Aldolase antibody activation of prodrugs of potent aldehyde-containing cytotoxics for selective chemotherapy

Article abstract of DOI:10.1002/chem.200400419

Prodrugs of potent aldehyde analogues of the anticancer drug doxorubicin (Dox) were synthesized. These prodrugs were efficiently activated by antibody 93F3 and no drug formation was observed in the absence of 93F3 in either phosphate buffered saline or cell culture media. In the presence of antibody 93F3, these prodrugs were activated and decreased the proliferation of human cancer cells in in vitro proliferation assays.

Full text of DOI:10.1002/chem.200400419

FULL PAPER
Aldolase Antibody Activation of Prodrugs of Potent Aldehyde-Containing
Cytotoxics for Selective Chemotherapy
Subhash C. Sinha,*[a] Lian-Sheng Li,[a] Shin-ichi Watanabe,[b] Eiton Kaltgrad,[b]
Fujie Tanaka,[b] Christoph Rader,[b, c] Richard A. Lerner,*[b] and Carlos F. Barbas, III*[b]
Abstract: Prodrugs of potent aldehyde analogues of the anticancer drug doxorubi-
cin (Dox) were synthesized. These prodrugs were efficiently activated by antibody
93F3 and no drug formation was observed in the absence of 93F3 in either phos-
phate buffered saline or cell culture media. In the presence of antibody 93F3,
these prodrugs were activated and decreased the proliferation of human cancer
cells in in vitro proliferation assays.
Keywords: aldolase · antibodies ·
chemotherapy
prodrugs
·
doxorubicin
·
Introduction
dolase mAb 38C2,[4,5] which can be humanized,[6] for the
prodrug activation. As described in previous studies, mAb
38C2 effectively catalyzed the activation of the prodrugs of
enediyne analogues,[7] camptothecin, etoposide, and doxoru-
bicin (Dox, 1).[8,9]
To achieve selective chemotherapy with potent chemothera-
peutics, new approaches, including antibody-directed
enzyme prodrug therapy (ADEPT), are being developed.[1]
Catalytic monoclonal antibodies (mAb)[2] were suggested as
catalysts for prodrug activation almost a decade ago.[3] In
principle, catalytic mAbs are superior to both systemically
expressed endogenous human enzymes and externally ex-
pressed non-mammalian enzymes. For example, a catalytic
mAb can have unique substrates that are not acted on by
natural enzymes, thereby increasing the chemical space
available for prodrug design. In addition, a humanized cata-
lytic mAb should be less immunogenic than a non-mamma-
lian enzyme. Hence, we have investigated the use of the al-
As a key constituent of chemotherapeutic regimens, Dox
has been used as the first treatment for a variety of can-
cers.[10] In vitro studies have revealed that it is toxic to most
cancer cells in the low micromolar or sub-micromolar range.
However, the use of Dox is limited by its systemic toxicity
and the ability of cancer cells to develop resistance to it. To
have better efficacy and to counter acquired resistance, a
number of Dox derivatives have been synthesized that are
two to three orders of magnitude more toxic than the parent
molecule 1. These potent derivatives include 2-pyrrolinodox-
orubicin (or pyrrolino-Dox, 2) and 1,3-tetrahydropyridino-
doxorubicin (or tetrahydropyridino-Dox, 3) (Scheme 1).[11]
The toxicity of 2 and 3 is believed to originate through the
Schiff base intermediates I and II. In an independent
study,[12] it was shown that the products obtained from ester-
ase hydrolysis of diacetoxy derivatives of 4 and 5, presuma-
bly intermediates I and II, were toxic to the cells with sever-
al orders of magnitude higher potency than Dox itself. They
[a] Prof. Dr. S. C. Sinha, Dr. L.-S. Li
The Skaggs Institute for Chemical Biology and
Department of Molecular Biology, The Scripps Research Institute
10550 North Torrey Pines Road, La Jolla, CA 92037 (USA)
Fax : (+1)858-784-8732
E-mail: subhash@scripps.edu
[b] Dr. S.-i. Watanabe, E. Kaltgrad, Prof. Dr. F. Tanaka,
Prof. Dr. C. Rader, Prof. Dr. R. A. Lerner, Prof. Dr. C. F. Barbas, III
The Skaggs Institute for Chemical Biology and
Departments of Molecular Biology and Chemistry
The Scripps Research Institute
10550 North Torrey Pines Road, La Jolla, CA 92037 (USA)
Fax : (+1)858-784-2583
E-mail: rlerner@scripps.edu
carlos@scripps.edu
[c] Prof. Dr. C. Rader
Present address:
Experimental Transplantation and Immunology Branch
Center for Cancer Research, National Cancer Institute
National Institutes of Health
9000 Rockville Pike, Bethesda, MD 20892-1907 (USA)
Chem. Eur. J. 2004, 10, 5467 – 5472
DOI: 10.1002/chem.200400419
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5467

FULL PAPER
7 and 8 could also be obtained by aldolase antibody-cata-
lyzed retro-aldol reactions[15,16] of prodrugs 9 and 10, respec-
tively, the products of which will then undergo intramolecu-
lar cyclization to afford the corresponding carbinolamine de-
