
mixture was washed with ethylacetate and 1M HCl and
then with saturated NaHCO3. The organic layer was
dried. The solvent was removed under reduced pressure,
and the product was purified by column chromato-
graphy on silica gel (ethylacetate/hexanes 30:70), yield-
ing compound 3 (1.18g, 82%).
M.S. FAB MNa+ expected: 752.28, found: 752.2
4.4. Synthesis of Camptothecin prodrug 7a
Linker 2a was synthesized as published previously (7).
Linker 3a was prepared in situ by the following proce-
dure. Trifluoroacetic acid (2 mL) was added to com-
pound 2a (360 mg, 1.0 mmol). After 2 min of stirring,
the excess of TFAwas removed under reduced pressure,
resulting of linker 3a. The crude product was used
directly for the next reaction. Thus, compound 5 (754
mg, 1.0 mmol) was added to a stirred solution of linker
3a in 5 mL of DMF, followed by the addition of 1 mL
triethylamine. The mixture was stirred for 1 h, and the
DMF was removed under reduced pressure. The
remaining residue was purified by column chromato-
graphy on silica gel (ethyl acetate/methanol 95:5),
yielding prodrug 7a (711 mg, 81%).
1H NMR (200MHz,CDCl3): d=8.26 (2H, d); 7.36
(11H, m); 5.23 (2H, s); 3.76 (2H, s);
Compound 5. Camptothecin (1.5 g, 4.3 mmol) and
PNP-chloroformate (2.6 g, 12.9 mmol) were dissolved in
methylene-chloride (45 mL) at 0ꢀ C, followed by the
addition of DMAP (3.18 g, 25.8 mmol). The resulting
clear solution was stirred at room temperature for 1 h.
The reaction was monitored by TLC (EtOAc 100%).
After completion, the mixture was diluted with 75 mL
of methylene-chloride and washed with 30 mL HCl
0.1 N. The organic layer was dried over magnesium sul-
fate, concentrated under reduced pressure to 10 mL,
and precipitated with ether. The precipitated solid was
filtered and dried to give crude compound 5 in the form
of yellow powder (2.5 g).
M.S. MALDI-FTMS MNa+ expected: 657.2531,
found: 657.2505.
4.5. HPLC assay for pro-CPT activation
(Prodrug 7) Astock solution of 3.5 mg/mL Penicillin-G-
o
Amidase in DMEM cell medium stored at 4 C was
used. All enzyme reactions were performed in DMEM
cell medium at 37o C in microfuge tubes. The reactions
were performed at a concentration of 250 mM of sub-
strate and 43 mM enzyme. Enzyme-catalyzed reactions
were monitored at 360 nm for CPT by RP-HPLC using
a Gradient of acetonitrile:water at 1 mL/min.
1H NMR (200MHz,CDCl3): d=8.39 (1H, s); 8.25 (1H,
d, J=8); 7.94 (1H, d, J=8); 7.85–7.81 (1H, m); 7.71–
7.67 (1H, m); 7.48 (2H, d, J=7); 7.14 (2H, d, J=7); 5.76
(1H, d, J=16); 5.35-5.2-32 (3H, m); 1.90–2.86 (2H, m);
1.11–1.03 (3H, m).
(Prodrug 7a), Astock solution of 5 mg/mL antibody
38C2 in PBS (pH 7.4) stored at 4ꢀ C was used. All
antibody reactions were performed in PBS (pH 7.4) at
37ꢀ C in microfuge tubes, typically at concentrations of
20–200 mM of substrate and 5 mM antibody. Antibody-
catalyzed reactions were monitored at 360 nm for CPT
by RP-HPLC using a ratio of 35:65 of acetonitrile:water
at 1 mL/min.
Compound 6 was prepared by the following procedure. p-
nitrophenyl camptothecin carbonate 5 (710 mg, 1.4 mmol)
was dissolved in DMF, and mono-Boc-N,N0-dimethyl
ethylene-diamine (400 mg, 2.13 mmol) was added. The
mixture was stirred for 30 min, and the DMF was removed
under reduced pressure. The remaining residue was
purified by column chromatography on silica gel (ethyl
acetate, 100%), yielding Compound 6 (256 mg, 32%).
4.5.1. Cell lines. Human embryonic kidney (293T),
human cervix carcinoma (HeLa) cells and human hepa-
tocellular carcinoma (HepG2) were grown in DMEM
supplemented with 10% foetal calf serum and were
purchased from Biological industries, Bet- Haemek.
Breast adenocarcinoma (MCF-7), prostate cancer (Du
145), malignant melanoma (A375), and primary pancrea-
tic adenocarcinoma (BxPc-3) cell lines were obtained from
American Type Culture Collection (ATCC). EMEM
(2mM L-glutamine, 1 mM sodium pyruvate, 0.1 mM non-
essential aminoacid, 1.5 g/l sodium bicarbonate, and 10%
fetal bovine serum) medium was used for these cell lines.
1H NMR (200 MHz,CDCl3): d=8.40 (1H, s); 8.20 (1H,
d, J=8); 7.97–7.67 (3H, m); 5.65 (1H, d, J=10); 5.29
(4H, m) 3.8–2.83 (10H, m); 2.02–1.0 (14H, m).