rivatives and then dehydrate to produce iminium intermedi-
ates I and II (Scheme 1). Furthermore, unlike compounds 4
and 5, prodrugs 9 and 10 are expected to be more stable
under physiological conditions where esterase activity is
abundant. In addition, since substitution of the primary
amine functionality of doxorubicin decreases its toxicity in
general,[9] prodrugs 9 and 10 are expected to be less toxic in
comparison to the parent molecule 1, as well as to 2-pyrroli-
no-Dox (2, AN-201) and 1,3-tetrahydropyridino-Dox (3,
AN-205).[11]
Prodrugs 9 and 10 were designed as potential substrates
for aldolase mAb93F3. This aldolase antibody catalyzes
retro-aldol reactions of a wide variety of substrates with a
very high catalytic rate ((kcat/Km)/kun > 1013 mÀ1) and enan-
tioselectivity. We noted that in 93F3-catalyzed reactions,[14,16]
a b-substituted b-hydroxyethyl ethyl ketone was retro-aldol-
ized faster than the corresponding b-substituted b-hydrox-
yethyl methyl ketone. Hence, both prodrugs 9 and 10 were
designed to contain a b-hydroxyethyl ethyl ketone as the
aldol-triggered linker. Furthermore, only the “S” enantiom-
er (considering the alkyl ketone as second largest group) of
an aldol compound was consumed by mAb93F3 leaving the
“R” enantiomer intact. Therefore, in order to allow for the
highest possible activation using aldolase mAb93F3, pro-
drug 9 contained an enantiomerically pure linker with an S
stereochemistry.
Scheme 1. Proposed mechanism for the observed potent cytotoxicity of
Dox analogues, 2 and 3, through DNA-adduct formation via intermedi-
ates I and II from 4 and 5 by an esterase-catalyzed reaction or the antici-
pated activation of the prodrugs
9 and 10, by using an aldolase
The prodrugs were synthesized starting from Dox (1) and
(4S)-octan-4-(tert-butyldimethyl-silyloxy)-6-keto-1-aldehyde
(11) for prodrug 9 or 1-iodo-7-keto-nonan-5-ol (12) for pro-
drug 10 as shown in Scheme 2. The linkers, 11 and 12, were
prepared starting from racemic 5-hexene-1,2-epoxide, (Æ)-
13, and aldehyde 15,[17] respectively. The enantiomerically
pure epoxide (S)-13 was obtained by kinetic resolution of
the corresponding racemic epoxide (Æ)-13.[18] Compound
(S)-13 was then reacted with 1-buten-2-yl magnesium bro-
mide in the presence of catalytic CuI to afford alcohol 14a
that was then converted to TBS ether 14b. Both alkene
functions were cleaved by dihydroxylation by using OsO4
and NMO followed by treatment with Pb(OAc)4 to afford
the TBS-protected linker 11. Aldehyde 11 was treated with
1 to give the corresponding imine. The latter was reduced in
situ with NaCNBH3; subsequently the TBS group in the
product was deprotected by using a solution of HF·Py and
pyridine in THF to afford the desired prodrug 9. For the
synthesis of linker 12, aldehyde 15 underwent aldol reaction
with 2-butanone using LDA as a base. The TBS group in
16a was deprotected using HF·Py to give 16b and the pri-
mary alcohol was then converted to iodide 12. Finally, alky-
lation of the amine of Dox (1), by using iodide 12 afforded
prodrug 10.
mAb93F3, via aldehydes 7 and 8.
were also toxic to the Dox-resistant cells. By using 5, it was
also shown that the product caused DNA–DNA crosslinks
in HL-60 and HL-60/AMSA cell lines, which suggests that
this mechanism contributes to its marked potency.[13] It is be-
lieved that the intermediates I and II undergo nucleophilic
attack by a proximate base, such as the 2-amino group of a
guanine residue leading to the corresponding DNA adducts
(Scheme 1). The hydrolysis of the glycosides in the DNA ad-
ducts followed by subsequent reaction with an additional
DNA molecule then produce free doxorubicinol aglycone
and double-DNA adducts 6. Therefore, one could imagine
that a process, which can selectively generate the markedly
potent intermediates, I or II, from the corresponding less
toxic prodrugs will be highly useful for the ADEPT ap-
proach. In this communication, we describe the synthesis
and in vitro evaluation of the novel prodrugs of aldehyde-
based Dox-analogues that can be activated using catalytic
aldolase mAb93F3.[14]
Results and Discussion
Design and synthesis of prodrugs: In the study described
above,[12] intermediates I and II were obtained from the cor-
responding diacetoxy derivatives 4 and 5, via the corre-
sponding aldehydes 7 and 8. We anticipated that aldehydes
Prodrug activation: Prodrug 9 (500mm) was incubated with a
catalytic amount of 93F3 (34mm) in PBS buffer at 378C, and
the activation of prodrug was analyzed by LCMS. Reactions
in PBS buffer (pH 7.4) and 10% fetal calf serum in cell cul-
5468
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 5467 – 5472

Prodrugs
5467 – 5472
Scheme 2. Synthesis of prodrugs 9 and 10.
Figure 2. a) Activation of prodrug 9 and production of drug 2 using aldo-
lase mAb93F3. Shown are the consumption of 9 and production of 2
ture medium (pH 7.4)[19] were used to assess the uncatalyzed
reaction. The results are shown in Figures 1 and 2. It is evi-
dent from Figure 1 (data collected after 10 h of incubation
at 378C) that the product, assigned as 2 based on molecular
weight, is formed only in the 93F3-catalyzed reactions (A),
but not in the background reactions either with the PBS (B)
or with the cell culture medium (C). This clearly showed
that the activation of the prodrug 9 was 93F3-dependent.
The reaction with 93F3 produced about 35% yield of drug 2
with an approximate rate of 0.009minÀ1, before reaching a
plateau (see, Figure 2a).
~
^
over time in the 93F3-catalyzed reaction (9/cat and 2/cat , respective-
&
ly) and the consumption of 9 in the background reaction (9/BKG ). In
these experiments, compound 9 (500mm) was incubated at 378C with
93F3 (34mm) in PBS buffer or PBS buffer alone. The reaction mixtures
were periodically analyzed by using LCMS equipped with column, UV
detector (254 nm), and EI-MS. b) Comparison of the stability of prodrug
9 and Dox (1) in PBS buffer, and 10% fetal calf serum in cell culture
medium (“S” stands for serum). In these experiments, prodrug 9 (500mm)
or Dox (1) (500mm) were incubated at 378C in PBS buffer (pH 7.4) alone
&
^
*
(9: , 1: ) or in 10% fetal calf serum in cell culture medium (9: , 1:
~
), and analyzed by using LCMS as described above.
not observed in the absence of mAb93F3, it is evident that
prodrug 9 degrades by a pathway similar to that of the pa-
rental drug Dox, 1, which is known in the literature to de-
grade rapidly in plasma and blood.[20] In various cell culture
media, the half-life of Dox has been estimated to be only
3 h.[21] We compared the stability of Dox and prodrug 9 in
PBS and the cell culture media at 378C (Figure 2b). The de-
composition of Dox and prodrug 9 are comparable, which
suggests that the formation of intermediate I and II from
the activation of prodrugs 9 and 10 will be achieved only by
mAb93F3, and thus they may not be formed in vivo by en-
dogenous enzyme-mediated reactions.
Figure 1. Comparison of the chromatograms (B) from antibody 93F3-cat-
alyzed activation of prodrug 9 to produce drug 2, and background reac-
tions in A) PBS buffer (pH 7.4) and C) 10% fetal calf serum in cell cul-
ture medium. In these experiments, compound 9 (500mm) was incubated
at 378C with 93F3 (34mm) in given buffer for 10 h and the reaction mix-
tures were analyzed by using LCMS equipped with column, UV detector
(254nm), and EI-MS. The unconsumed prodrug 9 or the produced drug 2
is shown as percent on y axis. Retention times for 2 and 9 are 5.15 min
and 5.30 min, respectively, as shown on the x axis of the chromatogram-
s A)–C).
In vitro evaluations: Biological activities of prodrugs 9 and
10 were evaluated, in vitro, by tumor cell proliferation
assays using human Kaposiꢁs sarcoma (SLK) and breast
cancer (MDA-MB-435) cell lines. The experiments with pro-
drugs 9 and 10 both in the presence and absence of antibody
93F3 were carried out as reported earlier,[9] and the results
are shown in Figures 3 and 4. As shown in Figure 3, prolifer-
ation of human breast cancer MDA-MB-435 and human Ka-
posiꢁs sarcoma cells was inhibited by 50% in the presence of
prodrug 10 (2mm) and mAb93F3 (1mm). Doxorubicin also
As shown in Figure 2a and b, the concentration of pro-
drug 9 also decreased in the absence of mAb93F3, especial-
ly in the cell culture medium. Since the formation of 2 was
Chem. Eur. J. 2004, 10, 5467 – 5472
www.chemeurj.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5469

S. C. Sinha, R. A. Lerner, C. F. Barbas, III, et al.
FULL PAPER
Figure 4. Growth inhibition of human Kaposiꢁs sarcoma (SLK) cell line
by prodrug 9 in the presence and absence of 1mm mAb93F3. 9: &,
~
.
9+93F3:
the linkers of prodrugs 9 and 10 were selectively activated
using the proficient aldolase mAb93F3. Prodrugs 9 and 10
effectively inhibited the proliferation of cancer cells upon
activation with mAb93F3 as observed by in vitro evaluation
by using human breast cancer (MDA-MB-435) and Kaposiꢁs
sarcoma (SLK) cell lines. These characteristics suggest that
prodrugs 9 and 10 are promising candidates for in vivo
study.