MS(FAB):C30H34N4O7 [M+Na]+ 585.2
4.3. Camptothecin prodrug 7
Trifluoroacetic acid (2 mL) was added to compound 6
(50 mg, 0.089 mmol). After 2 min of stirring, the excess
of TFAwas removed under reduced pressure. The crude
product was used directly for the next reaction and was
dissolved in 3 mL of DMF. Compound 3 (37.43 mg,
0.089 mmol) was added to the stirred solution, followed
by the addition of 0.5 triethylamine. The mixture was
4.5.2. In vitro growth inhibition assays. The activity of
pro-CPTs against carcinoma cell lines was analyzed in
the present and absence of PGAand catalytic antibody

1866
N. Pessah et al. / Bioorg. Med. Chem. 12 (2004) 1859–1866
38C2. Cells were plated in 96-well plates at 2ꢂ103 cells
per well. CPT and pro-CPT were added 24 h later at
various concentrations in the presence and absence of
the enzymes. CPT was dissolved in DMSO and pro-
CPT in acetonitrile, diluted in 200 mL of EMEM or
DMEM. Quadruped wells were prepared for each drug
concentration and for the controls. Plates were incubated
at 37 C 5% CO2 for 72 h. After 72 h, the medium was
removed, and 200 mL of fresh medium was added to the
cells. The number of living cells was determined by Cell-
Titer 961 AQueous Non-Radioactive Cell Proliferation
Assay Kit (Promega). Briefly, 2.0 mL of novel tetra-
zolium compound (3-(4,5-dimethylthiazol-2-yl)-5-(3-car-
boxymethoxyphenyl)-2-(4-sulfophenyl)-2H-terrazolium,
inner salt MTS, and 100 mL of an electron coupling
reagent (phenazine methosulfate, PMS) were mixed. 20
mL of MTS/PMS solution was added to cells, which
then were incubated at 37 C 5% CO2 for 3 h. The con-
version of MTS into aqueous soluble formazan is accom-
plished by dehydrogenase enzymes found in metabolically
active cells. The quantity of formazan product was mea-
sured by the intensity of 490 nm absorbance on ELISA
plate reader. IC50 values were calculated from interpola-
tion of logaritmic dose-A490 curves. Cell proliferation
assay was repeated twice for most of the cell lines, and
the mean value and standard deviation were calculated.
2. Schroeder, U.; Bernt, K. M.; Lange, B.; Wenkel, J.; Jikai,
J.; Shabat, D.; Amir, R.; Huebener, N.; Niethammer,
A. G.; Hagemeier, C.; Wiebusch, L.; Gaedicke, G.;
Wrasidlo, W.; Reisfeld, R. A.; Lode, H. N. Blood 2003,
102, 246.
3. Ast, G. Curr. Pharm. Des. 2003, 9, 455.
4. Wall, M. E.; Wani, M. C.; Cook, C. E.; Palmar, K. H.;
McPhail, A. T.; Sim, G. A. J. Am. Chem. Soc. 1966, 88,
3888.
5. Hertzberg, R. P.; Caranfa, M. J.; Hecht, S. M. Biochem-
istry 1989, 28, 4629.
6. Hsiang, Y. H.; Liu, L. F.; Wall, M. E.; Wani, M. C.;
Nicholas, A. W.; Manikumar, G.; Kirschenbaum, S.;
Silber, R.; Potmesil, M. Cancer Res. 1989, 49, 4385.
7. Zhao, H.; Lee, C.; Sai, P.; Choe, Y. H.; Boro, M.; Pendri,
A.; Guan, S.; Greenwald, R. B. J. Org. Chem. 2000, 65,
4601.
8. Leu, Y. L.; Roffler, S. R.; Chern, J. W. J. Med. Chem.
1999, 42, 3623.
9. Greenwald, R. B.; Pendri, A.; Conover, C.; Gilbert, C.;
Yang, R.; Xia, J. J. Med. Chem. 1996, 39, 1938.
10. Greenwald, R. B.; Pendri, A.; Conover, C. D.; Lee, C.;
Choe, Y. H.; Gilbert, C.; Martinez, A.; Xia, J.; Wu, D.;
Hsue, M. Bioorg. Med. Chem. 1998, 6, 551.
11. Shabat, D.; Rader, C.; List, B.; Lerner, R. A.;
Barbas, C. F., III Proc. Natl. Acad. Sci. U.S.A. 1999,
96, 6925.
12. Vrudhula, V. M.; Senter, P. D.; Fischer, K. J.; Wallace,
P. M. J. Med. Chem. 1993, 36, 919.
13. Shabat, D.; Lode, H.; Pertl, U.; reisfeld, R. A.; Rader, C.;
Lerner, R. A.; Barbas, C. F., III Proc. Natl. Acad. Sci.
U.S.A. 2001, 98, 7528.
Acknowledgements
14. Wagner, J.; Lerner, R. A.; Barbas, C. F., III Science
(Washington, D. C.) 1995, 270, 1797.
15. Duncan, R.; Gac-Breton, S.; Keane, R.; Musila, R.; Sat,
Y. N.; Satchi, R.; Searle, F. J. Control Release 2001, 74,
135.
This work was supported by a grant from Reckanati
Foundation to GAand DS, and in part by the Israel
Science Foundation to GA.
16. Satchi, R.; Connors, T. A.; Duncan, R. Br. J. Cancer
2001, 85, 1070.
17. Satchi-Fainaro, R.; Hailu, H.; Davies, J. W.; Summer-
ford, C.; Duncan, R. Bioconjug. Chem. 2003, 14, 797.
18. Satchi-Fainaro, R.; Wrasidlo, W.; Lode, H. N.; Shabat,
D. Bioorg. Med. Chem. 2002, 10, 3023.
References and notes
1. Wrasidlo, W.; Schroder, U.; Bernt, K.; Hubener, N.;
Shabat, D.; Gaedicke, G.; Lode, H. Bioorg. Med. Chem.
Lett. 2002, 12, 557.