Figure 3. Growth inhibition of a) human breast cancer cell line (MDA-
MB-435) and b) human Kaposiꢁs sarcoma (SLK) cell line by prodrug 10
in the presence and absence of 1mm mAb93F3. The growth inhibition in
Experimental Section
^
the presence of Dox (1) is shown for comparison. 1: , 10: &, 10+93F3:
General methods: TLC was performed on glass sheets precoated with
silica gel (Merck, Kieselgel 60, F254, Art. 5715). Column chromatograph-
ic separations were performed on silica gel (Merck, Kieselgel 60, 230–400
mesh, Art. No. 9385) under pressure. All commercially available reagents
were used without further purification. Solvents were either used as pur-
chased or distilled by using common practices where appropriate. All re-
actions were carried out under dry argon.
~
.
showed toxicity identical to the mixture of prodrug 10 (2mm)
and mAb93F3 (1mm). In contrast, 2mm prodrug 10 in the ab-
sence of mAb93F3 was nontoxic.
Compound (S)-13: A mixture of (Æ)-13 (1.96 g, 20 mmol), (S,S)-salen
CoIIIOAc catalyst (136 mg, 0.2 mmol), and H2O (198 mL, 11 mmol) was
stirred at room temperature for 24 h. The product was purified by distil-
lation. 1H NMR (500 MHz, CDCl3): d = 5.84 (m, 1H), 5.06 (m, 1H),
5.01 (m, 1H), 2.94 (m, 1H), 2.76 (dd, J=5.1, 4.1 Hz, 1H), 2.49 (dd, J=
5.1, 2.9 Hz, 1H), 2.26–2.21 (m, 2H), 1.70–1.60 ppm (m, 2H).
Next, we studied the effect of prodrug 9 in the presence
and absence of mAb93F3 using human Kaposiꢁs sarcoma
cells (SLK) and compared the results using prodrug 10. The
IC50 of prodrug 9 was found to be 0.06mm (Figure 4) in the
presence of mAb93F3, whereas the IC50 of prodrug 10 was
only 2mm (Figure 3) under the same conditions. These re-
sults are consistent with the higher toxicity of 2 in compari-
son to 3. It is noteworthy that drug 2 and prodrug 9 generate
intermediate I, whereas drug 3 and prodrug 10 produce in-
termediate II. In both cases, the prodrugs were less toxic in
the absence than in the presence of mAb93F3. Prodrugs 9
and 10 alone were 20–30 times less toxic than in the pres-
ence of mAb93F3, which is comparable to prodrug-to-drug
toxicity ratios of other doxorubicin prodrugs.[9]
Compound 14a:
A
solution of 1-buten-2-yl magnesium bromide
(12.76 mL, 0.75m in THF), prepared from 2-bromo-1-butene (2.02 g,
15 mmol) and Mg (0.4 g, 16.5 mmol) in dry THF (17 mL), was added to a
stirred suspension of CuI (155 mg, 0.8 mmol) in anhydrous THF (5 mL)
under argon at À108C. Then the yellow suspension was further cooled to
À208C, a solution of epoxide 13 (313 mg, 3.19 mmol) in dry THF (5 mL)
was added dropwise. The resulting mixture was stirred from À20 to 08C
for 2 h, it was worked up by using aqueous solution of NH4Cl and Et2O,
the organic layer was washed with NH4OH and brine, dried over MgSO4.
After filtration, the solvents were removed under vacuum, and the resi-
due was purified over silica gel by using EtOAc/hexanes to afford pure
14a (460 mg, 94%). 1H NMR (500 MHz, CDCl3): d
= 5.89–5.81 (m,
In conclusion, we have synthesized novel prodrugs, 9 and
10, of two Dox analogues, pyrrolino-Dox and tetrahydropyr-
idino-Dox and evaluated their toxicities in cell culture
assays. These prodrugs, unlike many other prodrugs, con-
tained short retro-aldol activated linker. Unlike other pro-
drugs designed to work together with aldolase antibod-
ies,[7,9,22] coupling of other reactions together with the retro-
aldol reaction is not required for drug activation. Moreover,
1H), 5.05 (m, 1H), 4.97 (m, 1H), 4.90 (s, 1H), 4.83 (s, 1H), 3.76–3.71 (m,
1H), 2.28–1.98 (m, 6H), 1.60–1.55 (m, 2H), 1.05 ppm (t, J=7.7 Hz, 3H);
[a]D =À10.9 (c = 1.0, CHCl3).
Compound 14b: TBSCl (590 mg, 3.9 mmol) was added to a solution of
14a (460 mg, 3.0 mmol) and imidazole (408 mg, 6.0 mmol) in dry DMF
(1 mL) at 0 C. After the mixture was stirred at this temperature for 12 h,
it was worked up by using water and Et2O. The combined organic layer
was washed with brine, dried over MgSO4, and filtered. The solvents
were removed under vacuum, and the residue was purified over silica gel
5470
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 5467 – 5472

Prodrugs
5467 – 5472
by using EtOAc/hexane 1:50 to afford pure 14b (790 mg, 98%). 1H
NMR (300 MHz, CDCl3): d = 5.87–5.74 (m, 1H), 5.03–4.91 (m, 2H),
4.76 (s, 1H), 4.72 (s, 1H), 3.80 (m, 1H), 2.27–1.99 (m, 6H), 1.63–1.42 (m,
2H), 1.03 (t, J=7.2 Hz, 3H), 0.90 (s, 9H), 0.06 (s, 3H), 0.04 ppm (s, 3H);
MS: m/z: 269 [M+Na]+; [a]D =À11.1 (c = 1.0, CHCl3).
(dd, J=2.6, 17.6 Hz, 1H), 2.58 (dd, J=9.2, 17.6 Hz, 1H), 2.53 (q, J=
7.3 Hz, 2H), 1.65–1.42 (m, 6H), 1.14 (t, J=7.3 Hz, 3H), 0.97 (s, 9H),
0.12 ppm (s, 6H); 13C NMR (125 MHz, CDCl3): d = 212.2, 67.4, 62.8,
48.5, 36.5, 36.1, 32.4, 25.7, 21.6, 7.3, À5.5 ppm; MALDI-FTMS: m/z:
calcd for C15H32O3SiNa: 311.2013; found 311.2003 [M+Na]+.
Compound 11: OsO4 (0.2m in toluene, 0.1 mL, 0.02 mmol) and NMO
(50% w/w in H2O, 0.23 mL, 1.1 mmol) were added to a solution of 14b
(100 mg, 0.37 mmol) in acetone/water (3:1, 4 mL) at room temperature.
After the mixture was stirred for 12 h at this temperature, excess oxidants
were destroyed by using 10% solution of Na2S2O3 and then the resulting
mixture was extracted with EtOAc. The combined organic layer was
dried over anhydrous Na2SO4. The insoluble materials were filtered out
and the solvents were removed under vacuum. The residue was passed
over a short bed of silica gel (EtOAc), and the product (tetrol) was taken
to next step without further purification.
Compound 16b: HF·Py (2 mL) was added to a solution of 16a (770 mg,
2.67 mmol) in THF (10 mL) at À208C. The mixture was stirred at this
temperature for 2 h, and then neutralized using NaHCO3 solution and ex-
tracted with EtOAc. The combined organic layer was washed with brine,
dried over MgSO4, filtered, and concentrated under reduced pressure.
The residue was purified over silica gel (hexane/EtOAc 3:1) to afford
1
16b (300 mg, 65%). H NMR (400 MHz, CDCl3): d = 4.05 (m, 1H), 3.65
(t, J=6.3 Hz, 2H), 3.2 (br, 1H), 2.61 (dd, J=2.9, 17.6 Hz, 1H), 2.51 (dd,
J=9.1, 17.6 Hz, 1H), 2.45 (q, J=7.3 Hz, 2H), 1.80–1.37 (m, 7H),
1.06 ppm (t, J=7.3 Hz, 3H); 13C NMR (100 MHz, CDCl3): d = 212.4,
67.3, 61.9, 48.8, 36.6, 36.0, 32.0, 21.4, 7.3 ppm.
Lead tetraacetate (576 mg, 1.3 mmol) was added in portions to a solution
of the above-described tetrol in CH2Cl2 (5 mL) at 08C. After the mixture
was stirred for 2 h, it was worked up using aqueous Na2S2O5 and EtOAc.
The combined organic layer was dried over anhydrous Na2SO4. The in-
soluble materials were filtered out and the solvents were removed under
vacuum. The residue was purified over silica gel (hexane/EtOAc 4:1), to
Compound 12: Iodine (349 mg, 1.37 mmol) was added in portion to a so-
lution of 16b (171 mg, 0.98 mmol), PPh3 (386 mg, 1.47 mmol) and pyri-
dine (0.24 mL, 2.94 mmol) in benzene (20 mL) at room temperature. The
mixture was stirred under reflux for 30 min. After cooling to room tem-
perature, the mixture was filtered by Celite and washed with EtOAc. The
filtrate was concentrated and the residue was purified over silica gel
(hexane/EtOAc 20:1) to afford iodide 12 (215 mg, 77%); 1H NMR
(500 MHz, CDCl3): d = 4.02 (m, 1H), 3.18 (t, J=7.0 Hz, 2H), 3.15 (br,
1H), 2.60 (dd, J=2.6 Hz, 17.6 Hz, 1H), 2.50 (dd, J=9.2 Hz, 17.6 Hz,
1H), 2.45 (q, J=7.3 Hz, 2H), 1.88–1.77 (m, 2H), 1.60–1.35 (m, 4H),
1.05 ppm (t, J=7.3 Hz, 3H); 13C NMR (125 MHz, CDCl3): d = 212.7,
67.2, 48.4, 36.7, 35.1, 33.2, 26.4, 7.5, 6.8 ppm.
1
afford pure aldehyde 11 (60 mg, 65%). H NMR (300 MHz, CDCl3): d =
9.75 (s, 1H), 4.24 (dt, J=11.1, 6.0 Hz, 1H), 2.61 (dd, J=15.6, 6.3 Hz,
1H), 2.50 (dd, J=7.5, 1.8 Hz, 1H), 2.45 (m, 1H), 2.43 (dd, J=15.6,
5.7 Hz, 1H), 1.92–1.68 (m, 2H), 1.02 (t, J=7.5 Hz, 3H), 0.86 (s, 9H), 0.06
(s, 3H), 0.01 ppm (s, 3H); 13C NMR (75 MHz, CDCl3): d = 209.4, 201.7,
67.8, 49.5, 39.6, 37.6, 29.6, 25.9, 18.0, 7.6 ppm; MS: m/z: 241 [M+H]+,
263 [M+Na]+; [a]D =17.8 (c = 0.7, CHCl3).
Prodrug 9: A solution of NaBH3CN (1m in THF) (24 mL, 0.67 equiv) was
added to a stirred solution of Dox-hydrochloride (20 mg, 0.035 mmol)
and aldehyde 11 (27.7 mg, 0.102 mmol) in CH3CN/H2O (2:1, 5 mL). The
mixture was stirred at room temperature in the dark for 2 h. The reaction
mixture was diluted with water and extracted repeatedly (10”10 mL)
with a mixture of CHCl3/MeOH 5:1. The combined organic layer was
dried over Na2SO4. After filtration, the solvents were removed under
vacuum, and the residue was purified by using PTLC (CH2Cl2/MeOH
5:1) to afford the TBS ether protected 9 (5.8 mg, 22%). 1H NMR
(500 MHz, CDCl3): d = 8.01 (d, J=7.3 Hz, 1H), 7.78 (t, J=8.1 Hz, 1H),
7.39 (d, J=8.1 Hz, 1H), 5.54 (brs, 1H), 5.30 (brs, 1H), 4.77 (s, 2H), 4.13
(m, 1H), 4.08 (s, 3H), 4.01 (q, J=6.7 Hz, 1H), 3.79 (brs, 1H), 3.49 (s,
1H), 3.26 (d, J=18.7 Hz, 1H), 3.00 (d, J=18.7 Hz, 1H), 2.99 (m, 1H),
2.70 (m, 2H), 2.56 (dd, J=15.4, 7.0 Hz, 1H), 2.44–2.36 (m, 4H), 2.16 (m,
1H), 1.95 (m, 1H), 1.77 (m, 1H), 1.59 (m, 2H), 1.46 (m, 2H), 1.35 (d, J=
6.6 Hz, 3H), 0.98 (t, J=7.3 Hz, 3H), 0.78 (s, 9H), À0.01 (s, 3H),
À0.04 ppm (s, 3H); MS: m/z: 800 [M+H]+, 798 [MÀH]À, 834 [M+Cl]À.
A solution of HF·Py (0.01 mL) in THF/pyridine (4:1, 1 mL) was added to
a solution of the product described above in THF (0.5 mL) at 08C. The
mixture was stirred at this temperature for 12 h, and then neutralized
using NaHCO3 solution and extracted 5” with CH2Cl2/MeOH 5:1. Sol-
vents were removed under vacuum and the residue was purified by using
PTLC (CH2Cl2/MeOH 5:1) to afford pure 9 (1.4 mg, 52%). 1H NMR
(500 MHz, CDCl3): d = 8.04 (dx, J=7.7 Hz d,), 7.81 (t, J=7.7 Hz, 1H),
7.41 (d, J=8.5 Hz, 1H), 5.56 (d, 1H), 5.28 (m, 1H), 4.76 (s, 2H), 4.10 (s,
3H), 4.04 (q, J=6.6 Hz, 1H), 3.87 (s, 1H), 3.46 (s, 1H), 3.43 (m, 1H),
3.28 (dd, J=19.0 Hz, 1H), 3.10 (m, 1H), 3.04 (d, J=19.0 Hz, 1H), 2.84
(m, 1H), 2.76 (m, 1H), 2.68 (dd, J=17.6, 9.2 Hz, 1H), 2.51 (dd, J=17.2,
3.3 Hz, 1H), 2.46 (m, 2H), 2.36 (dt, J=14.7 Hz, 1H), 2.17 (dd, 1H),
1.95–1.60 (m, 4H), 1.55 (m, 2H), 1.30 (d, J=7.4 Hz, 3H), 1.04 ppm (t,
J=7.4 Hz, 3H); MS: m/z: 686 [M+H]+, 708 [M+Na]+, 684 [MÀH]À, 720
[M+Cl]À.
Prodrug 10: Iodide 12 (147 mg, 0.517 mmol) was added to a solution of
Dox-hydrochloride (10.0 mg, 0.0172 mmol) in DMF (0.2 mL) followed by
iPr2NEt (6.0 mL, 0.034 mmol), and the mixture was stirred overnight at
room temperature for 16 h. The mixture was purified over silica gel
(CH3Cl/MeOH 10:1) to afford 10 (7.2 mg, 60%). 1H NMR (600 MHz,
CD3OD): d = 7.94 (d, J=7.8 Hz, 1H), 7.83 (t, J=7.8 Hz, 1H), 7.57 (d,
J=7.8 Hz, 1H), 5.47 (s, 1H), 5.10 (s, 1H), 4.74 (d, J=18 Hz, 1H), 4.70
(d, J=18 Hz, 1H), 4.60 (brs, 1H), 4.27 (q, J=6.6 Hz, 1H), 4.04 (s, 3H),
3.96 (m, 1H), 3.80 (s, 1H), 4.49 (m, 1H), 3.10 (d, J=19 Hz, 1H), 2.97 (d,
J=19 Hz, 1H), 2.96–2.70 (m, 3H), 2.56–2.40 (m, 4H), 2.35 (d, J=15 Hz,
1H), 2.20–1.95 (m, 3H), 1,70–1.30 (m, 6H), 1.30 (d, J=6 Hz, 3H),
0.97 ppm (t, J=7.3 Hz, 3H); MALDI-FTMS: m/z: calcd for
C36H45NO13Na: 722.2783; found 722.2810 [M+Na]+.
In vitro cell growth assay: Briefly, human breast cancer cells (MDA-MB-
435) and Kaposiꢁs sarcoma cells (SLK) were plated at a density of 5”103
cells per well in 96-well tissue culture plates and maintained in culture.
After 24 h, the media was gently removed from the 96-well plates and all
wells were washed 2” with media, without disturbing the cells. Prodrugs
were added immediately after washing. For the antibody experiments,
prodrug and 93F3IgG were mixed just before adding the activated pro-
drug solution to the cells. The final concentration of antibody in all solu-
tions was 1 mm. Each concentration of prodrug added in triplicate. After
prodrug addition, the cells were maintained at 378C in 5% CO2 for 1 h.
After incubation, 20 mL of the Promega Substrate MTS (2H-tetrazolium,
5-[3-carboxymethoxy)phenyl]-3-(4,5-dimethyl-2-thiazolyl)-2-(4-sulophen-
yl)-inner salt (9Cl)) were directly added to every well of the plate, which
already contained 100 mL of the prodrug (+antibody) sample. The MTS
substrate is quickly converted to a red formazan product in the presence
of lactate dehydrogenase, which is released from living cells. After 1 and
2 h, the absorbance at 490 nm was recorded in an ELISA plate reader to
quantify the number of surviving cells. Six wells without cells had been
kept blank throughout the entire experiment except for the addition of
the MTS substrate, and the A490 values from these wells was averaged
and subtracted from every other well. Three wells containing untreated
cells and three wells containing cells with only antibody added were aver-
aged, and the A490 values were set as 100% cell survival for comparison
with addition of prodrug and addition of prodrug+antibody, respectively.
The standard deviation for each triplicate experiment was also calculated
after correction of background and is reported in Figures 3 and 4.
Compound 16a: Butanone (4.5 mL, 50.8 mmol) in THF (5 mL) was drop-
wise added to a solution of LDA (1.9m in THF, 26.7 mL, 50.8 mmol) in
dry THF (30 mL) at À1008C. After the mixture was stirred for 15 min,
aldehyde 15 (10.0 g, 46.2 mmol) in THF (2 mL) was added and stirred for
3 h at À788C. The mixture was quenched using aqueous solution of
NH4Cl and extracted with EtOAc. The combined organic layer was
washed with brine, dried over MgSO4, filtered, and concentrated under
reduced pressure. The residue was purified over silica gel (hexane/
1
EtOAc 5:1) to afford 16a (5.06 g, 38%). H NMR (500 MHz, CDCl3): d
= 4.12 (m, 1H), 3.69 (t, J=6.3 Hz, 2H), 3.15 (d, J=2.6 Hz, 1H), 2.68
Chem. Eur. J. 2004, 10, 5467 – 5472
www.chemeurj.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5471

S. C. Sinha, R. A. Lerner, C. F. Barbas, III, et al.
FULL PAPER
[9] a) D. Shabat, C. Rader, B. List, R. A. Lerner, C. F. Barbas, III, Proc.
Natl. Acad. Sci. USA 1999, 96, 6925–6930. b) D. Shabat, H. N. Lode,
U. Pertl, R. A. Reisfeld, C. Rader, R. A. Lerner, C. F. Barbas, III,
Proc. Natl. Acad. Sci. USA 2001, 98, 7528–7533.
[10] J. Jassem, Semin. Oncol. 2003, 30 (2, Suppl. 3), 11–14.
[11] A. Nagy, P. Armatis, A. V. Schally, Proc. Natl. Acad. Sci. USA 1996,
93, 2464–2469.
Acknowledgement
This study was supported by The Skaggs Institute for Chemical Biology
and in part by the NIH (CA27489). We thank Dr. B. Mueller for provid-
ing us with the human breast cancer cell line, MDA-MB-435.
[12] D. Farquhar, A. Cherif, E. Bakina, J. A. Nelson, J. Med. Chem. 1998,
41, 965–972.
[13] L. A. Zwelling, E. Altshuler, A. Cherif, D. Farquhar, Cancer Res.
1991, 51, 6704–6707.
[1] a) K. D. Bagshawe, S. K. Sharma, P. J. Burke, R. G. Melton, R. J.
Knox, Curr. Opin. Immunol. 1999, 11, 579–583; b) K. N. Syrigos,
A. A. Epenetos, Anticancer Res. 1999, 19, 605–613; c) G. Xu, H. L.
McLeod, Clin. Cancer Res. 2001, 7, 3314–3324.
[14] G. Zhong, R. A. Lerner, C. F. Barbas, III, Angew. Chem. 1999, 111,
3957–3960; Angew. Chem. Int. Ed. 1999, 38, 3738–3740.
[15] G. Zhong, D. Shabat, B. List, J. Anderson, S. C. Sinha, R. A. Lerner,
C. F. Barbas, III, Angew. Chem. 1998, 110, 2609–2612; Angew.
Chem. Int. Ed. 1998, 37, 2481–2484.
[16] S. C. Sinha, J. Sun, G. Miller, C. F. Barbas, III, R. A. Lerner, Org.
Lett. 1999, 1, 1623–1626.
[17] S. J. Danishefsky, W. H. Pearson, J. Org. Chem. 1983, 48, 3865–3866.
[18] a) M. Tokunaga, J. F. Larrow, F. Kakiuchi, E. N. Jacobsen, Science
1997, 277, 936–938; b) M. E. Furrow, S. E. Schaus, E. N. Jacobsen, J.
Org. Chem. 1998, 63, 6776–6777.
[19] Dulbeccoꢁs modified eagle medium.
[20] N. Laubrock, G. Hempel, P. Schulze-Westhoff, G. Wurthwein, S.
Flege, J. Boos, Chromatographia 2000, 52, 9–13.
[21] E. J. Pavlik, D. E. Kenady, J. R. Van Nagell, Jr., M. B. Hanson, E. S.
Donaldson, S. Casper, D. Garrett, D. Smith, K. Keaton, R. C. Flani-
gan, Cancer Invest. 1984, 2, 449–458.
[22] M. Shamis, H. N. Lode, D. Shabat, J. Am. Chem. Soc. 2004, 126,
1726–1731.
[2] a) P. G. Schultz, J. Yin, R. A. Lerner, Angew. Chem. 2002, 114,
4607–4618; Angew. Chem. Int. Ed. 2002, 41, 4427–4437; b) F.
Tanaka, Chem. Rev. 2002, 102, 4885–4906.
[3] a) H. Miyashita, Y. Karaki, M. Kikuchi, I. Fujii, Proc. Natl. Acad.
Sci. USA 1993, 90, 5337–5340; b) D. A. Campbell, B. Gong, L. M.
Kochersperger, S. Yonkovich, M. A. Gallop, P. G. Schultz, J. Am.
Chem. Soc. 1994, 116, 2165–2166; c) P. Wentworth, A. Datta, D.
Blakey, T. Boyle, L. J. Partridge, G. M. Blackburn, Proc. Natl. Acad.
Sci. USA 1996, 93, 799–803.
[4] a) J. Wagner, R. A. Lerner, C. F. Barbas, III, Science 1995, 270,
1797–1800; b) F. Tanaka, C. F. Barbas, III, “Antibody-catalyzed
Aldol Reactions” in Modern Aldol Reactions, Vol 1, (Ed.: R. Mahr-
wald), Wiley-VCH, Weinheim, Germany, 2004, p. 273–310.
[5] F. Tanaka, C. F. Barbas, III, J. Immunol. Methods 2002, 269, 67–79.
[6] a) C. Rader, J. M. Turner, A. Heine, D. Shabat, S. C. Sinha, I. A.
Wilson, R. A. Lerner, C. F. Barbas, III, J. Mol. Biol. 2003, 332, 889–
899. b) F. Tanaka, R. Fuller, H. Shim, R. A. Lerner, C. F. Barbas III,
J. Mol. Biol. 2004, 335, 1007–1018.
[7] S. C. Sinha, L.-S. Li, S. Dutta, C. Rader, R. A. Lerner, Proc. Natl.
Acad. Sci. USA 2004, 101, 3095–3099.
Received: April 30, 2004
[8] C. Rader, B. List, Chem. Eur. J. 2000, 6, 2091–2095.
Published online: September 20, 2004
5472
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 5467 – 5472